United States
Environmental Protection
Agency
Integrated Science Assessment
for Oxides of Nitrogen and Sulfur — Ecological Criteria
EPA/600/R-08/082F
December 2008
-------
December 2008
EPA/600/R-08/082F
Integrated Science Assessment
for Oxides of Nitrogen and Sulfur —
Ecological Criteria
National Center for Environmental Assessment-RTP Division
Office of Research and Development
U.S. Environmental Protection Agency
Research Triangle Park, NC
-------
Disclaimer
This document has been reviewed in accordance with U.S. Environmental Protection Agency
policy and approved for publication. Mention of trade names or commercial products does not constitute
endorsement or recommendation for use.
-------
Executive Summary
Integrated Science Assessment
Oxides of Nitrogen and Sulfur
Ecological Criteria
December 2008 ¦ EPA/600/R-08/082F
U.S. Environmenta! Protection Agency | Office of Research and Development | National Center for Environmental Assessment
http://cfpub.epa.gov/ncea/cfm/recordisplay.cfm?deid=201485
Introduction
his Integrated Science Assessment
(ISA) is a synthesis and evaluation of
M the most policy-relevant science
that forms the scientific foundation for the
review of the secondary (welfare-based)
national ambient air quality standards
(NAAQS) for oxides of nitrogen (NOx) and
oxides of sulfur (SOx). The Clean Air Act defi-
nition of welfare effects includes, but is not
limited to, effects on soils, water, wildlife,
vegetation, visibility, weather, and climate,
as well as effects on man-made materials,
economic values, and personal comfort
and well-being. The current secondary
NAAQS for SOx, set in 1973, is a 3-hour aver-
age of 0.5 ppm1 sulfur dioxide (SO2), not to
be exceeded
more than
once per year.
The secondary
NAAQS for
NOx is identi-
cal to the pri-
mary standard
set in 1971: an
annual average not to exceed 0.053 ppm
nitrogen dioxide (NO2). The current secon-
dary NAAQS were set to protect against di-
rect damage to vegetation by exposure to
gas-phase SOx or NOx.
Scope
7
his ISA is focused on ecological
effects resulting from current deposi-
tion of compounds containing nitro-
gen (N) and sulfur (S). Acidification, N nutri-
ent enrichment and effects of sulfate (SO42
on methylation of mercury (Hg) are high-
lighted in the document. The following fig-
ure illustrates the scope of the document.
1 ppm= 1000 ppb
Atmospheric cycle
of N oxides
See figure 2-16
Ambient Air
Concentration
Sunlight
Dissolution
2H+ +SO42-
~ H++N03
Atmospheric cycle
of S compounds
See figure 2-18
Oxidation
S02 H2SO4
NO* ~HNOa
Wet Deposition
H*. NH4+, NO3-, S042
t/li
VOC
[
Foliar and
nutrient effects
Dry deposition
NOx, NHX, SO*
Deposition
C and N cycle in
terrestrial ecosystem
See figure 3-36
Acidification of water + Eutrophication
Hg cycle in
aquatic ecosystem
See figure 3-59
Ecological
Effect
C, N and P cycle in
aquatic ecosystem
See figure 3-40
-------
ISA for Oxides of Nitrogen arid Sulfur
Executive Summary
Both N and S contribute to acidification of
ecosystems. This ISA considers several
chemical forms that contribute to acidifying
deposition, including gases and particles
derived from SOx, NOx, and reduced nitro-
gen (NHx).
Deposition of N contributes to N-nutrient en-
richment and eutrophication. An assess-
ment of the complex ecological effects of
atmospheric N deposition requires consid-
eration of many different chemical forms of
reactive N (NrJ. For this reason, the ISA in-
cludes evaluation of data on the most
common reduced inorganic forms of N,
ammonia (NH3) and ammonium (NH4+); on
oxidized inorganic forms including nitric ox-
ide (NO) and NO2, nitrate (NOT), nitric acid
(HNO3), and nitrous oxide (N2O); and on or-
ganic N compounds including peroxyacetyl
nitrate (PAN).
Other welfare effects addressed in the ISA
include effects of SO42" deposition on Hg
methylation, along with evidence related to
direct exposure to gas-phase NOx and SOx.
The key conclusions of the ISA follow.
-------
ISA for Oxides of Nitrogen arid Sulfur
Executive Summary
Current concentrations and deposition
V?/ mbient annual NOx and SOx con-
centrations as reported in
the routine national networks have
decreased substantially owing to controls
enacted since the 1970s. NOx decreased
-35% in the period 1990-2005, to current an-
nual average concentrations of -15 ppb.
Emissions of SOx have been substantially re-
duced in recent years: annual average
ambient SOx concentrations have de-
creased -50% in the period 1990-2005 and
now stand at -4 ppb for both aggregate
annual and 24-hour average concentra-
tions nation-wide.
Emitted NOx, SOx, NHx and other pollutants
can be transported vertically by convection
into the upper part of the mixed layer on
one day, then transported overnight in a
layer of high concentrations. Once pollut-
ants are lofted to the middle and upper tro-
posphere, they typically have a much
longer lifetime and, with the generally
stronger winds at these altitudes, can be
transported long distances from their source
regions. The length scale of this transport is
highly variable owing to differing chemical
and meteorological conditions encoun-
tered along the transport path.
-------
ISA for Oxides of Nitrogen and Sulfur
Executive Summary
Numerical chemical-transport models
(CTMs) are the prime tools for computing
emissions and interactions among pollutants
like NOx, SOx, and NHx, their transport and
transformation including production of sec-
ondary aerosols like ammonium nitrate and
ammonium sulfate, the evolution of particle
size distributions, the resulting atmospheric
concentrations and the deposition of these
pollutants to the surface. CTMs are driven by
calculated emissions for primary species
such as NOx, SOx, NH3, and primary particu-
late matter, and by the meteorological
fields produced by other numerical predic-
tion models. As such, CTMs are the chief
means of relating emitted pollutants with
deposited ones.
The emitted, transported, and transformed
pollutants reach the surface where they can
have ecological effects largely through
deposition. Direct and indirect wet and dry
deposition to specific locations like water-
sheds depend on air pollutant emissions and
concentrations in the airshed above the wa-
tershed, but the shape and areal extent of
the airshed is quite different from that of the
watershed owing to the transport and trans-
formation of emitted pollutants described
above.
Deposition is spatially heterogeneous across
the U.S. In the years 2004-2006, routine na-
tional monitoring networks reported mean S
deposition in the U.S. highest east of the Mis-
sissippi River with the highest reported depo-
sition, 21 kg S/ha/yr, in the Ohio River Valley
where most recording stations reported
three-year averages for this period of more
than 10 kg S/ha/yr. Numerous other stations
in the eastern U.S. reported S deposition
greater than 5 kg S/ha/yr. Data are sparse
for the central U.S. between the 100th me-
ridian and the Mississippi River; but, where
available, deposition values there were
lower than in most of the eastern U.S., rang-
ing from 4 to over 5 kg S/ha/yr. Total S depo-
sition in the U.S. west of the 100th meridian is
lower than in the East or upper Midwest, ow-
ing to lower densities of high-emitting
sources in the West. In the years 2004-2006,
all routine recording stations in the West re-
ported less than 2 kg S/ha/yr and many re-
ported less than 1 kg S/ha/yr. S was primarily
deposited in the form of wet SO42-, followed
by a smaller proportion of dry SO2, and a
much smaller proportion of dry SO42-.
Expanding urbanization, agricultural intensi-
fication, and industrial production during the
previous 100 years have produced a nearly
10-fold increase in total N deposited from
the atmosphere compared to pre-industrial
levels. NOx, chiefly from fossil fuel combus-
tion, often dominates total N pollution in the
U.S. and comprises from 50 to 75% of current
4
-------
ISA for Oxides of Nitrogen anci Ann.
total N atmospheric deposition. This wet and
dry atmospheric N deposition is spatially
heterogeneous, too, owing to the influence
of meteorology, transport, precipitation pat-
terns and land use.
For 2004-2006, routine national monitoring
networks reported the highest mean N
deposition totals in the U.S. in the Ohio River
Valley, specifically in the states of Indiana
and Ohio, with values greater than 9 kg
N/ha/yr. N deposition was lower in other
parts of the East, including the Southeast
and in northern New England. In the central
U.S., the highest N annual average deposi-
tion totals were on the order of 6 to 7 kg
N/ha/yr. Measured concentrations and in-
ferred deposition totals were dominated by
wet NOs" and NH/ species, followed by
dry HNO3, dry NH/, and dry N03~. NH3 is not
yet measured routinely in any national net-
works; however, smaller-scale intensive
monitoring and numerical air quality model-
ing both indicate that it may account for
more than 80% of the dry reduced N deposi-
tion total.
The thin coverage of monitoring sites in
many locations, especially in the rural West,
means that limited data exist on deposition
totals in a large number of potentially sensi-
tive places. Numerical modeling experi-
ments can help fill in these data gaps and
suggest that local and even regional areas
of high ambient concentration and deposi-
tion exist where measured data are un-
available. Model-predicted values for N
deposition in some regions of the Adiron-
dacks in New York are greater than 20 kg
N/ha/yr; other model estimates as high as
32 kg N/ha/yr have been made for a region
of southern California, where more than half
of that total was predicted to come from
NO and NO2.
The ISA concludes that the national-scale
networks routinely monitoring N deposition
are inadequate to characterize both the full
range of reduced and oxidized forms of N
deposition and the substantial regional het-
erogeneity across the landscape of the U.S.
Although S and N deposition in most areas
of the U.S. occurred as wet deposition,
there were some exceptions, including parts
of California where N deposition was primar-
ily dry.
5
-------
ISA for Oxides of Nitrogen and Sulfur
Executive Summary
Ecological effects of acidification
~~ W he effects of acidifying deposition
m on ecosystems have been well
§ studied over the past several dec-
ades and vulnerable areas have been iden-
tified in the U.S. The wealth of data has led
to the development of widely used ecologi-
cal models for predicting soil and surface
water acidification. Regional and ecosys-
tem vulnerability to acidification results from
inherent sensitivity and exposure to acidify-
ing deposition.
Sensitivity of terrestrial and aquatic ecosys-
tems to acidification from S and N deposi-
tion is regional and predominantly governed
by surficial geology. Other factors contribut-
ing to the sensitivity of soils and surface wa-
ters to acidifying deposition include topog-
raphy, vegetation, soil chemistry, land use,
and hydrologic flowpath.
most sensitive to terrestrial effects from acidi-
fying deposition include forests in the Adi-
rondack Mountains of New York, the Green
Mountains of Vermont, the White Mountains
of New Hampshire, the Allegheny Plateau of
Pennsylvania, and high-elevation forest
ecosystems in the central and southern Ap-
palachians. While studies show some recov-
ery of surface waters, there are widespread
areas of ongoing depletion of exchange-
able base cations in forest soils in the north-
eastern U.S., despite recent decreases in
acidifying deposition.
In aquatic systems, consistent and coherent
evidence from multiple studies of many
species shows that acidification can cause
the loss of acid-sensitive species, and that
more species are lost with greater acidifica-
tion. These effects are linked to changes in
Soil acidification is a natural process,
but is often accelerated by acidifying
deposition, which can decrease con-
centrations of exchangeable base
cations in soils. Biological effects of
acidification on terrestrial ecosystems
are generally attributable to Al toxic-
ity and decreased ability of plant
roots to take up base cations. Areas
The evidence is sufficient to infer a causal
relationship between acidifying deposition
and effects on:
(1) biogeochemistry related to terrestrial and
aquatic ecosystems;
(2) biota in terrestrial and aquatic ecosystems.
6
-------
ISA for Oxides of Nitrogen and Sulfur
Executive Summary
surface water chemistry, including concen-
trations of SO42", NOs", inorganic Al, and cal-
cium (Ca2+), surface water pH, sum of base
cations, acid neutralizing capacity (ANC),
vertebrates, and fish species richness. These
effects on species richness may also affect
ecosystem services, such as biodiversity and
cultural services such as fishing and tourism.
Although both N and S deposition can
cause terrestrial and aquatic acidification, S
deposition is generally the primary cause of
chronic acidification, with secondary con-
tributions from N deposition. Following de-
creases in S deposition in the 1980s and
1990s, one quarter to one third of the
chronically acidic lakes and streams in the
U.S. were no longer acidic during baseflow
in the year 2000. A number of lakes and
streams, however, remain acidic even
though wet SO42- deposition has decreased
by as much as 30% since 1989. N deposition,
which has also decreased in the years since
1990 in most places in the U.S. with routine
monitoring, is the primary cause of episodic
acidification which, despite its short duration,
has been shown to cause long-term bio-
logical effects.
Many of the surface waters most sensitive to
acidification in the U.S. are found in the
Examples of biogeochemical indicators of
effects from acidifying deposition on ecosystems
Ecosystem
Biogeochemical Indicator
Terrestrial
• Soil base saturation
• Inorganic Aluminum concentration
in soil water
• Soil carbon-to-nitrogen ratio
Aquatic
• Sulfate
• Nitrate
• Base cations
• Acid neutralizing capacity
• Surface water inorganic Aluminum
• pH
Examples of biological indicators of
effects from acidifying deposition on ecosystems
Indicator
Measure
Terrestrial Ecosystems
Red Spruce
Sugar
Maple
• Percent dieback of canopy
trees
• Dead basal area, crown
vigor index, fine twig die-
back
Aquatic Ecosystems
Fishes, zoo- • Presence/absence
plankton, crus- • Fish condition factor
taceans, rotifers • Biodiversity
and base cation surplus. These effects are
also influenced by historical inputs to these
systems. Decreases in ANC and pH and in-
creases in inorganic Al concentration con-
tribute to declines in zooplankton, macroin-
7
-------
Executive Summary
ISA for Oxides of Nitrogen and Sulfur
0
1
a
CD
£
(/)
Q>
JD
E
3
z
14-
12-
10-
8-
6-
4-
2-
0-
-200
-100
100 200
ANC(peq/L)
300
—r~
400
500
Number of fish species per lake vs. acidity status, expressed
as acid neutralizing capacity (ANC), for Adirondack lakes.
Source: Sullivan et nL ?00Ab
Adirondack
Mountains.
Northeast, the Southeast, and the moun-
tainous West, In the West, acidic surface
waters are rare and the extent of chronic
surface water acidification that has oc-
curred to date has been limited. However,
episodic acidification does occur. In both
the mountainous West and the Northeast,
the most severe acidification of surface wa-
ters generally occurs during spring snowmelt.
The ISA highlights evidence from two well-
studied areas to provide more detail on
how acidification affects ecosystems: The
Adirondacks (NY) and Shenandoah Na-
tional Park (VA). In the Adirondacks, the cur-
rent rates of N and S deposition exceed the
amount that would allow recovery of the
most acid sensitive lakes. In the Shenan-
doah, past SO42- has accumulated in the soil
and is slowly released from the soil into
stream water where it causes acidification,
making parts of this
region sensitive to even
the current lower
deposition loadings.
Numerical models spe-
cifically calibrated to
these locations and
conditions suggest that
the number of acidic
streams will increase
even under current
deposition loads.
Regions of the northern and eastern U.S. that contain appreciable numbers of lakes
and streams sensitive to deleterious effects from acidifying deposition.
Source: Stoddard et aL, 2003
8
-------
ISA for Oxides of Nitrogen and Sulfur
Executive Summary
Ecological effects of nitrogen
rhere are many well-studied effects
of N deposition on ecosystems and
some vulnerable areas have been
identified in the U.S. However, the full extent
of ecosystem vulnerability is still unknown.
Substantial empirical information from spe-
cific ecosystems and for specific endpoints
is available, but given the complexity of the
N cycle, a broadly applicable and well-
tested predictive model of the ecological
effects of N deposition is not yet available.
Though the sensitivity of ecosystems to N
deposition
deposition across the U.S. varies, a large
body of evidence clearly demonstrates a
relationship between N deposition and a
broad range of ecological effects.
The contribution of N deposition to total N
load varies among ecosystems. Atmos-
pheric N deposition is the main source of
new N to most headwater streams, high
elevation lakes, and low-order streams. At-
mospheric N deposition contributes to the
total N load in terrestrial, wetland, freshwa-
ter, and estuarine ecosystems that receive N
through multiple pathways (i.e. biological N-
fixation, agricultural land runoff and waste
water effluent). There are multiple biogeo-
chemical indicators of N deposition effects.
Examples of biogeochemical indicators of
effects from reactive N deposition on ecosystems
Ecosystem
Biogeochemical Indicator
• NO3 leaching
• Nitrification
• Denitrification
Terrestrial
• N2O emissions
and
• CH4 emissions
Wetland
• Soil C:N ratio
• Foliar / plant tissue [N], C:N,
N:magnesium, N:phosphorus
• Soil water [NO3 ] and [NH4 ]
• Chlorophyll a
Freshwater
• Water [NOs ]
and
• Dissolved inorganic N
Estuarine
• Dissolved oxygen
• N:P
The evidence is sufficient to infer a causal rela-
tionship between N deposition, to which NOx
and NHx contribute, and the alteration of the
following:
(1) biogeochemical cycling of N and carbon
(C) in terrestrial, wetland, freshwater aquatic,
and coastal marine ecosystems;
(2) biogenic flux of methane (CH4), and N2O in
terrestrial and wetland ecosystems;
(3) species richness, species composition, and
biodiversity in terrestrial, wetland, freshwater
aquatic and coastal marine ecosystems.
9
-------
ISA for Oxides of Nitrogen and Sulfur
Executive Summary
In terrestrial ecosystems, the onset of NC>3~
leaching is one of the best documented
biogeochemical indicators that an ecosys-
tem receives more N than it uses and is able
to retain. N removal by ecosystems is a
valuable ecosystem service regulating wa-
ter quality. When atmospheric deposition of
N impairs the ability of terrestrial and
aquatic ecosystems to retain and remove
N, NOs" leaching occurs and the degrada-
tion of water quality can occur. The onset of
leaching was calculated to occur with
deposition levels between 5.5 and 10 kg
N/ha/yr for sensitive eastern forests. In the
mixed conifer forests of the Sierra Nevada
and San Bernardino mountains, the onset of
increased NC>3~ leaching was calculated to
be 17 kg N/ha/yr. Several studies in the
Rocky Mountains indicate that the capacity
of alpine catchments to retain N was ex-
ceeded at levels greater than 5-10 kg
N/ha/yr.
N deposition alters the biogenic sources and
sinks of two greenhouse gases (GHGs), ChU
and N2O, in terrestrial and wetland ecosys-
tems, resulting in increased emissions to the
atmosphere. Non-flooded upland soil is the
largest biological sink and takes up about
6% of atmospheric CH4. N addition de-
creases CH4 uptake in coniferous and de-
ciduous forests, and N addition increases
CH4 production in wetlands. Soil is the larg-
est source of N2O, accounting for 60% of
global emissions. N deposition increases the
biogenic emission of N2O in coniferous for-
est, deciduous forests, grasslands, and wet-
lands. Although N addition can cause a
general stimulation of biogenic CH4 and
N2O emissions from soils, it is difficult to gen-
eralize a dose-response relationship be-
tween the amount of N addition and the
changes in GHG flux on a large heteroge-
neous landscape. This is because GHG pro-
duction is influenced by multiple environ-
mental factors (e.g., soil, vegetation and
climate), which vary greatly over small spa-
tial and temporal scales.
N is often the most limiting nutrient to growth
in ecosystems. N deposition thus often in-
creases primary productivity, thereby alter-
ing the biogeochemical cycling of C. N
Examples of biological indicators of
effects from N deposition on ecosystems
Ecosystem
Biological Indicators
Terrestrial and
•
Altered community composi-
Wetlands
tion, biodiversity and/or popu-
lation decline. Taxa affected
include: diatoms, lichen, my-
corrhizae, moss, grasses and
other herbaceous plants
•
Plant root: shoot ratio
•
Terrestrial plant bio-
mass/production
Freshwater
•
Phytoplankton bio-
and Estuarine
mass/production
•
Toxic or nuisance algae
blooms
•
Submerged aquatic vegetation
•
Fauna from higher trophic lev-
els
10
-------
ISA for Oxides of Nitrogen arid Sulfur
Executive Summary
deposition can cause changes in ecosys-
tem C budgets. However, whether N depo-
sition increases or decreases ecosystem C-
sequestration remains unclear. A limited
number of studies suggest that N deposition
may increase C-sequestration in some for-
ests, but has no apparent effect on C-
sequestration in non-forest ecosystems.
In terrestrial ecosystems, N deposition can
accelerate plant growth and change C al-
location patterns (e.g. shoot:root ratio),
which can increase susceptibility to severe
fires, drought, and wind damage. These ef-
fects have been shown in studies con-
ducted in the western U.S. and Europe. The
alteration of primary productivity can also
alter competitive interactions among plant
species. The increase in growth is greater for
some species than others, leading to possi-
ble shifts in population dynamics, species
composition, community structure, and in
few instances, ecosystem type.
There are numerous sensitive terrestrial biota
and ecosystems that are affected by N
deposition. Acidophytic lichens are among
the most sensitive terrestrial taxa to N depo-
sition, with adverse effects occurring with
exposures as low as 3 kg N/ha/yr in the
Pacific Northwest and southern California.
The onset of declining biodiversity in grass-
lands has been estimated to be 5 kg
N/ha/yr in Minnesota and the European Un-
ion. Altered community composition of al-
pine ecosystems in the Rocky Mountains
and forest encroachment into temperate
grasslands in Southern Canada is estimated
to be 10 kg N/ha/yr.
The productivity of many freshwater ecosys-
tems is N-limited. N deposition can alter
species assemblages and cause eutrophi-
cation of aquatic ecosystems to the extent
that N is the growth-limiting nutrient. In the
Rocky Mountains, deposition loads of ap-
proximately 1.5-2 kg N/ha/yr are reported to
alter species composition in the diatom
communities in some freshwater lakes, an
indicator of impaired water quality.
11
-------
ISA for Oxides of Nitrogen and Sulfur
Executive Summary
In estuarine ecosystems, N from atmospheric
and non-atmospheric sources contributes to
increased phytoplankton and algal produc-
tivity, leading to eutrophication. Estuary eu-
trophication is an ecological problem indi-
cated by water quality deterioration,
resulting in numerous adverse effects includ-
ing hypoxic zones, species mortality, and
harmful algal blooms. The calculated con-
tribution of atmospheric deposition to total
N loads can be as high as 72% in estuaries.
The Chesapeake Bay is an example of a
large, well-studied, and severely eutrophic
estuary that is calculated to receive as
much as 30% of its total N load from the at-
mosphere.
Examples of quantified relationships between
deposition levels and ecological effects
Kg
N/ha/yr
Ecological effect
Altered diatom communities in
high elevation freshwater lakes
~1.5 and elevated N in tree leaf tissue
high elevation forests in the west-
ern U.S.
3.1
Decline of some lichen species in
the western U.S.
Altered growth and coverage of
alpine plant species in the western
U.S.
Onset of decline of species rich-
ness in grasslands of the U.S. and
U.K.
Onset of nitrate leaching in Eastern
forests of the U.S.
Multiple effects in tundra, bogs
and freshwater lakes in Europe
Multiple effects in arctic, alpine,
5-15 subalpine and scrub habitats in
Europe
5.5 - 10
5-10
Other welfare effects:
Mercury methylation
Hg is highly neurotoxic and once methy-
lated, principally by S-reducing bacteria, it
can be taken up by microorganisms, zoo-
plankton and macroinvertebrates, and
concentrated in higher trophic levels, in-
cluding fish eaten by humans. In 2006, 3,080
fish consumption advisories were issued be-
cause of methylmercury (MeHg), and as of
July 2007, 23 states had issued statewide
advisories. The production of meaningful
amounts of MeHg requires the presence of
S042- and Hg, and where Hg is present, in-
creased availability of SCU2- results in in-
creased
production
of MeHg.
The
amount of
MeHg
produced
varies with
oxygen content, temperature, pH, and sup-
ply of labile organic C. Watersheds with
conditions known to be conducive to Hg
methylation can be found in the northeast-
ern U.S. and southeastern Canada, but bi-
otic Hg accumulation has been widely ob-
served in other regions that have not been
studied as extensively, and where a different
set of conditions may exist.
The evidence is sufficient to
infer a causal relationship be-
tween S deposition and in-
creased Hg methylation in
wetlands and aquatic envi-
ronments.
12
-------
ISA for Oxides of Nitrogen and Sulfur
Executive Summary
Other welfare effects:
Direct phytotoxic
Acute and chronic exposures to SO2 have
phytotoxic effects on vegetation which in-
clude foliar injury, decreased photosynthe-
sis, and decreased growth. Acute exposures
to NO2, NO, PAN, and HNO3 cause plant
foliar injury and decreased growth. How-
ever, the majority of studies have been per-
formed at concentrations of these gas-
phase species above current ambient con-
ditions observed in the U.S. Consequently,
there is little evidence that current concen-
Conclusion
The main effects of N and S pollution assessed in the ISA are acidification, N en-
richment, and Hg methylation. Acidification of ecosystems is driven primarily by
deposition resulting from SOx, NOx, and NHx pollution. Acidification from the
deposition resulting from current emission levels causes a cascade of effects
that harm susceptible aquatic and terrestrial ecosystems, including slower
growth and injury to forests and localized extinction of fishes and other aquatic
species. In addition to acidification, atmospheric deposition of reactive N re-
sulting from current NOx and NHx emissions along with other non-atmospheric
sources (e.g., fertilizers and wastewater), causes a suite of ecological changes
within sensitive ecosystems. These include increased primary productivity in
most N-limited ecosystems, biodiversity losses, changes in C cycling, and eutro-
phication and harmful algal blooms in freshwater, estuarine, and ocean eco-
systems. In some watersheds, additional SO42" from atmospheric deposition in-
creases Hg methylation rates by increasing both the number and activity of S-
reducing bacteria. Methylmercury is a powerful toxin that can bioaccumulate
to toxic amounts in food webs at higher trophic levels (e.g. bass, perch, otters,
or kingfishers).
13
trations of gas-phase S or N oxides are high
enough to cause phytotoxic effects. One
excep-
tion is
that
some
studies
indicate
that cur-
rent
HNO3 concentrations may be contributing
to the decline in lichen species in the Los
Angeles basin.
The evidence is sufficient to
infer a causal relationship
between exposure to SO2,
NO, NO2, PAN, and HNOs
and injury to vegetation.
-------
Contents
List of Tables xi
List of Figures xv
Acronyms and Abbreviations xxiii
Authors, Contributors, Reviewers xxix
Project Team xxxiii
Clean Air Scientific Advisory Committee xxxvi
Chapter 1. Introduction 1-1
1.1. Scope 1-1
1.2. History of the NOxand SOx Review 1-3
1.3. History of the Current Review 1-4
1.4. Development of the Integrated Science Assessment 1 -5
1.5. Organization of the Integrated Science Assessment 1-5
1.6. Causality Framework 1-6
1.6.1. First Step: Determination of Causality 1-7
1.6.2. Second Step: Evaluation of Ecological Response 1-8
Chapter 2. Source to Deposition 2-1
2.1. Introduction 2-1
2.2. Sources and Emissions of Tropospheric NOx 2-2
2.2.1. Major Anthropogenic Sources 2-2
2.2.2. Major Biogenic Sources 2-7
2.2.2.1. Soils 2-7
2.2.2.2. Live Vegetation 2-9
2.2.2.3. Biomass Burning 2-9
2.2.2.4. Lightning 2-9
2.2.3. Anthropogenic and Biogenic Sources of N2O 2-10
2.3. Sources and Emissions of Tropospheric SOx 2-11
2.3.1. Major Anthropogenic Sources 2-11
2.3.2. Major Biogenic Sources 2-16
2.4. NHx Emissions 2-17
2.5. Evaluating Emissions Inventories 2-20
2.5.1. Emissions for Historical Modeling 2-23
2.6. NOx-SOx-NHx Chemistry in the Troposphere 2-23
i
-------
2.6.1. Introduction 2-23
2.6.2. NOx Chemistry 2-25
2.6.2.1. NOx and Ozone Formation 2-30
2.6.2.2. Multiphase Interactions 2-30
2.6.3. SOx Chemistry 2-32
2.6.3.1. Multi-Phase SOx Chemistry 2-34
2.6.4. Multiphase NOx, SOx, and NHx Interactions 2-38
2.6.5. Transport-Related Effects on Chemistry 2-41
2.7. Sampling and Analysis Techniques 2-43
2.7.1. Methods for Relevant Gas-Phase N Species 2-43
2.7.1.1. NO and N02 2-43
2.7.1.2. Total NOy 2-47
2.7.1.3. Nitric Acid 2-47
2.7.1.4. Other Nitrates 2-49
2.7.1.5. Ammonia 2-49
2.7.2. Methods for Relevant Gas-Phase Sulfur Species 2-53
2.7.2.1. Positive Interference 2-53
2.7.2.2. Negative Interference 2-54
2.7.2.3. Other Methods 2-54
2.7.3. Methods for Relevant Aerosol-Phase Nitrogen and Sulfur Species 2-54
2.7.3.1. Artifacts 2-56
2.7.3.2. Other Methods 2-58
2.8. Methods to Compute NOx and SOx Concentrations, Chemical Interactions, and Deposition 2-59
2.8.1. Chemical Transport Models 2-59
2.8.1.1. Global Scale 2-61
2.8.1.2. Regional Scale 2-62
2.8.1.3. Sub-Regional Scale 2-63
2.8.1.4. Modeling Effects of Convection for Chemical Transport 2-64
2.8.2. Computed Deposition 2-65
2.8.2.1. Deposition Forms 2-66
2.8.2.2. Methods for Estimating Dry Deposition 2-67
2.8.2.3. Factors Affecting Dry Deposition Rates and Totals 2-68
2.8.2.4. Nitrogen Deposition and Flux with Biota 2-70
2.8.3. Air Quality Model Evaluation 2-73
2.8.3.1. Ground-based Comparisons of Photochemical Dynamics 2-74
2.8.3.2. Predicted Chemistry for Nitrates and Related Compounds 2-82
2.8.3.3. Evaluating Deposition with CTMs 2-88
2.8.3.4. Regional CTM Performance 2-89
2.8.4. Computing Atmospheric Deposition to Specific Locations 2-93
2.8.5. Policy Relevant Background Concentrations of NOx and SOx 2-103
2.9. Ambient Monitoring and Reported Concentrations of Relevant Nitrogen and Sulfur Species 2-109
2.9.1. Routine Air Monitoring Networks in North America 2-110
2.9.1.1. Pollutant Categories 2-114
2.9.2. Intensive Field Campaigns 2-116
2.9.3. Satellite-Based Air Quality Observing Systems 2-118
2.9.3.1. Satellite Coverages 2-122
2.9.3.2. Measurement Issues 2-122
2.9.4. European Air Monitoring Networks 2-122
2.9.5. Ambient Concentrations of Relevant N Compounds 2-125
2.9.5.1. NO and N02 2-125
2.9.5.2. NOy and NOz 2-137
2.9.5.3. Nitro-PAHs 2-142
2.9.5.4. Ammonia 2-142
2.9.5.5. NH4NO3 2-146
2.9.6. Ambient Concentrations of Relevant Sulfur Compounds 2-151
2.9.6.1. SO2 and SO42" Near Urban Areas 2-151
2.9.6.2. SO2 and SO42" in Rural and Remote Areas 2-167
2.10. Deposition of Nitrogen and Sulfur Species Across the Landscape 2-175
2.10.1. Nitrogen 2-177
2.10.1.1. Example of NO2 and HNO3 Deposition and Flux Data from Harvard Forest 2-179
ii
-------
2.10.2. Sulfur 2-185
2.11. Summary 2-188
2.11.1. Emissions and Atmospheric Concentrations 2-188
2.11.2. Field Sampling and Analysis 2-189
2.11.3. Nitrogen and Sulfur Deposition 2-189
Chapter 3. Ecological and Other Welfare Effects 3-1
3.1. Introduction to Ecological Concepts 3-1
3.1.1. Critical Loads as an Organizing Principle for Ecological Effects of Atmospheric Deposition 3-1
3.1.2. Ecosystem Scale, Function, and Structure 3-2
3.1.3. Ecosystem Services 3-3
3.2. Ecological Effects of Acidification 3-3
3.2.1. Effects on Major Biogeochemical Processes 3-4
3.2.1.1. Soil Acidification 3-5
3.2.1.2. Sulfur Accumulation and SO42" Leaching 3-7
3.2.1.3. Nitrogen Accumulation and NO3" Leaching 3-8
3.2.1.4. Base-Cation Leaching 3-9
3.2.1.5. Aluminum Leaching 3-10
3.2.1.6. Episodic Acidification 3-11
3.2.2. Terrestrial Ecosystems 3-16
3.2.2.1. Chemical Effects 3-16
3.2.2.2. Summary of Biogeochemistry and Chemical Effects 3-20
3.2.2.3. Biological Effects 3-20
3.2.2.4. Summary of Biological Effects 3-28
3.2.3. Aquatic Ecosystems 3-29
3.2.3.1. Chemical Effects 3-29
3.2.3.2. Summary of Biogeochemistry and Chemical Effects 3-45
3.2.3.3. Biological Effects 3-47
3.2.3.4. Summary of Biological Effects 3-54
3.2.4. Most Sensitive and Most Affected Ecosystems and Regions 3-55
3.2.4.1. Characteristics of Sensitive Ecosystems 3-55
3.2.4.2. Extent and Distribution of Sensitive Ecosystems 3-56
3.2.4.3. Levels of Deposition at Which Effects are Manifested 3-62
3.2.4.4. Acidification Case Study #1: Adirondack Region of New York 3-68
3.2.4.5. Acidification Case Study #2: Shenandoah National Park, Virginia 3-78
3.2.5. Ecosystem Services 3-84
3.3. Nutrient Enrichment Effects from Nitrogen Deposition 3-84
3.3.1. Reactive Nitrogen and the Nitrogen Cascade 3-85
3.3.2. Nitrogen Enrichment Effects on N Cycling 3-87
3.3.2.1. Terrestrial Ecosystems 3-88
3.3.2.2. Wetland Ecosystems 3-99
3.3.2.3. Freshwater Aquatic Ecosystems 3-101
3.3.2.4. Estuarine and Coastal Marine Ecosystems 3-106
3.3.2.5. Summary of N Effects on Biogeochemical Cycling of N and Associated Chemical Indicators 3-109
3.3.3. N Deposition Effects on Productivity and C Budgets 3-110
3.3.3.1. Terrestrial Ecosystems 3-110
3.3.3.2. Freshwater Aquatic 3-122
3.3.3.3. Estuarine and Marine 3-125
3.3.3.4. Summary of Nitrogen Effects on Carbon Cycling 3-130
3.3.4. Biogenic Trace Gases: Nitrous Oxide, Methane, Nitric Oxide and VOCs 3-131
3.3.4.1. Methane 3-131
3.3.4.2. Nitrous Oxide 3-134
3.3.4.3. Nitric Oxide and VOCs 3-137
3.3.4.4. Volatile Organic Compounds (VOCs) 3-138
3.3.4.5. Summary of N Effects on Biogenic Trace Gases 3-138
3.3.5. Species Composition, Species Richness and Biodiversity 3-139
3.3.5.1. Terrestrial Ecosystem Biodiversity 3-139
3.3.5.2. Transitional Ecosystems 3-152
iii
-------
3.3.5.3. Freshwater Aquatic Ecosystems 3-155
3.3.5.4. Estuarine and Marine Ecosystems 3-157
3.3.3.5. Summary of N Effects on Species Composition, Species Richness and Biodiversity 3-159
3.3.6. N Deposition Effects on NO3" Toxicity 3-161
3.3.7. Critical Loads and Other Quantified Relationships between Deposition Levels and Ecological Effects 3-161
3.3.7.1. Empirical Critical Loads for Europe 3-161
3.3.7.2. Empirical Critical Loads for U.S. 3-163
3.3.8. Characterization of Sensitivity and Vulnerability 3-167
3.3.8.1. Extent and Distribution of Sensitive and Vulnerable Ecosystems 3-167
3.3.8.2. Case Study: Alpine and Subalpine Communities of the Eastern Slope of the Rocky Mountains 3-173
3.3.8.3. Case Study: Chesapeake Bay 3-176
3.3.8.4. Case Study: San Bernardino 3-177
3.3.9. Ecosystem Services 3-179
3.4. Other Welfare Effects 3-182
3.4.1. Non-Acidification Effects of Sulfur 3-182
3.4.1.1. Biological Role of Sulfur 3-182
3.4.1.2. Cycling and Storage of Sulfur 3-185
3.4.1.3. Export of Sulfur 3-188
3.4.1.4. Sulfur and Methylation of Mercury 3-190
3.4.1.5. Summary of S and Methylation of Mercury 3-196
3.4.1.6. S Nutrient Enrichment Case Study: Interactive Effects of S and Hg in Little Rock Lake, Wl 3-197
3.4.2. Direct Phytotoxic Effects of Gaseous N and S on Vegetation 3-197
3.4.2.1. Direct Phytotoxic Effects of SO2 on Vegetation 3-198
3.4.2.2. Direct Phytotoxic Effects of NO, N02 and PAN 3-200
3.4.2.3. Direct Phytotoxic Effects of HNO3 3-202
3.4.2.4. Summary of Phytotoxic Effects of Gaseous Nitrogen and Sulfur on Vegetation 3-203
Chapter 4. Summary and Conclusions 4-1
4.1. Source to Deposition 4-1
4.1.1. Chemical Families and Constituent Species 4-1
4.1.2. Transport and Transformation 4-1
4.1.3. Emissions and Atmospheric Concentrations 4-2
4.1.4. Deposition 4-3
4.1.5. Field Sampling and Analysis 4-4
4.2. Acidification 4-4
4.2.1. Terrestrial 4-5
4.2.1.1. Biogeochemistry and Chemical Effects 4-5
4.2.1.2. Biological Effects 4-6
4.2.1.3. Regional Vulnerability and Sensitivity 4-7
4.2.2. Aquatic 4-7
4.2.2.1. Biogeochemistry and Chemical Effects 4-8
4.2.2.2. Biological Effects 4-11
4.2.2.3 Regional Vulnerability and Sensitivity 4-12
4.2.3. Ecosystem Services 4-13
4.3. Nitrogen Nutrient Enrichment 4-13
4.3.1. Terrestrial 4-14
4.3.1.1. Biogeochemical Effects 4-15
4.3.1.2. Species Richness, Composition and Biodiversity 4-16
4.2.3. Transitional 4-19
4.3.2.1. Biogeochemical Effects 4-19
4.3.2.2. Biological Effects 4-21
4.3.2.3. Regional Vulnerability and Sensitivity 4-21
4.3.3. Freshwater Aquatic 4-21
4.3.3.1. Biogeochemical Effects 4-22
4.3.3.2. Biological Effects 4-22
4.3.3.3. Regional Vulnerability and Sensitivity 4-23
4.3.4. Estuarine Aquatic 4-23
4.3.4.1. Biogeochemical Effects 4-23
iv
-------
4.3.4.2. Biological Effects 4-25
4.3.4.3. Regional Vulnerability and Sensitivity 4-25
4.3.5. Ecosystem Services 4-26
4.4. Direct Phytotoxic Effects 4-26
4.4.1. Sulfur Dioxide 4-26
4.4.2. NO, N02 and PAN 4-27
4.4.3. HNOs 4-27
4.5. Mercury Methylation 4-27
Annex A. Ecosystem Monitoring and Models A-1
A.1. Introduction A-1
A.2. Ecosystem Monitoring A-2
A.2.1. Environmental Monitoring and Assessment Program A-2
A.2.2. Surface Water Chemistry Monitoring A-4
A.2.2.1. TIME Project A-4
A.2.2.2. Long-Term Monitoring Project A-5
A.2.3. USGS Monitoring Programs A-7
A.2.3.1. National Water Quality Assessment Program A-7
A.2.3.2. Hydrologic Benchmark Network A-8
A.2.3.3. New York City Water Quality Network A-8
A.2.3.4. Catskill Long-Term Monitoring Sites A-9
A.2.3.5. Buck Creek, New York A-9
A.2.4. NSF Long-Term Ecological Research Network A-9
A.2.4.1. Hubbard Brook Experimental Forest A-13
A.2.4.2. Coweeta A-17
A.2.4.3. Walker Branch A-18
A.2.5. Water, Energy, and Biogeochemical Budgets Program A-18
A.2.5.1. Sleepers River A-19
A.2.5.2. Loch Vale A-21
A.2.6. Other Monitoring Programs A-22
A.2.6.1. Bear Brook A-22
A.2.6.2. Shenandoah Watershed Study A-23
A.2.6.3. Fernow A-25
A.2.6.4. National Ecological Observatory Network A-26
A.3. Modeling A-27
A.3.1. Principal Ecosystem Models Used in the U.S. A-27
A.3.1.1. MAGIC A-27
A.3.1.2. NuCM A-30
A.3.1.3. PnET-BGC A-31
A.3.1.4. DayCent-Chem A-34
A.3.1.5. SPARROW A-35
A.3.1.6. WATERSN A-39
A.3.2. Additional Effects Models Used Widely in Europe A-44
A.3.2.1. The Very Simple Dynamic Model A-44
A.3.2.2. SMART A-44
A.3.2.3. SAFE A-45
A.3.3. Other Models A-45
Annex B. Acidification Effects B-1
B.1. Effects on Biogeochemical Processes along Acidification Pathways B-1
B.1.1. Atmospheric Deposition and Canopy Interaction B-1
B.1.2. Interactions with Soil B-1
B.1.2.1. Sulfur Retention and Release B-1
B.1.2.2. Base Cation Depletion B-3
B.1.2.3. Aluminum Mobilization B-5
v
-------
B.1.2.4. Soil Acidification B-6
B.1.2.5. Nitrogen Saturation B-7
B.1.2.6. Nitrate Leaching B-11
B.1.3. Interactions with Transitional Ecosystems B-12
B.1.3.1. Sulfur Storage and Release in Transitional Ecosystems B-12
B.1.3.2. Organic Acidity in Transitional Ecosystems and Downstream Surface Waters B-12
B.2. Factors That Determine Ecosystem Sensitivity B-16
B.2.1. Transitional Ecosystems B-16
B.2.1.1. Wetlands and Peatlands B-16
B.2.1.2. Ponds B-17
B.2.2. Streams and Lakes B-17
B.2.3. Other Types of Ecosystems B-18
B.3. Distribution and Extent of Ecosystem Effects B-18
B.3.1. Terrestrial Ecosystems B-18
B.3.2. Transitional Ecosystems B-20
B.3.3. Aquatic Ecosystems B-20
B.3.3.1. Status of Surface Waters - Regional Overview B-21
B.3.3.2. Recent Changes in Surface Water Chemistry B-22
B.3.4. Regional Assessments B-24
B.3.4.1. Northeastern Surface Waters B-24
B.3.4.2. Southeastern Surface Waters B-34
B.3.4.3. Upper Midwest B-39
B.3.4.4. West B-41
B.3.4.5. Temporal Variability in Water Chemistry B-46
B.4. Effects on Biota B-54
B.4.1. Types of Effects of Acidification on Biota B-55
B.4.1.1. Individual Condition Factor B-55
B.4.1.2. Species Composition B-57
B.4.1.3. Taxonomic Richness B-57
B.4.1.4. Community Structure B-62
B.4.1.5. Indices of Ecological Effects B-63
B.4.2. Timing of Effects B-63
B.4.2.1. Life Stage Differences in Sensitivity B-63
B.4.2.2. Biological Effects of Episodes B-65
B.4.2.3. Timing of Recovery from Acidification B-68
B.4.3. Effects by Ecosystem Type B-71
B.4.3.1. Terrestrial Ecosystems B-71
B.4.3.2. Aquatic Ecosystem B-73
B.5. Effects on Watersheds and Landscapes B-80
B.5.1. Interactions among Terrestrial, Transitional, and Aquatic Ecosystems B-80
B.5.2. Interactions with Land Use and Disturbance B-81
B.5.2.1. Timber Harvest B-83
B.5.2.2. Insect Infestation B-84
B.5.3. Wind or Ice Storm Damage B-85
B.5.3.1. Fire B-86
B.5.3.2. Multiple Stress Response B-86
B.6. Ecological indicators of acidification B-86
B.6.1. Biological Indicators B-86
B.6.1.1. Phytoplankton B-88
B.6.1.2. Zooplankton B-88
B.6.1.3. Benthic Invertebrates B-89
B.6.1.4. Fish B-91
B.6.1.5. Amphibians B-94
B.6.1.6. Fish-Eating Birds B-95
Annex C. Nutrient Enrichment Effects from Nitrogen C-1
C.1. Effects on Biogeochemical Pathways and Cycles C-1
vi
-------
C.1.1. Nitrogen Cycling in Terrestrial Ecosystems C-1
C.1.1.1. Nitrogen Deposition Effects on DON Leaching C-1
C.1.1.2. Interactions Between Snow Melt and Nitrate Leaching C-1
C.1.1.3. Denitrification: NO and N2O Flux C-1
C.1.1.4. Climate and N2O Interactions C-2
C.1.2. Nitrogen Cycling in Transitional Ecosystems C-3
C.1.2.1. Denitrification: Measurement Techniques C-3
C.1.2.2. Nitrogen Deposition Effects on Methane C-3
C.1.3. Nitrogen Cycling in Estuarine Ecosystems C-4
C.1.3.1. Denitrification and Anammox in Estuarine Ecosystems C-4
C.1.3.2. Nitrogen Budgets C-4
C.1.4. Timing of Chemical Change C-6
C.1.4.1. Interannual Change: Nitrate Leaching C-6
C.1.4.2. Episodic Change C-6
C.1.4.3. Reversibility of Impacts C-7
C.1.5. Tables Supporting Cross Ecosystem Evaluation of N2O, CH4 and CO2 Flux C-8
C.2. Terrestrial Ecosystems C-15
C.2.1. General C Cyling C-16
C.2.2. Forest Growth Interactions with Herbivores C-16
C.2.3. Southern California Coniferous Forest C-16
C.2.4. Boreal Forests C-17
C.2.5. Alpine C-17
C.2.6. Arctic Tundra C-20
C.2.7. Arid Land C-20
C.2.8. Lichens C-20
C.3. Transitional Ecosystems C-22
C.4. Aquatic Ecosystems C-23
C.4.1. History of Evaluating Nitrogen Enrichment in Freshwater Aquatic Ecosystems C-23
C.4.2. Interactions between Nitrogen and P loading C-24
C.4.3. Aquatic Species Affected C-25
C.4.3.1. Phytoplankton and Plants C-26
C.4.3.2. Seasonal Nitrogen Input and Cyanobacteria C-27
C.4.3.3. Nitrate Toxicity: Invertebrates C-27
C.4.4. Nitrate Toxicity: Amphibians and Fish C-28
C.5. Estuary and Coastal Ecosystems C-32
C.5.1. Interacting Factors with Productivity C-33
C.5.2. Hydrology Interactions with Phytoplankton Biomass C-33
C.6. Watersheds, Landscapes and Disturbance C-33
C.6.1. Interactions among Terrestrial, Transitional, and Aquatic Ecosystems C-33
C.6.2. Interactions with Land Use and Disturbance C-38
C.6.3. Timber Harvest and Fire C-39
C.6.4. Insect Infestation and Disease C-40
C.6.5. Urbanization C-41
C.6.6. Agriculture C-42
C.6.7. Other Disturbances C-42
C.6.8. Multiple Stress Response C-43
Annex D. Critical Loads D-1
D.1. Background D-1
D.1.1. The Critical Load Process D-2
D.1.2. Organization of this Annex D-3
D.2. Definitions and Conceptual Approach D-4
D.2.1. Critical Load Definitions D-4
D.2.2. Critical Load Analysis Procedures D-5
D.2.3. Target Load Definition D-7
D.3. Time Frame of Response D-8
vii
-------
D.3.1. Steady State Critical Loads D-9
D.3.2. Dynamic Critical Loads D-9
D.3.3. Receptor Responses D-10
D.3.4. Deposition Schedules D-10
D.3.5. Long-Term Implications D-13
D.3.6. Steady State and Dynamic Critical Loads: Complementary Information D-14
D.4. Calculation of Critical Loads D-14
D.4.1. Empirical Models D-15
D.4.2. Acidification Effects of Sulfur and Nitrogen D-15
D.4.3. Nutrient Effects of Nitrogen D-15
D.4.4. Process-Based Models D-16
D.4.5. Steady State Models D-16
D.4.6. Dynamic Models D-16
D.5. Use of Critical Loads in the U.S. - Current Status D-18
D.5.1. Current Recommendations on Critical Loads Uses in the U.S. D-18
D.5.2. Questions and Limitations Regarding Critical Loads Uses in the U.S. D-19
D.5.3. Critical Loads Research and Monitoring Needs D-19
D.5.3.1. Emissions and Deposition D-19
D.5.3.2. Soils D-20
D.5.3.3. Surface Waters D-20
D.5.3.4. Biological Effects D-20
D.5.3.5. Critical Loads Models D-21
Annex E. Effects of NOy, NHx, and SOx on Structures and Materials E-1
E.1. Introduction E-1
E.2. Environmental Exposures of Materials E-2
E.2.1. Mechanisms of Materials Damage E-2
E.2.2. Deposition Processes E-2
E.2.3. Chemical Interactions of Nitrogen and Sulfur Oxide Species E-3
E.2.4. Materials Damage Experimental Techniques E-3
E.3. Effects on Dyes and Textiles E-4
E.3.1. Fading of Dyes E-4
E.3.2. Degradation of Textile Fibers E-4
E.4. Effects on Plastics and Elastomers E-4
E.5. Effects on Metals E-5
E.5.1. Role of NOy, NHx, and SOx in the Corrosion Process E-9
E.5.2. Effect on Economically Important Metals E-10
E.5.3. Effects on Electronics E-11
E.6. Effects on Paints E-11
E.7. Effects on Stone and Concrete E-12
E.8. Effects of NOx on Paper and Archival Materials E-15
E.9. Costs of Materials Damage from NOy, NHx, and SOx E-16
E.10. Summary E-16
Annex F. Valuation of the Environmental Effects of N and S (non-materials) F-1
F.1. Introduction F-1
F.1.1. Valuation in the Context of NOx and SOx F-1
F.1.1.1. Ecosystem Services F-1
F.1.1.2. Use of the Valuation Literature to Define Adversity F-3
F.1.1.3. Methods for Selecting Literature for this Assessment F-4
F.2. Conceptual Framework F-6
viii
-------
F.2.1. Taxonomy of Values for Environmental Goods and Services F-6
F.2.2. Welfare Economics F-7
F.2.3. Benefit Estimation Approaches F-9
F.3. Valuation of Forests and Terrestrial Ecosystems F-11
F.3.1. Use Values F-11
F.3.2. Total Values F-14
F.3.3. National-Scale Valuation F-15
F.3.4. Valuation of Degrees of Injury F-15
F.3.5. Limitations and Uncertainties F-17
F.4. Valuation of Transitional Ecosystems F-18
F.4.1. Use and Non-Use Values F-18
F.4.2. Limitations and Uncertainties F-20
F.5. Valuation of Aquatic Ecosystems F-20
F.5.1. Acidification F-21
F.5.2. Eutrophication F-22
F.5.2.1. Recreation F-22
F.5.2.2. Commercial Fisheries F-24
F.5.2.3. Water Clarity F-25
F.5.3. Avoided Costs F-25
F.5.4. Limitations and Uncertainties F-25
F.6. Summary F-26
Glossary G-1
References R-1
ix
-------
List of Tables
Table 1-1. Aspects to aid in judging causality. 1-6
Table 1-2. Weight of evidence for causal determination. 1-8
Table 2-1. Emissions of NOx, NH3, and SO2 in the U.S. by source and category, 2002. 2-2
Table 2-2. Total and non-EGU SO2 emissions densities for selected U.S. counties, 2001. 2-13
Table 2-3. Relative contributions of various gas and aqueous phase reactions to aqueous NO3 formation within a sunlit cloud,
10 minutes after cloud formation. 2-31
Table 2-4. Atmospheric lifetimes (t) of SO2 and reduced S species with respect to reaction with OH, NO3, and CI radicals. 2-33
Table 2-5. Relative contributions of various reactions to the total S(IV) oxidation rate within a sunlit cloud, 10 minutes after
cloud formation. 2-37
Table 2-6. Satellite instruments used to retrieve tropospheric NO2 columns. 2-46
Table 2-7. Verified ambient NH3 monitors. 2-50
Table 2-8. Performance characteristics of the 7 U.S. EPA ETV tested NH3 methods. 2-51
Table 2-9. Site codes and locations of passive NH3 samplers in the U.S. EPA and Lake Michigan Air Directors and Illinois
State Water Survey Consortium Project. 2-52
Table 2-10. Atmospheric N loads relative to total N loads in selected great waters* 2-94
Table 2-11. Natural and anthropogenic sources of atmospheric N compounds. 2-96
Table 2-12. Characteristics of oxidized-nitrogen airsheds. 2-97
Table 2-13. Characteristics of principal airsheds for reduced-N (Red-N) deposition. 2-98
Table 2-14. Major routine operating air monitoring networks. 2-110
Table 2-15. Air monitoring networks/campaigns for non-routine special intensive studies conducted since the mid-1990s. 2-117
Table 2-16. Satellite-based air quality observing systems.14 2-119
Table 2-17. Key atmospheric chemistry and dynamics data sets at the NASA Goddard DAAC. 2-121
Table 2-18. International and European air monitoring programs. 2-124
Table 2-19. Ambient NH3 concentrations summarized by study. 2-143
Table 2-20. Number of monitors in California and San Diego County, 2005. 2-152
Table 2-21. Number of monitors in Ohio and Cuyahoga County, 2005. 2-152
Table 2-22. Regional distribution of SO2 and SO42" ambient concentrations, averaged for 2003-2005. 2-159
Table 2-23. Distributions of temporal averaging inside and outside CMSAs. 2-159
Table 2-24. Range of mean annual SO2 concentrations and Pearson correlation coefficients in urban areas having at least four
regulatory monitors, 2003-2005. 2-160
Table 2-25. Regional changes in wet and dry N and S atmospheric concentrations and deposition, 1989-1991 and 2003-2005. 2-190
Table 3-1. An example of the matrix of information that must be considered in the definition and calculation of critical loads
(see discussion in text). 3-2
Table 3-2. Examples of chemical indicators of effects from acidifying deposition to terrestrial ecosystems. 3-16
Table 3-3. Example biological effects indicators in terrestrial ecosystems. 3-21
Table 3-4. Examples of chemical indicators of effects from acidifying deposition to aquatic ecosystems. 3-30
Table 3-5. Estimates of change in number and proportion of acidic surface waters in acid-sensitive regions of the North and
East, based on applying current rates of change in Gran ANC to past estimates of population characteristics from
probability surveys. 3-37
xi
-------
Table 3-6. Regional trend results for long-term monitoring lakes and streams for the period 1990 through 2000. 3-38
Table 3-7. Model projections of surface water SO42- and associated ANC, shown as changes between dates, for Adirondack
and Shenandoah streams. 3-44
Table 3-8. General summary of biological changes anticipated with surface water acidification, expressed as a decrease in
surface water pH. 3-64
Table 3-9. Studies that either did or did not yield evidence that acidifying deposition affected certain species of birds. 3-67
Table 3-10. Observed relationships between zooplankton species richness and lakewater ANC in the Adirondack Mountains. 3-74
Table 3-11. Summary of biogeochemical indicators of N deposition to terrestrial ecosystems. 3-93
Table 3-12. Effects of fire on nutrient concentrations in forests in Nevada and California. 3-97
Table 3-13. Summary of N cycling studies for wetlands. 3-100
Table 3-14. Summary of studies on the effects of N deposition on freshwater aquatic ecosystems. 3-105
Table 3-15. Summary of N effects on forest carbon cycling. 3-113
Table 3-16. Summary of additional evidence of N effects on productivity of freshwater ecosystems. 3-124
Table 3-17. Summary of N effects on grassland biodiversity. 3-142
Table 3-18. Summary of N effects on arid and semi-arid ecosystems. 3-144
Table 3-19. Summary of N effects on desert ecosystems. 3-147
Table 3-20. Summary of N effects on lichens. 3-149
Table 3-21. Summary of N effects on alpine ecosystems. 3-151
Table 3-22. Summarized responses of coastal marshes ecosystem to N fertilization. 3-154
Table 3-23. N effects on species composition and biodiversity 3-157
Table 3-24 Biological indicators for the effects of elevated N deposition and related empirical critical loads for major ecosystem
types (according to the eunis classification) occurring in Europe. 3-162
Table 3-25. Summary of dose-response curves for N deposition and ecological indicators. 3-164
Table 3-26. Changes in aquatic ecosystems associated with elevated N loadings in the Western U.S. 3-171
Table 3-27. Primary Goods and Services Provided by Ecosystems 3-181
Table 3-28. Summary of recent studies of SO2 exposure to plants. 3-204
Table 4-1. Chemical indicators of acidification to terrestrial ecosystems. 4-6
Table 4-2. Chemical indicators of acidification in surface water. 4-10
Table 4-3. Indicators of estuarine eutrophication. 4-25
Table 4-4. Summary of N deposition levels and corresponding ecological effects. 4-28
Table A-1. LTER site locations and basic site description information. A-10
Table A-2. Current long-term monitoring data sets developed through the Hubbard Brook Ecosystem Study A-14
Table A-3. Study watersheds at HBEF. A-15
Table A-4. Parameter estimates, probability levels, and regression results of parametric and bootstrap regressions of total
nitrogen at 414 national stream quality accounting network stations on basin attributes, for the Chesapeake Bay
total nitrogen SPARROW model. A-37
Table A-5. Effect of spatial referencing on measures of regression model performance for predicting total N flux using the
sparrow model. A-38
Table A-6. Parameter estimates, probability levels, and regression results for the Chesapeake Bay total N SPARROW model A-39
Table A-7. Summary of N retention rates used in recent WATERSN studies. A-42
Table A-8. Some examples of models that could contribute to development of a better understanding of the ecological efforts
of atmospheric S and N deposition, but that are not explicitly addressed in this annex. A-45
xii
-------
Table B-1. N-saturated forests in North America, including estimated N inputs and outputs. B-96
Table B-2 Summary of measured ANC, pH, and Al concentrations compared with reference values in the six high-interest
areas. B-97
Table B-3 Sources of data and sample sizes for datasets analyzed by Stoddard et al. (2003), along with estimates of the
condition of surface waters in each region in the 1980s. B-98
Table B-4 Estimates of change in number and proportion of acidic surface waters in acid-sensitive regions of the North
and East, based on applying current rates of change in Gran ANC to past estimates of population characteristics
from probability surveys. B-99
Table B-5 Regional trend results for long-term monitoring sites for the period 1990 through 2000. B-99
Table B-6 Slopes of trends in Gran ANC in acidic, low, and moderate ANC lakes and streams, 1990-2000. B-100
Table B-7 Changes in key chemical characteristics during periods of record in Maine aquatic systems. B-100
Table B-8 Projected changes (|Jeq/L) in median values of streamwater chemistry at the regional modeling sites from 1995 to
2040 in each of the three emissions control strategies, stratified into two segments of the SAMI region (northeast
and southwest) and by physiographic province. B-100
Table B-9. Population estimates of water chemistry percentiles for selected lake populations in the western U.S.3 B-101
Table B-10. Population estimates of the percentage of lakes in selected subregions of the West with ANC and NO3" within
defined ranges. B-102
Table B-11. Median streamwater ANC and watershed area of streams in Shenandoah National Park that have water chemistry
and fish species richness data. B-102
Table B-12. Reference levels for the Acidic Stress Index based on logistic regression offish presence as a function of the
sensitive intermediate and tolerant ASI values for brown bullhead, brook trout, lake trout, and common shiner. B-103
Table B-13. General summary of biological changes anticipated with surface water acidification, expressed as a decrease in
surface water pH. B-103
Table B-14. Estimated percentage of Adirondack lakes with and Acidic Stress Index exceeding the reference levels for effects
on fish populations, based on diatom-inferred historical (pre-industrial) chemistry and present-day measured and
inferred acid-base chemistry. B-104
Table B-15. Estimated percentage of Adirondack lakes with acid-base chemistry unsuitable for fish population survival, based
on diatom-inferred historical (pre-industrial) chemistry and present-day measured and inferred acid-base chemistry. B-104
Table B-16. Estimated percentage of the lakes in the Northeast and Upper Midwest, ELS/NSWS target population
with an Acidic Stress Index exceeding the reference levels for fish populations defined in Table C-12. B-105
Table B-17. Estimated percentage of lakes in the Northeast, ELS/NSWS target populations with acid-base chemistry unsuitable
for fish population survival. B-105
Table B-18. Distribution of acidic stress index values among the NSS-1 Target populations for the mid-Appalachian region. B-106
Table B-19. Distribution of acidic stress index values among the NSS-1 target populations for the interior Southeast region. B-107
Table B-20. Comparison of solution and tissue chemistries at threshold treatment levels where significant impacts on tree
growth or nutrient content were first observed. B-108
Table B-21. Overview of selected major processes by which landscape change can alter drainage water acid-base chemistry B-111
Table B-22. Observed relationships between zooplankton species richness (R) and lakewater ANC. B-111
Table B-23. Threshold response of increased mortality of fish to lowpH listed from least sensitive to most sensitive. B-112
Table B-24. Threshold values of pH for various taxa and effects. B-113
Table B-25. Threshold values of Al for various species and effects (form of Al not specified for most studies). B-114
Table B-26. The effects of increasing Ca2+to ameliorate low pH and high Al. B-115
Table B-27. Brook trout acidification response categories developed by Bulger et al. (2000) for streams in Virginia. B-115
Table B-28. Partial listing of bioassays demonstrating decreased fish survival in waters with low pH and (or) elevated aluminum. B-116
Table B-29. Mills et al., 1987. Shows effect of various pH on fish forage fish and lake trout. B-117
xiii
-------
Table B-30. Range of minimum pH of fish species occurrence in 11 lakesurveys. B-117
Table B-31. Studies3 that either did (yes) or did not (no) yield evidence that acidic deposition affected certain species of birds B-119
Table B-32. Predicted habitat suitability for lakes in the Algona Model Dataset B-119
Table B-33. Summary statistics of biological data layers for mercury (Hg) concentrations in fish and wildlife (pg/g) in the
northeastern U.S. and southeastern Canada. B-120
Table B-34. Mercury concentrations in avian eggs and tissues and related effects. B-120
Table C-1. Estimated percent of total N load to Delaware Bay and Hudson River/Raritan Bay contributed
by atmospheric deposition. C-5
Table C-2 The study site, experimental condition, ecosystem type, N form, amount of N addition and citations is presented for
all studies used in NEE, EC, Cm uptake, ChU emission and N2O emission meta analyses. C-8
Table C-3. Principal Air Quality Indicator Lichen Species in Oregon and Washington C-21
Table C-4. Contribution of fens to support of plant species diversity in selected states. C-22
Table C-5. Summary of additional evidence for N limitation on productivity of freshwater ecosystems. C-25
Table C-6. Summary of effects of N enrichment on aquatic biota in freshwater ecosystems. C-28
Table C-7. Essential ecological attributes and reporting categories. C-34
Table C-8. Primary goods and services provided by ecosystems. C-36
Table C-9. Ecological effects of N deposition described for study sites in the Western U.S. C-44
Table D-1 An example of the matrix of information that must be considered in the definition and calculation of critical loads. D-5
Table D-2. Biological indicators for the effects of elevated N deposition and related empirical critical loads for major ecosystem
types (according to the Eunis classification) occurring in Europe. D-21
Table E-1. Studies on corrosive effects of NOy/NH3/SOx effects on metals. E-6
Table E-2. Studies on corrosive effects of NOy/NH3/SOx on stone. E-13
Table F-1. Commonly adopted environmental valuation methods based on revealed or state preferences. F-27
Table F-2. Economic effects of ozone and other pollutants on agriculture, as reported in the 1996 ozone criteria document. F-28
Table F-3. Economic effects of ozone on marketable benefits from forests. F-29
Table F-4. Forecasted average values for select activities, per day per person in 1996. F-30
Table F-5. Typical impacts of specific pollutants on the visual quality of forests. F-30
Table F-6. Economic valuation studies related to forest aesthetics. F-31
Table F-7. Summary of the monetized estimates of the annual value of forest quality changes F-33
Table F-8. Estimated value of avoiding forest damage in the U.S. F-33
Table F-9. Ecological wetland functions, economic goods and services, types of value, and applicable valuation methods. F-34
Table F-10. Economic valuation studies related to acidification and eutrophication in aquatic ecosystems. F-35
xiv
-------
List of Figures
Figure 1-1. Replace Biogeochemical cycles of NOx and SOx. 1-2
Figure 2-1. 2001 county-level total U.S. NOx (NO and NO2) emissions. 2-4
Figure 2-2. 2001 county-level total U.S. NOx (NO and NO2) emissions densities (tons per square mile). 2-5
Figure 2-3. 2001 county-level total U.S. NOx (NO and NO2) emissions densities (tons per square mile) from electric-generating
utilities (EGUs).1 2-5
Figure 2-4. 2001 county-level total U.S. NOx (NO and NO2) emissions densities (tons per square mile) from on-road mobile
sources. 2-6
Figure 2-5. 2001 county-level total U.S. SO2 emissions. 2-12
Figure 2-6. 2001 county-level total U.S. SO2 emissions densities (tons per square mile).1 2-12
Figure 2-7. 2001 county-level SO2 emissions densities (tons per square mile) from EGUs. 2-13
Figure 2-8. 2001 county-level SO2 emissions densities (tons per square mile) from on-road mobile sources. 2-14
Figure 2-9. 2001 county-level SO2 emissions densities (tons per square mile) from off-road mobile and other transportation
sources. 2-14
Figure 2-10. State-level SO2 emissions, 1990-2005. 2-15
Figure 2-11. 2001 county-level total U.S. NH3 emissions. 2-18
Figure 2-12. 2001 county-level total U.S. NH3 emissions densities.1 2-18
Figure 2-13. 2001 county-level NH3 emissions densities from on-road mobile sources. 2-19
Figure 2-14. 2001 county-level NH3 emissions densities from EGUs. 2-19
Figure 2-15. 2001 county-level NH3 emissions densities from miscellaneous and biogenic sources. 2-20
Figure 2-16. Atmospheric cycle of reactive oxidized N species. 2-24
Figure 2-17. The combined NOx + SOx + NHx system showing how atmospheric fates and lifetimes of reduced and oxidized N
components are linked. 2-25
Figure 2-18. Atmospheric cycle of S compounds. 2-34
Figure 2-19. Comparison of aqueous-phase oxidation paths. 2-36
Figure 2-20. RH effects on deliquescence and efflorescence points for a NaCI+ Na2S04 particle, indicating deliquescence at
-72%relative humidity and re-crystallization at-52% RH. 2-38
Figure 2-21. Predicted isolines of particulate NO3 concentrations (|jg/m3) as a function of total HNO3 and NH3 at 293 K and 80%
relative humidity, and with 25 |jg/m3 SO42" and 2 |jg/m3 total CI". 2-40
Figure 2-22. Predicted particulate NO3 concentration as a function of RH for a typical environment. 2-40
Figure 2-23. Tropospheric NO2 column estimates (molecules N02/cm2) retrieved from the SCIAMACHY satellite instrument for
2004-2005. 2-47
Figure 2-24. Average ambient NH3 concentrations from the NH3 passive samplers trial network, 2007-2008. 2-52
Figure 2-25. Schematic of the resistance-in-series analogy for atmospheric deposition. Function of wind speed, solar radiation,
plant characteristics, precipitation/moisture, and soil/air temperature. 2-65
Figure 2-26. The relationship between particle diameter and deposition velocity for particles. 2-69
Figure 2-27. 8 km southeast U.S. CMAQ domain zoomed over Tampa Bay, FL. 2-74
Figure 2-28. 2 km southeast U.S. CMAQ domain zoomed over Tampa Bay, FL. 2-74
Figure 2-29. Hourly averages for May 1-31,2002. CMAQ 8 km and 2 km results and measured concentrations of NO (top), NO2
(middle), and total NOx (bottom). 2-75
xv
-------
Figure 2-30.
Mav 2002 dailv concentrations and 8 km CMAQ predictions for ethene at Svdnev. FL.
2-76
Figure 2-31.
Mav 2002 dailv concentrations and 8 km CMAQ predictions for isoprene at Svdnev. FL.
2-77
Figure 2-32.
Observed hourlv PM? s concentrations at St. Petersburg FL and results from CMAQ 8 km.
2-78
Figure 2-33.
Observed and modeled ratios of O3 to NOx.
2-79
Figure 2-34.
Observed and CMAQ 8 km and 2 km predicted formaldehvde concentrations.
2-80
Figure 2-35.
Hourly concentrations of hydrogen peroxide, observed and predicted by CMAQ 8 km and 2 km, May 1-31,2002 at
Svdnev. FL.
2-81
Figure 2-36.
Hourly and CMAQ-UCD-predicted total pN03 concentrations at Tampa Bay, FL, and observations at Sydney, FL
Mav 1-31. 2002.
2-83
Figure 2-37.
Hourlv and CMAQ-predicted HNO3 concentrations at Svdnev. FL. Mav 1-31. 2002.
2-84
Figure 2-38.
Hourly and CMAQ-UCD-predicted ratio of HNO3 to total NO3 at Tampa Bay, FL and observations at Sydney, FL,
Mav 1-31. 2002.
2-85
Figure 2-39.
CMAQ-UCD predicted fractions and totals of total NO3 for days in May 2002 with measurements in Tampa Bay, FL..
2-86
Figure 2-40.
Hourlv and CMAQ-predicted NH3 concentrations at Svdnev. FL. Mav 1-31.2002.
2-87
Figure 2-41.
Scatter plot of total nitrate (HNO3 plus PNO3) wet deposition (mg N/m2/yr) of the model mean versus measurements
for the North American Deposition Proaram (NADP) network.
2-88
Figure 2-42.
Scatter plot of total SO42" wet deposition (mg S/m2/yr) of the model mean versus measurements for the National
Atmospheric Deposition Proaram (NADP) network.
2-89
Figure 2-43.
CMAQ modeling domains for the OAQPS risk and exposure assessments: 36 km outer parent domain in black; 12
km western U.S. (WUS) domain in red: 12 km eastern U.S. (EUS) domain in blue.
2-89
Figure 2-44.
12-km EUS Summer sulfate PM, each data point represents a paired monthly averaged (June/July/August)
observation and CMAQ prediction at a particular IMPROVE. STN. and CASTNet site.
2-90
Figure 2-45.
12-km EUS Winter nitrate PM, each data point represents a paired monthly averaged
(December/Januarv/Februarv) observation and CMAQ prediction at a particular IMPROVE and STN site.
2-91
Figure 2-46.
12-km EUS Winter total nitrate (HNO3 + total PNO3), each data point represents a paired monthly averaged
(December/Januarv/Februarv) observation and CMAQ prediction at a particular CASTNet site.
2-91
Figure 2-47.
12-km EUS annual sulfate wet deposition, each data point represents an annual average paired observation and
CMAQ prediction at a particular NADP site.
2-92
Figure 2-48.
12-km EUS annual nitrate wet deposition, each data point represents an annual average paired observation and
CMAQ prediction at a particular NADP site.
2-92
Figure 2-49.
Principal airsheds and watersheds NOx for these estuaries:
2-93
Figure 2-50.
Tvpical surface laver cell and total column structure and processes represented in CMAQ.
2-98
Figure 2-51.
CMAQ vs. measured air concentrations from east-coast sites in the IMPROVE, CSN (labeled STN), and CASTNet
networks in the summer of 2002: (riaht) sulfate and (left) ammonium.
2-99
Figure 2-52.
Comparison of CMAQ-predicted and NADP-measured NH4+wet deposition.
2-100
Figure 2-53.
CMAQ-predicted (red symbols and lines) and 12-h measured (blue symbols and lines) NH3 and SO42- surface
concentrations at hiah and low concentration cells in North Carolina in Julv 2004.
2-101
Figure 2-54.
Surface cell (layer 1) analysis of the sensitivity of NHx deposition and transport to the change in NH3 Vd in CMAQ.
2-101
Figure 2-55.
Total column analysis for NH3 (left) and NHx (right) showing modeled NH3 emissions, transformation, and transport
throuahout the mixed laver and up to the free troposphere.
2-102
Figure 2-56.
Range of influence (where 50% of emitted NH3 deposits) from the high concentration Sampson County cell in the
June 2002 CMAQ simulation of Vd sensitivities.
2-102
Figure 2-57.
Areal extent of the change in NHx range of influence as predicted by CMAQ for the Sampson County high
concentration cell (center of ranae circles) in June 2002 usina the base case and sensitivitv case Vd.
2-103
Figure 2-58.
Annual mean concentrations of NO2 (ppb) in surface air over the U.S. in the present-day (upper panel) and policy
relevant backaround (middle panel) MOZART-2 simulations.
xvi
2-104
-------
Figure 2-59. Annual mean concentrations of SO2 (ppb) in surface air over the U.S. in the present-day (upper panel) and policy
relevant background (middle panel) MOZART-2 simulations. 2-105
Figure 2-60. Annual mean concentrations of wet and dry deposition of NHNO3, NH4NO3, NOx, HO2NO2, and organic nitrates
(mg N/m2/yr) in surface air over the U.S. in the present-day (upper panel) and policy relevant background (middle
panel) MOZART-2 simulations. 2-107
Figure 2-61. Annual mean concentrations of SOx deposition (SO2 + pSCU) (mg S/m2/yr) in surface air over the U.S. in the
present-day (upper panel) and policy relevant background (middle panel) MOZART-2 simulations. 2-108
Figure 2-62. July mean soil NO emissions (upper panels; 1 x 109 molecules/cm2/s) and surface PRB NOx concentrations (lower
panels; ppt) over the U.S. from MOZART-2 (left) and GEOS-Chem (right) model simulations in which anthropogenic
O3 precursor emissions were set to zero in North America. 2-109
Figure 2-63. Aggregate map of most routine U.S. monitoring stations. 2-112
Figure 2-64. Trends in regional chemical composition of PM2.5 aerosols based on urban speciation sites and averaged over the
entire 2006 sampling period. 2-113
Figure 2-65. Original 3-tiered NCore design (left) and proposed site locations for Level 2 multiple pollutant sites. 2-113
Figure 2-66. Maps illustrating coverage of PM2.5 FRM and FEM and O3 network (left); and PM2.5 continuous samplers (right). 2-115
Figure 2-67. Routinely operating North American precipitation and surface water networks. 2-116
Figure 2-68. Correlation surfaces between MODIS AOD and hourly PM2.5 surface sites from April-September 2002. 2-123
Figure 2-69. Location of ambient-level NO2 monitors for NAAQS compliance in 2007. Shaded states have NO2 monitors;
unshaded states have none. 2-125
Figure 2-70. Ambient concentrations of NO2 measured at all monitoring sites located within Metropolitan Statistical Areas
(MSAs) in the U.S. from 2003 through 2005. 2-126
Figure 2-71. Monthly average NO2 concentrations (ppb) for January 2002 (left panel) and July 2002 (right panel) calculated by
CMAQ (36 X 36 km horizontal resolution). 2-127
Figure 2-72. Nationwide trend in NO2 concentrations. 2-127
Figure 2-73. Time series of 24-h average NO2 concentrations at individual monitoring sites in Atlanta, GA from 2003 through
2005. A natural spline function (with 9 degrees of freedom) was fit and overlaid to the data (dark solid line). 2-128
Figure 2-74. Time series of 24-h average NO2 concentrations at individual monitoring sites in New York City from 2003 through
2005. A natural spline function (with 9 degrees of freedom) was fit and overlaid to the data (dark solid line). 2-129
Figure 2-75. Time series of 24-h average NO2 concentrations at individual monitoring sites in Chicago, IL from 2003 through
2005. 2-130
Figure 2-76. Time series of 24-h average NO2 concentrations at individual monitoring sites in Baton Rouge, LA from 2003
through 2005. 2-131
Figure 2-77. Time series of 24-h average NO2 concentrations at individual monitoring sites in Houston, TX from 2003 through
2005. 2-132
Figure 2-78. Time series of 24-h average NO2 concentrations at individual monitoring sites in Los Angeles, CA from 2003
through 2005. 2-133
Figure 2-79. Time series of 24-h average NO2 concentrations at individual monitoring sites in Los Angeles, CA from 2003
through 2005. 2-134
Figure 2-80. Time series of 24-h average NO2 concentrations at individual monitoring sites in Riverside, CA from 2003 through
2005. A natural spline function (with 9 degrees of freedom) was fit and overlaid to the data (dark solid line). 2-135
Figure 2-81. Time series of 24-h average NO2 concentrations at individual monitoring sites in Riverside, CA from 2003 through
2005. 2-136
Figure 2-82. Mean hourly NO2 concentrations on weekdays and weekends measured at two sites in Atlanta, GA. 2-137
Figure 2-83. Measured O3 (ppb by volume) versus PAN (ppt by volume) in Tennessee, including (a) aircraft measurements, and
(b, c, and d) suburban sites near Nashville. 2-138
Figure 2-84. Ratios of PAN to NO2 observed at Silwood Park, Ascot, Berkshire, U.K. from July 24 to August 12,1999. 2-139
xvii
-------
Figure 2-85. Ratios of HNO2 to NO2 observed in a street canyon (Marylebone Road) in London, U.K. from 11 a.m. to midnight
during October 1999. Data points reflect 15-min average concentrations of HONO and NO2. 2-140
Figure 2-86. Annual average gas-phase HNO3 concentrations, 2004-2006. 2-141
Figure 2-87. Annual average gas-phase NO3 concentrations, 2004-2006. 2-142
Figure 2-88 County-scale NH3 emissions densities from the CMU inventory model. 2-144
Figure 2-89. Ambient NH3 concentration as a function of county scale NH3 emissions. 2-145
Figure 2-90. Estimated county-scale ambient NH3 concentrations. 2-145
Figure 2-91. IMPROVE network measured annual averaged NH4NO3 concentration for 2000 (left) and for 2004 (right). 2-147
Figure 2-92. IMPROVE and CSN (labeled STN) monitored mean NH4NO3 concentrations for 2000 through 2004. 2-147
Figure 2-93. Regional and local contributions to annual average PM2.5 by pN03 for select urban areas based on paired
IMPROVE and CSN monitoring sites. 2-148
Figure 2-94. Maps of spatial patterns for average annual pN03 measurements (top), and for NH3 emissions for April 2002 from
the WRAP emissions inventory (bottom). 2-149
Figure 2-95. CMAQ simulation of January monthly averaged pN03 concentration using 1996 emissions (left), and for a 50%
decrease in NH3 emissions (right). 2-150
Figure 2-96. Particulate NO3 source attribution by region using CAMx modeling for six western remote area monitoring sites 2-150
Figure 2-97. Criteria pollutant monitor locations (A) and SO2 monitor locations (B), California, 2005. 2-153
Figure 2-98. Criteria pollutant monitor locations (A) and SO2 monitor locations (B), Ohio, 2005. 2-154
Figure 2-99. Criteria pollutant monitor locations (A) and SO2 monitor locations (B), Arizona, 2005. 2-155
Figure 2-100. Criteria pollutant monitor locations (A) and SO2 monitor locations (B), Pennsylvania, 2005. 2-156
Figure 2-101. Criteria pollutant monitor locations (A) and SO2 monitor locations (B), New York, 2005. 2-157
Figure 2-102. Criteria pollutant monitor locations (A) and SO2 monitor locations (B), Massachusetts, 2005. 2-158
Figure 2-103. Steubenville, OH, 2003-2005. 2-162
Figure 2-104. Philadelphia, 2003-2005. 2-163
Figure 2-105. Los Angeles, 2003-2005. 2-164
Figure 2-106. Riverside, CA, 2003-2005. 2-165
Figure 2-107. Phoenix, 2003-2005. 2-166
Figure 2-108. Annual mean ambient SO2 concentration, 1989 through 1991 (top), and for 2003 through 2005 (bottom). 2-168
Figure 2-109. Annual mean ambient SO42" concentration, 1989 through 1991 (top), and 2003 through 2005 (bottom). 2-169
Figure 2-110. IMPROVE network measured annual averaged pS04 concentration for 2000 (top) and for 2004 (bottom). 2-170
Figure 2-111. IMPROVE mean (NH4)2S042~ concentrations for 2000 through 2004. 2-171
Figure 2-112. IMPROVE and CSN (labeled STN) monitored mean (NH4)2S042~ concentrations for 2000 through 2004. 2-171
Figure 2-113. Regional and local contributions to annual average PM2.5 by pS04 for select urban areas based on paired
IMPROVE and CSN monitoring sites. 2-172
Figure 2-114. Contributions of the Pacific Coast area to the (NH4)2S04 (|jg/m3) at 84 remote-area monitoring sites in western U.S.
based on trajectory regression. Dots denote locations of the IMPROVE aerosol monitoring sites. 2-173
Figure 2-115. pS04 source attribution by region using CAMx modeling for six western remote area monitoring sites. 2-174
Figure 2-116. Trends, 1990-2005 in S (left) and N (right) deposition for 34 sites in the eastern U.S. 2-175
Figure 2-117. Total average yearly wet and dry inorganic N deposition, excepting NH3, for 2004-2006 (top) and 1989-1991
(bottom). 2-176
Figure 2-118. Total average yearly inorganic N deposition by species, excepting NH3, for 2004-2006 (top) and 1989-1991
(bottom). 2-177
xviii
-------
Figure 2-119. NO3" concentration in NADP wet deposition samples, 2004-2006. 2-178
Figure 2-120. Average NO3" concentration in NADP wet deposition samples, 2004-2006. 2-179
Figure 2-121. Diel cycles of median concentrations (upper panels) and fluxes (lower panels) for the northwest clean sector, left
panels) and southwest (polluted sector, right panels) wind sectors at Harvard Forest, April-November, 2000, for
NO, NO2, and O3/IO. 2-180
Figure 2-122. Simple NOx photochemical canopy model outputs. 2-181
Figure 2-123. Hourly (dots) and median nightly (pluses) NO2 flux vs. concentration, with results of least squares fit on the hourly
data (curve). 2-182
Figure 2-124. Averaged profiles at Harvard Forest give some evidence of some NO2 input near the canopy top from light-
mediated ambient reactions, or emission from open stomates. 2-183
Figure 2-125. Summer (June-August) 2000 median concentrations (upper panels), fractions of NOy (middle panels), and fluxes
(lower panels) of NOy and component species separated by wind direction (northwest on the left and southwest on
the right). 2-184
Figure 2-126. Total average yearly wet and dry S deposition for 2004-2006 (top) and 1989-1991 (bottom). 2-186
Figure 2-127. Total average yearly S deposition by species for 2004-2006 (top) and 1989-1991 (bottom). 2-187
Figure 3-1. Illustration of major fluxes of ions associated with S-driven acidification of drainage water. 3-4
Figure 3-2. Diagram illustrates soil horizons commonly found. 3-6
Figure 3-3. Results of an in situ bioassay during a period of episodic acidification in Buck Creek, Adirondack Mountains, in
spring 1990. 3-13
Figure 3-4. Relationship between mean summer acid neutralizing capacity (ANC) and the mean of minimum spring ANC
values at long-term monitoring lake and stream sites in New England, the Adirondacks, and the Northern
Appalachian Plateau. 3-14
Figure 3-5. Diagram based on Fenn et al. (2006) shows indicators of forest physiological function, growth and structure that are
linked to biogeochemical cycles through processes that control rates of Ca supply. 3-18
Figure 3-6. Distribution of red spruce (rose) and sugar maple (green) in the eastern U.S. 3-22
Figure 3-7. Mean (+ standard error bars) of current-year red spruce needle winter injury in reference and calcium-addition
watersheds and among crown classes, expressed as foliar injury (A) and bud mortality (B). 3-23
Figure 3-8. Conceptual diagram outlining the current understanding of sugar maple decline. 3-25
Figure 3-9. Native range of flowering dogwood (Cornus florida) (dk. gray) and the documented range of dogwood anthracnose
in the eastern U.S. (red). 3-26
Figure 3-10. Surface water alkalinity in the conterminous U.S. 3-30
Figure 3-11. Summary of regional trends in surface water chemistry from 1990 to 2000 in regions covered by the Stoddard et al.
(2003) report. 3-33
Figure 3-12. Concentration of inorganic Al in Adirondack streams as a function of the calculated base cation surplus. 3-35
Figure 3-13. F-factors calculated from PnET-BGC model results for the period 1850 to 1980 as a function of simulated ANC in
1980 for 44 EMAP lakes in the Adirondack region of New York. 3-40
Figure 3-14. Median and range of projected change in ANC (|Jeq/L) of Adirondack lakes for 50-year MAGIC simulations versus
median future change in sulfur deposition (kg S/ha/yr) for each deposition scenario. 3-42
Figure 3-15. Number of fish species as a function of mean stream ANC among 13 streams in Shenandoah National Park,
Virginia. 3-52
Figure 3-16. Number offish species per lake versus acidity status, expressed as ANC, for Adirondack lakes. 3-53
Figure 3-17. Regions of the northern and eastern U.S. that contain appreciable numbers of lakes and streams that are sensitive
to acidification from acidifying deposition. 3-58
Figure 3-18. Spatial patterns in predicted wet SO42" and NO3" deposition in the Adirondack Park during the period 1988 to 1999. 3-69
Figure 3-19. Measured wet deposition of sulfur at the Huntington Forest NADP/NTN monitoring station. 3-70
Figure 3-20. Estimated time series of S deposition at one example watershed in the southwestern Adirondack Mountains. 3-71
xix
-------
Figure 3-21. Cumulative distribution functions of selected major ions (|Jeq/L), calculated ANC of lakewater (|Jeq/L), and B
horizon soil %base saturation for the MAGIC and PnET-BGC models.
Figure 3-22. Time series data for SO42", NO3", base cations [Ca plus Mg], Gran ANC, pH, and DOC in one example of long-term
monitoring in Darts Lake in the Adirondack Park. 3-76
Figure 3-23. Mean rates of change in solute concentration in 16 lakes of the Adirondack Long-Term Monitoring (ALTM) program
from 1982 to 2000. 3-77
Figure 3-24. Simulated cumulative frequency distributions of lakewater ANC at three dates for the population of Adirondack
lakes, based on MAGIC model simulations reported by Sullivan et al., 2006b. 3-78
Figure 3-25. Wet sulfur deposition for the period of record at the Big Meadows NADP/NTN monitoring station in Shenandoah
National Park. 3-79
Figure 3-26. Length-adjusted condition factor (K), a measure of body size in blacknose dace (Rhinichthys atratulus) compared
with mean stream pH among 11 populations (n = 442) in Shenandoah National Park. 3-82
Figure 3-27. Illustration of the N cascade showing the movement of the human-produced Nr as it cycles through the various
environmental reservoirs in the atmosphere, terrestrial ecosystems, and aquatic ecosystems. 3-86
Figure 3-28. N cycle (dotted lines indicated processes altered by N saturation). 3-87
Figure 3-29. Schematic illustration of the response of temperate forest ecosystems to long-term, chronic N additions. 3-89
Figure 3-30. Surface water NO3" concentrations as a function of N deposition at the base of each watershed in summer and
spring. 3-90
Figure 3-31. a) N export in stream water as a function of N deposition at the base of sampled watersheds. 3-91
Figure 3-32. Mean annual NO3" concentrations in 230 lakes and streams across the northeastern U.S. 3-102
Figure 3-33. NO3" concentrations in high-elevation lakes in western North America. 3-103
Figure 3-34. A conceptualization of the relationship between overall eutrophic conditions, associated eutrophic symptoms, and
influencing factors (N loads and susceptibility). 3-106
Figure 3-35. Estimated anthropogenic N inputs to the estuaries of the northeastern U.S., in kg N/ha/yr. 3-108
Figure 3-36. Interactions between the C and N cycles. 3-111
Figure 3-37. Mean 5-year radial increment from 31,606 core samples from Picea abies during the period 1945 to 1996 for three
atmospheric N deposition zones (high, medium, and low wet N-deposition in 1990), respectively. 3-115
Figure 3-38. Effects of N addition on forest ecosystem C content. 3-118
Figure 3-39. Effects of N addition on NEE of non-forest ecosystems. 3-121
Figure 3-40. N cycle in freshwater ecosystem. 3-122
Figure 3-41. Description of the eutrophic symptoms included in the national estuary condition assessment. 3-127
Figure 3-42. A high Chi a rating was observed in a large number of the nation's estuaries. 3-128
Figure 3-43. Frequency of hypoxia in Long Island Sound, 1994 to 2002. 3-129
Figure 3-44. Effects of N addition on biogenic CH4 emission. 3-132
Figure 3-45. Effects of N addition on biological CH4 uptake. 3-133
Figure 3-46. Effects of N addition on biological CH4 uptake. 3-134
Figure 3-47. Effects of N addition on biogenic N2O emission. 3-135
Figure 3-48. Effects of N addition on biogenic N2O emission. 3-136
Figure 3-49. The relationship between N2O emission and N deposition. 3-137
Figure 3-50. Diatom assemblage sediment patterns in Emerald Lake, WY. 3-156
Figure 3-51. Microscopic counts of phytoplankton species composition in the Neuse River Estuary, NC following 36-h in situ
bioassays to manipulate available forms of N. 3-159
Figure 3-52. Map of the western U.S. showing the primary geographic areas where N deposition effects have been reported. 3-168
Figure 3-53. Map of location of wetlands in the eastern U.S. 3-169
xx
-------
Figure 3-54. Overall eutrophication condition on a national scale. 3-172
Figure 3-55. Changes in plant species composition associated with N addition treatments in an alpine dry meadow of the
Colorado Front Range. 3-175
Figure 3-56. Diagram of multiple factors contributing to forest susceptibility to wildfire. 3-178
Figure 3-57. Diagram of relationships of human actions, N loading, and ecosystem services. 3-179
Figure 3-58. Representation of the S cycle in forest ecosystems. 3-186
Figure 3-59. Simplified cycle of mercury, showing the role of sulfur. 3-191
Figure 3-60. (A) SO42" and (B) MeHg concentrations as a function of time in sediment slurries made from Quabbin Reservoir
littoral sediments. 3-192
Figure 3-61. MeHg produced in sediment cores incubated two weeks under artificial lake water containing 3-1040 |jM Na2S04. 3-194
Figure 3-62. The microarchitecture of a dicot leaf. 3-198
Figure A-1. Location of acid-sensitive regions of the northern and eastern U.S. A-3
Figure A-2. Location LTM sites used in the 2003 Surface Water report. A-5
Figure A-3. Long-term record of SO42" concentration in streamwater and precipitation at Watershed 6 of HBEF. A-16
Figure A-4. Location of sampling stations in Sleepers River watershed, Vermont. A-20
Figure A-5. SWAS-VTSS Program Study Sites. A-24
Figure A-6. Conceptual structure of the MAGIC model showing major pools and fluxes included in simulation of effects of S and
N deposition. A-28
Figure A-7. The PnET-BGC model illustrating the compartments and flow paths of carbon and nutrients (C/Nut) within the
model. A-33
Figure A-8. Mathematical form of the SPARROW model. A-36
Figure A-9. Schematic diagram of the WATERSN approach to estimate the contribution made by different N sources to the total
N inputs an estuary. A-41
Figure A-10. WATERSN model estimates of anthropogenic N inputs to the estuaries of the northeastern U.S., in kilograms per
hectare per year. A-43
Figure B-1. MAGIC model hindcast estimates of pre-industrial pH versus diatom-inferred pH for 33 statistically selected
Adirondack lakes: B-14
Figure B-3. Location and percentage of acidic surface waters in U.S. high-interest subpopulations with respect to acidic
deposition effects. B-22
Figure B-4. Measured concentration of SO42" in selected representative lakes and streams in six regions of the U.S. during the
past approximately 15 years. B-29
Figure B-5. Summary of regional trends in surface water chemistry from 1990 through 2000 in regions covered by the Stoddard
et al. (2003) report. B-30
Figure B-6. Estimated time series of S deposition at one example watershed in the SW Adirondack Mountains used by Sullivan
et al. (2006b) as input to the MAGIC model for projecting past and future changes in lakewater chemistry
attributable to acidic deposition. B-32
Figure B-7. Simulated cumulative frequency distributions of lakewater ANC at three points in time for the population of
Adirondack lakes. B-33
Figure B-8. Map showing simulated changes in streamwater ANC from 1995 to 2040 in response to the SAMIA2 emmissions
control strategy, representing existing emissions control regulations. B-37
Figure B-9. Major geomorphic units and locations of lakes sampled in the Western Lake Survey. Those areas known to contain
sensitive lake resources are shaded with cross-hatching. B-42
Figure B-10. Estimated percent changes in the total deposition of sulfur, reduced nitrogen, and nitrate-nitrogen at MAGIC
modeling sites from 1995 to 2040 under each of the emissions control strategies B-43
Figure B-11. Relationship between mean summer and spring ANC values at LTM sites in New England, the Adirondacks, and
the Northern Appalachian Plateau. B-49
xxi
-------
Figure B-12. Minimum streamwater ANC sampled at each site during each year versus median spring ANC for all samples
collected at that site during that spring season. B-50
Figure B-13 Relationship between ANC and runoff for streamwater samples collected at intensively studied sites in Shenandoah
National Park. B-52
Figure B-14. Decrease in ANC and pH and increase in dissolved aluminum in response to a sharp increase in streamflow in
three watersheds within Shenandoah National Park during a hydrological episode in 1995. B-53
Figure B-15. Length-adjusted condition factor (K), a measure of body size in blacknose dace (Rhinichthys atratulus) compared
with mean stream pH among 11 populations (n = 442) in Shenandoah National Park. B-56
Figure B-16. Mean residual number of species per lake for lakes in Ontario, by pH interval. B-59
Figure B-17. Number of fish species per lake or stream versus acidity statues, expressed either as pH or ANC. B-60
Figure B-18. Number of fish species among 13 streams in Shenandoah National Park. Values of ANC are means based on
quarterly measurements, 1987-94. B-61
Figure B-19. Life stages of brook trout. B-64
Figure B-20. Example model application. B-71
Figure B-21. Time series data for SO42", NO3", base cations [Ca2+ + Mg2+], Gran ANC, pH, and DOC in example Long Term
Monitoring Lakes and streams that have relatively low ANC. B-76
Figure C-1. Schematic representation of the response of vegetation to nutrient addition. C-16
Figure C-2. Distribution of alpine vegetation in three western regions that are in close proximity to urban and agricultural
sources of atmospheric N emissions C-18
Figure C-3. Number of nationally rare species versus standing crop in each of 401 quadrants from wetlands in Ontario,
Quebec, and Nova Scotia. C-22
Figure C-4. Sample stressors and the essential ecological attributes they affect. C-36
Figure C-5. Linkages among various ecosystem goods and services (food, water, biodiversity, forest products) and other
driving forces (climate change). C-38
Figure C-6. Effect of watershed defoliation by the gypsy moth caterpillar on NO3" flux in streamwater. C-41
Figure D-1. Conceptual patterns of pollutant deposition effects on a chemical indicator and a corresponding biological indicator
during increasing and decreasing deposition. D-11
Figure D-2. Pollutant deposition patterns for defining the temporal parameters of dynamic critical loads analyses. D-12
Figure F-1. Illustration chart of the assessment. F-2
Figure F-2. Reviewed studies by ecosystem addressed. F-5
Figure F-3. Reviewed studies by ecological endpoint. F-6
Figure F-4. Taxonomy of values for environmental goods and services. F-7
Figure F-5. Linkages from emissions to forest aesthetics. F-13
Figure F-6. Geographic distribution of forested areas historically affected by air pollution. F-14
xxii
-------
Acronyms and Abbreviations
ACCENT Atmospheric Composition Change: the
European Network of excellence
AIRMoN Atmospheric Integrated Research Monitoring
Network
AIRS Atmospheric Infrared Sounder (instrument)
Al aluminum
Al3+ aluminum ion
Ali inorganic aluminum
Aln+ aluminum ion
Alo organic aluminum
AI(OH)3 aluminum hydroxide
ALSC Adirondack Lake Survey Corporation
ALTM Adirondack Long Term Monitoring
AMD acid mine drainage
ANC acid neutralizing capacity
AOD aerosol optical depth
AQCD Air Quality Criteria Document
AQEG Air Quality Expert Group
AQI Air Quality Index
AQS Air Quality System (database)
Ar argon
ARS Agricultural Research Service
As arsenic
ASI Acid Stress Index
asl above sea level
ATMOS Atmospheric Trace Molecule Spectroscopy
ATTILA type of Lagrangian model
AUSPEX Atmospheric Utility Signatures, Predictions,
and Experiments
AVIRIS Airborne Visible and Infrared Imaging
Spectrometer
Ba barium
BBW Bear Brook Watershed
BBWM Bear Brook Watershed, Maine
BC black carbon
BCS base-cation surplus
BGC Biogeochemical (model)
B-IBI benthic index of biological integrity
BMPs best management practices
BNF bacterial nitrogen fertilization
Br bromine
Br bromine ion
Br2 molecular bromine
BrCI bromine chloride
BrO bromine oxide
BUV Backscatter Ultraviolet Spectrometer
BUVD Beneficial Use Values Database
C carbon; concentration
12c
13C
Ca
Ca
Ca2+
CAA
CAAA
CAAAC
CaCb
CaC03
CALIPSO
Ca(N03)2
Ca(OH)2
CAPMoN
CaS04-2H20
CASTNet
CB4
Cd
CEC
CENTURY
CFCs
CG
Chi a
CH4
C2H4
C2H6
CsHs
CH3CHO
CH3C(0)
CH3C(0)00
CH2I2
CH20
CH300H
CH3-S-CH3
CHs-S-H
(CH3)2SO
CH3SO3H
CH3-S-S-CH3
Ci
CL
CI
Cl"
Cl2
carbon-12, stable isotope of carbon
carbon-13, stable isotope of carbon
ambient air concentration
calcium
calcium ion
Clean Air Act
Amendments to the Clean Air Act
Clean Air Act Advisory Committee
calcium chloride
calcium carbonate
Cloud-Aerosol Lidar and Infrared Pathfinder
Satellite Observation (satellite)
calcium nitrate
calcium hydroxide
Canadian Air and Precipitation Monitoring
Network
gypsum
Clean Air Status and Trends Network
Carbon Bond 4 (model)
cadmium
cation exchange capacity
model that simulates carbon, nitrogen,
phosphorus, sulfur, and water dynamics in the
soil-plant system at monthly intervals over time
scales of centuries and millennia
chlorinated fluorocarbons
cloud-to-ground (lightning flash)
chlorophyll a
methane
ethene
ethane
isoprene
acetaldehyde
acetyl radical
acetyl peroxy radical
diiodomethane
formaldehyde
methyl hydroperoxide
dimethylsulfide, DMS
methyl mercaptan
dimethyl sulfoxide, DMSO
methanesulfonic acid
dimethyl disulfide, DMDS
interstitial air concentration
critical load
chlorine
chlorine ion
molecular chlorine
xxiii
-------
CLaMS type of Lagrangian model
CloudSat NASA Earth observation satellite
CINO2 nitryl chloride
CMAQ Community Multiscale Air Quality (modeling
system)
CMSA consolidated metropolitan statistical area
CO carbon monoxide
CO2 carbon dioxide
CO3" carbonate
CONUS continental U.S.
CPUE catch per unit effort
CRREL U.S. Army Cold Regions Research and
Engineering Laboratory
CS Consumer surplus
CS2 carbon disulfide
CSS coastal sage scrub (ecosystem)
CTM chemical transport model
Cu copper
CV contingent valuation
CVM contingent valuation method
A, 6 delta, difference; change
DayCent model for daily biogeochemistry for forest,
grassland, cropland, and savanna systems
DayCent-Chem combination of DayCent-Chem and PHREEQC
models
DC dichotomous choice
DDRP Direct Delayed Response Project
DDT Damage Delay Time
DECOMP decomposition model based on soil-plant
system dynamics
DEP Department of Environmental Protection
DIC dissolved inorganic carbon
DIN dissolved inorganic nitrogen
DMDS dimethyl disulfide, CH3-S-S-CH3
DMS dimethyl sulfide, CH3-S-CH3
DMSO dimethylsulfoxide
DNDC Denitrification-Decomposition (model)
DO dissolved oxygen
DOC dissolved organic carbon
DON dissolved organic nitrogen
EBB East Bear Brook
EC elemental carbon
EEAs Essential Ecological Attributes
ELA Experimental Lakes Area
ELS Eastern Lakes Survey
EMAP Environmental Monitoring and Assessment
Program
EMEFS Eulerian Model Evaluation Field Study
EMEP Co-operative Programme for Monitoring and
Evaluation of the Long-range Transmission of
Air Pollutants in Europe
EMF ectomycorrhizal fungi
EOS Earth Observation System
EPA U.S. Environmental Protection Agency
EPT Ephemeroptera-Plecoptera-Tricoptera (index)
ERP Episodic Response Project
ESA European Space Agency
EVRI Environmental Valuation Reference Inventory
F flux
F" fluorine ion
FAB First-order Acidity Balance model
FACE free-air CO2 enrichment (studies)
Fe iron
FeP04 iron phosphate
FeS iron sulfide
F-factor fraction of the change in mineral acid anions
that is neutralized by base cation release
FHM Forest Health Monitoring
FIA Forest Inventory and Analysis (program)
FISH Fish in Sensitive Habitats (project)
FLEXPART type of Lagrangian model
ForSAFE three-component model using nitrogen, carbon
cycling, and soil chemistry
FRM Federal Reference Method
FTIR Fourier Transform Infrared Spectroscopy
FW2 black carbon soot
Fx flux
YN2O5 reaction potential coefficient for N2O5
GAW Global Atmospheric Watch (program)
GCE Goddard Cumulus Ensemble (model)
GDP gross domestic product
GEOS Goddard Earth Observing System
GEOS-Chem Goddard Earth Observing System (with global
chemical transport model)
GEOS-1 DAS Goddard Earth Observing System Data
Assimilation System
GFED Global Fire Emissions Database
GHG greenhouse gas
GOES Geostationary Operational Environmental
Satellites
GOME Global Ozone Monitoring Experiment
gs stomatal conductance
GtC global ton carbon
Gton global ton
GWP global warming potential
H hydrogen; hydrogen atom
2H hydrogen-2, deuterium, stable isotope of
hydrogen
H- proton, hydrogen ion; relative acidity
ha hectare
HAPs hazardous air pollutants
HBEF Hubbard Brook Experimental Forest
HBES Hubbard Brook Ecosystem Study
HBN Hydrologic Benchmark Network
HC hydrocarbon
HCHO formaldehyde
HCI hydrochloric acid
xxiv
-------
Hg mercury
HNO2, HONO nitrous acid
HNOs, HOONO nitric acid
HNO4 pernitric acid
HO2 hydroperoxyl radical
H2O2 hydrogen peroxide
HO2NO2 peroxynitric acid
HOBr hypobromous acid
HOCI hypochlorous acid
HOX hypohalous acid
HP hedonic pricing
HSO3" bisulfate ion
HSO4" sulfuric acid ion
H2S hydrogen sulfide
H2SO3 sulfurous acid
H2SO4 sulfuric acid
hv photon with energy at wavelength v
I iodine
I2 molecular iodine
IA Integrated Assessment
IADN Integrated Atmospheric Monitoring Deposition
Network
IC intracloud (lightning flash)
ILWAS Integrated Lake-Watershed Acidification Study
IPC International Cooperative Programme
lEc Industrial Economicsym
NASA International Institute for Applied Systems
Analysis
IMPROVE Interagency Monitoring of Protected Visual
Environments
INO3 iodine nitrate
INTEX-NA Intercontinental Chemical Transport
Experiment - North America
10 iodine oxide
IPCC Intergovernmental Panel on Climate Change
IPCC-AR4 Intergovernmental Panel on Climate Change
4th Assessment Report
IPCC-TAR Intergovernmental Panel on Climate Change
3rd Assessment Report
IQR interquartile range
IR infrared
ISA Integrated Science Assessment
J flux from a leaf, deposition flux (g/cm/second)
JPL Jet Propulsion Laboratory
JRGCE Jasper Ridge Global Climate Change
Experiment
K potassium
K- potassium ion
Ka dissociation constant
Kb dissociation constant
KH Henry's Law constant in M/atm (M-atrrr1)
KNO3 potassium nitrate
Kw ion product of water
LAF Lake Acidification and Fisheries
LAR leaf-area ratio
LB laboratory bioassay
LC0.01 lethal concentration at which 0.01 % of exposed
animals die
LD33 lethal dose at which 33% of exposed animals
die
LDH lactic acid dehydrogenase
LIDAR Light Detection and Ranging (remote sensing
system)
LIF laser-induced fluorescence
LIMS Limb Infrared Monitor of the Stratosphere
LOD limit of detection
LP long-path
LRTAP Long Range Transport of Air Pollution
LTER Long-Term Ecological Research (program)
LTM Long-Term Monitoring (project)
M air molecule
MA Millennium Ecosystem Assessment
MAGIC Model of Acidification of Groundwater in
Catchments (model)
MAHA Mid-Atlantic Highlands Assessment of streams
MAQSIP Multiscale Air Quality Simulation Platform
(model)
MAT moist acidic tundra
MAX-DOAS multiple axis differential optical absorption
spectroscopy
MBL marine boundary layer
MDN Mercury Deposition Network
MeHg methylmercury
MEM model ensemble mean
|jeq microequivalent
Mg magnesium
Mg2+ magnesium ion
MIMS membrane inlet mass spectrometry
MM5 National Center for Atmospheric
Research/Penn State Mesoscale Model,
version 5
Mn manganese
MOBILE6 Highway Vehicle Emission Factor Model
MODIS Moderate Resolution Imaging
Spectroradiometer
MOPITT Measurement of Pollution in the Troposphere
M OZAIC Measurement of Ozone and Water Vapor by
Airbus In-Service Aircraft
MOZART Model for Ozone and Related Chemical
Tracers
MPAN peroxymethacrylic nitrate
MSA metropolitan statistical area
Mt million tons
N nitrogen
N, n number of observations
14N nitrogen-14, stable isotope of nitrogen
15N nitrogen-15, stable isotope of nitrogen
xxv
-------
N2 molecular nitrogen; nonreactive nitrogen
NA not available; insufficient data
Na sodium
Na+ sodium ion
NAAQS National Ambient Air Quality Standards
NaCI sodium chloride
NADP National Atmospheric Deposition Program
Na2Mo04 sodium molybdate
NAMS National Air Monitoring Stations
NANI Net anthropogenic nitrogen inputs
NAPAP National Acid Precipitation Assessment
Program
NASQAN National Stream Quality Accounting Network
NARSTO program formerly known as North American
Regional Strategy for Atmospheric Ozone
NAS National Academy of Sciences
NASA National Aeronautics and Space Administration
Na2S04 sodium sulfate
NASQAN National Stream Quality Accounting Network
NATTS National Air Toxics Trends (network)
NAWQA National Water Quality Assessment (program)
NCore National Core Monitoring Network
NEE net ecosystem exchange
NEG/ECP New England Governors and Eastern
Canadian Premiers
NEI National Emissions Inventory
NEON National Ecological Observatory Network
NEP net ecosystem productivity
NFI net factor income
NH3 ammonia
NH2 amino (chemical group)
NH4+ ammonium ion
NH4CI ammonium chloride
NH4NO3 ammonium nitrate
(NH4)2S04 ammonium sulfate
NHx category label for NH3 plus NH4"*
NHy total reduced nitrogen
Ni nickel
NILU Norwegian Institute for Air Research
NITREX Nitrogen saturation Experiments
nitro-PAH nitro-polycyclic aromatic hydrocarbon
NLCD National Land Cover Data
NMOC nonmethane organic compound
NO nitric oxide
NO2 nitrogen dioxide
NO2" nitrite
NO3" nitrate
N2O nitrous oxide
N2O5 dinitrogen pentoxide
NOAA U.S. National Oceanic and Atmospheric
Administration
NOAA-ARL U.S. National Oceanic and Atmospheric
Administration Air Resources Laboratory
NOAEL no-observed-adverse-effect level
NOEC no-observed-effect concentration
NOx sum of NO and NO2
NOy sum of NOx and NOz; odd nitrogen species;
total oxidized nitrogen
NOz sum of all inorganic and organic reaction
products of NOx (HONO, HNO3, HNO4, organic
nitrates, particulate nitrate, nitro-PAHs, etc.)
NPOESS National Polar-orbiting Operational
Environmental Satellite System
NPP net primary production
NPS National Park Service
Nr reactive nitrogen
NRC National Research Council
NS nonsignificant
NSF National Science Foundation
NSS National Stream Survey
nss non-sea salt
NSTC National Science and Technology Council
NSWS National Surface Water Survey
NTN National Trends Network
NuCM nutrient cycling model
02 molecular oxygen
03 ozone
160 oxygen-16, stable isotope of oxygen
180 oxygen-18, stable isotope of oxygen
190 oxygen-19, radioactive isotope of oxygen
OC organic carbon
OCO Orbiting Carbon Observatory
OCS carbonyl sulfide
0(1D) electronically excited oxygen atom
OH hydroxyl radical
OMI Ozone Monitoring Instrument
0(3P) ground-state oxygen atom
P phosphorus
P, p probability value
Pi 1st percentile
Ps 5th percentile
P95 95th percentile
P99 99th percentile
PAHs polycyclic aromatic hydrocarbons
PAMS Photochemical Assessment Monitoring
Stations
PAN peroxyacetyl nitrate
PANs peroxyacyl nitrates
PARASOL Polarization and Anisotropy of Reflectances for
Atmospheric Sciences coupled with
Observations from a Lidar (satellite)
Pb lead
PBL planetary boundary layer
PC payment card
PCBs polychlorinated biphenyl compounds
pH relative acidity
xxvi
-------
P(HN03) production of nitric acid
PHREEQC model for soil and water geochemical
equilibrium
PIRLA Paleocological Investigation of Recent Lake
Acidification (projects)
pKa dissociation constant
PM particulate matter
PM2.5 particulate matter with aerodynamic diameter
of #2.5 |jm
PM10 particulate matter with aerodynamic diameter
#10 |jm
PM10-2.5 particulate matter with aerodynamic diameter
between 10 and 2.5 |jm
PM-CAMx Comprehensive Air Quality Model with
extensions and with particulate matter
chemistry
PnET Photosynthesis and EvapoTranspiration
(model)
PnET-BGC Photosynthesis and EvapoTranspiration-
Biogeochemical (model)
PnET-CN Photosynthesis and EvapoTranspiration model
of C, water, and N balances
PnET-N-DNDC Photosynthesis and EvapoTranspiration-
Denitrification-Decomposition (model)
PNO3" particulate nitrate
P(03) production of O3
PO4", PO43" phosphate
POPs persistent organic pollutants
ppb parts per billion
PPN peroxypropionyl nitrate
ppt parts per trillion
PRB policy relevant background
PRE-STORM Preliminary Regional Experiment for STORM
PROFILE model using soil mineralogy as input
PS producer surplus
PSO42" particulate sulfate
P(S042") production of sulfate
Q flow rate; discharge
Q10 temperature coefficient
QAPP Quality Assurance Project Plan
R generic organic group attached to a molecule
R2 coefficient of determination
r2 correlation coefficient
Ra aerodynamic resistance
Rb boundary layer resistance
Rc internal resistance
RADM Regional Acid Deposition Model
RAMS Regional Atmospheric Modeling System
RAPS Regional Air Pollution Study
RCOO-s strongly acidic organic anions
RC(0)00 organic peroxy radical
RDT Recovery Delay Time
REMAP Regional Environmental Monitoring and
Assessment Program
RH relative humidity
RLTM Regional Long-Term Monitoring
RMCC Research and Monitoring Coordinating
Committee
RMSE root mean squared error
RO2 organic peroxyl; organic peroxy
RONO2 organic nitrate
RO2NO2 peroxynitrate
RP revealed preferences
RRx lognormal-transformed response ratio
RuBisCO ribulose-1,5-bisphosphate
carboxylase/oxygenase
s second
S sulfur
32S sulfur-32, stable isotope of sulfur
34S sulfur-34, stable isotope of sulfur
35S sulfur-35, radioactive isotope of sulfur
SAA sum of mineral acid anion concentrations
SAFE Soil Acidification in Forest Ecosystems (model)
SAMAB Southern Appalachian Man and the Biosphere
(program)
SAMI Southern Appalachian Mountains Initiative
SAO Smithsonian Astrophysical Observatory
SAPRAC Statewide Air Pollution Research Center
SBC sum of base cation concentrations
SBUV Solar Backscatter Ultraviolet Spectrometer
SC safe concentration
SCAQS Southern California Air Quality Study
SCIAMACHY Scanning Imaging Absorption Spectrometer for
Atmospheric Chartography
Se selenium; standard error
SEARCH Southeastern Aerosol Research and
Characterization Study (monitoring program)
Si silicon
SIP State Implementation Plan
S JAQS San Joaquin Valley Air Quality Study
SLA specific leaf area
SLAMS State and Local Air Monitoring Stations
SMART Simulation Model for Acidification's Regional
Trends (model)
SMB Simple Mass Balance (model)
SO sulfur monoxide
502 sulfur dioxide
503 sulfur trioxide
SO32" sulfite
SO42" sulfate ion
S2O disulfur monoxide
SONEX Subsonics Assessment Ozone and Nitrogen
Oxides Experiment
SOS Southern Oxidant Study
SOSH State of Science/Technology (report)
SOx sulfur oxides
SP stated preferences
xxvii
-------
SPARROW SPAtially Referenced Regressions on
Watershed Attributes (model)
Sr strontium
86Sr strontium-86, stable isotope of strontium
87Sr strontium-87, stable isotope of strontium
SRB sulfate-reducing bacteria
SRP soluble reactive phosphorus
SSWC Steady State Water Chemistry (model)
STE stratospheric-tropospheric exchange
STN Speciation Trends Network
SUM06 seasonal sum of all hourly average
concentrations > 0.06 ppm
SVOC semivolatile organic compound
SWAS Shenandoah Watershed Study
T, t tau, atmospheric lifetime
T time; duration of exposure
TAF Tracking and Analysis Framework (model)
T air air temperature
TAMM Timber Assessment Market Model
TAR Third Assessment Report
TC total carbon; travel cost
TCM travel cost method
TDLAS Tunable Diode Laser Absorption Spectrometer
Tg teragram
TIME Temporally Integrated Monitoring of
Ecosystems (program)
TN total nitrogen
TOMS Total Ozone Mapping Spectrometer
TOR tropospheric ozone residual
TP total phosphorus
TRACE-P Transport and Chemical Evolution over the
Pacific
TSI timber-stand improvement
TSS total suspended solids
Twater water temperature
UMD-CTM University of Maryland Chemical Transport
Model
UNECE United Nations Economic Commission for
Europe
USDA U.S. Department of Agriculture
USFS U.S. Forest Service
USGS U.S. Geological Survey
UV ultraviolet
UV-A ultraviolet radiation of wavelengths from 320 to
400 nm
UV-B ultraviolet radiation of wavelengths from 280 to
320 nm
Vd deposition rate, deposition velocity (cm/s)
VOC volatile organic compound
VSD Very Simple Dynamic (soil acidification model)
VTSSS Virginia Trout Stream Sensitivity Study
WARMS Waterfowl Acidification Response Modeling
System
WATERSN
Watershed Assessment Tool for Evaluating
Reduction Scenarios for Nitrogen
WBB
West Bear Brook
WEBB
Water, Energy, and Biogeochemical Budgets
WFPS
water-filled pore space
WGE
Working Group on Effects
WLS
Western Lakes Survey
WMO
World Meteorological Organization
WMP
Watershed Manipulation Project
WSA
Wadeable Stream Assessment (survey)
wt %
percent by weight
WTA
willingness-to-accept
WTP
willingness-to-pay
XNOs
nitrate halogen-X salt
XO
halogen-X oxide
Zn
zinc
ZnO
zinc oxide
xxviii
-------
Authors, Contributors, Reviewers
Authors
Dr. Tara Greaver (NOx and S0X Project Manager)—National Center for Environmental Assessment,
Office of Research and Development, U.S. Environmental Protection Agency, Research Triangle Park,
NC
Dr. Jeffrey R. Arnold—National Center for Environmental Assessment, Office of Research and
Development, U.S. Environmental Protection Agency, Research Triangle Park, NC
Dr. Jill S. Baron—U.S. Geological Survey, Natural Resource Ecology Laboratory, Colorado State
University, Fort Collins CO
Dr. Jana Compton—National Health and Environmental Effects Research Laboratory, Office of Research
and Development, U.S. Environmental Protection Agency, Corvallis, OR
Dr. Bernard J. Cosby, Jr.—Department of Environmental Sciences, University of Virginia,
Charlottesville, VA
Dr. Ila Cote—National Center for Environmental Assessment, Office of Research and Development,
U.S. Environmental Protection Agency, Research Triangle Park, NC
Ms. Rebecca Daniels— National Health and Environmental Effects Research Laboratory, Office of
Research and Development, U.S. Environmental Protection Agency, Research Triangle Park, NC
Dr. Jean-Jacques B. Dubois—National Center for Environmental Assessment, Office of Research and
Development, U.S. Environmental Protection Agency, Research Triangle Park, NC
Dr. Christine L. Goodale—Department of Ecology and Evolutionary Biology, Cornell University,
Ithaca, NY
Dr. Alan T. Herlihy—Department of Fisheries & Wildlife, Oregon State University, Corvallis, OR
Dr. Jeffrey D. Herrick—National Center for Environmental Assessment, Office of Research and
Development, U.S. Environmental Protection Agency, Research Triangle Park, NC
Dr. Alan J. Krupnick—Resources for the Future, Washington, DC
Dr. Kathleen Fallon Lambert—Ecologic: Analysis and Communications, Woodstock, VT
Dr. Gregory B. Lawrence—U.S. Geological Survey Troy, NY
Dr. Lingli Liu—National Center for Environmental Assessment, Office of Research and Development,
U.S. Environmental Protection Agency, Research Triangle Park, NC
Dr. Todd C. McDonnell—E&S Environmental Chemistry, Inc., Corvallis, OR
Dr. Kristopher Novak—National Center for Environmental Assessment, Office of Research and
Development, U.S. Environmental Protection Agency, Research Triangle Park, NC
Dr. Joseph Pinto—National Center for Environmental Assessment, Office of Research and Development,
U.S. Environmental Protection Agency, Research Triangle Park, NC
xxix
-------
Dr. Mary A. Ross—National Center for Environmental Assessment, Office of Research and
Development, U.S. Environmental Protection Agency, Research Triangle Park, NC
Mr. Rich Scheffe—Office of Air Quality Planning and Standards, U.S. Environmental Protection Agency,
Research Triangle Park, NC
Dr. Juha Siikamaki—Resources for the Future, Washington, DC
Dr. Timothy J. Sullivan—E&S Environmental Chemistry, Inc., Corvallis, OR
Dr. Helga Van Miegroet—Department of Wildland Resources, Department of Watershed Sciences, Utah
State University, Logan, UT
Dr. Paul F. Wagner—National Center for Environmental Assessment, Office of Research and
Development, U.S. Environmental Protection Agency, Research Triangle Park, NC
Contributors
Dr. Robin L. Dennis—National Effects Research Laboratory, Office of Research and Development, U.S.
Environmental Protection Agency, Research Triangle Park, NC
Dr. Russell Dickerson—Dept. of Atmospheric and Oceanic Sciences, University of Maryland, College
Park, MD
Dr. Tina Fan— Environmental and Occupational Health Sciences Institute/ University of Medicine and
Dentistry: New Jersey, Piscataway, NJ
Dr. Arlene Fiore—Geophysical Fluid Dynamics Laboratory/NOAA, Princeton, NJ
Dr. Larry Horowitz—Geophysical Fluid Dynamics Laboratory/NOAA, Princeton, NJ
Dr. William Keene—Dept. of Environmental Sciences, University of Virginia, Charlottesville, VA
Dr. Randall Martin—Dept. of Physics, Dalhousie University, Halifax, Nova Scotia, Canada
Dr. William Munger—Center for Earth and Planetary Physics, Harvard University, Cambridge, MA
Dr. Sandy Sillman—Dept. of Atmospheric and Oceanic Sciences, University of Michigan, Ann Arbor,
MI
Dr. John T. Walker—National Risk Management Research Laboratory, Office of Research and
Development, U.S. Environmental Protection Agency, Research Triangle Park, NC
Reviewers
Dr. Lawrence Band—Department of Geography, UNC Chapel Hill, Chapel Hill, NC
Dr. Tamara Blett—National Park Service, Lakewood, CO
Dr. Jana Compton—National Health and Environmental Effects Research Laboratory, Office of Research
and Development, U.S. Environmental Protection Agency, Corvallis, OR
Dr. Russell Dickerson—Department of Chemistry & Chemical Physics, University or Maryland,
College Park, MD
xxx
-------
Dr. Dave A. Evans—Office of Policy, Economics, and Innovation, Washington, DC
Dr. Dale Evarts—Office of Air and Radiation, Office of Air Quality Planning and Standards,
U.S. Environmental Protection Agency, Research Triangle Park, NC
Dr. Jacqueline Geoghegan—Department of Economics, Clark University, Worcester, MA
Dr. D. Alan Hansen—Electric Power Research Institute, Palo Alto, CA
Dr. Brooke Hemming—National Center for Environmental Assessment, Office of Research and
Development, U.S. Environmental Protection Agency, Research Triangle Park, NC
Dr. Brian Heninger—Office of Policy, Economics, and Innovation, Washington, DC
Dr. William Hogsett—National Health and Environmental Effects Research Laboratory, Office of
Research and Development, U.S. Environmental Protection Agency, Corvallis, OR
Dr. Bryan Hubbell—Office of Air and Radiation, Office of Air Quality Planning and Standards,
U.S. Environmental Protection Agency, Research Triangle Park, NC
Dr. Gary Lear—Office of Air and Radiation, Office of Atmospheric Programs Washington, DC
Dr. John Lehrter—National Health and Environmental Effects Research Laboratory, Office of Research
and Development, U.S. Environmental Protection Agency, Gulf Breeze, FL
Dr. Jason Lynch—Office of Air and Radiation, Office of Atmospheric Programs, U.S. Environmental
Protection Agency, Washington, DC
Dr. William Malm—National Park Service, Fort Collins, CO
Dr. Steve McNulty—USDA Forest Service, Raleigh, NC
Ms. Connie Meacham—National Center for Environmental Assessment, Office of Research and
Development, U.S. Environmental Protection Agency, Research Triangle Park, NC
Mr. Greg Miller—Office of Policy, Economics, and Innovation, Washington, DC
Dr. James Morris—Belle W. Baruch Institute for Marine & Coastal Sciences, University of South
Carolina, Columbia, SC
Dr. William Munger—School of Engineering and Applied Science Harvard University, Cambridge, MA
Dr. Knute Nadelhoffer—University of Michigan Biological Station, Pellston, MI
Dr. Christine Negra—The Heinz Center, Washington, DC
Dr. William Orem—U.S. Geological Survey, Reston, VA
Dr. Ellen Porter—National Park Service, Denver, CO
Dr. Anne Rea—Office of Air and Radiation, Office of Air Quality Planning and Standards,
U.S. Environmental Protection Agency, Research Triangle Park, NC
Dr. Paul Ringold—National Health and Environmental Effects Research Laboratory, Office of Research
and Development, U.S. Environmental Protection Agency, Corvallis, OR
Ms. Vicki Sandiford—Office of Air and Radiation, Office of Air Quality Planning and Standards,
U.S. Environmental Protection Agency, Research Triangle Park, NC
xxxi
-------
Dr. William Showers—Dept. of Marine, Earth and Atmospheric Sciences, North Carolina State
University, Raleigh, NC
Dr. Gail Tonnesen—University of California, Denver, CO
Dr. John Vandenberg (Division Director) —National Center for Environmental Assessment-RTP
Division, Office of Research and Development, U.S. Environmental Protection Agency, Research
Triangle Park, NC
Dr. Lisa Wainger—TN and Associates, Hollywood, MD
Mr. Randy Waite—Office of Air and Radiation, Office of Air Quality Planning and Standards, U.S.
Environmental Protection Agency, Research Triangle Park, NC
Ms. Debra Walsh (Deputy Division Director)— National Center for Environmental Assessment-RTP
Division, Office of Research and Development, U.S. Environmental Protection Agency, Research
Triangle Park, NC
Dr. Shaun Watmough—Dept. of Environmental and Resource Studies, Trent University, Peterborough,
Ontario
Ms. Lydia Wegman—Office of Air and Radiation, Office of Air Quality Planning and Standards
U.S. Environmental Protection Agency, Research Triangle Park, NC
Dr. Jason West—Dept. of Environmental Sciences & Engineering, University of North Carolina,
Chapel Hill, NC
Dr. William Wilson—National Center for Environmental Assessment, Office of Research and
Development, U.S. Environmental Protection Agency, Research Triangle Park, NC
xxxii
-------
Project Team
Executive Direction
Dr. Ila Cote (Acting Division Director)—National Center for Environmental Assessment-RTP Division,
Office of Research and Development, U.S. Environmental Protection Agency, Research Triangle Park,
NC
Dr. John Vandenberg (Division Director) —National Center for Environmental Assessment-RTP
Division, Office of Research and Development, U.S. Environmental Protection Agency, Research
Triangle Park, NC
Ms. Debra Walsh (Deputy Division Director)— National Center for Environmental Assessment-RTP
Division, Office of Research and Development, U.S. Environmental Protection Agency, Research
Triangle Park, NC
Dr. Mary A. Ross (Branch Chief)—National Center for Environmental Assessment, Office of Research
and Development, U.S. Environmental Protection Agency, Research Triangle Park, NC
Scientific Staff
Dr. Tara Greaver (NOx and SOx Project Manager)—National Center for Environmental Assessment,
Office of Research and Development, U.S. Environmental Protection Agency, Research Triangle Park,
NC
Dr. Jeffrey R. Arnold—National Center for Environmental Assessment, Office of Research and
Development, U.S. Environmental Protection Agency, Research Triangle Park, NC
Dr. Jean-Jacques Dubois—Oak Ridge Institute for Science and Education, Postdoctoral Research Fellow
to National Center for Environmental Assessment, Office of Research and Development, U.S.
Environmental Protection Agency, Research Triangle Park, NC
Dr. Jeffrey D. Herrick—National Center for Environmental Assessment, Office of Research and
Development, U.S. Environmental Protection Agency, Research Triangle Park, NC
Dr. Lingli Liu— Oak Ridge Institute for Science and Education, Postdoctoral Research Fellow to
National Center for National Center for Environmental Assessment, Office of Research and
Development, U.S. Environmental Protection Agency, Research Triangle Park, NC
Dr. Kristopher Novak—National Center for Environmental Assessment, U.S. Environmental Protection
Agency, Research Triangle Park, NC
Dr. Paul F. Wagner—National Center for Environmental Assessment, Office of Research and
Development, U.S. Environmental Protection Agency, Research Triangle Park, NC
xxxiii
-------
Technical Support Staff
Ms. Ellen Lorang—National Center for Environmental Assessment, Office of Research and
Development, U.S. Environmental Protection Agency, Research Triangle Park, NC
Ms. Connie Meacham—National Center for Environmental Assessment, Office of Research and
Development, U.S. Environmental Protection Agency, Research Triangle Park, NC
Ms. Deborah Wales—National Center for Environmental Assessment, Office of Research and
Development, U.S. Environmental Protection Agency, Research Triangle Park, NC
Mr. Richard Wilson—National Center for Environmental Assessment, Office of Research and
Development, U.S. Environmental Protection Agency, Research Triangle Park, NC
xxxiv
-------
Clean Air Scientific Advisory Committee
NOx and SOx Secondary NAAQS Review Panel
Dr. Armistead (Ted) Russell* (Chair)—Georgia Power Distinguished Professor of Environmental
Engineering, Environmental Engineering Group, School of Civil and Environmental Engineering,
Georgia Institute of Technology, Atlanta, GA
Dr. Praveen Amar, Director, Science and Policy, Northeast States for Coordinated Air Use Management,
Boston, MA
Dr. Andrzej Bytnerowicz, Senior Scientist, Pacific Southwest Research Station, USDA Forest Service,
Riverside, CA
Ms. Lauraine Chestnut, Managing Economist, Stratus Consulting Inc., Boulder, CO
Dr. Ellis B. Cowling*, Emeritus Professor, Colleges of Natural Resources and Agriculture and Life
Sciences, North Carolina State University, Raleigh, NC
Dr. Douglas Crawford-Brown*, Professor and Director, Department of Environmental Sciences and
Engineering, Carolina Environmental Program, University of North Carolina at Chapel Hill, Chapel Hill,
NC
Dr. Charles T. Driscoll, Jr., Professor, Environmental Systems Engineering, College of Engineering and
Computer Science, Syracuse University, Syracuse, NY
Dr. Paul J. Hanson, Distinguished R&D Staff Member, Environmental Sciences Division, Oak Ridge
National Laboratory, Oak Ridge, TN
Dr. Rudolf Husar, Professor and Director, Mechanical Engineering, Engineering and Applied Science,
Center for Air Pollution Impact and Trend Analysis, Washington University, St. Louis, MO
Dr. Dale Johnson, Professor, Department of Environmental and Resource Sciences, College of
Agriculture, University of Nevada, Reno, NV
Dr. Donna Kenski*, Director, Lake Michigan Air Directors Consortium, Rosemont, IL
Dr. Naresh Kumar, Senior Program Manager, Environment Division, Electric Power Research Institute,
Palo Alto, CA
Dr. Myron Mitchell, Distinguished Professor and Director of Council on Hydrologic Systems Science,
College of Environmental and Forestry, State University of New York, Syracuse, NY
Mr. Richard L. Poirot, Environmental Analyst, Air Pollution Control Division, Department of
Environmental Conservation, Vermont Agency of Natural Resources, Waterbury, VT
Mr. David J. Shaw, Director, Division of Air Resources, New York State Department of Environmental
Conservation, Albany, NY
Dr. Kathleen Weathers, Senior Scientist, Cary Institute of Ecosystem Studies, Millbrook, NY
* Members of the statutory Clean Air Scientific Advisory committee (CASAC) appointed by the United States (U.S.) Environmental Protection
Agency (EPA) Administrator
xxxvi
-------
Chapter 1. Introduction
This Integrated Science Assessment (ISA) synthesizes and evaluates the most policy-relevant
science to help form the scientific foundation for the review of the secondary (welfare-based) National
Ambient Air Quality Standards (NAAQS) for oxides of nitrogen (NOx) and sulfur oxides (SOx). The
Clean Air Act (CAA) definition of welfare effects includes, but is not limited to, effects on soils, water,
wildlife, vegetation, visibility, weather, and climate, as well as effects on materials, economic values, and
personal comfort and well-being.
The intent of the ISA, according to the CAA, is to "accurately reflect the latest scientific
knowledge expected from the presence of [a] pollutant in ambient air" (U.S. Code, 1970a, 1970b). It
includes scientific research from atmospheric sciences, exposure and deposition, biogeochemistry,
hydrology, soil science, marine science, plant physiology, animal physiology, and ecology conducted at
multiple scales (e.g., population, community, ecosystem, landscape levels). Key information and
judgments formerly found in the Air Quality Criteria Documents (AQCDs) for NOx and SOx are
included; Annexes provide a more detailed discussion of the most pertinent scientific literature. Together,
the ISA and Annexes serve to update and revise the last NOx and SOx AQCDs that were published in
1993 and 1982, respectively.
As discussed in the Integrated Plan for the Review of the Secondary NAAQS for Nitrogen Dioxide
and Sulfur Dioxide (U.S. EPA, 2007a) a series of policy-relevant questions frames this review of the
scientific evidence used to provide a scientific basis for evaluation of the secondary NAAQS for N02
(0.053 parts per million [ppm], annual average) and S02 (0.5 ppm, 3-h average). The framing questions
considered are:
1. What are the known or anticipated welfare effects influenced by ambient NOx and SOx? For
which effects is there sufficient information available to be useful as a basis for considering
distinct secondary standards?
2. What is the nature and magnitude of ecosystem responses to NOx and SOx that are understood to
have known or anticipated adverse effects? What is the variability associated with these responses
(including ecosystem type, climatic conditions, environmental effects, and interactions with other
environmental factors and pollutants)?
3. To what extent do the current standards provide the requisite protection for the public welfare
effects associated with NOx and SOx?
4. Which biotic species are most vulnerable to the adverse effects of NOx and SOx air pollution?
How is adversity defined?
5. What ecosystems are most sensitive to NOx and SOx pollution?
6. How does NOx and SOx pollution impact ecosystem services?
7. What are the most appropriate spatial and temporal scales to evaluate impacts on ecosystems?
8. What is the relationship between ecological vulnerability to NOx and SOx pollution and
variations in current meteorology or gradients in climate?
1.1. Scope
The U.S. EPA is integrating the science assessment for these two criteria air pollutants due to their
combined effects on atmospheric chemistry, deposition processes, and public welfare effects. The focus of
1-1
-------
this assessment is primarily on effects related to the deposition of nitrogen (N)- and sulfur (S)-containing
compounds. Ecological effects from acidification and N-nutrient enrichment have been studied most
extensively in the ecological literature. An assessment of the complex ecological effects of N deposition
requires consideration of multiple forms of N. Thus, this assessment includes evaluation of data on
inorganic reduced forms of N (e.g., ammonia [NH3] and ammonium ion |NH4|). inorganic oxidized
forms (e.g., NOx, nitric acid [HN03], nitrous oxide [N20], nitrate |NO, |). and organic N compounds
(e.g., urea, amines, proteins, nucleic acids). In addition to acidification and N-nutrient enrichment, other
welfare effects related to deposition of N- and S-containing compounds are discussed, such as SOx
interactions with mercury (Hg) methylation. In addition, this assessment includes evidence related to
direct ecological effects of gas-phase NOx and SOx since the direct effects of gas-phase SOx on
vegetation formed a primary basis for the initial establishment of the secondary NAAQS for S02. The
contribution of gas-phase NOx as greenhouse gases (GHG), particularly N20, is considered, chie fly in the
response of soils to reactive nitrogen (Nr) enrichment.
A review of the particulate matter (PM) NAAQS is underway. Recent data on the welfare effects of
airborne particulate NOx and SOx in the ambient air will be evaluated in the PM ISA. These effects
include visibility impairment, soiling and damage to materials, and effects of ambient PM on climate.
(For more information, see http://www.epa.gOv/ttn/naaqs/standards/pm/s pm index.html.)
Gas-phase and particulate N0X and S0X compounds can affect ecosystems and alter numerous
linked biogeochemical cycles. A simplified diagram of the combined N0X and S0X cycle is presented in
Figure 1-1. The ISA includes additional figures that provide more detail on the interactions among
biogeochemical cycles, and the locations of those figures are indicated in the diagram. These figures
include atmospheric cycling, interactions between the N cycle and carbon (C), the N cycle and
phosphorous (P), and the S cycle and Hg.
Ambient Air
Concentration
Deposition
Ecological
Effect
Figure 1-1. Biogeochemical cycles of NOx and SOx.
Dry deposition
NO,, NH„, SO,
C and N cycle in
terrestrial ecosystem
See figure 3-36
Hg cycle in
aquatic ecosystem
See figure 3-59
Oxidation
SO2 ~ H2SO4-
N0X *HN03 -
Dissolution
-+> 2H+ +S042-
-> H++N03-
Atmospheric cycle
of N oxides
See figure 2-16
Atmospheric cycle
of S compounds
See figure 2-18
Wet Deposition
Acidification of water + Eutrophication
C, N and P cycle in
aquatic ecosystem
See figure 3-40
Sunlight
VOC
I
Foliar and
nutrient effects
1-2
-------
1.2. History of the NOx and SOx Review
Nitrogen Oxides
In 1971, U.S. EPA promulgated identical primary and secondary NAAQS for N02: 0.053 ppm as
an annual average (36 FR 8186). The scientific bases for these NAAQS were provided in the AQCD for
NOx(U.S. EPA, 1971).
In 1984, U.S. EPA proposed to retain these standards (49 FR 6866), and after the public comment
period, finalized that decision in 1985 (50 FR 25532); the scientific basis for this review was provided by
the 1982 AQCD forNOx (U.S. EPA, 1982a).
In 1991, U.S. EPA released an updated draft AQCD for the Clean Air Scientific Advisory
Committee (CASAC) and public review and comment (56 FR 59285). CASAC reviewed the document
and concluded it "provides a scientifically balanced and defensible summary of current knowledge of the
effects of this pollutant and provides an adequate basis for U.S. EPA to make a decision as to the
appropriate NAAQS forN02" (Wolff, 1993).
The U.S. EPA also prepared a draft Staff Paper that summarized and integrated the key studies and
scientific evidence contained in the revised AQCD and identified the critical elements to be considered in
the review of the N02 NAAQS. In September 1995, U.S. EPA finalized the Staff Paper, Review of the
National Ambient Air Quality Standards for Nitrogen Dioxide: Assessment of Scientific and Technical
Information (U.S. EPA, 1995b). The Administrator made a final determination that no revisions to the
primary and secondary NAAQS for N02 were appropriate at that time (61 FR 52852, October 8, 1996).
The level for both the existing primary and secondary NAAQS for N02 remains 0.053 ppm (equivalent to
100 micrograms per cubic meter of air | (ig/nr11) in annual arithmetic average, calculated as the arithmetic
mean of the 1-h N02 concentrations.
Sulfur Oxides
Based on the 1970 SOx AQCD (U.S. Department of Health, Education and Welfare, 1970), U.S.
EPA promulgated primary and secondary NAAQS for S02, under Section 109 of the CAA on April 30,
1971 (36 FR 8186). The secondary standard was set at 0.02 ppm in an annual arithmetic mean and a 3-h
average of 0.5 ppm, not to be exceeded more than once per year. These standards were established solely
based on vegetation effects evidence. In 1973, revisions made to Chapter 5 "Effects of Sulfur Oxide in the
Atmosphere on Vegetation" of the SOx AQCD (U.S. EPA, 1973), indicated that it could not properly be
concluded that the reported vegetation injury resulted from the average S02 exposure over the growing
season rather than from short-term peak concentrations. U.S. EPA, therefore, proposed (38 FR 11355) and
then finalized a revocation of the annual mean secondary standard (38 FR 25678).
In 1979, U.S. EPA announced that it was revising the SOx AQCD concurrently with the PM review
and would produce a combined PM SOx AQCD. Following review of the draft revised criteria document
in August 1980, CASAC concluded acidic deposition was atopic of extreme scientific complexity
because of the difficulty in establishing firm quantitative relationships among (a) emissions of relevant
pollutants (e.g., S02 and NOx), (b) formation of acidic wet and dry deposition products, and (c) effects on
terrestrial and aquatic ecosystems. CASAC also noted that acidic deposition involves, at a minimum,
several different criteria pollutants (i.e., SOx, NOx, and the fine particulate fraction of suspended
particles). CASAC recommended that any document on this subject should address both wet and dry
deposition, because dry deposition was believed to account for at least half of the total acid deposition
problem.
The U.S. EPA proposed not to revise the existing primary and secondary standards on April 26,
1988 (53 FR 14926). Regarding the secondary S02 NAAQS, the U.S. EPA Administrator concluded that
based upon then-current scientific understanding of the acidic deposition problem, it would be premature
and unwise to prescribe any regulatory control program at that time, and when the fundamental scientific
uncertainties had been reduced through ongoing research efforts, U.S. EPA would draft and support an
1-3
-------
appropriate set of control measures. On May 22, 1996, U.S. EPA's final decision, that revisions of the
NAAQS for SOxwere not appropriate at that time, was announced in the Federal Register (61 FR 25566).
Acidic Deposition Assessments
Based upon their conclusions from the AQCD review discussed above, CASAC recommended that
a separate, comprehensive document on acidic deposition be prepared before any regulatory consideration
for the control of acidic deposition. CASAC also suggested that a discussion of acidic deposition be
included in the AQCDs for both NOx and PM-SOx. Following CASAC closure on the criteria document
for S02 in 1981, U.S. EPA's Office of Air Quality Planning and Standards (OAQPS) published a Staff
Paper (U.S. EPA, 1982c); it did not, however, directly address this issue. U.S. EPA followed CASAC
guidance and subsequently prepared the following documents: The Acidic Deposition Phenomenon and
Its Effects: Critical Assessment Review Papers, Volumes I andII (U.S. EPA, 1984a, 1984b) and The
Acidic Deposition Phenomenon and Its Effects: Critical Assessment Document (U.S. EPA, 1985). These
documents, though they were not considered criteria documents and did not undergo CASAC review,
represented the most comprehensive summary of relevant scientific information completed by the U.S.
EPA at that point.
Assessment of the ecological effects of NOx and SOx has been conducted under the U.S. EPA acid
precipitation control program. In the 1990 CAA Amendments (CAAA), Title IV was to reduce emissions
of S02 and NOx from fossil fuel-burning power plants to protect ecosystems suffering damage from acid
deposition and to improve air quality. The National Acid Precipitation Assessment Program (NAPAP) has
periodically assessed and reported to Congress on the implementation of the Acid Rain Program, recent
scientific knowledge surrounding acid deposition and its effects, and the reduction in acid deposition
necessary to prevent adverse ecological effects. These assessments were to be reported to Congress
quadrennially, beginning in 1996. The most recent in this series of reports is the National Acid
Precipitation Assessment Program Report to Congress: An Integrated Assessment that was submitted to
Congress in 2005 (NAPAP, 2005).
The 1990 CAAA also required U.S. EPA to conduct a study on the feasibility and effectiveness of
an acid deposition standard or standards to protect "sensitive and critically sensitive aquatic and terrestrial
resources." In 1995, the U.S. EPA submitted to Congress its report titled Acid Deposition Standard
Feasibility Study: Report to Congress (U.S. EPA, 1995a) in fulfillment of this requirement. The Acid
Deposition Standard Feasibility Study Report to Congress concluded establishing acid deposition
standards for S and N deposition may at some point in the future be technically feasible although
appropriate deposition loads for these acidifying chemicals could not defined with reasonable certainty at
that time.
1.3. History of the Current Review
U.S. EPA's National Center for Environmental Assessment in Research Triangle Park, NC
announced the official initiation of the current periodic review of air quality criteria for NOx on
December 9, 2005 (70 FR 73236), and for SOx on May 15, 2006 (71 FR 28023), with a call for
information. A workshop to inform the Agency's review of the secondary standards for these two
pollutants was held on July 17-19, 2007 (72 FR 11960). The first ISA external review draft was published
in December 2007 (72 FR 72719) and reviewed by CASAC at a public meeting on April 2-3, 2008. The
second ISA external review draft was published in August 2008 (73 FR 46908) and reviewed by the
CASAC at a public meeting on October 1-2, 2008. This final ISA (December 2008) includes revisions to
address comments from CASAC and the public.
1-4
-------
1.4. Development of the Integrated Science Assessment
An extensive search and review of the literature is the initial step in preparing the ISA. Additional
publications were identified by U.S. EPA scientists in a variety of disciplines. In addition to peer-reviewed
literature, previous U.S. EPA reports and materials identified in reviewing reference lists were examined.
Further publications have been identified through the peer review process by CASAC, other experts, and
the public. The focus of this ISA is on literature published since the 1993 NOx AQCD and the 1982 SOx
AQCD. Key findings and conclusions from the 1993 and 1982 reviews are discussed in conjunction with
recent studies. In addition, analyses of air quality and emissions data, and studies on atmospheric
chemistry, transport, and fate of these emissions were scrutinized.
Emphasis was placed on studies that evaluated effects near ambient levels and studies that consider
NOx and SOx as components of a complex mixture of air pollutants. Studies conducted in any country
that contribute significantly to the knowledge base were considered for inclusion. In evaluating
quantitative exposure-response relationships, emphasis was placed on findings from studies conducted in
the U.S. and Canada as having ecological and climatic conditions most relevant for review of the
NAAQS. In assessing the relative scientific quality of studies reviewed here and to assist in interpreting
the findings, the following were considered:
1. To what extent are the aerometric data/exposure metrics of adequate quality and sufficiently
representative to serve as credible exposure indicators?
2. Were the study populations well-defined and adequately selected to allow for meaningful
comparisons between study groups?
3. Were the ecological assessment endpoints reliable and policy-relevant?
4. Were the statistical analyses used appropriately and properly performed and interpreted?
5. Were likely important covariates (e.g., potential confounders or effect modifiers) adequately
controlled or taken into account in the study design and statistical analyses?
6. Were the reported findings consistent, biologically plausible, and coherent in terms of consistency
with other known facts?
These guidelines provide benchmarks for evaluating various studies and for focusing on the highest
quality studies in assessing the body of environmental effects evidence. Detailed critical analysis of all
NOx and SOx environmental effects studies, especially in relation to the above considerations, is beyond
the scope of the ISA and Annexes. Studies providing qualitative or quantitative information on exposure-
response relationships for the environmental effects associated with current ambient air concentrations of
NOx and SOx or deposition levels likely to be encountered in the U.S. were considered most relevant.
1.5. Organization of the Integrated Science Assessment
This ISA has four chapters. Chapter 1 provides background information on the purpose of the
document, explains how policy-relevant scientific studies are identified and selected for inclusion, and
introduces the causality framework used in U.S. EPA's assessments. Chapter 2 presents fundamental and
applied atmospheric science data to support assessing the environmental exposures and effects associated
with N and S oxides. Information relevant to the review of the welfare effects of NOx and SOxis
integrated and evaluated in Chapter 3. Findings are organized into three categories: ecological effects of
acidification, ecological effects ofN nutrient pollution, and other welfare effects, which address several
minor welfare effects, including gas-phase foliar toxicity and the role of S in Hg methylation. Finally,
summary and conclusions are found in Chapter 4. Supplementary Annexes provide additional details.
1-5
-------
1.6. Causality Framework
U.S. EPA uses a two-step approach to evaluate the scientific evidence on welfare effects of criteria
pollutants, similar to the approach it uses for health effects. The steps address two general policy-relevant
questions:
1. Given the total body of evidence, what, if any, are the welfare effects of NOx and SOx?
2. Can levels of exposure at which welfare effects of concern occur be defined?
The first step determines the weight of evidence in support of causation, and characterizes the
strength of any resulting causal classification. The second step includes further evaluation of the
quantitative evidence with respect to concentration-response relationships and the levels, duration, and
pattern of exposures at which effects are observed.
The most widely cited aspects of causality in public health were articulated by Sir Austin Bradford
Hill (1965), and have been widely used (e.g., IARC, 2006; Samet and Bodurow, 2008). Several
adaptations of the Hill aspects have been used in aiding causality judgments in the ecological sciences
(Adams, 2003; Buck et al., 2000; Collier, 2003; Fox, 1991; Gerritsen et al., 1998). Based on these
adaptations, the U.S. EPA uses eight aspects in judging causality (see Table 1-1). The broad national scale
of this assessment differs from the site-specific scale of ecological assessment for which applications of
the Hill aspects have been published. The following aspects were developed to meet the scope of this
ISA:
Table 1-1. Aspects to aid in judging causality.
¦ CONSISTENCY of the observed association. The inference of causality is strengthened when the
same association between agent and effect is observed across similar, independent studies. The
reproducibility of findings constitutes one of the strongest arguments for causality. If there are
discordant results among comparable investigations, possible reasons such as differences in
exposure, confounding factors, and the power of the study are considered.
¦ STRENGTH of the observed association. The finding of large, well-demarcated effects increases
confidence that the association is causal. However, given a truly causal agent, a small magnitude
in the effect could follow from a lower level of exposure, a lower potency, or the prevalence of
other agents causing similar effects. While large effects support causality, modest effects,
therefore, do not preclude it.
¦ SPECIFICITY of the observed association. The effect is only observed after exposure to that agent,
and the agent produces only that effect. Hill (1965), and subsequent authors, consider specificity
a weak aspect. At the scale of ecosystems, as in epidemiology, complexity is such that single
agents causing single effects, and single effects following single causes, are extremely unlikely.
The absence of specificity cannot be used to exclude causality, especially at those scales.
However, if specificity can be demonstrated, as in some laboratory or other experimental studies,
it does add strong support to causality.
¦ TEMPORALITY of the observed association. Evidence of a temporal sequence between the
introduction of an agent and appearance of the effect constitutes another argument in favor of
causality.
¦ GRADIENT. A clear exposure-response relationship (e.g., increasing effects associated with
greater exposure) strongly suggests cause and effect.
1-6
-------
¦ PLAUSIBILITY. A credible ecological basis for the observed association adds strength to an
inference of causality. A proposed mechanistic linking between an effect, and exposure to the
agent, is an important source of support for causality, especially when data establishing the
existence and functioning of those mechanistic links are available. A lack of biological
understanding, however, is not sufficient reason to reject causality.
¦ EXPERIMENTAL evidence. Controlled exposure to the agents provides results that support the
proposed causal relationship. The practical limits on control as the number of potential interacting
factors increases are such that the most compelling experiments can only be conducted at the
scale of a laboratory, growth chamber, or, at most, mesocosm. Therefore, since a judgment of
causality derived from experimental evidence often cannot be extended very far beyond the scale
at which the experiment was conducted, experimental evidence is generally only one element of
the information that comes to bear in determining causality at the ecosystem, regional, or greater
scales.
¦ COHERENCE. Given the scale and complexity of the environment and of ecosystems,
determinations of causality are usually based on many lines of evidence, considered jointly.
Evidence may be drawn from a variety of experimental approaches (e.g., greenhouse, laboratory,
field) and subdisciplines of ecology (e.g., community ecology, biogeochemistry,
paleological/historical reconstructions). The coherence of the available sources is a critical aspect
of assessing the strength of a causal association. The coherence of evidence from various fields,
and at various scales, greatly adds to the strength of an inference of causality.
While these aspects provide a framework for assessing the evidence, they are not simple formulas
or fixed rules of evidence leading to conclusions about causality (Hill, 1965). The aspects in Table 1-1
cannot be used as a strict checklist, but rather to determine the weight of the evidence for inferring
causality. In particular, the absence of one or more of the aspects does not automatically exclude a study
from consideration (e.g., see discussion in CDC, 2004). For example, one cannot simply count the
number of studies reporting statistically significant or nonsignificant results, and reach credible
conclusions about the relative weight of the evidence and the likelihood of causality. Rather, the aspects
are an important part of the assessment, whose goal is to produce an objective appraisal of the evidence,
and is informed by peer and public comment and advice, including weighing of alternative views on
controversial issues.
1.6.1. First Step: Determination of Causality
In this ISA, U.S. EPA evaluated publications available since the previous NAAQS reviews. This
evaluation builds upon evidence available and conclusions drawn in the previous reviews to draw
conclusions on the causal relationships between relevant pollutant exposures and welfare outcomes. A
five-level hierarchy is used to classify the weight of evidence for causation, as assessed by the reviewing
group with input from peers, CASAC, and the public. After integration of the evidence from all relevant
disciplines or types of studies (laboratory studies, ecosystem experiments, simulation models and
observational studies), the weight of evidence in support of causality is expressed using one of the five
descriptors (see Table 1-2). In this multi-pollutant assessment, the effects may be due to a combination of
pollutants (e.g., in acidifying deposition or N deposition). To the extent possible, U.S. EPA will identify
the pollutants that are "significant contributing factors" to the relationship being evaluated.
1-7
-------
Table 1-2.
Weight of evidence for causal determination.
Relationship Description
Causal relationship Evidence is sufficient to conclude that there is a causal relationship between relevant pollutant exposure
and the outcome. Causality is supported when an association has been observed between the pollutant
and the outcome in studies in which chance, bias, and confounding could be ruled out with reasonable
confidence. Controlled exposure (laboratory or small- to medium-scale field studies) provides the
strongest evidence for causality, but the scope of inference may be limited. Generally, determination is
based on multiple studies conducted by multiple research groups, and evidence that is considered
sufficient to infer a causal relationship is usually obtained from the joint consideration of many lines of
evidence that reinforce each other.
Evidence is sufficient to conclude that there is a likely causal association between relevant pollutant
exposures and the outcome. That is, an association has been observed between the pollutant and the
outcome in studies in which chance, bias and confounding are minimized, but uncertainties remain. For
example, field studies show a relationship, but suspected interacting factors cannot be controlled, and
other lines of evidence are limited or inconsistent. Generally, determination is based on multiple studies
in multiple research groups.
Suggestive of a causal
Evidence is suggestive of an association between relevant pollutant exposures and the outcome, but
relationship
chance, bias and confounding cannot be ruled out. For example, at least one high-quality study shows an
association, but the results of other studies are inconsistent.
Inadequate to infer a causal
The available studies are of insufficient quality, consistency or statistical power to permit a conclusion
relationship
regarding the presence or absence of an association between relevant pollutant exposure and the
outcome.
Suggestive of no causal Several adequate studies, examining relationships between relevant exposures and outcomes, are
relationship consistent in failing to show an association between exposure and the outcome at any level of exposure.
Likely to be a causal
relationship
1.6.2. Second Step: Evaluation of Ecological Response
Beyond judgments regarding causality are questions relevant to characterizing exposure
concentration response and risk to ecosystems (e.g., the levels and loads of pollution at which ecological
effects occur). Such questions include:
1. What elements of the ecosystem (e.g., types, regions, taxonomic groups, populations, functions)
appear to be affected, and/or are more susceptible to effects?
2. Under what exposure conditions (amount or concentration, duration and pattern) are effects seen?
3. What is the shape of the concentration-response or exposure-response relationship?
Causal and likely causal claims typically characterize how the probability of ecological effects
changes in response to exposure. The ecological scale at which those quantitative considerations are valid
is a concern. Initially, responses are evaluated within the range of observation, but ecological data for
concentration-response analyses are often not available at the national or even regional scale. They are,
therefore, typically presented site by site. Where greenhouse or animal ecotoxicological studies are
available, they may be used to aid in characterizing concentration-response relations, particularly those
relative to mechanisms of action and characteristics of sensitive biota.
1-8
-------
Chapter 2. Source to Deposition
This chapter provides fundamental and applied atmospheric science data to support assessing the
environmental exposures and effects associated with N and S oxides. More specifically, these data relate
to N and S emissions sources and rates, atmospheric transformation and transport, total atmospheric
loadings, measurement and modeling techniques, and deposition issues relevant to this review of the
NAAQS. These data are prologue for the detailed descriptions of the evidence of environmental effects
from N and S oxides that follow in Chapter 3, and a source of information to help interpret those effects
when integrated with these data on atmospheric concentrations and exposures.
2.1. Introduction
As noted in Chapter 1, the definition of NOx appearing in the NAAQS enabling legislation differs
from the one used by atmospheric scientists and air quality control experts. The atmospheric sciences
community defines NOx as the sum of NO and N02. However, in the Federal Register Notice (FRN)
(October 8, 1996) forthe "National Ambient Air Quality Standards forN02: Final Rule" (61 FR 52852),
the term "nitrogen oxides" was used to "describe the sum of NO, N02, and other oxides of nitrogen." This
ISA uses the legal, rather than the technical definition; hence, the terms "oxides of nitrogen" and
"nitrogen oxides" here refer to all forms of oxidized N compounds, including NO, N02, and all other
oxidized N-containing compounds transformed from NO and N02.1 Additionally, because some of the
constituent members of the NOx family of chemical species interact with particulate-phase chemical
species and change phase themselves, the chemistry, concentrations, and deposition of particulate N
compounds are also considered in this assessment.
Oxides of sulfur (SOx) is defined here to include sulfur monoxide (SO), sulfur dioxide (S02 [the
largest component of SOx and the U.S. EPA Criteria Air Pollutant]), sulfur trioxide (S03), and disulfur
monoxide (S20). Of these, only S02 is present in the lower troposphere at concentrations relevant for
environmental considerations. Moreover, some gas-phase sulfur oxides interact with particles and change
phase themselves, just as do some constituent members of the N family of gas-phase chemical species;
hence, particulate-phase S compounds are also assessed here.
NH3 is included in this ISA both because its oxidation can be a minor source of NOx and because it
is the precursor for ammonium ion (NH/), which plays a key role in neutralizing acidity in ambient
particles produced from N02 and S02 and in cloud, fog, and rain water. (NH3 and NH/ are conventionally
grouped together under the category label NHX.) Excess NH3 is also an actor in nitrification of aqueous
and terrestrial ecosystems, participating alone and together with NOx in the N cascade (Galloway et al.,
2003). Additionally, NH3 is involved in the ternary nucleation of new particles and reacts with gas-phase
HN03 to form ammonium nitrate (NH4N03), a major component of N deposition in many areas of the
contiguous U.S. (CONUS).
1 This follows usage in the Clean Air Act, Section 108(c): "Such criteria [for oxides of nitrogen] shall include a discussion of nitric and nitrous
acids, nitrites, nitrates, nitrosamines, and other carcinogenic and potentially carcinogenic derivatives of oxides of nitrogen." (U.S. Code 1970a)
By contrast, within the air pollution research and control communities, the terms "oxides of nitrogen" and "nitrogen oxides" are restricted to refer
only to the sum of NO and NO2, and this sum is commonly abbreviated as NOx- The category label used by this air pollution research and control
community for the sum of all oxidized N compounds, including those listed in Section 108(c), is NOy-
2-1
-------
2.2. Sources and Emissions of Tropospheric NOx
Tropospheric NOx emissions sources can be anthropogenic, resulting from human activity, or
biogenic and natural, resulting from the activity of non-human organisms, though sometimes with the
addition of human activities, as with production from livestock or agriculture, and from other smaller
miscellaneous non-biological sources. However, anthropogenic sources contribute substantially more
mass than biogenic ones. The anthropogenic and biogenic sources of NOx are described in detail and their
emissions totals are provided below.
2.2.1. Major Anthropogenic Sources
Anthropogenic NOx emissions are dominated by fossil fuel combustion sources that release NOx
predominantly in the form of NO with variable amounts of N02. In 2002, anthropogenic NOx emissions
in the U.S. totaled 23.19 teragram/year (Tg/yr). Table 2-1 lists fractions and totals from anthropogenic
NOx sources collected for the 2002 National Emissions Inventory (NEI) (U.S. EPA, 2006a).
Table 2-1. Emissions of NOx, NH3, and SO2 in the U.S. by source and category, 2002.
2002 Emissions (Tg/yr)
NOx1
NHs
so2
2002 Emissions (Tg/yr)
NOx1
NHs
SO2
Total All Sources
23.19
4.08
16.87
Liquid Waste
0.01
Fuel Combustion Total
9.11
0.02
14.47
Other
0.04
Fuel Combustion Electrical Utilities
5.16
<0.01
11.31
Internal Combustion
1.15
<0.01
0.01
Coal
4.50
<0.01
10.70
Fuel Combustion Other
0.80
<0.01
0.63
Bituminous
2.90
8.04
Commercial / Institutional Coal
0.04
<0.01
0.16
Subbituminous
1.42
2.14
Commercial / Institutional Oil
0.08
<0.01
0.28
Anthracite & Lignite
0.18
0.51
Commercial / Institutional Gas
0.25
<0.01
0.02
Other
<0.01
Misc. Fuel Combust. (Exc. Resident.)
0.03
<0.01
0.01
Oil
0.14
<0.01
0.38
Residential Wood
0.03
<0.01
Residual
0.13
0.36
Residential Other
0.36
0.16
Distillate
0.01
0.01
Distillate Oil
0.06
0.15
Gas
0.30
<0.01
0.01
Bituminous/Subbituminous
0.26
<0.01
Natural
0.29
Other
0.04
<0.01
Process
0.01
Industrial Process Total
1.10
0.21
1.54
Other
0.05
<0.01
0.21
Chemical & Allied Product Mfg
0.12
0.02
0.36
Internal Combustion
0.17
<0.01
0.01
Organic Chemical Mfg
0.02
<0.01
0.01
Fuel Combustion Industrial
3.15
<0.01
2.53
Inorganic Chemical Mfg
0.01
<0.01
0.18
Coal
0.49
<0.01
1.26
Sulfur Compounds
0.17
Bituminous
0.25
0.70
Other
0.02
Subbituminous
0.07
0.10
Polymer & Resin Mfg
<0.01
<0.01
<0.01
Anthracite & Lignite
0.04
0.13
Agricultural Chemical Mfg
0.05
0.02
0.05
Other
0.13
0.33
Ammonium Nitrate/Urea Mfg.
<0.01
Oil
0.19
<0.01
0.59
Other
0.02
Residual
0.09
0.40
Paint, Varnish, Lacquer, Enamel Mfg
0.00
0.00
Distillate
0.09
0.16
Pharmaceutical Mfg
0.00
0.00
Other
0.01
0.02
Other Chemical Mfg
0.03
<0.01
0.12
Gas
1.16
<0.01
0.52
Metals Processing
0.09
<0.01
0.30
Natural
0.92
Non-Ferrous Metals Processing
0.01
<0.01
0.17
Process
0.24
Copper
0.04
Other
<0.01
Lead
0.07
Other
0.16
<0.01
0.15
Zinc
0.01
Wood/Bark Waste
0.11
Other
<0.01
2-2
-------
2002 Emissions (Tg/yr)
NOx1
NHs
so2
2002 Emissions (Tg/yr)
NOx1
NHs
SO2
Ferrous Metals Processing
0.07
<0.01
0.11
Highway Vehicles
8.09
0.32
0.30
Metals Processing
0.01
<0.01
0.02
Light-Duty Gas Vehicles & Motorcycles
2.38
0.20
0.10
Petroleum & Related Industries
0.16
<0.01
0.38
Light-Duty Gas Vehicles
2.36
0.10
Oil & Gas Production
0.07
<0.01
0.11
Motorcycles
0.02
0.00
Natural Gas
0.11
Light-Duty Gas Trucks
1.54
0.10
0.07
Other
0.01
Light-Duty Gas Trucks 1
1.07
0.05
Petrol. Refineries & Related Industries
0.05
<0.01
0.26
Light-Duty Gas Trucks 2
0.47
0.02
Fluid Catalytic Cracking Units
<0.01
0.16
Heavy-Duty Gas Vehicles
0.44
<0.01
0.01
Other
<0.01
0.07
Diesels
3.73
<0.01
0.12
Asphalt Manufacturing
0.04
0.01
Heavy-Duty Diesel Vehicles
3.71
Other Industrial Processes
0.54
0.05
0.46
Light-Duty Diesel Trucks
0.01
Agriculture, Food, & Kindred Products
0.01
<0.01
0.01
Light-Duty Diesel Vehicles
0.01
Textiles, Leather, & Apparel Products
<0.01
<0.01
<0.01
Off-Highway
4.49
<0.01
0.46
Wood, Pulp & Paper, Publish. Prods.
0.09
<0.01
0.10
Non-Road Gasoline
0.23
<0.01
0.01
Rubber & Misc. Plastic Products
<0.01
<0.01
<0.01
Recreational
0.01
Mineral Products
0.42
<0.01
0.33
Construction
0.01
Cement Mfg
0.24
0.19
Industrial
0.01
Glass Mfg
0.01
Lawn & Garden
0.10
Other
0.10
0.09
Farm
0.01
Machinery Products
<0.01
<0.01
<0.01
Light Commercial
0.04
Electronic Equipment
<0.01
<0.01
<0.01
Logging
<0.01
Transportation Equipment
<0.01
<0.01
Airport Service
<0.01
Miscellaneous Industrial Processes
0.01
0.05
0.02
Railway Maintenance
<0.01
Solvent Utilization
0.01
<0.01
<0.01
Recreational Marine Vessels
0.05
Degreasing
<0.01
<0.01
<0.01
Non-Road Diesel
1.76
<0.01
0.22
Graphic Arts
<0.01
<0.01
<0.01
Recreational
0.00
Dry Cleaning
<0.01
<0.01
<0.01
Construction
0.84
Surface Coating
<0.01
<0.01
<0.01
Industrial
0.15
Other Industrial
<0.01
<0.01
<0.01
Lawn & Garden
0.05
Nonindustrial
<0.01
Farm
0.57
Solvent Utilization
<0.01
Light Commercial
0.08
Storage & Transport
<0.01
<0.01
0.01
Logging
0.02
Bulk Terminals & Plants
<0.01
<0.01
<0.01
Airport Service
0.01
Petrol. & Petrol. Product Storage
<0.01
<0.01
<0.01
Railway Maintenance
<0.01
Petrol. & Petrol. Product Transport
<0.01
<0.01
<0.01
Recreational Marine Vessels
0.03
Service Stations: Stage II
<0.01
<0.01
Aircraft
0.09
0.01
Organic Chemical Storage
<0.01
<0.01
<0.01
Marine Vessels
1.11
0.18
Organic Chemical Transport
0.01
<0.01
Diesel
1.11
Inorganic Chemical Storage
<0.01
<0.01
<0.01
Residual Oil
Inorganic Chemical Transport
<0.01
<0.01
Other
Bulk Materials Storage
0.01
<0.01
<0.01
Railroads
0.98
0.05
Waste Disposal & Recycling
0.17
0.14
0.03
Other
0.32
<0.01
0.00
Incineration
0.06
<0.01
0.02
Liquefied Petroleum Gas
0.29
Industrial
Compressed Natural Gas
0.04
Other
<0.01
Miscellaneous
0.39
3.53
0.10
Open Burning
0.10
<0.01
<0.01
Agriculture & Forestry
<0.01
3.45
<0.01
Industrial
<0.01
Agricultural Crops
<0.01
Land Clearing Debris
Agricultural Livestock
2.66
Other
<0.01
Other Combustion
0.08
0.10
Public Operating Treatment Works
<0.01
0.14
<0.01
Health Services
Industrial Waste Water
<0.01
<0.01
<0.01
Cooling Towers
Treatment, Storage, Disposal Facility
<0.01
<0.01
<0.01
Fugitive Dust
Landfills
<0.01
<0.01
<0.01
Other
Industrial
<0.01
Natural Sources
3.10
0.03
Other
<0.01
1 Emissions are expressed in terms of NO2. Note: Subcategory values may not sum to
Other
<0.01
<0.01
<0.01
Transportation Total
12.58
0.32
0.76
2-3
-------
Of this total, emissions from all types of transportation accounted for -56% of NOx, or 12.58 Tg,
with on-road highway vehicles representing the major mobile source component, 8.09 Tg. Roughly one-
half of these on-road emissions have diesel engine sources and one-half have gasoline engine sources.
(Sawyer et al. [2000] reviewed in detail the factors associated with NOx emissions by mobile sources.)
The next largest source category, electric-generating utilities (EGUs), accounted for -22%, or 5.16 Tg of
total NOx in 2002. Stationary engines, non-road vehicles, and industrial facilities also emit NOx, but
because they are fewer in number or burn less fuel, their mass contributions to total NOx are less than
transportation and EGUs.
The values in Table 2-1 are U.S. national averages and so may not reflect differences in the relative
contributions of NOx sources to ambient mass loadings at any particular location; hence, these values are
not likely to be useful predictors of any particular localized environmental exposures to NOxf As a partial
refinement of scale, county-level NOx emissions are depicted in Figure 2-1 1A further refinement appears
in Figure 2-2, where the same 2001 NOx emissions data are plotted as area-normalized intensities in tons
per square mile. This normalized emissions intensity base is also used to show the separate contributions
from EGUs and on-road mobile sources in Figures 2-3 and 2-4, respectively.
2001 County Emissions (1000 Tons per Yeor) of Nitrogen Oxides
0 >0-0.54 0.54-1.2 1.2-2.5
I 2.5-6 H 6-16 16+
Source: U.S. EPA (2006a)
Figure 2-1. 2001 county-level total U.S. NOx (NO and NO2) emissions.1
1 Range values: White, 0 or no reported value; Blue, from the smallest non-zero to the 10th percentile value; Green, from above the 10th to the 25th
percentile; Yellow, from above the 25th to the 50th percentile; Pink, from above the 50th to the 75th percentile; Red, from above the 75th to the 90th
percentile; Brown, from above the 90th percentile to the highest reported value.
2-4
-------
2001 County Emissions Density (Tons per sq.mi.) of Nitrogen Oxides
0 >0-0.79 0.79-1.8 1.B-4.1
4.1-10 IB 10-30 H 30+
Source US EPA Office of fik and Rodtolkm. N£l DctobOM Thursday, July 10. 2008
Source: U.S. EPA (2006a)
Figure 2-2. 2001 county-level total U.S. N0X (NO and N02) emissions densities (tons per square mile).1
2001 County Emissions Density (Tons per sq.mi.) of Nitrogen Oxides
O >0-0.0022 __ 0 0022-0 02S 0.025-0.3
0 3-5 4 5.4-25 H 25+
Source: U.S. EPA (2006a)
Figure 2-3. 2001 county-level total U.S. NOx (NO and NO2) emissions densities (tons per square mile) from
electric-generating utilities (EGUs).1
1 Range values: White, 0 or no reported value; Blue, from the smallest non-zero to the 10th percentile value; Green, from above the 10th to the 25th
percentile; Yellow, from above the 25th to the 50th percentile; Pink, from above the 50th to the 75th percentile; Red, from above the 75th to the 90th
percentile; Brown, from above the 90th percentile to the highest reported value.
2-5
-------
2001 County Emissions Density (Tons per sq.mi.) of Nitrogen Oxides
0 >0-0.2 0.2-D.65 0.6S-1.7
1.7-4.1 ¦ 4.1 — 10 ¦! 10+
Source: U.S. EPA (2006a)
Figure 2-4. 2001 county-level total U.S. NOx (NO and NO2) emissions densities (tons per square mile) from
on-road mobile sources.1
Emissions of NOx from combustion are derived from both fuel N and atmospheric N. Combustion-
zone temperatures >-1300 K are required to fix atmospheric N2 by the reaction.
N2+02 >2 NO
Reaction 1
Below this temperature, NO can be formed from fuel N by the reaction
CaHbOcNd + 02 -> xCOt + yH20 + zNO
Reaction 2
Both Reaction 1 and Reaction 2 have temperature dependencies and van with concentrations of hydroxyl
radical (OH), hydroperoxy radical (H02), and 02.
The N content in fossil fuels and its specific chemical form vary strongly with source type, fuel,
engine emissions controls, and running conditions. N content in fuel stocks ranges from 0.05% by weight
(wt %) in light distillates such as diesel fiiel (Samet and Bodurow, 2008) to 1.5 wt % in heavy fuel oils,
and from 0.5 to 2.0 wt % in coal, as surveyed by the United Kingdom (U.K. AQEG, 2004) Air Quality
Expert Group.
On-road mobile source emissions constitute the largest type of emissions from all transportation
sources. Significant variability attaches to these emissions. For example, the ratio ofN02 to total NOx in
1 Range values: White, 0 or no reported value; Blue, from the smallest non-zero to the 10th percentile value; Green, from above the 10th to the 25th
percentile; Yellow, from above the 25th to the 50th percentile; Pink, from above the 50th to the 75th percentile; Red, from above the 75th to the 90th
percentile; Brown, from above the 90th percentile to the highest reported value.
2-6
-------
exhaust gases in primary emissions ranges from 1 to 3% from gasoline engines tested on dynamometers
(Heeb et al., 2008; Hilliard and Wheeler, 1979). On the other hand, some European studies have reported
N02-to-NOx ratios > 15% from gasoline vehicles based on integrated measurements from Tedlar bags
(Lenner, 1987; Soltic and Weilenmann, 2003). However, subsequent studies suggesting that NO-to-N02
conversion will occur within a bag sample of diluted exhaust if not properly handled have led groups
performing these measurements to revise their measurement techniques to avoid use of Tedlar bag
samples (Alvarez et al., 2008). As a result, dynamometer-based measurements generally indicate that in
the absence of post-tailpipe transformation, N02 comprises, at most, only a few percent of the total NOx
in current-generation gasoline engine exhaust.
The emissions ratio of N02to NOx ranges between 5 and 12% from heavy-duty diesel truck
engines, although some emission control devices used for diesel engines in Europe increase the fraction of
exhaust NOx emitted as N02 to >20% (Carslaw and Beevers, 2005; Carslaw, 2005; Carslaw and Carslaw,
2007; Kessler et al., 2006). In the U.S., on-road experiments with diesel engines propelling heavy buses
in congested urban areas like New York City have shown that engines equipped with emissions control
devices similar to those in the European studies increased the N02-to-NOx ratio from -10% before
addition of the new controls to -30% after controls were added (Shorter et al., 2005). In a second type of
experiment in a different setting, Kittelson et al. (2006) used an on-road laboratory to sample exhaust
plumes of a truck equipped with the European-style emissions control device under highway cruise
conditions and found the N02-to-NOx ratios for this exhaust under highway cruise conditions ranged
from 59 to 70%. The wide range revealed by comparing these two studies illustrates the significant
differences in NOx exhaust under different conditions of engine load and ambient temperature.
As for other combustion sources, N02-to-NOx emissions ratios for compressed natural gas engines
range between 5 and 10%, and between 5 and 10% from most stationary sources. The N02-to-NOx ratios
in emissions from turbine jet engines are as high as 35% during taxi and takeoff (U.S. EPA, 2006a).
In addition to NO and N02, mobile sources emit other forms of oxidized N including nitrous acid
(HN02); measured ratios of HN02 to NOx range from a low of 0.3% in the Caldecott Tunnel, San
Francisco, CA (Kirchstetter and Harley, 1996), up to as much as 0.5 and 1.0% in studies in the U.K.
(U.K. AQEG, 2004).
Marine transport represents an additional source of NOx in the U.S., especially for coastal cities
with large ports, but constitutes a larger source in Europe where it is expected to represent more than 60%
of land-based NOx sources (U.K. AQEG, 2004).
The anthropogenic sources of NOx are distributed with height such that some, like on-road mobile
sources, are nearer to ground level than others, like the emissions stacks from EGUs and some industrial
emitters. Emissions height is an important consideration because the prevailing winds aloft are generally
stronger than those at the surface. The result is that emissions from elevated sources can be distributed
over a wider area than those emitted at the surface and hence can be diluted to lower mixing ratios than
those emitted nearer their sources.
2.2.2. Major Biogenic Sources
2.2.2.1. Soils
Nitrification and denitrification processes in soils produce two gas-phase intermediates, NO and
N20, which can evolve from soil microbes before reaching their reaction endpoint, N2. N20 is not among
the nitrogen oxides important for urban and regional air quality either for human health concerns or
environmental effects because its reaction potential on these spatio-temporal scales in the troposphere is
insignificant. As a result, NO from soil metabolism is the prime, but not exclusive, form of atmospheric
NOx from the biosphere relevant to this ISA.
2-7
-------
Biogenic NOx emissions are predominately the result of incomplete bacterial denitrification and
nitrification processes, as described above. Denitrification is a reduction process performed by particular
groups of heterotrophic bacteria having the ability to use nitrate ion (N03 ) as an electron acceptor during
anaerobic respiration, thereby converting N03 in soils and water to gas-phase forms (Firestone and
Davidson, 1989). At low 02 concentrations, these microbial communities may use N03 . nitrite (N02 ), or
N20 as alternative electron acceptors to 02 (Davidson and Schimel, 1995).
The basic outlines of these reaction pathways are known, but uncertainty remains concerning the
conditions favoring production of the various products of the N03 transformations. Groups of aerobic
bacteria use most NH44" in soils as an energy source, oxidizing it to N02 and then N03 . Oxidized N
products from nitrification may undergo denitrification and thus also drive production of NOx. Some
bacteria are known to be nitrifiers and denitrifiers and can change depending on environmental
conditions, including high loadings of exogenous N.
Soil emissions of NOx can be increased by agricultural practices and activities, including the use of
synthetic and organic fertilizers, production of N-fixing crops, cultivation of soils with high organic
content, and the application of livestock manure to croplands and pasture. All of these practices directly
add exogenous N to soils, of which a portion will then be converted to NO or N20 on the pathway to full
conversion to N2. Additionally, indirect additions of N to soils can also result in NOx emissions from
agricultural and non-agricultural systems. Indirect additions include processes by which atmospheric NOx
is deposited directly to a region or N from applied fertilizer or manure volatilizes to NH3 and is oxidized
to NOx and then is ultimately re-deposited onto soils as NH4NO3, HN03, or NOx (U.S. EPA, 2006c).
N metabolism in soils is strongly dependent on soil substrate concentrations and physical
conditions. Where N is limiting, it is efficiently retained and little gas-phase N is released; where N is in
excess of demand, N emissions increase. As a consequence, soil NO emissions are highest from fertilized
agricultural lands and tropical soils (Davidson and Kingerlee, 1997; Williams et al., 1992). In addition,
temperature, soil moisture, and 02 concentrations control both the rates of reaction and the partitioning
between NO and N20. In flooded soils where 02 concentrations are low, N20 is the dominant soil N gas;
as soils dry, more 02 diffuses in and NO emissions increase. In very dry soils, microbial activity is
inhibited and emissions of both N20 and NO decrease.
Emission rates of NO from cultivated soils depend largely on fertilization levels and soil
temperature. Production of NO from agriculture results from the oxidation of NH3 emitted both by
livestock and by soils after fertilization with NH4NO3. Estimates of biogenic N emissions are far less
certain than those of anthropogenic emissions sources. Uncertainty on the order of a factor of 3 or more is
introduced by the variation within biomes to which fertilizer is applied, such as between shortgrass and
tallgrass prairie for example (Davidson and Kingerlee, 1997; Williams et al., 1992; Yienger and Levy,
1995). The contribution of soil emissions to the global NOx budget is approximately 10% (Finlayson-Pitts
and Pitts, 2000; Seinfeld and Pandis, 1998; Van Aardenne et al., 2001), but NOx emissions from fertilized
fields are highly variable. Soil NO emissions can be estimated from the fraction of the applied fertilizer N
emitted as NOx, for example, but the flux depends strongly on land use type and temperature. Estimates
of globally averaged fractional-applied N lost as NO vary from a low of 0.3% (Skiba et al., 1997) up to
2.5% (Yienger and Levy, 1995).
The spatial scales of these N fluxes are also significant. Local contributions to soil NOx can be
much greater than the global average, particularly in summer, and especially where corn is grown
extensively. Approximately 60% of total NOx emitted by soils in the U.S. occurs in the central corn belt.
Nitrification of fertilizer NH3 to N03 in aerobic soils appears to be the dominant pathway to soil NOx
emissions, but the mass and chemical form of N applied to soils, the vegetative cover, the temperature and
soil moisture characteristics, and the agricultural practices such as tillage all influence the amount of
fertilizer N converted and released as NOx. On sub-national scales these emissions can be large and
highly variable. Williams et al. (1992) estimated that NOx from soils in Illinois was -25% of the total
NOx emissions from industrial and commercial processes in that state. In Iowa, Kansas, Minnesota,
Nebraska, and South Dakota—states with smaller human populations than Illinois—soil emissions may,
in fact, dominate the NOx budget.
2-8
-------
Emissions of NOx from soils often peak in summer when ozone (03) formation is also at a
maximum. The significance of agricultural emission sources of NO and NH3 among other air pollutants
was described in detail in a recent National Research Council report (NRC, 2002). That report
recommended immediate implementation of best management practices to control these emissions, and
called for additional research to quantify the magnitude of emissions and the effects of agriculture on air
quality. The effects of such changes in management practice can be dramatic: Civerolo and Dickerson
(1998) reported that the use of no-till cultivation techniques on a fertilized cornfield in Maryland reduced
NO emissions by a factor of 7.
2.2.2.2. Live Vegetation
Extensive work on N inputs from the atmosphere to forests was conducted in the 1980s as part of
the Integrated Forest Study, summarized by Johnson and Lindberg (1992b). As noted below and in
Chapter 4, our understanding of N02 exchange with vegetation suggests that N02 should be emitted from
foliage when ambient concentrations are below the compensation point of ~1 ppb. However, Lerdau et al.
(2000) noted that current understanding of the global distribution of NOx is not consistent with the large
source that would be expected in remote forests if N02 emissions were significant when atmospheric
concentrations were below the 1 ppb compensation point.
2.2.2.3. Biomass Burning
During biomass burning, N is derived mainly from fuel N and not from atmospheric N2, since
temperatures required to fix atmospheric N2 are likely to be found only in the flaming crowns of the most
intense boreal forest fires. N is present in plants mostly as amine (NH2) groups in amino acids. During
combustion, N is released in many forms, mostly unidentified and presumably as N2, leaving very little N
remaining in the fuel ash. Emissions of NOx are estimated to be -0.2 to 0.3% of the total biomass burned
(e.g., Andreae, 1991; Radke et al., 1991). The most abundant NOx species in biomass burning plumes is
NO, emissions of which account for -10 to 20% of the total fuel N loadings (Lobert et al., 1991); other
N-containing species such as N02, nitriles, and NH3 together account for a similar amount. Westerling
et al. (2006) noted that the frequency and intensity of wildfires in the western U.S. increased substantially
since 1970, lending added importance to consideration of all NOx emissions from this sector.
2.2.2.4. Lightning
Annual global production of NO by lightning is the most uncertain source of atmospheric N. In the
last decade, literature values of the global average production rate ranged from 2 to 20 Tg N/yr. Most
recent estimates, however, are in the range of 3 to 8 Tg N/yr. This large and persistent uncertainty stems
from several factors: a wide range of as much as two orders of magnitude in NO production rates per
meter of flash length; uncertainty over whether cloud-to-ground (CG) and intracloud (IC) flashes produce
substantially different NO levels; the global average flash rate; and the ratio of IC to CG flashes.
Estimates of the NO concentration produced per flash have been made from theoretical
considerations (e.g., Price et al., 1997), laboratory experiments (e.g., Wang et al., 1998), and field
experiments (Huntrieser et al., 2002, 2007; Stith et al., 1999), and with a hybrid method of cloud-
resolving model simulations, observed lightning flash rates, and measurements of NO concentrations in
cloud anvils (DeCaria et al., 2000, 2005; Ott et al., 2007). A series of midlatitude and subtropical
thunderstorm events were simulated with the model of DeCaria et al. (2005) and the derived NO
production per CG flash was, on average, 500 moles/flash, while production per IC flash was
425 moles/flash on average (Ott et al., 2007). The hybrid method had earlier been used by Pickering et al.
(1998) who showed that only -5 to 20% of the total NO produced by lightning in a given storm exists in
2-9
-------
the planetary boundary layer (PBL) at the end of a thunderstorm event, thereby reducing its importance as
a direct emissions source to the urban and regional troposphere.
2.2.3. Anthropogenic and Biogenic Sources of N2O
N20 has an atmospheric lifetime (x) of-114 years, resulting from its having effectively no
chemistry in the lower troposphere on urban and regional scales. The chief N20 loss pathway with a
quantum yield of ~1 is the photodissociation process
N20 > N2 + OCD)
Reaction 3
driven by the short wavelength UV present only in the stratosphere.
However, N20 is also a GHG with a global warming potential (GWP) on the conventional 100-
year time horizon of -296; i.e., 1 molecule of N20 is nearly 300 times more effective at trapping heat in
the atmosphere than 1 molecule of carbon dioxide (C02) over a 100-year period (IPCC, 2001b). The high
GWP of N20 results from its combination of direct and indirect radiative forcing climate effects in the
stratosphere. By comparison, the primary climate effects of NO and N02 are indirect and result from their
role in promoting the production of 03 (P(03)) in the troposphere and, to a lesser degree, in the lower
stratosphere where NOx has positive radiative forcing effects. Additional complications for calculating
NOx GWPs ensue owing to the fact that NOx emissions from high-altitude aircraft are also likely to
decrease methane (CH4) concentrations, a negative radiative forcing effect (IPCC, 1996), and that
particulate nitrate (pN03 ) transformed from NOx also has negative radiative forcing effects. U.S. EPA
does not calculate GWPs for total NOx or for SOx or for the other atmospheric constituents for which no
agreed-upon method exists to estimate the contributions from these gases that are short-lived in the
atmosphere, have strong spatial variability, or have only indirect effects on radiative forcing.
Thus, because there are no tropospheric reactions or effects to consider, N20 is not a significant
component of NOx for this ISA review of the NOx and SOx secondary effects related to the NAAQS.
However, the role of N20 as an intermediate product along with NO from the complex soil metabolism
described in Section 2.2.2.1 means that a brief description of its emissions strengths and its component
part of the total budget of U.S. GHGs will be useful, and so appears just below.
N20 is a contributor to the total U.S. GHG budget, with 6.5% of total GHG on a Tg C02
equivalents basis (C02e) in 2005 (U.S. EPA, 2007b). C02, by comparison, accounted for 83.9% in the
same year, and CH4 for 7.4% (U.S. EPA, 2007b). Although atmospheric concentrations of N20 have
increased globally by -18% to a current value of -315 ppb due to western industrialization since the year
1750 C.E. (Hofmann et al., 2004), there is considerable interannual variation in N20 emissions which
remains largely unexplained (IPCC, 2001a). N20 emissions in the U.S., for example, decreased by 2.8%,
or 13.4 Tg C02e, between 1990 and 2005 (U.S. EPA, 2007a).
N20 is produced by biological processes occurring in the soil and water, as described in
Section 2.2.2 above, and by a variety of anthropogenic activities in the agricultural, energy, industrial, and
waste management sectors. The chief anthropogenic activities producing N20 in the U.S. are agricultural
soil management, fuel combustion in motor vehicles, manure management, production of adipic acid
(nylon) and HN03, wastewater treatment, and stationary fuel combustion.
N20 emissions from anthropogenic activities in the U.S. were 386.7 Tg C02e/yr between 1990 and
2004 (U.S. EPA, 2007a). These emissions resulted from the fuel combustion, industrial practices, and
stimulation of biogenic sources through agricultural practices listed above. In 2005, N20 emissions from
mobile sources were 38.0 Tg C02e, or -8% of the U.S. N20 emissions total (U.S. EPA, 2007a). In the
period between 1990 and 1998, control technologies on mobile sources reduced on-road vehicle NO and
N02 emissions at the expense of increasing N20 emissions by 10%. The overall reduction in N20 mobile
2-10
-------
source emissions between 1998 and 2005 (when totals were last available), however, has been 13% owing
to more efficient controls used after 1998.
Biogenic production of N20 stimulated through soil management accounted for >75% of all U.S.
N20 emissions in 2005 (U.S. EPA, 2007a). N20 emissions from these sources have shown no significant
long-term trend because the biogenic emitters are highly sensitive to the concentrations and forms of N
applied to soils, and these applications have been largely constant (U.S. EPA, 2007a).
Aquatic sources of N20 may also be stimulated by environmental conditions. In some ocean areas,
large areas of surface water can become depleted in 02, allowing active denitrification in open water, and
potentially increasing N20 emissions as described in Section 2.2.2. In addition, oceanic N20 can also
arise from denitrification in marine sediments, particularly in nutrient-rich areas like estuaries.
2.3. Sources and Emissions of Tropospheric SOx
Emissions of S02, the chief component of SOx, are due mostly to combustion of fossil fuels by
EGUs and industrial processes, with transportation-related sources making smaller but significant
contributions.
2.3.1. Major Anthropogenic Sources
Table 2-1 shows that for 2002, fossil fuel combustion at EGUs accounted for -66% of total S02
emissions in the U.S., or 11.31 Tg of the total 16.87 Tg. All transportation sources accounted for -5% of
the total U.S. S02 emissions in 2002, or 0.76 Tg. On-road vehicles produced -40% of the transportation-
related total S02 emissions in 2002, with off-road diesel and marine traffic together accounting for the
remainder. Thus, most S02 emissions originate from point sources having well-known locations and
identifiable fuel streams.
Since nearly all S in fuels is released in volatile components, either S02 or S03, during combustion,
total S emissions from these point sources can be computed from the known S content in fuel stocks with
greater accuracy than can total NOx emissions from point sources. However, just as for the NOx
emissions totals described above, total SOx emissions estimates are national-scale averages and so cannot
accurately reflect the contribution of local sources to selected environmental exposures to SOx at specific
locations and times. To refine those national estimates, county-level average S02 emissions for 2001 are
shown in Figure 2-5; and normalized emissions intensities per square mile like those shown above for
NOx are shown for S02 in Figure 2-6.
Figure 2-6 illustrates the west-to-east increasing gradient in S02 emissions densities, with most
counties east of the Mississippi River in warmer colors (greater emissions densities) than most counties in
the West. The upper end of the S02 emissions density distribution represented here includes many
counties in the eastern U.S.—primarily in the Ohio River Valley—with 2001 S02 emissions densities
significantly greater than 20. Examples of these high densities (in tons per square mile) are Hillsborough
County, FL, 80; Grant County, WV, 156; Indiana County, PA, 190; Washington County, OH, 273; and
Armstrong County, PA, 292. In these counties, S02 emissions were due mostly to EGU fuel combustion,
as shown in Table 2-2. For the non-EGU emissions densities and the total S02 densities in Figure 2-6, the
upper end of the density distribution compresses a wide range; see Table 2-2. Thus, for the five counties
considered above, non-EGU emissions were <5% of total S02 emissions in Washington County, OH,
and <1% in Indiana County, PA, Armstrong County, PA, and Grant County, WV. Hillsborough County,
FL, is an exception, where 17% of the 2001 S02 emissions density came from non-EGU sources, the
largest of which was chemical and allied product manufacturing.
2-11
-------
2001 County Emissions (1000 Tons per Year) of Sulfur Dioxide
0 >0-0.05 ¦ 0.05-0.11 0.11-0.31
0.31-1.5 1.5-12 12+
Figure 2-5. 2001 county-level total U.S. S02 emissions.1
Source: U.S. EPA (2006a)
2001 County Emissions Density (Tons per sq.mi.) of Sulfur Dioxide
0 __ >0-0.069 0.069-0.17 0.17-0.54
0.54-2.7 ¦ 2.7-20 20+
Source: U.S. EPA (2006a)
Figure 2-6. 2001 county-level total U.S. SO2 emissions densities (tons per square mile).1
1 Range values: White, 0 or no reported value; Blue, from the smallest non-zero to the 10th percentile value; Green, from above the 10th to the 25th
percentile; Yellow, from above the 25th to the 50th percentile; Pink, from above the 50th to the 75th percentile; Red, from above the 75th to the 90th
percentile; Brown, from above the 90th percentile to the highest reported value.
2-12
-------
2001 County Emissions Density (Tons per sq.rni.) of Sulfur Dioxide
0 >0-0.000098 0.000098-0.0011 O.OOIt-0.045
0.045-13 ¦ 13-57 57+
Source: U.S. EPA (2006a)
Figure 2-7. 2001 county-level SO2 emissions densities (tons per square mile) from EGLIs.1
Although on-road mobile sources in 2001 contributed <5% to S02 emissions totals on the national
scale, their fraction of county-level emissions densities varies widely. Generally, however, on-road mobile
source S02 emissions reflect the west-to-east increasing gradient m the densities of both total S02
emissions and U.S. population, as shown in Figure 2-8. In areas such as Wayne County, MI, and Bronx
County, NY, for example, 2001 S02 emissions densities from on-road mobile sources were 3 and 8.8 tons
per square mile out of totals of 98 and 160 tons per square mile, total S02, respectively. In other areas like
Dallas County, TX, and DeKalb County, GA, however, the on-road fraction of total S02 emissions
densities in 2001 was substantially greater: 1.5 out of the total 4 .1 tons per square mile in Dallas County,
and 3.5 out of the total 6.5 tons per square mile in DeKalb County.
Table 2-2. Total and non-EGU SO2 emissions densities for selected U.S. counties, 2001.
County
SO2 Emissions Density (tons/mile2)
Non-EGU Emissions Density Fraction (%)
Hillsborough, FL
80
17
Grant, WV
156
<1
Indiana, PA
190
<1
Washington, OH
273
<5
Armstrong, PA 292 <1
'Range values: White. 0 or no reported value; Blue, from the smallest non-zero to the 10th percentile value; Green, from above the 10th to the 25th
percentile; Yellow, from above the 25th to the 50th percentile; Pink, from above the 50th to the 75th percentile; Red, from above the 75th to the 90th
percentile; Brown, from above the 90th percentile to the highest reported value.
2-13
-------
2Q01 County Emissions Density {Tons per sq.mi.) of Sulfur Dioxide
0 >0-0.0065 0.0065-0.022 0.022-0.05?
0,057-0.14 Hi 0,14-0,32 IH 0,32+
Source: U.S. EPA (2006a)
Figure 2-8. 2001 county-level SO2 emissions densities (tons per square mile) from on-road mobile
sources
2001 County Emissions Density (Tons per sq.mi.) of Sulfur Dioxide
0 >0-0.016 0.016-0.041 0.041 -0.087
0.007-0.18 IH 0.18-0.5 ¦ 0.5+
Source: U.S. EPA (2006a)
Figure 2-9. 2001 county-level S02 emissions densities (tons per square mile) from off-road mobile and
other transportation sources.1
'Range values: White, 0 or no reported value; Blue, from the smallest non-zero to the 10th percentile value; Green, from above the 10th to the 25th
percentile; Yellow, from above the 25th to the 50th percentile; Pink, from above the 50th to the 75th percentile; Red, from above the 75th to the 90th
percentile; Brown, from above the 90th percentile to the highest reported value.
2-14
-------
An additional source of S02 emissions of concern in particular locations not immediately obvious
from national-scale averages and totals are transit and in-port activities in areas with substantial shipping
traffic (Wang et al., 2007). Because of the importance of these S02 emissions, the ports of Long Beach
and Los Angeles, CA, for example, are part of a Sulfur Emissions Control Area in which S contents of
fuels are not to exceed 1.5%. Figure 2-9 shows S02 emissions densities combined for all non-road
transportation-related emitters in which coastal areas with ports and shipping routes, such as the
Mississippi River, are easily discerned. In Los Angeles County, CA, for example, off-road transportation
including shipping and port traffic contributed 1.4 of the total 4.1 tons of S02 per square mile in 2001; in
King County (including the city of Seattle), WA, the off-road transportation fraction was 42% of the total
S02 emissions density, or 1.2 of the total 2.8 tons per square mile. Emissions density data at finer scales
more specific to the ports are not available in the routine emissions inventories and some confusion
attends estimates of the actual S02 loads from these sources. Modeling studies by Vutukuru and Dabdub
(2008) for southern California ports, for example, have shown that ships contribute <2 ppb to the 24-h
average S02 concentration in Long Beach, CA, in 2002 and <0.5 ppb farther inland.
S02 data collected from the State and Local Air Monitoring Stations (SLAMS) and National Air
Monitoring Stations (NAMS) networks show that the decline in S02 emissions following controls placed
on electric generating utilities in the previous 15 years has improved air quality. There has not been a
single monitored exceedance of the S02 annual ambient air quality standard in the U.S. since 2000,
(U.S. EPA, 2006e). U.S. EPA trends data (www.epa.gov/airtrends) reveal that the national composite
average S02 annual mean ambient concentration decreased by -48% from 1990 to 2005, with the largest
single-year reduction coming in 1994-1995, the Acid Rain Program's (ARP) first operating year (U.S.
EPA, 2006e). Figure 2-10 depicts data for S02 emissions in the CONUS m these years that reflect this
reduction using individual state-level totals. Note that SOx emissions have changed over this period both
temporally and spatially, with some areas in the southeast U.S. such as North Carolina and Georgia
realizing increased SOx emissions since 2000. SOx emissions from the largest emitters in the states of the
Ohio River Valley, however, have mostly decreased, in some cases by very large fractions.
BO:
Hi S02 Emissions in 1990
H S02 Emissions in 1995
I I SO2 Emissions in 2000
Hi SO2 Emissions in 2005
Scale: Largest bar equals
2.2 million tons of S02
emissions in Ohio, 1990
Figure 2-10. State-level SO2 emissions, 1990-2005.
2-15
-------
These trends in emissions data are consistent with the trends in the observed ambient
concentrations from the Clean Air Status and Trends Network (CASTNet). Following implementation of
the Phase I controls on ARP sources between 1995 and 2000, significant reductions in S02 concentrations
and ambient S042 concentrations were observed at CASTNet sites throughout the eastern U.S.
2.3.2. Major Biogenic Sources
Emissions of SOx from natural sources are small compared to industrial emissions within the U.S.
(see Table 2-1). However, important exceptions occur locally as the result of volcanic activity, wildfires,
and in certain coastal zones as described above.
The major biogenic sources of S02 are volcanoes, biomass burning, wildfires, and dimethylsulfide
(DMS) oxidation over the oceans. Although S02 constitutes a relatively minor fraction of 0.005% by
volume of total volcanic emissions (Holland, 1978), concentrations in volcanic plumes can range from
several to tens of ppm. The ratio of hydrogen sulfide (H2S) to S02 is highly variable in volcanic gases,
typically <1, as in the Mount St. Helens eruption in the Washington Cascade Range (46°20'N, 122°18'W,
summit 2549 m asl) (Turco et al., 1983). However, in addition to being degassed from magma, H2S can be
produced if ground waters, especially those containing organic matter, come into contact with volcanic
gases. In this case, the ratio of H2S to S02 can be >1. H2S produced this way would more likely be
emitted through side vents than through eruption columns (Pinto et al., 1989). Primary particulate sulfate
(pS04) is a component of marine aerosol and is also produced by wind erosion of surface soils.
Since 1980, the Mount St. Helens volcano has been a variable source of S02. Its major effects came
in the explosive eruptions of 1980, which primarily affected the northern part of the mountainous western
half of the U.S. The Augustine volcano near the mouth of the Cook Inlet in southwestern Alaska
(59°36'N, 153°43'W, summit 1252 m asl) has had variable S02 emissions since its last major eruptions in
1986. Volcanoes in the Kamchatka peninsula of the eastern region of Siberian Russia do not significantly
affect surface S02 concentrations in northwestern North America. The most serious effects from volcanic
S02 in the U.S. occur on the island of Hawaii. Nearly continuous venting of S02 from Mauna Loa and
Kilauea produces S02 in such large amounts that >100 km downwind of the island, levels of S02 can
exceed 30 ppb (Thornton and Bandy, 1993).
Emissions of S02 from burning vegetation are generally in the range of 1 to 2% of the biomass
burned (see e.g., Levine et al., 1999). S is bound in amino acids in vegetation, and -50% of this
organically-bound S is released during combustion, leaving the remainder in the ash (Delmas, 1982).
Gas-phase emissions are mainly in the form of S02, with much smaller amounts of H2S and carbonyl
sulfide (OCS). The ratio of reduced S species such as H2S to more oxidized forms such as S02, increases
as the fire conditions change from flaming to smoldering phases of combustion because emissions of
reduced species are favored by lower temperatures and decreased 02 availability.
S02 is also produced by the photochemical oxidation of reduced S compounds such as DMS, H2S,
carbon disulfide (CS2), OCS, methyl mercaptan (CH3-S-H), and dimethyl disulfide (CH3-S-S-CH3). The
sources for these compounds are mainly biogenic (see Table 2-1). Emissions of reduced S species are
associated typically with marine organisms living either in pelagic or coastal zones and with anaerobic
bacteria in marshes and estuaries. Emissions of DMS from marine plankton represent the largest single
source of reduced S species to the atmosphere (Berresheim et al., 1995). Other sources such as wetlands
and terrestrial plants and soils account for <5% of the DMS global flux, with most of this coming from
wetlands.
Other than OCS, which is lost mainly by photolysis with a t of ~6 months, SOx species are lost
mainly by reaction with OH and N03 and are relatively short-lived, with x ranging from a few hours to a
few days. Reaction with N03 at night most likely represents the major loss process for DMS and methyl
mercaptan. Although the mechanisms for the oxidation of DMS are not known with certainty, excess
S042 in marine aerosol appears related mainly to production of S02 from the oxidation of DMS. Because
2-16
-------
OCS is relatively long-lived, it can survive oxidation in the troposphere and be transported upward into
the stratosphere. Crutzen (1976) proposed that its oxidation to S042 in the stratosphere serves as the
major source of the stratospheric aerosol layer. However, Myhre et al. (2004) proposed that S02
transported upward from the troposphere by deep convection is the most likely source, since the flux of
OCS is too small to account for current atmospheric loadings. In addition, in situ measurements of the
isotopic composition of S in stratospheric S042 do not match those of OCS (Leung et al., 2002). Thus, in
addition to biogenic OCS, anthropogenic S02 emissions could be important precursors to the formation of
the stratospheric aerosol layer.
The coastal and wetland sources of DMS have a dormant period in the fall and winter from plant
senescence. Marshes die back in fall and winter, so DMS emissions from them are lower, and lower light
levels in winter at mid-to-high latitudes lessen phytoplankton growth also tend to lower DMS emissions.
Western coasts at mid-to-high latitudes have lower actinic flux to drive photochemical production and
oxidation of DMS. Freezing at mid and high latitudes affects the release of biogenic S gases, particularly
in the nutrient-rich regions around Alaska. Transport of S02 from regions of biomass burning seems to be
limited by heterogeneous losses that accompany convective processes that ventilate the surface layer and
the lower boundary layer (Thornton et al., 1996).
Reduced S species are also produced by several anthropogenic industrial sources: DMS is used in
petroleum refining; and in petrochemical production processes to control the formation of coke and CO;
to control dusting in steel mills; in a range of organic syntheses; as a food flavoring component; and can
also be oxidized by natural or artificial means to dimethyl sulfoxide, a widely-used industrial solvent.
2.4. NHx Emissions
NH3 can be emitted from or deposited to soils, water, or vegetation depending on the ratio of the
atmospheric NH3 concentration to the compensation point of the underlying surface. The compensation
point, x, generally is governed by the form, concentration, and acidity of N at the surface of exchange,
and hence changes over time as these variables change. For most of the year, large areas of the U.S. are
very near the nominal % of 1 (ig/m3, with the result that the NH3 air-surface flux is very often highly
dynamic. Figure 2-11 and Figure 2-12 show county-level annual total NH3 emissions for 2001 in tons,
and the spatially normalized county-level emissions in tons per square mile, respectively.
Total emissions of NH3 on a national scale show a strikingly different pattern from those of NOx or
S02 as comparison of these figures to their NOx and SOx analogs above illustrates. Anthropogenic NH3
emissions from mobile sources are small since the three-way catalysts used in motor vehicles emit only
small amounts of NH3 as a byproduct during the reduction of NOx; in 2002, this totaled -8% of the
national NH3 total of -3.7 metric short tons. Stationary combustion sources including EGUs make even
smaller contributions to emissions of NH3 because their efficient combustion favors NOx formation and
NH3 is produced during combustion largely by inefficient, low-temperature burning. In 2002, the total
from all stationary source fuel combustion processes amounted to <2% of total NH3 emissions and
chemical production added only -0.7% more. Hence, NH3 emissions totals are dominated by biogenic
production from agriculture, chiefly from livestock management and fertilizer applications to soils. In
2002, these sources accounted for -86% of U.S. total emissions.
As with NOx and SOx emissions, however, these national-scale emissions totals obscure important
variability at finer scales. To illustrate this point, Figures 2-13 through 2-15 show county-level NH3
emissions densities separately for emissions from on-road mobile sources, EGUs, and miscellaneous and
biogenics, respectively.
2-17
-------
2001 County Emissions (Tons per Year) of Ammonia
>0-120
1300-2400
120-320
2400+
Source: U.S. EPA (2006a)
Figure 2-11. 2001 county-level total U.S. NH3 emissions.
2001 County Emissions Density (Tons per sq.mi.) of Ai
mmonia
>0-0.18
2.1-3.7
0.18-0.46
3.7 +
Source: U.S. EPA (2006a)
Figure 2-12. 2001 county-level total U.S. NH3 emissions densities.
1 Range values: White, 0 or no reported value; Blue, from the smallest non-zero to the 10th percentile value; Green, from above the 10th to the 25th
percentile; Yellow, from above the 25th to the 50th percentile; Pink, from above the 50th to the 75th percentile; Red, from above the 75th to the 90th
percentile; Brown, from above the 90th percentile to the highest reported value.
2-18
-------
0
0.05-0.1
001 County Emissions Density (Tons per sq.mi.) of Ammonia
0.019-0.05
>0-0.0057
0.12-0.34
0.0057-0.019
0.34 +
Source: U.S. EPA (2006a)
Figure 2-13. 2001 county-level NH3 emissions densities from on-road mobile sources.
2001 County Emissions Density (Tons per sq.mi.) of Amr
0.0023-0.009
>0-0.00008
0.009-0.067
0.00008-0.00045
0.067+
0.00045-0.0023
Source: U.S. EPA (2006a)
Figure 2-14. 2001 county-level NH3 emissions densities from EGUs.1
1 Range values: White, 0 or no reported value; Blue, from the smallest non-zero to the 10th percentile value; Green, from above the 10th to the 25th
percentile; Yellow, from above the 25th to the 50th percentile; Pink, from above the 50th to the 75th percentile; Red, from above the 75th to the 90th
percentile; Brown, from above the 90th percentile to the highest reported value.
2-19
-------
2001 County Emissions Density (Tons per sq.mi.) of Ammonia
I 0.16-0.37 0.37-0.8
-1.i
>0-0.16
1.8-3.2
Source: U.S. EPA (2006a)
Figure 2-15. 2001 county-level NH3 emissions densities from miscellaneous and biogenic sources.1
2.5. Evaluating Emissions Inventories
Emissions inventories are very complex and highly changeable conjoined forms built from
measurements and production and transfer rates, some measured directly, others indirectly, and others
merely assumed, combined with model predictions. National-scale emissions inventories like the ones
illustrated in county-level maps here have uncertainties embedded in them owing to unknown emission
factors, unknown and varying emission rates, generalized or depleted profiles, and the like. Substantial
effort is applied at national, state, and local scales to test these terms in the final emissions totals, and to
assure their quality.
One means for evaluating emissions inventories has been to compare predictions in the inventories
to measured long-term trends or to ratios of pollutants in ambient air. Comparisons of emissions model
predictions with observations have been performed in a number of environments. Very often emissions
inventories for NOx and SOx are evaluated in relation to CO emissions because the low reactivity of CO
on urban and regional scales means it can be treated as largely conserved. Using the distinction between
mobile sources which emit NOx and CO but little S02, and power plants which emit NOx and S02 but
little CO, Stehr et al. (2000) evaluated emissions estimates for the eastern U.S. Results indicated that coal
combustion contributes 25 to 35% of the total area NOx emissions in rough agreement with the U.S. EPA
NEI (2006a). Studies using ratios of CO concentrations to NOx concentrations, and concentrations of
nonmethane organic compounds (NMOC) to NOx carried out in the early 1990s in tunnels and ambient
air indicated that emissions of CO and NMOC were systematically underestimated in emissions
inventories at that time. More details are available in the 2000 CO AQCD (U.S. EPA, 2000a).
These reconciliation studies depend on the assumption that NOx emissions are predicted correctly
by emissions factor models which are merely mean and aggregate descriptions of the highly variable U.S.
'Range values: White. 0 or no reported value; Blue, from the smallest non-zero to the 10th percentile value; Green, from above the 10th to the 25th
percentile; Yellow, from above the 25th to the 50th percentile; Pink, from above the 50th to the 75th percentile; Red. from above the 75th to the 90th
percentile; Brown, from above the 90th percentile to the highest reported value.
2-20
-------
mobile source fleet. Roadside remote sensing data have indicated that >50% of non-methane
hydrocarbons (NMHC) and CO emissions are produced by less than 10% of vehicles (Stedman et al.,
1991), typically the poorly maintained super-emitters.
Parrish et al. (1998) and Parrish and Fehsenfeld (2000) proposed methods to derive emission rates
by examining measured ambient ratios among individual volatile organic compounds (VOCs), NOx, and
NOY. Typically, strong correlations exist among measured values for these species because emission
sources are geographically co-located, even when individual sources are different. Correlations can be
used to derive emissions ratios between species, including adjustments for the effect of photochemical
aging. Examples of this type include using correlations between CO and NOY (e.g., Parrish et al., 1991),
between individual VOC species and NOY(Goldan et al., 1995, 2000) and among various VOC species
(McKeen and Liu, 1993; McKeen et al., 1996). Many of these studies were summarized in Parrish et al.
(1998), Parrish and Fehsenfeld (2000), and Trainer et al. (2000).
Other methods for emissions evaluation exist. Buhr et al., (1992) derived emission estimates from
principal component analysis (PCA) and other statistical methods. Goldstein and Schade (2000) also used
species correlations to identify the relative effects of anthropogenic and biogenic emissions. Chang et al.
(1996, 1997), and Mendoza-Dominguez and Russell (2000) used inverse modeling to derive emission
rates in conjunction with results from chemical-transport models (CTMs).
A decadal field study of ambient CO at a rural site in the eastern U.S. (Hallock-Waters et al., 1999)
indicated a downward trend consistent with the downward trend in estimated emissions over the period
1988 to 1999 (U.S. EPA, 2000e), even when the global downward trend was taken into account.
Measurements at two urban areas in the U.S. confirmed the decrease in CO emissions (Parrish et al.,
2002). That study also indicated that the ratio of CO to NOx emissions decreased by approximately a
factor of 3 over 12 years. NEI estimates (U.S. EPA, 1997b) indicated a much smaller decrease in this
ratio, suggesting that NOx emissions from mobile sources may have been underestimated or increasing or
both. Parrish et al. (2002) concluded that 03 photochemistry in U.S. urban areas may have become more
NOx-limited over the past decade. (See Section 2.6.2.1 for a discussion of NOx and its role in enhancing
and limiting 03 formation.)
Results from these recent emissions evaluation studies have been mixed, with some studies
showing agreement to within ± 50% (U.S. EPA, 2000e). However, Pokharel et al. (2002) employed
remotely sensed emissions from on-road vehicles and fuel use data to estimate emissions in Denver. Their
calculations indicated a continual decrease in CO, hydrocarbons (HC), and NO emissions from mobile
sources over the 6-year study period, 1996 through 2001. Inventories based on the ambient data were 30
to 70% lower for CO, 40% higher for HC, and 40 to 80% lower for NO than those predicted by the
MOBILE6 on-road mobile source emissions model. (See http://www.epa.gov/otaq/m6.htm for
information on MOBILE6).
Satellite data also have proved useful for optimizing estimates of N02 emissions (Jaegle et al.,
2005; Leue et al., 2001; Martin et al., 2003). Satellite-borne instruments such as the Global Ozone
Monitoring Experiment (GOME) (see, e.g., Martin et al., 2003, and references therein) and the Scanning
Imaging Absorption Spectrometer for Atmospheric Chartography (SCIAMACHY) retrieve tropospheric
N02 columns that can be combined with model-derived x of NOx to yield emissions of NOx.
Top-down inference of NOx emission inventories from the satellite observations of N02
concentrations columns by mass balance requires at minimum three pieces of information: the retrieved
tropospheric N02 column; the tropospheric NOx-to-N02 ratio in the columns; and the NOx x against
reaction losses to stable chemical reservoirs. (See the discussion of these chemical reservoirs in Section
2.6.) A photochemical model has been used to provide information on the latter two pieces of information.
The method is most often applied to land surface emissions, excluding lightning. Tropospheric N02
columns are largely insensitive to lightning NOx emissions since most of the lightning NOx in the upper
troposphere is present as NO at the time of the satellite measurements (Ridley et al., 1996) owing to the
slower reactions of NO with 03 at the altitude where lightning production is most prevalent.
Using satellite data, Bertram et al. (2005) found clear signals in the SCIAMACHY observations of
short, intense NOx pulses following springtime fertilizer application and subsequent precipitation over
2-21
-------
agricultural regions of the western U.S. For the agricultural region in north-central Montana, they
calculated an annual SCIAMACHY top-down estimate that is 60% greater than a commonly-used model
of soil NOx emissions by Yienger and Levy (1995).
Jaegle et al. (2005) applied additional information on the spatial distribution of emissions and fire
activity to partition NOx emissions into sources from fossil fuel combustion, soils, and biomass burning.
Global a posteriori estimates of soil NOx emissions were 68% larger than the a priori estimates. Large
increases were found for the agricultural region of the western U. S. during summer, increasing total U.S.
soil NOx emissions by a factor of 2.
Martin et al. (2006) retrieved tropospheric N02 columns for May 2004 to April 2005 from the
SCIAMACHY satellite instrument to derive top-down NOx emissions estimates via inverse modeling
with the GEOS-Chem global chemical transport model. (See http://www.as.harvard.edu/ctm/geos/ for
more information on GEOS-Chem.) The top-down emissions were combined with a priori information
from a bottom-up emissions inventory with error weighting to achieve an improved a posteriori estimate
of the global distribution of surface NOx emissions. Their a posteriori inventory improved GEOS-Chem
simulations of NOx, peroxyacetyl nitrate (PAN), and HN03 as compared against airborne in situ
measurements over and downwind of New York City. Their a posteriori inventory also showed lower
NOx emissions from the Ohio River Valley in summer than winter, reflecting recent controls on NOx
emissions from EGUs there. Their a posteriori global inventory was highly consistent with the NEI 1999
(http://www.epa.gov/ttn/chief/net/1999inventorv.html') (R2 = 0.82, bias = 3%); however, it was 68%
greater than a recent inventory by Streets et al. (2003) for East Asia for the year 2000.
Significant uncertainties attach to estimates of the magnitude and spatial and temporal variability of
NH3 emissions. A strong seasonal pattern should be evident in NH3 emissions profiles to correspond with
the overwhelmingly agricultural sources of NH3 and strong seasonal temperature differences in NH3
volatility, for example, but this pattern has not appeared in previous emissions factors and inventories.
The magnitude of these temporal differences is large: Heber et al. (2001) showed that NH3 flux from two
swine finishing buildings were -70% higher in June than in fall and winter months, and Aneja et al.
(2000) found fluxes from hog waste lagoons -80 to 90% higher in summer as compared to winter.
The value of inverse modeling techniques using large-scale Eulerian air quality models (AQMs)
has been successfully demonstrated for several aspects of emissions inventories; see, for example,
Mendoza-Dominguez and Russell (2000, 2001a, b). Gilliland (2001), Gilliland et al. (2001), and Pinder
et al. (2006) have worked extensively with Kalman filter inverse modeling and the U.S. EPA Community
Multiscale Air Quality (CMAQ) modeling system (Byun and Ching, 1999), to reduce uncertainties
specifically in NH3 emissions. NH3 is an especially good case for emissions estimate evaluation with
inverse modeling techniques because the modeled response in NH44" wet deposition is strongly linear with
changes in NH3 emissions. Correcting the NH3 emissions estimates was also shown to be an essential step
for reasonable model predictions of other N compounds (Gilliland et al., 2003). Results can be highly
significant. For example, the a posteriori R value of CMAQ predictions against measured wet NH/
concentrations from the National Atmospheric Deposition Program (NADP) sites in the U.S. was 0.98,
increased from the a priori value of 0.12. Pinder et al. (2004) provided the first farm-level model for NH3
emissions from dairy cattle, and this has been coupled with the seasonally varying fertilizer inventory for
NH3 from Goebes et al. (2003) and with the inverse modeling results of Gilliland et al. (2003) to correct
the NEI NH3 emissions totals. The estimate of Gilliland et al. (2003) was that the annual NEI NH3 was
-37% too high to optimize the modeled wet NH44" concentration. Following earlier work by Gilliland and
others in this vein, U.S. EPA (2006a), in fact, reported its intention to decrease total NH3 emissions in the
NEI by 23% by altering emissions factors for nondairy cows and swine.
Holland et al. (2005) estimated wet and dry deposition of NHX based on measurements over the
CONUS and reported that NH3 emissions in the 1999 NEI were underestimated by a factor of -2 or 3.
Possible reasons for this error included under-representation of deposition monitoring sites in populated
areas and the neglect of offshore transport in the NEI. The use of fixed deposition velocities (Vd) not
reflective of local conditions at the time of measurement introduces additional uncertainty into estimates
of dry deposition to which NH3 is particularly sensitive.
2-22
-------
2.5.1. Emissions for Historical Modeling
Rigorous emissions inventories require careful analysis of very large data sets on fuel use and types
and activity patterns for them to be a reliable basis for atmospheric concentration and deposition
calculations. Sections 2.2 to 2.5 make clear that even with current best estimates for these input data,
present-day emissions inventories can be substantially in error.
However, estimates of biological effects from long-term acidifying deposition require computation
of historical emissions. Such historical estimates of emissions generally use some divisions of economic
sectors together with estimates of fuel type and S content, for example. Schopp et al. (2003) have pursued
this method for estimating acid deposition in Europe between 1880 and 2030 using the International
Institute for Applied Systems Analysis (HASA) Regional Acidification and Information System (RAINS)
model; see Alcamo et al. (1990) and Asman et al. (1988) for descriptions and applications of RAINS.
Mylonda (1996) used similar tools and methods to compute estimates of S02 emissions and atmospheric
concentrations and deposition in Europe showing that emissions, which peaked in Europe in the 1960s
and 1970s had increased by a factor of 10 since the 1880s. Similarly, modeled concentrations and
deposition increased by factors of 2 to 6, depending on location in the same time period.
Historical emissions and concentration estimates are often not resolved to more than decadal
resolution because of the very large uncertainties inherent in their computation, which very often includes
no information on variability in the meteorological state of the atmosphere or its oxidizing capacity, both
of which change overtime. Spatial resolution, however, has extended from continental scale, like those
using RAINS, up to global scale estimates and down to subregional ones. A survey of methods and
cautions for interpretation was made by Galloway (1995).
Global estimates like the ones by Lefohn et al. (1999) use methods similar to those at the
continental scale, though with more specific factors on S content in fuels, for example, where those are
known. Lefohn et al. (1999), for example, estimated the global average anthropogenic S emissions in
1850 to be <2% of current levels. At finer scales, Driscoll et al. (2001b) used U.S. EPA (2000e) estimates
of emissions trends between 1900 and 1998 together with biogeochemical modeling to estimate historical
loadings to the northeast U.S. to calculate potential recovery times in affected lakes and streams.
Among the finest-scale historical estimates for N emissions and deposition are those by Bowen and
Valiela (2001) for Cape Cod, MA. Bowen and Valiela (2001) computed decadal changes in several
oxidized and reduced nitrogen species for the time period between 1910 and 1995 to conclude that total N
deposition at Cape Cod has been increasing at the rate of 0.26 kg N/ha/decade.
2.6. NOx-SOx-NHx Chemistry in the Troposphere
2.6.1. Introduction
NOx, VOCs, and CO are precursors in the formation of 03 and other elements of photochemical
smog and PM. The role of NOx in producing 03, the factors controlling P(03) efficiency and methods for
calculating 03 from its NOx and VOC precursors were all reviewed in Section 2.2 of the 2006 03 AQCD
(U.S. EPA, 2006b) and are available in numerous texts including Jacob, 2000; Jacobson, 2002; and
Seinfeld and Pandis, 1998. More specific details on the chemistry and transformation of SOx can be
found in the 2008 SOx ISA (U.S. EPA 2008b). Hence, those topics are only briefly recounted here with
special reference to the secondary NOx and SOx NAAQS.
2-23
-------
Long range transport to remote
regions at low temperatures
-NO
HNO
PANs
HO
hv, OH
NO,
-N2°5
hv, M
HONO
PAH
nitro-PAHs
~
"°3-<
HO.
NH
NO
MPP
.,n R-C=C-R
no3 _—y RONO
nitrosamines,
^nitro-phenols, etc.
IN
MPP
Inorganic
Nitrates
HO
ro:
RO
NO
NO •
Oo
deposition
deposition
emissions
Figure 2-16. Atmospheric cycle of reactive oxidized N species. NQy refers to all the species shown within
the inner and outer box: NOx to NO and NO2 (in the inner box), and NOz to all the species
outside of the inner box. IN refers to inorganic particulate species (e.g., sodium [Na+], calcium
[Ca2+]), MPP to multiphase processes, hv to a solar photon, and R to an organic radical.
Particle-phase RONO2 are formed from the species shown on the right side. For the purposes
of this document, "NOx" is defined as the group of all N-containing compounds inside the
large dashed-line box, the same group generally termed "N0Y" by atmospheric scientists.
Important compounds, reactions, and cycles are schematized in Figure 2-16. Figure 2-16 also
illustrates that N02, itself an oxidant, can react to form other photochemical oxidants including organic
nitrates (R0N02) like the PANs shown in Figure 2-16, and can react with toxic compounds like the
polycyclic aromatic hydrocarbons (PAHs) to form nitro-PAHs, some of which demonstrate greater
toxicity than either reactant alone. N02 can also be further transformed to HN03 and can contribute in
that form to the acidity of cloud, fog, and rain water and can form ambient NO; particles (pNO-. ).
The only gas-phase forms of SOx of interest in tropospheric chemistry are SO . S03, and H2SQ4.
S03 can be emitted from the stacks of power plants and factories; however, it reacts extremely rapidly
with H20 in the stacks or immediately after release into the atmosphere to form sulfuric acid (H2S04).
H2S04 in turn mainly condenses onto existing particles when particle loadings are high, or nucleates to
form new particles under lower concentration conditions. Thus, of the gas-phase SOx species, only S02 is
emitted in the tropospheric boundary layer at concentrations of concern for environmental exposures.
NH3, the gas-phase precursor for NH4 . plays a key role in neutralizing acidity in ambient particles
and in cloud, fog, and rain water. NH3 is also involved in the ternary nucleation of new particles and
reacts with gas-phase P1N03 to form \ l l4\0;. and with SO_f to form ammonium bisulfate (NH4HS04)
and ammonium sulfate ((NH4)2S04), three significant components of N and S deposition across the
landscape. The NOx-SOx-NHx cycles and phase-changes are schematized in Figure 2-17.
2-24
-------
Gas-phase products
H02
Secondary Organic
aerosol ^
CO, RH, RCHO
OH
NO
N02
RCOk^/ *
PAN f NO3
NO,W Mr
WH3
Droplet Phase
Source: Adapted from Meng and Seinfeld (1996) and Warneck (1999a); used with permission), by R. Mathurand R. Dennis, U.S. EPA /
Office of Research and Development (ORD) / National Environmental Research Laboratory (NERL)/Atmospheric Modeling Division.
Figure 2-17. The combined NOx + SOx + NHx system showing how atmospheric fates and lifetimes of
reduced and oxidized N components are linked.
2.6.2. NOx Chemistry
As described in Section 2.2, NOx is emitted by combustion sources mainly as NO with ~5 to 10%
N02. The rapid photochemical cycle in the troposphere linking NO and N02 involves photolysis of N02
by UV-A radiation to yield NO and a ground-state oxygen atom, O( P).
N02 + hv > NO + 0(3P)
Reaction -
This ground state oxygen atom, O( P). can then combine with molecular oxygen (02) to form 03; and,
colliding with M, any molecule from the surrounding air (M = N2, 02, etc.), the newly formed 03
molecule transfers excess energy and is stabilized
OCP) + 02 + M > 03 + M
Reaction 5
NO and 03 react to re-form N02.
2-25
-------
NO + O, > N02 + 02
Reaction 6
Reaction 6 is responsible for 03 decreases and N02 increases found near sources of NO (like highways or
power plant plumes). The falloff of N02 with distance from a road depends on: wind speed and direction;
the local structure of turbulent mixing; temperature through the temperature dependence of Reaction 6;
and the amount of UVA-flux through Reaction 4.
Oxidation of reactive VOCs leads to formation of reactive radical species that allow conversion of
NO to N02 without participation of 03 as in Reaction 6.
N0 ho2, ro2 > NQ^
Reaction 1
03, therefore, can accumulate as N02 photolyzes as in Reaction 4, followed by Reaction 5. Specific
mechanisms for the oxidation of a number of VOCs were discussed in the 2006 03 AQCD (U.S. EPA,
2006b).
It is often convenient to speak about families of chemical species defined in terms of members that
interconvert rapidly on time scales shorter than those for formation or destruction of the family as a
whole. For example, an odd oxygen (Ox) family can be defined as
Ox = £(0(JP) + 0(D) + O, + N02)
Equation 1
In much the same way, NOx is sometimes referred to as odd nitrogen. Hence, production of Ox
occurs by Equation 1, while the sequence of reactions given by Reactions 4-6 represents no net
production of Ox. (Definitions of species families and methods for constructing families are discussed in
Jacobson [1999] and references therein). Other families including N-containing species used later in this
chapter are
NOz=YJ HN02 + HN03 + HN04 + NO; + 2N205 + PAN(CH3CH0-00-N02)
+ other organic nitrates + halogen nitrates + particulate nitrates
Equation 2
and
NOy = NOx + NOz
Equation 3
and
HOx = OH + H02
Equation 4
NOz refers to the sum of all oxidation products of NOx without the original NO and N02.
The reaction of N02 with 03 leads to formation of the N03 radical.
N02 + (I > NO, + 02
Reaction 8
2-26
-------
However, N03 radical reacts rapidly, having a x of ~5 seconds during the photochemically most active
period of the day near local solar noon, by two pathways (Atkinson et al., 1992):
N03 + hv > NO + O2Q0%)
Reaction 9
Aro3 + hv > N02 + O(3P\90%)
Reaction 10
Because of this, N03 concentrations remain low during daylight hours but can increase after sunset
to nighttime concentrations of <5 x 107 to 1 x 1010 molecules/cm3 or <2 to 430 parts per trillion (ppt) over
continental areas influenced by anthropogenic emissions of NOx (Atkinson et al., 1986). At night, N03,
rather than OH, is most often the primary oxidant in polluted tropospheric systems. Moreover, N03 can
combine with N02 to form dinitrogen pentoxide (N205)
N(l + N02 N205
Reaction 11
and N205 then both photolyzes and thermally decomposes back to N02 and N03 during the day; however,
N205 concentrations can accumulate during the night to ppb levels in polluted urban atmospheres.
The tropospheric chemical removal processes for NOx include reaction of N02 with the OH radical
and hydrolysis of N205 in aqueous aerosol solutions if there is no organic coating. Both of these reactions
produce HN03
OH + N02 > HNO-,
Reaction 12
N205 H2°w > 2HN03
Reaction 13
The gas-phase reaction of OH with N02 (Reaction 12) is one of the major and ultimate removal
processes for NOx in the troposphere. This reaction removes OH and N02 in one step and competes with
HC for OH in areas characterized by high NOx concentrations such as urban centers. The x for conversion
of NOx to HN03 in the PBL at 40°N latitude ranges from ~4 hours in July to -16 hours in January. The
corresponding range in x at 25°N latitude is between 4 and 5 hours, while at 50°N latitude, HN03 x ranges
from about 4 to 20 hours (Martin et al., 2003).
In addition to gas-phase HN03, Golden and Smith (2000) have shown on the basis of theoretical
studies that pernitrous acid (H02NO) is also produced by the reaction of N02 and OH; however, this
production channel most likely represents a minor yield of-15% at the surface (JPL, 2003). Pernitrous
acid will also thermally decompose and photolyze. Gas-phase HN03 formed from Reaction 12 and
Reaction 13 undergoes wet and dry deposition to the surface and uptake by ambient aerosol particles.
2-27
-------
In addition to uptake on particles and in cloud droplets, HN03, photolyzes and reacts with OH
HN(l + hv > OH + N02
Reaction 14
HN03 + hv > O + HN02
Reaction 15
HN03 + hv > H + N03
Reaction 16
and
HN(l + OH > N(l + H20
Reaction 17
Margitan and Watson (1982) established that Reaction 14 has a quantum yield of ~1, with only very small
contributions from the two other possible photolytic pathways (Reactions 15 and 16). The x of HN03 with
respect to these two reactions is long enough for HN03 to act somewhat as a reservoir species for NOx
during long-range transport, contributing in this way to N02 levels in areas remote from the source region
of the NOx that formed this HN03.
The contribution of Reaction 13 to HN03 formation is highly uncertain during both winter and
summer. The importance of Reaction 13 could be much higher during winter than during summer because
of the much lower concentration of OH and the enhanced stability of N205 due to lower temperatures and
UV flux. Geyer and Piatt (2002) concluded that Reaction 16 constituted about 10% of the removal of
NOx at a site near Berlin, Germany during spring and summer. However, Dentener and Crutzen (1993)
estimated this to be 20% in summer and 80% in winter. A modeling study by Tonnesen and Dennis (2000)
reported; 16 to 31% of summer HN03 production from Reaction 13. Recent work in the northeastern U.S.
(Brown et al., 2006a, b; Frost et al., 2006a) indicates that this reaction proceeds at a faster rate in power
plant plumes than in urban plumes.
OH also reacts with NO to produce nitrous acid (HN02).
OH + NO M ) HN02
Reaction 18
In sunlight, HN02 is rapidly photolyzed back to the original reactants.
HN02 + hv > OH + NO
Reaction 19
Reaction 18 is, however, a negligible source of HN02, which is formed mainly by multiphase processes.
At night, heterogeneous reactions of N02 in aerosols or at Earth's surface result in accumulation of HN02
(Harris et al., 1982; Jacob, 2000). Harris et al., (1982) suggested that photolysis of HN02 at sunrise could
provide an important early-morning source of OH necessary to drive P(03).
2-28
-------
H02 can react with N02 to produce pernitric acid (HN04)
H02 + N02 + M > HN04 + M
Reaction 20
which can thermally decompose and photolyze back to its original reactants. The acids formed in these
gas-phase reactions are all water soluble; thus, they can be incorporated into cloud droplets and in the
aqueous phase of particles.
A broad range of organic N compounds are directly emitted by combustion sources or formed in
the atmosphere from NOx emissions. Organic N compounds include the PANs, nitrosamines, nitro-PAHs,
and the more recently identified nitrated quinones. Oxidation of VOCs produces organic peroxy radicals
(R02). Reaction of R02 radicals with NO and N02 produces R0N02 and peroxynitrates (R02N02)
R02 + NO —^ R0N02
Reaction 21
ro2 + no2 —^ ro2no2
Reaction 22
Reaction 21 is a minor branch for the reaction of R02 with NO; the major branch produces RO and N02
as discussed in the next section. However, the R0N02 yield increases with carbon number (Atkinson,
2000).
The most important of these organic nitrates is PAN, the dominant member of the broader family of
peroxyacyl nitrate species (PANs) which includes peroxypropionyl nitrate (PPN) of anthropogenic origin
and peroxymethacrylic nitrate (MPAN) produced from isoprene oxidation. The PANs are formed by the
combination reaction of acetyl peroxy radicals with N02
CH3C(0)00 + N02 > CH3C(OpONq
Reaction 23
where the acetyl peroxy radicals are formed mainly during the oxidation of ethane (C2H6). Acetaldehyde
(CH3CHO) is formed as an intermediate species during the oxidation of ethane. CH3CHO can be
photolyzed or react with OH to yield acetyl radicals
CH3C{0)H + hv > CH3C(0) + H
Reaction 24
CH3C(0)H + OH > CH3C(0) + h2o
Reaction 25
Acetyl radicals then react with 02 to yield acetyl peroxy radicals.
ch3c(o) + o2 + m —> ch3 c(oyjo + m
Reaction 26
2-29
-------
However, acetyl peroxy radicals will react with NO in areas of high NO concentrations
CH3 (COyDO + NO > CH3(C0)0 + no2
Reaction 21
and the acetyl-oxy radicals will then decompose.
CH3(C0)0 > CH3 + C02
Reaction 28
Thus, formation of PAN is favored at conditions of high ratios of N02 to NO, which are most
typically found under low total NOx concentration conditions. The PANs both thermally decompose and
photolyze back to their reactants with x on the order of a few hours during warm sunlit conditions: x with
respect to thermal decomposition range from ~1 h at 298 K to -2.5 days at 273 K and up to several weeks
at 250 K. Thus, PANs can provide a reversible sink of NOx at cold temperatures and high solar zenith
angles, allowing release of N02 as air masses warm, particularly by subsidence from the free troposphere.
The PANs are also removed by uptake to vegetation (Sparks et al., 2003; Teklemariam and Sparks, 2004).
R02N02 produced by Reaction 22 are thermally unstable and most have very short x of <100 seconds
owing to thermal decomposition back to the original reactants. Thus they are not effective permanent
sinks of NOx, except at lower temepatures.
2.6.2.1. NOx and Ozone Formation
03 is unlike most other air pollution species whose rates of formation vary directly with the
emissions of their precursors in that P(03) changes nonlinearly with the concentrations of its precursors.
At the low NOx concentrations found in most environments ranging from remote continental areas to
rural and suburban areas downwind of urban centers, net P(03) increases with increasing NOx levels. At
the high NOx concentrations found in downtown metropolitan areas especially near busy streets and
roadways and in power plant plumes, net destruction of 03 by titration reaction with NO dominates.
Between these two regimes is a transition stage in which P(03) shows only a weak dependence on NOx
concentrations. In the high NOx concentration regime, N02 scavenges OH (Reaction 12) which would
otherwise oxidize VOCs to produce H02, which in turn would oxidize NO to N02 (Reaction 7). In the low
NOx concentration regime, VOC oxidation generates, or at least does not consume, free radicals, and
P(03) varies directly with NOx levels. Sometimes the terms VOC-limited and NOx-limited are used to
describe these two regimes; also, the terms NOx-limited and NOx-saturated are used (see Jaegle et al.,
2001). OH chemistry initiates HC oxidation and behaves similarly to that for 03 with respect to NOx
concentrations (Hameed et al., 1979; Pinto et al., 1993; Poppe et al., 1993; Zimmermann and Poppe,
1993). These considerations introduce a high degree of uncertainty into attempts to relate changes in 03
concentration to precursors emissions. Note that in a NOx-limited or NOx-sensitive regime, P(03) is not
insensitive to radical production or the flux of solar UV photons; rather, P(03) is simply more sensitive to
the NOx concentrations. For example, global tropospheric 03 is sensitive to CH4, even though the
troposphere is predominantly NOx-limited. More details on P(03) are given in the 2008 NOx ISA (U.S.
EPA, 2008a).
2.6.2.2. Multiphase Interactions
Warneck (1999a) constructed a box model describing the chemistry of the oxidation of N02
including the interactions of N species and minor processes in sunlit cumulus clouds. The relative
contributions of different reactions to the oxidation of N02 to N03 10 minutes after cloud formation are
2-30
-------
given in Table 2-3 where the last two columns show the relative contributions with and without transition
metal ions. Oxidation of N02 as delineated in Table 2-3 occurs mainly in the gas phase within clouds,
implying that gas-phase oxidation of N02 by OH predominates.
NOx and Halogen Chemistry Interactions
Four decades of observational data on 03 in the troposphere have revealed numerous anomalies not
fully explained by the gas-phase HOx-NOx photochemistry (described above). The best-known example
is the dramatic decrease in ground-level 03 concentration during polar sunrise due to multiphase catalytic
cycles involving inorganic Br and CI radicals (Barrie et al., 1988; Foster et al., 2001; Martinez et al.,
1999). Other examples of anomalies in tropospheric 03 at lower latitudes include 03 concentrations
<10 ppb in the marine boundary layer (MBL) and overlying free troposphere (FT) at times over large
portions of the tropical Pacific (Kley et al., 1996), as well as post-sunrise 03 concentration decreases over
the western subtropical Pacific Ocean (Nagao et al., 1999), the temperate Southern Ocean (Galbally et al.,
2000), and the tropical Indian Ocean (Dickerson et al., 1999).
Table 2-3. Relative contributions of various gas and aqueous phase reactions to aqueous N03 formation
within a sunlit cloud, 10 minutes after cloud formation.
Reaction
% of Total3
% of Totalb
GAS PHASE
OH + NO2 + M
57.7
67.4
AQUEOUS PHASE
N205(g) + H20
8.1
11.2
NO3- + CI-
<0.1
0.1
NO3- + S032"
0.7
1.0
NO3- + HCOO-
0.6
0.8
HNO4 + SO32-
31.9
20.5
HOONO + NOs"
0.8
<0.1
O3 + N02-
<0.1
<0.1
8 In the absence of transition metals.b In the presence of iron and copper ions.
Source: Adapted from Warneck (1999a). Used with permission.
The set of N reactions with aerosol salts in marine atmospheres sketched briefly here was reviewed
in detail by De Haan et al. (1999). This chemistry remains important not only for halogen cycling and
atmospheric oxidation reactions (Andreae and Crutzen, 1997), but also because through them N03 can
be shifted from gas-phase HN03 or from fine-mode aerosol after dissociation of NH4N03, for example, to
coarse-mode particles, thereby enhancing the potential for local N deposition to coastal regions. The
actual areal extent ofN deposition resulting from gas-to-particle N03 conversion, however, is a complex
function of local wind speeds, as shown by Pryor and Sorensen (2000). With moderate winds of 3.5-10
m/s (~8 to 22 mph), gas-phase HN03 Vd exceeded that of an average sodium nitrate (NaN03) particle,
whereas at higher and lower wind speeds the reverse was true. This means that as a result of gas-to-
particle N03 conversion, under commonly moderate winds, less N would be deposited locally and more
would be available for transport and deposition in a larger area of extent.
2-31
-------
2.6.3. SOx Chemistry
The only forms of monomeric SOx of interest in troposheric chemistry are S02 and S03. S03 can
be emitted from the stacks of power plants and factories; however, it reacts extremely rapidly with H20 in
the stacks or immediately after release into the atmosphere to form H2S04, which mainly condenses onto
existing particles when particle loadings are high; it can nucleate to form new particles under lower
concentration conditions. Thus, only S02 is present in the troposheric boundary layer at concentrations of
concern for environmental exposures.
Gas phase oxidation of S02 is initiated by the reaction
S02 + OH + M > HS03 + M
Reaction 29
followed by
HSCl + 02 > SCI + H02
Reaction 30
so3 + h2o —> h2soa
Reaction 31
Because the saturation vapor pressure of H2S04 is extremely low, it will be removed rapidly by
transfer to the aqueous phase of aerosol particles and cloud droplets. Depending on atmospheric
conditions and the concentrations of other ambient particles and gas-phase species that can participate in
new particle formation, it can also nucleate to form new particles. Rate coefficients for the reactions of
S02 with either H02 or N03 (JPL, 2003) are too low to be significant.
S02 is chiefly but not exclusively primary in origin; it is also produced by the photochemical
oxidation of reduced S compounds such as DMS, H2S, CS2, OCS, and methyl mercaptan, which are all
mainly biogenic in origin (Their sources are discussed in Section 2.3.2). Table 2-4 lists the x of reduced S
species with respect to reaction with various oxidants. Except for OCS, which is lost mainly by photolysis
with x ~6 months, these species are lost mainly by reaction with OH and N03 .
2-32
-------
Table 2-4. Atmospheric lifetimes (t) of SO2 and reduced S species with respect to reaction with OH, NO3,
and CI radicals.
Compound
OH
NO3
CI
kx10«
X
kx 1012
X
kx 1012
X
S02
1.6
7.2 d
NA
NA
CH3-S-CH3
5.0
2.3 d
1.0
1.1 h
400
29 d
H2S
4.7
2.2 d
NA
74
157 d
CS2
1.2
9.6 d
<0.0004
>116 day
<0.004
Nr
OCS
0.0019
17 y
<0.0001
>1.3 yr
<0.0001
Nr
CHs-S-H
33
8.4 h
0.89
1.2 h
200
58 d
CH3-S-S-CH3
230
1.2 h
0.53
2.1 h
NA
1Rate coefficients were taken from JPL Chemical Kinetics Evaluation No. 14 (JPL 2003). NA = Reaction rate coefficient not available. OH = 1 x 106/cm3. NO3 = 2.5 x 108/cm3
CI = 1 x 108/cm3. Nr = Rate coefficient too low to be relevant as an atmospheric loss mechanism. Rate coefficients were calculated at 298 K and 1 atmosphere.
Source: Seinfeld and Pandis (1998)
Because OCS is relatively long-lived in the troposphere, it can be transported upwards into the
stratosphere. Crutzen (1976) proposed that its oxidation serves as the major source of S042 in the
stratospheric Junge layer during periods when volcanic plumes do not reach the stratosphere. However,
the flux of OCS into the stratosphere is probably not sufficient to maintain this stratospheric aerosol layer.
Myhre et al. (2004) proposed instead that S02 transported upwards from the troposphere is the most likely
source since the upward flux of OCS is too small to sustain observed S042 loadings in the Junge layer.
In addition, in situ measurements of the isotopic composition of S do not match those of OCS (Leung
et al., 2002).
Reaction with N03 at night most likely represents the major loss process for DMS and methyl
mercaptan, although the mechanisms are not well understood. Initial attack by N03 and OH involves H
abstraction, with a smaller branch leading to OH addition to the S atom. The OH addition branch
increases in importance as temperatures decrease, becoming the major pathway below temperatures of
285 K (Ravishankara, 1997). The adduct may either decompose to form methane sulfonic acid (MSA) or
undergo further reactions in the main pathway to yield DMS (Barnes et al., 1991). Following H4"
abstraction from DMS, the main reaction products include MSA and S02. The ratio of MS A to S02 is
strongly temperature dependent, varying from 0.1 in tropical waters to 0.4 in Antarctic waters (Seinfeld
and Pandis, 1998). S042 in excess of that expected from sea salt aerosols is related mainly to production
of S02 from the oxidation of DMS. These transformations among atmospheric S compounds are
summarized in Figure 2-18.
2-33
-------
Tropopause
OTHER
, S('V) ,
OTHER
S(VI)
Source: Adapted from Berresheim et al. (1995). Used with permission.
Figure 2-18. Atmospheric cycle of S compounds.
2.6.3.1. Multi-Phase SOx Chemistry
The major S species m clouds are hydrogen sulfite (HSO ;; ) and the sulfite ion (SOv ). Both are
derived from the dissolution of S02 in water and are grouped together as S(IV); bisulfate ion (HS04 ) and
S042 are grouped together as S(VI). The chief species capable of oxidizing S(IV) to S(VI) in cloud water
are 03, H202 or organic peroxides, OH, and ions of transition metals such as iron (Fe), manganese (Mn)
and copper (Cu) in the presence of 02. The basic mechanism of aqueous-phase oxidation of S02 has long
been studied and can be found in numerous texts on atmospheric chemistry; see for example, Finlayson-
Pitts and Pitts (2000), Jacob (1999), Jacobson (2002), and Seinfeld and Pandis (1998). Following
Jacobson (2002), the steps involved in aqueous-phase oxidation of S02 can be summarized as: dissolution
of S02
S02(g) <¦ > SO2 (aq)
Reaction 32
and formation and dissociation of H2S03.
S02(aq) + H20(ctq) < > H2S(l < > H + HSO, < > 2H + SO:'
Reaction 33
2-34
-------
In the pH range commonly found in rainwater, pH 2 to 6, the most important reaction converting
S(IV) to S(VI) is
HSO; + H20 + H+ < > S042~ + H20 + 2H+
Reaction 34
as SO ,2 is much less abundant than HSO , .
Another pathway to aqueous-phase oxidation of S(IV) is reaction with 03
hso; + 03 + OH > so; + H20 + 02
Reaction 35
But while the gas-phase reaction of S02 with 03 is slow, the aqueous phase analog of Reaction 35
is rapid, with rate coefficient increases up to a value of ~5 x 10"3 with increasing pH between 1 and 3
(Finlayson-Pitts and Pitts, 2000).
Major pathways for the aqueous-phase oxidation of S(IV) to S(VI) as a function of pH are shown
in Figure 2-19. For pH up to about 5.3, H202 is the dominant oxidant converting S(IV) to S(VI), while at
pH >5.3, 03 becomes dominant, followed by Fe(III), using the characteristic values found in Seinfeld and
Pandis (1998). However, differences in concentrations of oxidants result in differences in the pH at which
this transition occurs. Note that oxidation of S02 by 03 and 02 tends to be self-limiting: as S042 is
formed, pH decreases and rates of these reactions decrease. Higher pH levels are expected to be found
mainly in marine aerosols; however, in marine aerosols, the Cl-catalyzed oxidation of S(IV) may be more
important (Hoppel and Caffrey, 2005; Zhang and Millero, 1991). Because NH/ is so effective in
neutralizing acidity, it, too, affects the rate of oxidation of S(IV) to S(VI) and the rate of dissolution of
S02 in particles and cloud droplets.
Comparison of the relative rates of oxidation by gas-and-aqueous phase reactions by Warneck
(1999b) indicated that, on average, only -20% of S02 is oxidized by gas-phase reactions; the remainder is
oxidized by aqueous-phase reactions. Warneck's box model (1999a) describing the chemistry of the
oxidation of S02 and N02 included interactions of S species and minor processes in sunlit cumulus
clouds. The relative contributions of different reactions to the oxidation of S(IV) species to S(VI) 10
minutes after cloud formation are given in Table 2-5 with the last two columns showing the relative
contributions with and without transition metal ions. As can be seen from Table 2-5, S02 within a cloud
(gas + cloud droplets) is oxidized mainly by H202 in the aqueous phase, while gas-phase oxidation by OH
is small by comparison. A much smaller contribution in the aqueous phase is made by methyl
hydroperoxide (CH3OOH) because it is formed mainly in the gas phase and its Henry's Law constant is
several orders of magnitude smaller that of H202. After H202, HN04 is the major contributor to S(IV)
oxidation.
The values shown in Table 2-5 here and Table 2-3 above for N03 indicate that gas-phase oxidation
accounts for only -20% of S02 oxidation but -90% of N02 oxidation.
2-35
-------
1-10
w
NO,
Z
,-12
w
/
4,m
"D
i
1-14
,-16
0
1
2
3
4
5
6
pH
Figure 2-19. Comparison of aqueous-phase oxidation paths. The rate of conversion of S(IV) to S(VI) is
shown as a function of pH. Concentrations assumed are: [S02(g)] = 5 ppb; [N02(g)] = 1 ppb;
[H202(g)] = 1 ppb; [03(g)] = 50 ppb; [Fe(lll)(aq)] = 0.3 |jM; [Mn(M)(aq)] = 0.3 \M.
In areas away from strong pollution sources, the S02 x is ~7 days, based on measurements of the
rate constant for Reaction 29 (JPL, 2003) and a nominal OH concentration of 106 molecules/cm3.
However, the mechanism of S02 oxidation at a particular location depends on local environmental
conditions. For example, near stacks, oxidants such as OH are depleted and almost no S02 is oxidized in
the gas phase. Farther downwind, as the plume is diluted with background air, the gas phase oxidation of
S02 increases in importance. Finally, even farther downwind when conditions in the plume can become
more oxidizing than in background air, the S02 oxidation rate can exceed that in background air.
S02 in the PBL is also removed from the atmosphere by dry deposition to moist surfaces, resulting
in an atmospheric x with respect to dry deposition of ~1 day to 1 week. Wet deposition of S naturally
depends on the variable nature of rainfall, but in general results in a x of ~7 days, too. These two
processes, oxidation and deposition, lead to an overall x of S02 in the atmosphere of 3 to 4 days.
2-36
-------
Table 2-5. Relative contributions of various reactions to the total S(IV) oxidation rate within a sunlit
cloud, 10 minutes after cloud formation.
Reaction
% of Total3
% of Totalb
GAS PHASE
0H + S02
3.5
3.1
AQUEOUS PHASE
03 + HSO3-
0.6
0.7
O3 + SOs2"
7.0
8.2
H2O2 + S032"
78.4
82.1
CHsOOH + HSO3-
0.1
0.1
HNO4 + HSO3-
9.0
4.4
HOONO + HSOs-
<0.1
<0.1
HSOs" + HSO3-
1.2
<0.1
SOs" + SOs2"
<0.1
<0.1
HSOs" + Fe2+
0.6
aln the absence of transition metals.bln the presence of iron and copper ions.
Source: Adapted from Warneck (1999a). Used with permission.
Multiphase chemical transformations involving inorganic halogenated compounds effect changes in
the multiphase cycling of SOx in ways analogous to their effects on NOx. Oxidation of DMS by BrO
produces dimethyl sulfoxide (DMSO) (Barnes et al., 1991; Toumi, 1994), and oxidation by CI leads to
formation of S02 (Keene et al., 1996). DMSO and S02 are precursors for MSA and H2S04. In the MBL,
virtually all H2S04 and MSA vapor condenses onto existing aerosols or cloud droplets, which
subsequently evaporate thereby contributing to aerosol growth and acidification. Unlike MSA, H2S04
also has the potential to produce new particles (Korhonen et al., 1999; Kulmala et al., 2000) which in
marine regions is thought to occur primarily in the free troposphere.
Excepting H2S04, inorganic particles are solid at low relative humidity (RH), and their composition
determines their deliquescence thresholds for forming saturated aqueous solutions. Crystallization is not
simply the reverse of deliquescence but is a process subject to hysteresis; see the sodium chloride (NaCl)
+ Na2S04 example in Figure 2-20, and Tang and Munkelwitz (1993) for deliquescence RH points of other
inorganic particles.
Particles with several components behave similarly to the example of Na2S04 but with more
complex curves and generally have deliquescence RH points below those of their constituent components
(Wexlerand Seinfeld, 1991).
Saiz-Lopez et al. (2004) estimated that observed levels of BrO at Mace Head Atmospheric
Research Station in Ireland would oxidize (CH3)2S ~6 times faster than OH and thereby substantially
increase P(H2S04) and other condensable S species in the MBL. S02 is also scavenged by deliquesced
aerosols and oxidized to H2S04 in the aqueous phase by several strongly pH-dependent pathways
(Chameides and Stelson, 1992; Keene and Savoie, 1998; Vogt et al., 1996). Model calculations indicate
that oxidation of S(IV) by 03 dominates in fresh, alkaline sea salt aerosols, whereas oxidation by
hypohalous acids, primarily HOC1, dominates in moderately acidic solutions.
2-37
-------
4.5
• deliquescence
o efflorescence
predicted
measured
3.0
2.5
2.0
Q_
0.50
0.65
0.60
0.65
0.70
0.75
0.80
0.85
0.90
Relative Humidity
Source: Ansari and Pandis (2000)
Figure 2-20. RH effects on deliquescence and efflorescence points for a NaCI+ Na2S04 particle, indicating
deliquescence at ~72% relative humidity and re-crystallization at ~52% RH. Points are
measurements from Tang et al., (1997); solid line is aerosol thermodynamic model prediction
of Ansari and Pandis (2000).
Additional non-sea salt (nss) pS04 is generated by S02 oxidation in cloud droplets (Clegg and
Toumi, 1998). Ion-balance calculations indicate that most of the nss pS04 in short-lived sea salt size
fractions accumulates in acidic aerosol solutions or in acidic aerosols processed through clouds or both
(see, e.g., Keene et al., 2004). The production, cycling, and associated radiative effects of S-containing
aerosols in marine and coastal air are regulated in part by chemical transformations involving inorganic
halogens (Von Glasow et al., 2002b). These transformations include: dry-deposition fluxes of nss pS04 in
marine air dominated, naturally, by the sea salt size fractions (Huebert et al., 1996; Turekian et al., 2001);
HC1 phase partitioning that regulates sea salt pH and associated pH-dependent pathways for S(IV)
oxidation (Keene et al., 2002; Pszenny et al., 2004); and potentially important oxidative reactions with
reactive halogens for (CH3)2S and S(IV). However, both the absolute magnitudes and relative importance
of these processes in MBL S cycling are poorly understood.
2.6.4. Multiphase NOx, SOx, and NHx Interactions
Figure 2-17 above illustrated the central role NH3 can play in the atmospheric chemistry of NOx
and SOx. This results in part from its being the most common soluble base in the atmosphere and from a
range of its possible chemical reactions. OH attack on NH3 proceeds by
NH3 + OH > NH2 + H20
Reaction 36
2-38
-------
The fate of the NH2 radical is not known with certainty, but in polluted atmospheres can be
NH 2 + () -, > NH, NHO, NO
Reaction 37
nh2 + no2 —> n2
Reaction 38
nh2 + no2 —> n2o + h2o
Reaction 39
However, with typical OH concentrations of 1 to 2 x 106 molecules/cm3, the x of NH3 against the initial
reaction is -30 to 70 days, sufficiently long that this is a small sink compared to NH3 uptake by cloud
droplets where it is reduced to NH4+
NH, (g) + H20 < > NH3 ¦ H20(aq) < > NH4 + OH
Reaction 40
Gas-phase NH3 also reacts with gas-phase HN03 to form particulate NH4N03
NH3 + HN03 < > NH4N03
Reaction 41
and with aqueous-phase H2S04 to form particulate and aqueous-phase NH4HSO4 and (NFL02SO4.
NH, (g) + H2S04(I) > NH4HS04(s,l)
Reaction 42
NH3 (g) + NH4HS04(l) > (NH4)2S04(s,l)
Reaction 43
These products are of special note because submicron NH4HSO4 and (NH^SC^ can act as cloud
condensation nuclei, but the H2S04 + H20 system does not readily undergo nucleation without addition of
NH4+ (Coffman and Hegg, 1995; Kulmala et al., 2007). These two product species are also of interest
because Reaction 42 and Reaction 43 are not reversible under typical ambient conditions, while
Reaction 41, resulting in creation ofNH4N03, is reversible. The NH4N03 is in a thermodynamic
equilibrium with NH/ and HN03 in the gas phase such that lower temperatures shift the equilibrium
toward greater production of NH4N03. Higher RH also shifts the equilibrium toward liquid-phase
NH4N03. For these reasons, and because gas-phase NH3 will neutralize S042 preferentially first,
NH4N03 can only form when an excess of gas-phase NH3 first exists.
Along with HN03 and H2S04, NH3 can be limiting in the formation of secondary atmospheric
particles containing N03 and S042 . Measurements and thermodynamic models of free and condensed-
phase precursors have been used to predict the limiting reactant under different atmospheric conditions
(Blanchard et al., 2000; Dennis et al., 2001; Watson et al., 1994). Figure 2-21 shows results from one
application of this technique from Blanchard et al. (2000) with isopleths of N03 concentrations as
predicted from total HN03 and NH3 with 25 |_ig/m3 S042 and 2 |_ig/m3 total CI". Formation of pN03 is
limited by total NH3 availability, but not HN03, where isolines are vertical. N03 exists predominately in
2-39
-------
the condensed phase where isolines are horizontal and formation is not limited by NH3 there. These
relationships have been confirmed in field measurements like those reported for NCh aloft in Arnold and
Luke (2007).
0 10 20 30 40
Total Ammonia Concentration, ng/m3
Source: Adapted from Bl an chard et al. (2000).
Figure 2-21. Predicted isolines of particulate NO3 concentrations (pg/m3) as a function of total HNO3 and
NH3 at 293 K and 80% relative humidity, and with 25 pg/m3 SO42" and 2 pg/m3 total Ch.
CO
6"
D)
=1
aT
(3
5-
o
CO
o
CD
< 3
NH3(g) + HN03(g) « NH4NO
efflorescence
branch
V
» I I t >—4
x A deliquescence
2^
L
branch
I • I ¦ I ¦
35 40 45 50 55 60 65 70 75
Relative Humidity (%)
Source: Pandis (2004) in McMurry et al. (2004)
Figure 2-22. Predicted particulate N03 concentration as a function of RH for a typical environment. Actual
measured values depend on aging characteristics of the particle.
2-40
-------
Changing RH and particle water content also change the partitioning between gas and condensed phases
for semivolatile species like N03~; see Figure 2-22 for the example ofNH4N03.
The phase partitioning of NH3 with deliquesced aerosol solutions is controlled primarily by the
thermodynamic properties of the system
NH3(g)^^[NH3(aq)]^^[NH4+ ] + Kw/[H+]
Reaction 44
where KH and Kb are the temperature-dependent Henry's Law and dissociation constants 62 M/atm, 1.8 x
1CT5 M, respectively, for NH3; and Kw is the ion product of water, 1.0 x 10 14 M (Chameides, 1984). For a
given NHX concentration, increasing aqueous concentrations of particulate H will shift the partitioning of
NH3 towards the condensed phase. Consequently, under the more polluted conditions characterized by
higher pS04 concentration, ratios of gas-phase NH3 to particulate NH/ decrease (Smith et al., 2007). It
also follows that in marine air, where aerosol acidity varies substantially as a function of particle size,
NH3 partitions preferentially to the more acidic submicron size fractions (Keene et al., 2004; Smith et al.,
2007).
Because the dry Vd of gas-phase NH3 to the surface is substantially greater than that for the
submicron pS04 fractions with which most NH/ is associated, dry deposition fluxes of total NH3 are
dominated by the gas-phase fraction (Russell et al., 2003; Smith et al., 2007). Consequently, partitioning
with acidic pS04 effectively increases the x of total NH3 against dry deposition.
This shift has important consequences for NH3 cycling in the atmosphere and potential ecological
effects. In coastal New England during summer, air transported from rural eastern Canada contains
relatively low concentrations of nss pS04 and total NH3 (Smith et al., 2007). Under these conditions, the
roughly equal partitioning of total NH3 between gas and particulate phases sustains substantial dry
deposition fluxes to the coastal ocean of total NH3 with a median value of 10.7 |_iM/m2/day. In contrast,
heavily polluted air transported from the industrialized midwestern U.S. contains median concentrations
of nss pS04 and total NH3, which are a factor of ~3 or greater more than that median value for clean air.
Under these conditions, >85% of the total NH3 partitions to the acidic pS04 size fractions, and,
consequently, the median dry-deposition flux of total NH3 is 30% lower than that under the cleaner,
northerly flow regime. The relatively longer atmospheric x of total NH3 against dry deposition under more
polluted conditions implies that, on average, total NH3 would accumulate to higher atmospheric
concentrations under these conditions and also be subject to atmospheric transport over longer distances.
Consequently, the importance of NHX removal via wet deposition would also increase. Because of the
inherently sporadic character of precipitation, greater heterogeneity may exist in NH3 deposition fields
leading to similar heterogeneity in any potential responses in sensitive ecosystems downwind of major S
emissions regions.
2.6.5. Transport-Related Effects on Chemistry
Convective processes and small-scale turbulence transport pollutants both upward and downward
throughout the PBL and the FT. NOx, SOx, and NHX can be transported vertically by convection into the
upper part of the mixed layer on one day, then transported overnight as a layer of elevated mixing ratios,
sometimes by a nocturnal low-level jet, and then entrained into a growing convective boundary layer
downwind and brought back to the surface.
Because NO and N02 are only slightly soluble, they can be transported over longer distances in the
gas phase than can more soluble species which are depleted by deposition to moist surfaces or taken up
more readily on aqueous surfaces of particles. During transport, emitted N species can be transformed
into reservoir species such as HN03, PANs, and N205. These species can then contribute to local NOx
concentrations in remote areas. For example, it is now well established that PAN decomposition provides
2-41
-------
a major source of NOx in the remote troposphere (Staudt et al., 2003). PAN decomposition in subsiding
air masses from Asia over the eastern Pacific could make an important contribution to 03 and NOx
enhancement in the U.S. (Hudman et al., 2004; Kotchenruther et al., 2001). Further details about
mechanisms for transporting 03 and its precursors were described at length in 2006 03 AQCD (U.S. EPA,
2006b).
Major episodes of high pollution concentrations in the eastern U.S. and in Europe are often
associated with slow-moving high-pressure systems. High-pressure systems during the warmer seasons
are associated with subsidence, resulting in warm, generally cloudless conditions with light winds. The
subsidence results in stable conditions near the surface, which inhibit or reduce the vertical mixing of
primary and secondary pollutants. Photochemical activity is enhanced under these conditions because of
higher temperatures and the availability of sunlight.
Crutzen and Gidel (1983), Gidel (1983), and Chatfield and Crutzen (1984) hypothesized that
convective clouds played an important role in rapid atmospheric vertical transport of trace species and
first tested simple parameterizations of convective transport in atmospheric chemical models. At nearly
the same time, evidence was shown of venting the boundary layer by shallow, fair weather cumulus
clouds (see, e.g., Greenhut et al., 1984; Greenhut, 1986). Field experiments were conducted in 1985,
which resulted in verification of the hypothesis that deep convective clouds are instrumental in
atmospheric transport of trace constituents (Dickerson et al., 1987). Once pollutants are lofted to the
middle and upper troposphere, they typically have a much longer chemical x and, with the generally
stronger winds at these altitudes, can be transported long distances from their source regions. Transport of
NOx from the boundary layer to the upper troposphere by convection tends to dilute concentrations and
extend the NOx x from <24 h to several days. Photochemical reactions occur during this long-range
transport. Pickering et al. (1990) demonstrated that venting of boundary layer NOx by convective clouds
(shallow and deep) causes enhanced P(03) in the FT. NOx at the surface can often increase P(03)
efficiency. Therefore, convection aids in the transformation of local pollution into a contribution to global
atmospheric pollution. Downdrafts within thunderstorms tend to bring air with less NOx from the middle
troposphere into the boundary layer. Lightning produces NO which is directly injected chiefly into the
middle and upper troposphere. The total global production of NO by lightning remains uncertain, but is
on the order of 10% of the total.
The first unequivocal observations of deep convective transport of boundary layer pollutants to the
upper troposphere were documented by Dickerson et al. (1987). Instrumentation aboard three research
aircraft measured CO, 03, NO, NOx, NOY, and HCs in the vicinity of an active mesoscale convective
system near the border of Oklahoma and Arkansas during the 1985 PRE-STORM experiment. Anvil
penetrations about two hours after cloud maturity found greatly enhanced mixing ratios inside the cloud
of all of the aforementioned species compared with outside it. NO mixing ratios in the anvil averaged 3 to
4 ppb, with individual 3-min observations reaching 6 ppb; boundary layer NOx was typically 1.5 ppb or
less outside the cloud. Therefore, the anvil observations represent a mixture of boundary layer NOx and
NOx contributed by lightning. Luke et al. (1992) summarized the air chemistry data from all 18 flights
during PRE-STORM by categorizing each case according to synoptic flow patterns. Storms in the
maritime tropical flow regime transported large amounts of CO, 03, and NOY into the upper troposphere
while the midtroposphere remained relatively clean. During frontal passages a combination of stratiform
and convective clouds mixed pollutants more uniformly into the middle and upper levels.
Thunderstorm clouds are optically very thick and, therefore, have major effects on radiative fluxes
and photolysis rates. Madronich (1987) provided modeling estimates of the effects of clouds of various
optical depths on photolysis rates. In the upper portion of a thunderstorm anvil, photolysis is likely to be
enhanced by a factor of 2 or more due to multiple reflections off the ice crystals. In the lower portion and
beneath the cloud, photolysis is substantially decreased. With enhanced photolysis rates, the NO-to-N02
ratio in the upper troposphere is driven to larger values than under clear-sky conditions.
Thunderstorm updraft regions, which contain copious amounts of water, are regions where efficient
scavenging of soluble species can occur (Balkanski et al., 1993). N02 itself is not very soluble; hence, wet
scavenging is not a major removal process for it. However, a major NOx reservoir species, HN03, is
2-42
-------
highly soluble. Very few direct field measurements of the amounts of specific trace gases, including
HN03, that are scavenged in storms are available. Pickering et al. (2001) used a combination of model
estimates of soluble species that did not include wet scavenging and observations of these species from
the upper tropospheric outflow region of a major line of convection observed near Fiji. Over 90% of the
NOx in the outflow air appeared to have been removed by the storm.
2.7. Sampling and Analysis Techniques
2.7.1. Methods for Relevant Gas-Phase N Species
Separate sections here on field-deployed measurement techniques focus on current methods and
promising new technologies so no attempt is made to cover development of these methods or methods no
longer in widespread use. Rather, the descriptions in this chapter concern chiefly the Federal Reference
Methods and Federal Equivalent Methods (FRM and FEM, respectively). More detailed discussions of
the FRM, FEM, and other, newer methods including issues about their field use are found in Clemitshaw
(2004); Edgerton et al. (2006, 2007); McClenny (2000); Parrish and Fehsenfeld (2000); and U.S. EPA
(1996a, 2008a, 2008b).
2.7.1.1. NO and N02
Chemiluminescence
NO can be measured reliably using the principle of gas-phase chemiluminescence (CL) induced by
the reaction of NO with 03 at low pressure. Modern commercial NOx analyzers have sufficient sensitivity
for adequate measurement in urban and many rural locations (U.S. EPA, 1993a, 2006b). Research grade
CL instruments have been compared under realistic field conditions to spectroscopic instruments with
results that indicate that both methods are reliable at concentrations relevant to smog studies to better than
15% when using a 95% confidence response time of 60 seconds (see Crosley, 1996). Near-source, urban,
and rural and remote concentrations of NO are routinely measured using CL. However, Cardelino and
Chameides (2000) reported that measured NO concentrations during the afternoon were frequently at or
below the operational limit of detection (LOD), ~1 ppb, of the regulatory NOx instruments even in large
metropolitan regions such as Washington, DC, Houston, TX, and New York, NY.
The FRM for N02 also makes use of this NO detection technique using a prerequisite step to
reduce N02 to NO on the surface of a molybdenum oxide (MoOx) substrate heated to -340 °C. Because
the FRM monitor cannot detect N02 directly, the N02 level is computed as the difference between the
sample passed over the heated MoOx substrate (the NOx total) and the sample not so reduced (the NO
alone). Reduction of N02 to NO on the MoOx substrate is not specific to N02, however; hence, the CL
analyzers are subject to unknown and varying levels of interferences produced by the presence in the
sample of other oxidized N compounds, the NOz species shown in the outer box of Figure 2-16. This
interference is often termed a "positive artifact" in the N02 concentration estimate since the presence of
NOz always results in an over-estimate in the reported measurement of the actual N02 concentration.
This interference by NOz compounds has long been known (see Crosley, 1996; Fehsenfeld et al.,
1987; McClenny et al., 2002; Nunnermacker et al., 1998; Parrish and Fehsenfeld, 2000; Rodgers and
Davis, 1989; Steinbacher et al., 2007; U.S. EPA, 2008a). These studies have relied on intercomparisons of
measurements using the FRM and other techniques for measuring N02. The sensitivity of the FRM to
potential interference by individual NOz compounds is variable and depends in part on characteristics of
2-43
-------
individual monitors, such as design of the instrument inlet, the temperature and composition of the
reducing substrate, and the interactions of atmospheric species with the reducing substrate.
Only recently have attempts been made to quantify systematically the magnitude and variability of
the interference by NOz species in ambient measurements of N02. Dunlea et al. (2007) found an average
of -22% of ambient N02 (~9 to 50 ppb) measured in Mexico City was due to interference from NOz
compounds; that is to say, the true ambient N02 concentration was -22% lower than what was reported at
monitors using the CL difference technique. Although comparable levels of N02 are found in many
locations in the U.S., the same comparison for distinct places in the U.S. is difficult to make because
significant uncertainty remains in determining the concentrations of the higher oxidation NOz products
since they are not routinely measured. Dunlea et al. (2007) compared N02 measured using the
conventional CL instrument with optical techniques. The main sources of interference were HN03 and
various RON02 which can be converted to NO on the catalyst with varying rates of efficiency. In this
study, the efficiency of conversion on the catalyst — that is, how much of the compound introduced to the
catalyst was converted to NO — was estimated to be -38% for HN03; -95% for PAN, and -95% for
other RON02. Peak interference (over-estimation) in the reported estimate of N02 concentrations from
the presence of NOz compounds of up to 50% was found during afternoon hours and was associated with
03 and NOz compounds such as HN03, and with multifunctional alkyl nitrates.
In a study in rural Switzerland, Steinbacher et al. (2007) compared measurements of N02
continuously measured using a conventional CL-NOx monitor and measurements in which N02 was
photolyzed to NO. They found that the conventional CL technique overestimated the reported N02
concentration, on average, by 10% during winter and 50% during summer.
Another approach to estimating the N02 measurement interference uses model calculations in
conjunction with known data on the reduction efficiencies of NOz species on the MoOx converters.
Lamsal et al. (2008) used the conversion efficiencies noted above along with output for NOY species from
the GEOS-Chem CTM to derive seasonal correction factors for the ambient monitoring data across the
U.S. These factors range from <10% in winter in the East to >80% in the West, with the highest values
found during summer in relatively unpopulated areas. Lamsal et al. (2008) also used these corrected data
to determine the feasibility of using satellite data to supplement the ground-based data. However, the
current generation of satellite monitors are in low earth orbit and so the N02 values are restricted to time
of satellite overpass in early afternoon local time.
Calculations using CMAQ for the mid-Atlantic region in a domain extending from Virginia to
southern New Jersey (see http://www.mde.state.md.us /Programs/AirPrograms/air_planning/index.asp)
were made at much higher spatial resolution than the GEOS-Chem simulations by Lamsal et al. (2008).
The daily average interference for an episode during the summer of 2002 estimated using model-derived
concentration fields for NOz species and using the conversion efficiencies for NOz species given above,
ranged from -20% in Baltimore to -80% in Madison, VA. Highest values were found during the
afternoon, when photochemical activity is highest and production and accumulation of the higher
oxidized NOz compounds is greatest; lowest values occurred during the middle of the night when
photochemistry stops. The model calculations showed episode averages of the NOz-to-N02 ratio ranging
from 0.26 to 3.6 in rural Virginia, with the highest ratios in rural areas and lowest in urban centers closer
to sources of fresh NOx emissions. (The capabilities of three-dimensional CTMs such as GEOS-Chem
and CMAQ and issues associated with their use are presented in Section 2.8 below.)
In summary, the current CL method of determining ambient NOx and then reporting N02
concentrations by subtraction of NO is subject to positive interference by N02 oxidation products, chiefly
HN03 and PAN as well as other oxidized N-containing compounds, though the magnitude of this positive
bias is largely unknown and can be rapidly changing. Measurements of these higher-order oxidation
products in urban areas are sparse. Concentrations of these oxidation products are expected to peak in the
afternoon because of the continued oxidation of N02 from NO emitted during the morning rush hours and
during conditions conducive to photochemistry in areas well downwind of sources, particularly during
summer.
2-44
-------
Within the urban core of metropolitan areas, where many of the ambient monitors are sited close to
strong NOx sources such as motor vehicles on busy streets and highways, these positive artifacts due to
N02 oxidation products are much smaller than outside the urban core farther downwind, and are
typically <10%. Conversely, the positive artifacts are larger in locations more distant from NOx sources
where N02 concentrations are lowest and can exceed 50% as noted above. Therefore, variable, positive
artifacts associated with measuring N02 using the FRM severely hamper its ability to serve as an accurate
and precise indicator of N02 concentrations at the typical ambient levels generally encountered in remote
areas outside of urban cores where they would be most relevant for the environmental exposures of
concern to a secondary NAAQS.
Other Methods
NO has also been successfully measured in ambient air with direct spectroscopic methods; these
include two-photon laser-induced fluorescence (TPLIF), tunable diode laser absorption spectroscopy
(TDLAS), and two-tone frequency-modulated spectroscopy (TTFMS). These were reviewed thoroughly
in the 2008 NOx ISA (U.S. EPA 2008a). The spectroscopic methods demonstrate excellent sensitivity and
selectivity for NO with detection limits on the order of 10 ppt for integration times of 1 min and
spectroscopic methods compare well with the CL method for NO in controlled laboratory air, ambient air,
and heavily polluted air (see Gregory et al., 1990; Kireev et al., 1999; Walega et al., 1984). However,
these spectroscopic methods remain research grade rather than field deployable due to their complexity,
size, and cost, but are essential for demonstrating that CL methods are reliable.
There are approaches to measuring N02 not affected by the artifacts mentioned above. For
example, N02 can be photolytically reduced to NO with an efficiency of -70% as used in Steinbacher
et al. (2007). Ryerson et al. (2000) developed a gas-phase CL method using a photolytic converter based
on a Hg lamp with increased radiant intensity in the region of peak N02 photolysis, 350 to 400 nm, and
producing conversion efficiencies of 70% or more in less than 1 second. Metal halide lamps with
conversion efficiency of -50% and accuracy on the order of 20% have been used (Nakamura et al., 2003).
Because the converter produces little radiation at wavelengths less than 350 nm, interferences from HN03
and PAN are minimal. This method requires additional development to ensure its cost effectiveness and
reliability for extensive field deployment.
Optical methods such as those using differential optical absorption spectroscopy (DOAS) or laser-
induced fluorescence (LIF) are also available. However, these particular methods are more expensive than
either the FRM monitors or photolytic reduction technique and require specialized expertise to operate.
Moreover, the DOAS obtains an area-integrated measurement rather than a point measurement. Cavity
attenuated phase shift (CAPS) monitors are an alternative optical approach potentially less costly than
DOAS or LIF (Kebabian et al., 2007). However, this technique is not highly specific to N02 and is
subject to interference by other species absorbing at 440 nm, such as the 1,2-dicarbonyl compounds. The
extent of this interference and the potential of the CAPS technique for extensive field deployment have
not been evaluated.
A DOAS system manufactured by OPSIS is designated as an FEM for measuring N02. DOAS
systems can also be configured to measure NO, HN02, and N03. Typical detection limits are 0.2 to 0.3
ppb for NO, 0.05 to 0.1 ppb forN02, 0.05 to 0.1 ppb for HN02, and 0.001 to 0.002 ppb forN03, at path
lengths of 0.2, 5, 5, and 10 km, respectively. The obvious advantage compared to fixed-point
measurements is that concentrations relevant to a much larger area are obtained, especially if multiple
targets are used. At the same time, any microenvironmental artifacts are minimized over the long path
integration. However, comparisons to other measurements made at point locations are difficult to
interpret. A major limitation in this technique had involved inadequate knowledge of absorption cross
sections. Harder et al. (1997) conducted an experiment in rural Colorado involving simultaneous
measurements of N02 by DOAS and by photolysis followed by CL. They found differences of as much as
110% in clean air from the west; but for N02 mixing ratios in excess of 300 ppt, the two methods agreed
to better than 10%. Stutz et al. (2000) cites two intercomparisons of note. NO was measured by DOAS,
2-45
-------
by photolysis of N02 followed by CL, and by LIF during July 1999 as part of the SOS in Nashville, TN.
On average, the three methods agreed to within 2%, with some larger differences likely caused by spatial
variability over the DOAS path. In another study in Europe, a multi-reflection set-up over a 15 km path
negated the problem of spatial averaging. Here agreement with the CL detector following photolytic
conversion was excellent (slope = 1.006 ± 0.005; intercept = 0.036 ± 0.019; r2 = 0.99) over a
concentration range from about 0.2 to 20 ppb.
Remote Sensing of NO2
The paucity of specific in situ N02 measurements motivates the inference of ground-level N02
concentrations from satellite measurements of tropospheric N02 columns. This prospect would take
advantage of the greater sensitivity of tropospheric N02 columns to NOx in the lower troposphere than in
the upper troposphere as in Section 2.2 above. Tropospheric N02 columns show a strong correlation with
in situ N02 measurements in northern Italy (Ordonez et al., 2006). Quantitative calculation of surface
N02 concentrations from a tropospheric N02 column requires information on the relative vertical profile.
Comparison of vertical profiles of N02 in GEOS-Chem versus in situ measurements over and downwind
of North America showed a high degree of consistency (Martin et al., 2004, 2006), suggesting that CTMs
could be used to infer the relationship between surface N02 concentrations and satellite observations of
the tropospheric N02 column.
Table 2-6 contains an overview of the three satellite instruments used to retrieve tropospheric N02
columns from measurements of solar backscatter. All three instruments are in polar sun-synchronous
orbits with global measurements in the late morning and early afternoon local time. The spatial resolution
of the measurement from SCIAMACHY is 7 times better than that from the GOME; and that from Ozone
Monitoring Instrument (OMI) is 40 times better than that from GOME.
Table 2-6. Satellite instruments used to retrieve tropospheric NO2 columns.
Instrument
Coverage
Typical U.S.
Measurement Time
Typical Resolution
(km)
Return Time
(days)1
Instrument
Overview
GOME
1995-2002
10:30-11:30 AM
320 x 40
3
Burrows etal. (1999)
SCIAMACHY
2002-
10:00-11:00 AM
30x60
6
Bovensmann et al.
(1999)
OMI
2004-
12:45-1:45 PM
13x24
1
Leveltetal. (2006)
1 Return time is reported here for cloud-free conditions. Note that due to precession of the satellite's orbit, return measurements are close to but not made over the same location. In
practice, clouds decrease observation frequency by a factor of 2.
Figure 2-23 shows tropospheric N02 columns retrieved from SCIAMACHY. Pronounced
enhancements are evident over major urban and industrial areas. The high degree of spatial heterogeneity
over the southwestern U.S. provides empirical evidence that most of the tropospheric N02 column is
concentrated in the lower troposphere. Tropospheric N02 columns are more sensitive to NOx in the lower
troposphere than in the upper troposphere (Martin et al., 2002). This sensitivity to NOx in the lower
troposphere is due to 25-fold decrease in the N02-to-NO ratio from the surface to the upper troposphere
(Bradshaw et al., 1999), driven by the temperature dependence of the NO + 03 reaction. Martin et al.
(2004) integrated in situ airborne measurements of N02 and found that during summer the lower mixed
layer contains 75% of the tropospheric N02 column over Houston and Nashville. However, it should be
noted that these satellite measurements are also sensitive to surface albedo and aerosol loading which
strongly vary locally.
2-46
-------
31 23456789 1(
SCIAMACHY Tropospheric NC>2 (10 15 molec cm "2)
Source: Martin et al. (2006). Reprinted with permission.
Figure 2-23. Tropospheric N02 column estimates (molecules NCh/cm2) retrieved from the SCIAMACHY
satellite instrument for 2004-2005.
2.7.1.2. Total NOY
Gold-catalyzed CO or H2 reduction, or conversion on heated MoOx. have been used for many years
to reduce total NOY to NO before detection by CL (Crosley, 1996; Fehsenfeld et al., 1987). Both
techniques offer generally reliable measurements, with response times on the order of 60 seconds and a
linear dynamic range demonstrated in field intercomparisons from -10 ppt to 10s of ppb. Under some
conditions, hydrogen cyanide (HCN), NH3, alkyl nitrates (RN02), and acetonitrile (CH3CN) can be
converted to NO; but at normal atmospheric concentrations and RH, and when converter temperature is
closely monitored, these are minor interferants. Thermal decomposition followed by LIF has also been
used for NO detection. In field comparisons, instruments based on these two principles generally showed
good agreement (Day et al., 2002) with experimental uncertainty estimated to be on the order of 15 to
30%.
Commercially available NOx monitors have been converted to NO monitors by moving the MoOx
converter to couple directly to the sample inlet. Because of losses on inlet surfaces and differences in the
efficiency of reduction of NOz compounds on the heated MoOx substrate, NOx measured by CL in
standard mode cannot be considered a surrogate for NO .. However, in settings close to relatively high-
concentration fresh emissions like those during urban rush hour, most of the NOy is present as NO<.
Reliable measurements of NO and N02. especially at the low concentrations observed in many areas
remote from sources remain crucial for evaluating the performance of CTMs; see Section 2-8 for
additional details on CTM application and evaluation.
2.7.1.3. Nitric Acid
A major issue to be considered when measuring NOx and NOy is the possibility that HN03, a
major component of NOY, is sometimes lost in inlet tubes and not measured (Luke et al., 1998; Parrish
and Fehsenfeld, 2000). This problem is especially important if measured NOY is used to identify N Ox-
limited versus NOx-saturated conditions. The problem is substantially alleviated, although not necessarily
completely solved, by using much shorter inlets on NOY monitors than on standard NOx monitors and by
2-47
-------
the use of surfaces less likely to take up HN03. The correlation between 03 and NOY differs for NOx-
limited versus NOx-saturated locations, but this difference is driven primarily by differences in the ratio
of 03 to HNO3. If HNO3 were omitted from the NOY measurements, then the measurements would
represent a severely biased estimate.
Accurate measurement of HN03 has presented a long-standing analytical challenge. To understand
why, it is useful to consider the major factors that control HN03 partitioning between the gas and
deliquesced-particulate phases in ambient air as depicted in Reaction 45.
HN03(g) < K" > [HN03(aq)\ < K" > [// ] + [NO, ]
Reaction 45
where Ku is the Henry's Law constant in M/atm and Ka is the acid dissociation constant in M. Thus, the
primary controls on HN03 phase partitioning are its thermodynamic properties, aerosol liquid water
content (LWC), solution pH, and reaction kinetics. Aerosol LWC and pH are controlled by the relative
mix of acids and bases in the system, the hygroscopic properties of condensed compounds, and
meteorological conditions, chiefly RH, temperature, and pressure.
In the presence of chemically distinct aerosols of varying acidities (e.g., supermicron,
predominantly sea salt, and submicron, predominantly pS04), HN03 should partition preferentially to the
less-acidic particles, and observations are consistent with this (see. e.g., Huebert et al., 1996; Keene et al.,
1998). The kinetics of this phase partitioning are controlled by atmospheric concentrations of HN03 vapor
and pN03, and the size distribution and x of the particles against deposition. Submicron diameter aerosols
typically equilibrate with the gas phase in seconds to minutes while supermicron aerosols require hours to
a day or more; see, e.g., Erickson et al. (1999) and Meng and Seinfeld (1996). Consequently, smaller
aerosol size fractions are typically close to thermodynamic equilibrium with respect to HN03, whereas
larger size fractions, for which x against deposition range from hours to a few days, are often
undersaturated (Erickson et al., 1999; Keene et al., 1998).
Methods used widely for measuring HN03 include standard filter packs configured with nylon or
alkaline-impregnated filters (see, e.g., Bardwell et al., 1990; Goldan et al., 1983), annular denuders like
U.S. EPA method IP-9, and mist chambers (Talbot et al., 1990), all typically joined to ion chromatography
detection. Intercomparisons of these measurement techniques by Hering et al. (1988), Tanner et al.
(1989), and Talbot et al. (1990) reported differences on the order of a factor of 2 or more. In part, this
variance is due to nonsystematic sampling error. When chemically distinct aerosols with different pH—
for example, sea salt and pS04—mix together on a bulk filter, the acidity of the bulk mixture will be
greater than that of the less-acidic aerosols with which most of the N03 is associated. This change in pH
may cause the bulk mix to be supersaturated with respect to HN03 leading to volatilization and, thus, to a
positive measurement bias in HN03 sampled downstream. Alternatively, when undersaturated
supermicron size fractions like sea salt accumulate on a bulk filter and chemically interact over time with
HNO3 in the sample air stream, scavenging may lead to a negative bias in the HN03 sampled
downstream. Because the magnitude of both effects will vary as functions of the overall composition and
thermodynamic state of the multiphase systems, the combined influence can cause net positive or net
negative measurement bias in data with unknown frequencies. Pressure drops across particle filters can
also lead to artifact volatilization and associated positive bias in HN03 concentrations measured
downstream.
Sensitive HN03 measurements based on the principle of chemical ionization mass spectroscopy
(CIMS) have become more common; see Huey et al. (1998), Furutani and Akimoto (2002), Mauldin et al.
(1998), and Neuman et al. (2002). The CIMS relies on selective formation of ions such as silicon
pentafluoride-nitric acid (Si Fs • HN03) or (HS04 • HN03) followed by detection via mass spectroscopy.
Two CIMS techniques and a filter pack technique were intercompared in Boulder, CO (Fehsenfeld et al.,
1998) where results indicated agreement to within 15% between the two CIMS instruments and between
the CIMS and filterpack methods under relatively clean conditions of 50-400 ppt HN03. In more polluted
2-48
-------
air, the filterpack technique generally yielded higher values than the CIMS, suggesting that interactions
between chemically distinct particles on bulk filters is a more important source of bias in polluted
continental air. Differences were also greater at lower temperature when pN03 accounted for a relatively
greater fraction of total N03 .
Three semi-continuous methods for detecting HN03 were tested against the annular denuder
filterpack (ADS) integrated collection technique at the Tampa Bay Regional Atmospheric Chemistry
Experiment (BRACE) Sydney research station -20 km downwind of the Tampa, FL, urban core (Arnold
and Luke, 2007). The semi-continuous instruments included: two slightly differing implementations of
the NOy-NOy* (total oxides of nitrogen minus that total denuded of HN03) denuder difference technique,
one from the NOAA Air Resources Lab (ARL), and one from Atmospheric Research and Analysis, Inc.
(ARA); the parallel plate wet diffusion scrubber online ion chromatography technique from Texas Tech
University (TTU); and the CIMS from the Georgia Institute of Technology (GIT). Integrated twelve hour
ADS samples were collected by the University of South Florida (USF). Results for 10 min samples
computed from the various higher sampling frequencies of each semi-continuous instrument showed good
agreement (R2 >0.7) for afternoon periods of the highest production and accumulation of HN03. Further,
agreement was within ± 30% for these instruments even at HN03 concentrations <0.30 ppb. The USF
ADS results were biased low, however, by 44%, on average, compared to the corporate 12-h aggregated
means from the semi-continuous methods, and by >600% for the nighttime samples; ADS results were
below the corporate mean maximum HN03 concentration by >30% as well. The four instruments using
semi-continuous methods, by contrast, were all within 10% of each other's 12-h mean mixing ratios.
While only ARA employed a formal minimum detection limit at 0.050 ppb, error analysis with the other
techniques established that at the same level of precision, TTU's effective limit was approximately the
same as ARA's, and that ARL's limit was 0.030 ppb; analysis for GIT showed no apparent effective limit
at the levels of HN03 encountered in this field study. The importance of sample inlet height for HN03
measurements was indirectly shown through comparison to previous fieldwork at this site when sample
inlet heights ranged from 1.5-10 m and produced systematic discrepancies in HN03 concentrations
correlated with height of more than a factor of 2.
2.7.1.4. Other Nitrates
Methods for sampling and analysis of RON02 in the atmosphere have been reviewed by Parrish
and Fehsenfeld (2000). PAN, PPN, and MPAN were typically measured using a gas chromatograph
followed by electron capture detectors (GC-ECD); see, e.g., Gaffney et al. (1998), although other
techniques such as Fourrier Transform InfraRed (FTIR) analysis can also be used. Field measurements
made using GC-ECD have reported a total uncertainty of ± 5 ppt + 15% (Gaffney et al., 1998; Roberts et
al., 1998). Additional descriptions of specific techniques for RON02 and some of the issues involved with
using data taken with them appear below in Section 2.10, accompanying descriptions of the methods used
routinely to monitor ambient air concentrations and deposition amounts of RON02, and in the 2008 NOx
ISA (U.S. EPA, 2008a).
2.7.1.5. Ammonia
Because NH3 plays a key role in the atmospheric chemistry of particle formation, several methods
have been developed for ambient and higher-level concentrations; see Allegrini et al. (1992); Appel et al.
(1988); Asman et al. (1998); Fehsenfeld et al. (1996); Genfa and Dasgupta (1989); Mennen et al. (1996);
Pryor et al. (2001); Schwab et al. (2007); Williams et al. (1992); and Wyers et al. (1993). Measurement of
NH3 is made difficult by its chemistry, whereby it forms strong H bonds with itself and water, and can be
lost either with hysteresis or irreversibly to many instrument surfaces. Moreover, the range of atmospheric
2-49
-------
NH3 concentrations extends over 4 or 5 orders of magnitude. Because of these challenges, many NH3
techniques remain research-grade with steep requirements of time, care, and technical experience.
U.S. EPA has proposed to include ambient NH3 measurements in its National Core (NCore)
monitoring network (U.S. EPA, 2005b) and this has motivated additional development and testing of NH3
monitors. The U.S. EPA Environmental Technology Verification (ETV) Program's Advanced Monitoring
Systems (AMS) Center, has verified the performance of seven ambient NH3 monitors for use at confined
animal feeding operations (CAFOs). In collaboration with the U.S. Department of Agriculture (USDA),
the AMS Center verified the seven ambient NH3 monitors (see Table 2-7) in two phases of testing, each at
separate CAFOs: Phase I was conducted at a swine finishing farm and Phase II at a cattle feedlot. These
sites were selected to provide realistic testing conditions and a wide range of NH3 concentrations.
Table 2-8 summarizes some of the performance data for the individual technologies. (The full verification
reports can be found at http://www.epa. gov/nrmrl/std/etv/vt-ams .html under the ambient NH3 sensors
category.)
Ambient NH3 monitors utilize a wide range of analytical methods, including direct detection by
spectroscopic techniques; indirect detection of NH3 using selective membrane permeation with
conductivity detection, catalytic conversion with CL detection; treatment with a chemical dopant followed
by ion mobility detection; and other techniques. Ambient NH3 monitors also can provide specialized
features valuable in specific uses, such as long-term monitoring or determining NH3 fluxes and emission
rates.
For example, monitors that collect NH3 concentration data with a frequency greater than 1 Hz can
be used with simultaneous three-dimensional windspeed and direction data to determine NH3 flux.
Alternatively, open-path monitors can be used to calculate emission rates from CAFOs, since these
monitors measure the average NH3 concentration over a 1 to 100 m path. Some monitors also are suitable
for long-term monitoring, since they can be operated without user intervention for weeks at a time.
Table 2-7. Verified ambient NH3 monitors.
Technology Name
Description
Aerodyne Research, Inc. QC-TILDAS
An infrared laser spectrometer, based on pulsed quantum cascade laser technology; continuous
measurement
Bruker Daltonics OPAG 22 Open-Path
A broadband, open-path, Fourier transform infrared spectrometer for remote sensing continuous
Gas Analyzer
measurement
Molecular Analytics lonPro-IMSNH3
An ion mobility spectrometer; continuous measurement
Analyzer
Omnisens SATGA310 NH3 Analyzer
A trace gas analyzer that uses photoacoustic spectrometry; continuous measurement
Pranalytica, Inc. Nitrolux™ 1000
A resonant photoacoustic spectrometer with a line-tunable CO2 laser; continuous measurement
Ambient NH3 Analyzer
Mechatronics Instruments BV
A single-point monitor composed of a membrane diffusion sampler, a detector block with a diffusion
AiRRmonia NH3 Analyzer
membrane, and two conductivity cells; continuous measurement
Thermo Electron Corp. Model 17C
ACL analyzer that uses NO and ozone (O3) reactions; time-averaged measurement
NH3 Analyzer
2-50
-------
Table 2-8.
Performance characteristics of the 7 U.S. EPA ETV tested NH3 methods.
Vendor3
Testing
Average
Relative
Accuracyb
Relative
Response
Time
(95%)
Linearity
Comparability
'C
Precisionb
Slope
Intercept
r2
Slope
Intercept
r2
A
Phase I &
Phase II
3.7 to
10.5%
0.3%
3 to 76 min
0.90 to
1.03
-24 to -0.6
1.000
0.86 to
1.20
-0.5 to 16
0.984 to
0.990
B
Phase I &
Phase II
2.4 to 34%
0.7 to 2.1%
8 to 20 min
1.02 to
1.28
-2.4 to 136
0.9957 to 0.9999
0.41 to
1.18
-1.4 to 58
0.538 to
0.9755
C
Phase I &
Phase II
10 to 44%
0.2 to 1.3%
1 to 32 min
0.716 to
1.25
-58.5 to 167
0.9854 to 0.9997
0.646 to
1.83
-6.7 to 21.6
0.9794 to
0.9842
D
Phase II
2.2%
0.9%
2 to 2.6 min
0.966
15.9
1.000
1.15
-4.1
0.994
E
Phase II
18.3%
1.0%
2.5 to 17
min
0.815
1.08
1.000
1.565
-16.5
0.994
F
Phase II
26%
1.8%
4 to 14 s
0.583
24.9
0.9144
NR
NR
NR
G
Phase I &
Phase II
4.7 to 10%
1.9 to 2.5%
0.8 to 66 s
0.840 to
0.962
-8.8 to 35
0.9989 to 0.9998
0.984 to
1.09
-9.5 to 14.4
0.9943 to
0.9982
a Because the ETV Program does not compare technologies, the performance results shown in this table do not identify the vendor associated with each result and are not in the same
order as the list of technologies in Table 2-7.b A result of 0% indicates perfect accuracy or precision.c The comparability of the verified technology with a standard reference method
was established by comparing the average NH3 sensors readings with time-integrated NH3 samples collected using citric-acid-coated denuders. The reference samples were collected
based on procedures described in the U.S. EPA Compendium Method 10-4.2, Determination of Reactive Acidic and Basic Gases and Acidity of Fine Particles (<2.5 pm). Comparability
between the NH3 sensors results and the reference method results with respect to ambient air was assessed by linear regression using the reference method NH3 concentrations as the
independent variable and results from the NH3 sensor as the dependent variable.
In addition to the evaluation by U.S. EPA, ETV, a laboratory-based intercomparison of real-time
ambient NH3 instruments was conducted and reported by Schwab et al. (2007) with seven instruments
using six methods. The tunable diode laser (TDL) absorption spectrometer, the wet scrubbing long-path
absorption photometer (LOPAP), the wet effusive diffusion denuder (WEDD), the ion mobility
spectrometer (IMS), the Nitrolux laser acousto-optical absorption analyzer, and a modified CL analyzer.
Schwab et al. (2007) reported that all instruments performed well and agreed to within -25% of the
expected calibration value, with the exception of the CL analyzer which suffered from problems related to
its MoOx conversion of NOz to NO. (Work with a modification of this technique has been continuing
with the Aerosol Research Inhalation Epidemiology Study (ARIES); see, e.g., Blanchard and Hidy, 2003.)
Instrument response time is known to be a crucial feature for ambient NH3 measurements, and
Schwab et al. (2007) showed response time to be sensitive to measurement history as well as the sample
handling materials. Shortest response was for the TDL; the Nitrolux and IMS and WEDD instruments had
unacceptably long time responses under some environmental conditions which rendered correlations
across instruments meaningless. The TDL and LOPAP reported values closest to delivered concentration
values; the IMS exhibited bias of ~ +25%; the Nitrolux bias was —25%. Schwab et al. (2007) concluded
that sub-ppb ambient NH3 measurements can be taken reliably with some of these instruments, but that
special care must still be exercised to ensure high-quality data.
These and other recent intercomparisons of ambient NH3 instruments have confirmed that no single
active sampling technique has yet been identified for automated, fast-response, low-concentration, high-
quality continuous data.
Passive samplers for ambient NH3 measurements, however, have been tested for multiple periods
of 14-day exposures over several years in the small network of sites in the U.S. Upper Midwest listed in
Table 2-9 and have performed well.
2-51
-------
Table 2-9. Site codes arid locations of passive NH3 samplers in the U.S. EPA and Lake Michigan Air
Directors and Illinois State Water Survey Consortium Project.
Site Code
Location
IL11
Bondville, IL
IN99
Indianapolis, IN
M196
Allen Park, Ml
MN18
Fernberg, MN
MN29
Blue Mounds S.P., MN
MN42
Great River Bluffs S.P., MN
OH02
Athens, OH
OH27
Cincinnati, OH
WI07 Mayville, Wl
In this test, the Radiello® samplers demonstrated excellent reproducibility across the three samples
taken at each site for each 14-day exposure period. Ambient NH3 concentration from each sampler and
the average value and standard deviation and percent relative standard deviation for each of the 14-day
exposure periods was reported at every site in 2007 and 2008 (data not shown).
IL11
IN99
MI96
-*¦— MN18
¦+— MN29
-«— MN42
OH02
OH27
-*— WI07
End Date
Figure 2-24. Average ambient NH3 concentrations from the NH3 passive samplers trial network, 2007-2008.
2-52
-------
Figure 2-24 shows the average ambient NH3 concentrations from each 14-day exposure period at
each site from the Radiello® samplers. The irregularly high concentrations at Blue Mounds State Park,
MN remains unexplained, but may be due to wintertime shifts in the location of a large bison herd nearer
to the sample site.
2.7.2. Methods for Relevant Gas-Phase Sulfur Species
Currently, ambient S02 is measured using instruments based on pulsed fluorescence. The UV
fluorescence monitoring method for atmospheric S02 was developed to improve upon the flame
photometric detection (FPD) method for S02, which in turn had displaced the pararosaniline wet chemical
method. The pararosaniline method is still the FRM for atmospheric S02, but is rarely used because of its
complexity and slow response, even in its automated forms. Both the UV fluorescence and FPD methods
are designated as FEMs by the U.S. EPA, but UV fluorescence has largely supplanted the FPD approach
because of the UV method's inherent linearity, sensitivity, and the absence of consumables, such as the H2
gas needed for FPD.
The LOD for a non-trace-level S02 analyzer is 10 ppb (CFR, 2006) However, most commercial
analyzers report operational detection limits of ~3 ppb. This concentration is very near the current
ambient annual average concentration of S02 of ~4 ppb.
S02 molecules absorb UV light at one wavelength and emit UV light at longer wavelengths. This
fluorescence involves excitation of the S02 molecule to a higher energy (singlet) electronic state. Once
excited, the molecule decays non-radiatively to a lower energy electronic state from which it then decays
to the original, or ground, electronic state by emitting a photon of light at a longer wavelength (i.e., lower
energy) than the original, incident photon. The process can be summarized by the following equations
S02 + hv] > S(X
Reaction 46
SO * > S02 + hv2
Reaction 47
where S02* represents the excited state of S02, hvi, and hv2 represent the energy of the excitation and
fluorescence photons, respectively, and hv2
-------
2.7.2.2. Negative Interference
Nonradiative deactivation (quenching) of excited S02 molecules can occur from collisions with
common molecules in air, including N, 02, and H20. During collisional quenching, the excited S02
molecule transfers energy, kinetically allowing the S02 molecule to return to the original lower energy
state without emitting a photon. Collisional quenching results in a decrease in the S02 fluorescence and
results in the underestimation of S02 concentration in the air sample. The concentrations of N2 and 02 are
constant in the ambient air, so quenching from those species at a surface site is also constant, but the
water vapor content of air can vary. Luke (1997) reported that the response of the detector could be
reduced by about 7% and 15% at water vapor mixing ratios of 1 and 1.5 mole percent (RH = 35 to 50%)
at 20-25 °C and 1 atm for a modified pulsed fluorescence detector. At very high S02 concentrations,
reactions between electronically excited S02 and ground state S02 to form S03 and SO are theoretically
possible (Calvert et al., 1978), but this possibility has not been examined.
2.7.2.3. Other Methods
A more sensitive S02 measurement method than the UV-fluorescence method was reported by
Thornton et al. (2002b) using an atmospheric pressure ionization mass spectrometer. The high
measurement precision and instrument sensitivity were achieved by adding isotopically labeled S02
(34S"'02) continuously to the manifold as an internal standard. Field studies showed that the method
precision was better than 10%.
S02 can also be measured by LIF at -220 nm (Matsumi et al., 2005). Because the laser wavelength
is alternately tuned to an S02 absorption peak at 220.6 and a trough at 220.2 nm and the difference signal
at the two wavelengths is used to extract the S02 concentration, the technique eliminates interference
from either absorption or fluorescence by other species and has high sensitivity: 5 ppt in 60 s.
S02 can also be measured by the same DOAS instrument that can measure N02; see the discussion
of DOAS in Section 2.7.1.1 above.
Photo-acoustic techniques have been employed for S02 detection, but generally have detection
limits sutable only for source monitoring (Gondal, 1997; Gondal and Mastromarino, 2001).
CIMS techniques for S02 have been shown to have high sensitivity, 10 ppt or better, with
uncertainty of -15% when a charcoal scrubber is used for zeroing and the sensitivity is measured with
isotopically labeled 34S02 (Hanke et al., 2003; Hennigan et al., 2006; Huey et al., 2004).
2.7.3. Methods for Relevant Aerosol-Phase Nitrogen and Sulfur
Species
S042 is commonly present in PM2 5. Most PM2 5 samplers have a size-separation device to separate
particles so that only those particles approximately 2.5 (.im or less are collected on the sample filter. Air is
drawn through the sample filter at a controlled flow rate by a pump located downstream of the sample
filter. The systems have two critical flow rate components for the capture of fine particulates: both the
flow of air through the sampler must be at a flow rate that ensures the size cut at 2.5 |_im: and the flow rate
must be optimized to capture the desired amount of particulate loading with respect to the analytical
method detection limits.
Using the system described above to collect pS04 sampling artifacts can introduce: positive
sampling artifacts for pS04, pN03, and pNH4 due to chemical reactions, and negative sampling artifacts
for N03 and NH4+ due to decomposition and evaporation.
Several traditional and new methods could be used to quantify elemental S collected on filters:
energy dispersive X-ray fluorescence, synchrotron induced X-ray fluorescence, proton induced X-ray
2-54
-------
emission (PIXE), total reflection X-ray fluorescence (XRF), and scanning electron microscopy. Energy
dispersive X-ray fluorescence (EDXRF) (Method 10-3.3, U.S. EPA, 1997; see U.S. EPA (2004) for
details) and PIXE are the most commonly used methods. Since sample filters often contain very small
amounts of particle deposits, preference is given to methods that can accommodate small sample sizes
and which require little or no sample preparation or operator time after the samples are placed into the
analyzer. XRF meets these needs and leaves the sample intact after analysis so it can be submitted for
additional examinations by other methods as needed. To obtain the greatest efficiency and sensitivity,
XRF typically places the filters in a vacuum which may cause volatile compounds to evaporate. As a
result, species that can volatilize such as NH4NO3 and certain organic compounds can be lost during the
analysis. The effects of this volatilization are important if the PTFE filter is to be subjected to subsequent
analyses of volatile species.
Polyatomic ions such as S042 . N03 . and NH4+ are quantified by methods such as IC or automated
colorimetry for NH/. All ion analysis methods require a fraction of the filter to be extracted in deionized
distilled water for S042 and Na2C03/NaHC03 solution for N03 and then filtered to remove insoluble
residues before analysis. The extraction volume should be as small as possible to avoid over-diluting the
solution and inhibiting the detection of the desired constituents at levels typical of those found in ambient
PM25 samples.
Continuous methods for the quantification of S compounds first remove gas-phase S from the
sample stream by a diffusion tube denuder followed by analysis of pS04 (Cobourn et al., 1978; Durham
et al., 1978; Huntzicker et al., 1978; Mueller and Collins, 1980; Tanner et al., 1980). Another approach is
to measure total S and gas-phase S separately by alternately removing particles from the sample stream.
pS04 is obtained as the difference between the total and gas-phase S (Kittelson et al., 1978). The total S
content is measured by a FPD by introducing the sampling stream into a fuel-rich, H2-air flame (e.g.,
Farwell and Rasmussen, 1976; Stevens et al., 1969) which reduces S compounds and measures the
intensity of the CL from electronically excited sulfur molecules (S2*). Sensitivities for pS04 as low as 0.1
(ig/m3 with time resolution ranging from 1 to 30 min have been reported. Continuous measurements of
pS04 content have also been obtained by on-line XRF analysis with resolution of 30 min or less (Jaklevic
et al., 1981). During a field-intercomparison study of five different S instruments, Camp et al. (1982)
found four out of five FPD systems agreed to ± 5% during a 1-week sampling period.
There are two major PM speciation ambient air-monitoring networks in the U.S.: the Chemical
Speciation Network (CSN), which now includes the former Speciation Trends Network (STN), and the
Interagency Monitoring of Protected Visual Environments (IMPROVE) network. The current CSN
samplers sample on a l-in-3 days cycle using three filters: Teflon for equilibrated mass and elemental
analysis including elemental S; a HN03 denuded nylon filter for ion analysis including N03 and S042 : a
quartz-fiber filter for elemental and organic carbon (EC and OC, respectively). The IMPROVE sampler,
which collects two 24-h samples per week, simultaneously collects one sample of PMi0 on a Teflon filter,
and three samples of PM2 5 on Teflon, nylon, and quartz filters. PM2 5 mass concentrations are determined
gravimetrically from the PM2 5 Teflon filter sample, which is also used to determine concentrations of
selected elements. The PM2 5 nylon filter sample, which is preceded by a denuder to remove acidic gases,
is analyzed to determine pN03 and pS04 concentrations. Finally, the PM2 5 quartz filter sample is
analyzed for OC and EC using the thermal-optical reflectance (TOR) method for IMPROVE and thermal-
optical transmittance (TOT) for CSN, though this network is in transition to TOR.
In a side-by-side comparison of two of the chief aerosol monitoring techniques, PM2 5 mass and the
major contributing species were moderately well correlated among the different methods with r >0.8
(Hains et al., 2007). Agreement was good for total mass, S042 , OC, total carbon (TC), and NH4+, while
N03 and black carbon (BC) showed less-good fits. Based on reported uncertainties, however, even daily
concentrations of PM2 5 mass and major contributing species were often significantly different at the 95%
confidence level. The older STN/CSN methods reported generally higher values of PM2 5 total mass and
of individual species than did the IMPROVE-like ones. Since these differences can only be partially
accounted for by known random errors, the authors concluded that the current uncertainty estimates used
for data analyzed with the older STN protocol may underestimate the actual uncertainty.
2-55
-------
2.7.3.1. Artifacts
The reaction of S02 and other acid gases with basic sites on glass fiber filters or with basic coarse
particles on the filter leads to the formation of nonvolatile pS04, pN03, and CI salts. These positive
artifacts lead to overestimates of concentrations of PM total mass and S042 and likely N03 as well.
These problems were largely overcome by changing to quartz fiber or Teflon filters and by the separate
collection of the PM2 5 fraction. However, the possible reaction of acidic gases with basic coarse particles
remains a possibility, especially with PMi0 and PM10-2.5 measurements. These positive artifacts could be
effectively eliminated by removing acid gases in the sampling line with denuders coated with NaCl or
Na2C03.
Positive sampling artifacts also occur during measurement of PNH4. The reaction of NH3 with acid
particles
2NH:, + H2S04 -> (NH4)2S04
Reaction 48
either during sampling or during transportation, storage, and equilibration could lead to an overestimation
of PNH4 concentrations. Techniques have been developed to overcome this problem, including using a
denuder coated with hydrofluoric, citric, or phosphoric acid to remove NH3 during sampling and to
protect the collected PM from NH3 (Brauer et al., 1991; Keck and Wittmaack, 2006; Koutrakis et al.,
1988a, 1988b; Possanzini et al., 1999; Suh et al., 1992, 1994; Winberry et al., 1999). Positive artifacts for
PNH4 can also develop during sample handling due to contamination with NH3 emitted directly from
human sweat, breath, and tobacco smoking to form (NHz^SC^ or NH4HSO4 if the filter is improperly
handled (Sutton et al., 2000).
Although pS04 is relatively stable on a Teflon filter, it is now well known that volatilization losses
of pN03 occur during sampling. For pN03, the effect on the accuracy of atmospheric measurements from
these volatilization losses is more significant for PM2 5 than for PMi0, partly because N03 contributes a
smaller fraction to PMi0 and partly because N03 is most often present in a non-volatile form such as
NaN03, in the coarse mode.
Sampling artifacts resulting from the loss of pN03 species represents a significant problem in areas
such as southern California that experience high total N03 loadings. Hering and Cass (1999) discussed
errors in PM2 5 mass measurements owing to the volatilization of pN03 using data from two field
measurement campaigns conducted in southern California: the Southern California Air Quality Study
(SCAQS) (Lawson, 1990); and the 1986 California Institute of Technology (CalTech) study (Solomon
et al., 1992). In both studies, side-by-side sampling of PM25 was conducted with one sampler collecting
particles directly onto a Teflon filter and a second using an MgO-coated denuder (Appel et al., 1981) to
remove gaseous HN03, followed by a nylon filter to absorb the evaporating HN03. In both studies, the
PM2 5 mass lost from NH4N03 volatilization represented a significant fraction of the total PM2 5 mass, and
these losses were greater during summer than fall: 17% (summer) versus 9% (fall) during SCAQS, and
21% (summer) versus 13% (fall) during CalTech. With regard to percentage loss of pN03, as contrasted to
percentage loss of mass discussed above, Hering and Cass (1999) found that the amount of pN03
remaining on the Teflon filter samples was, on average, 28% less than that on the HN03 denuded nylon
filters.
Hering and Cass (1999) also analyzed these data by extending the evaporative model developed by
Zhang and McMurry (1987). The extended model used by Hering and Cass (1999) takes into account the
dissociation of collected NH4N03 on Teflon filters into HN03 and NH3 via three mechanisms: scrubbing
of HN03 and NH3 in the sampler inlet. John et al. (1988) showed that clean PMi0 inlet surfaces serve as
an effective denuder for HN03; heating of the filter substrate above ambient temperature by sampling;
and pressure drop across the Teflon filter. For the sampling systems modeled, the flow-induced pressure
drop was measured to be less than 0.02 atm, and the corresponding change in vapor pressure was 2%, so
2-56
-------
losses driven by pressure drop were not considered to be significant in this work. Losses from Teflon
filters were found to be higher during the summer than during the winter, higher during the day compared
to night, and were reasonably consistent with modeled predictions.
Finally, during SCAQS (Lawson, 1990), particulate samples also were collected using a Berner
impactor and greased Tedlar substrates in size ranges from 0.05 to 10 (.un in aerodynamic diameter. The
Berner impactor PM2 5 N03 values were much closer to those from the denuded nylon filter than those
from the Teflon filter; the impactor N03 values were -2% lower than the nylon filter N03 for the fall
measurements and -7% lower for the summer measurements. When the impactor collection was
compared to the Teflon filter collection for nonvolatile species, pS04, the results were in agreement.
Chang et al. (2000) discussed reasons for reduced loss of N03 from impactors.
Brook and Dann (1999) observed much higher N03 losses during a study in which they measured
N03 in Windsor and Hamilton, Ontario, Canada, by three techniques: a single Teflon filter in a
dichotomous sampler; the Teflon filter in an ADS; and total N03 including both the Teflon filter and the
nylon back-up filter from the ADS. The Teflon filter from the dichotomous sampler averaged only 13% of
the total N03, whereas the Teflon filter from the ADS averaged 46% of total N03 . Considerable N03 was
lost from the dichotomous sampler filters during handling, which included weighing and XRF
measurement in a vacuum.
Kim et al. (1999) also examined pN03 sampling artifacts by comparing denuded and non-denuded
quartz and nylon filters during the PMi0 Technical Enhancement Program (PTEP) in the California South
Coast Air Basin. They observed N03 losses for most measurements; however, for a significant number of
measurements, they observed positive N03 artifacts. Kim et al. (1999) pointed out that random
measurement errors make it difficult to measure true amounts of N03 loss.
Diffusion denuder samplers, developed primarily to measure particle strong acidity (Koutrakis
et al., 1988a, 1992) also can be used to study N03 volatilization. Measurements were made with two
versions of the Harvard-U.S. EPA ADS (HEADS) for which HN03 vapor was removed by aNa2C03
coated denuder and the remaining pN03 was reported either as the sum of nonvolatile N03 collected on a
Teflon filter, and volatized N03 collected on a Na2C03 coated filter downstream of the Teflon filter (full
HEADS) or on a Nylon filter downstream of the Teflon filter (Nylon HEADS). The full HEADS
consistently underestimated the total N03 by -20% compared to the Nylon HEADS.
This comparison technique was then used to measure loss of N03 from Teflon filters in seven U.S.
cities by Babich et al. (2000) who found significant pN03 losses in Riverside, CA, Philadelphia, PA, and
Boston, MA, but not in Bakersfield, CA, Chicago, IL, Dallas, TX, or Phoenix, AZ, where measurements
were made only during winter.
Negative sampling artifacts due to decomposition and volatilization are also significant for PNH4,
more often when it appears as NH4N03 since (NELO2SO4 is much more stable. The presence and
deposition of NH4N03 is highly sensitive to environmental factors such as temperature, RH, acidity of
aerosols, as well as to filter type (Keck and Wittmaack, 2005; Spurny, 1999). Any change in these
parameters during the sampling period influences the position of the equilibrium between the particle and
gas phases. Keck and Wittmaack (2005) observed that at temperatures <0 °C, acetate-N03, quartz fiber,
and Teflon filters could properly collect pNEU, NH3, and Cl~; but at temperatures >0 °C, the salts were lost
from quartz fiber and Teflon filters, more so at higher temperatures and with no significant difference
between quartz fiber and Teflon filters. The salts were lost completely from denuded quartz fiber filters at
temperatures above -20 °C, and from non-undenuded quartz fiber and Teflon filters at temperatures above
-25 °C. It is anticipated that current sampling techniques underestimate PNH4 levels due to volatilization,
but fine particle mass contains many acidic compounds, and, as a consequence, a fraction of volatilized
NH4+ in the form of NH3 can be retained on the Teflon filter by reactions with them. Owing to these
positive and negative interference effects, the magnitude of pNEU remains largely unknown. However,
techniques have been applied to pNEU sampling to correct its concentrations due to evaporation using a
backup filter coated with hydrofluoric acids, citric acid, or phosphorous acids to absorb the evaporated
NH4 as NH3. Total NH4 concentration then is the sum of the pNEU collected on the Teflon filter and the
concentration of the NH3 collected on the backup filter.
2-57
-------
Volatile compounds can also leave the filter after sampling and before filter weighing or chemical
analysis. Losses ofN03 . NH/, and Cl~ from glass and quartz-fiber filters that were stored in unsealed
containers at ambient air temperatures for 2 to 4 weeks before analysis exceeded 50% (Witz et al., 1990).
2.7.3.2. Other Methods
An integrated collection and vaporization cell was developed by Stolzenburg and Hering (2000)
that provides automated, 10-min resolution monitoring of fine pN03. In this system, particles are
collected by a humidified impaction process and analyzed in place by flash vaporization and CL detection
of the evolved NOx. In field tests in which the system was co-located with two FRM samplers, the
automated pN03 sampler results followed the results from the FRM, but were offset lower. The system
also was co-located with a HEADS and a SASS speciation sampler (MetOne Instruments). In all these
tests, the automated sampler was well correlated to other samplers with slopes ranging from 0.95 for the
FRM to 1.06 for the HEADS, and correlation coefficients ranging from 0.94 to 0.996. During the
Northern Front Range Air Quality Study in Colorado (Watson et al., 1998), the automated N03 monitor
captured the 12-min variability in pN03 concentrations with a precision of approximately ± 0.5 (ig/m3
(Chow et al., 1998). A comparison with denuded filter measurements followed by IC analysis (Chow and
Watson, 1999) showed agreement within ± 0.6 (ig/m3 for most of the measurements, but exhibited a
discrepancy of a factor of 2 for the periods of high pN03 concentrations. More recent intercomparisons
took place during the 1997 Southern California Ozone Study (SCOS97) in Riverside, CA. Comparisons
with 14 days of 24-h denuder-filter sampling gave a correlation coefficient of 0.87 and showed no
significant bias. As currently configured, the system has a detection limit of 0.7 (ig/m3 and a precision of
0.2 (ig/m3.
The extent to which sampling artifacts for pNH4 have been adequately addressed in the current
networks is not clear. Recently, new denuder-filter sampling systems have been developed to measure
pS04, pN03, and pNH4 with an adequate correction of NH/ sampling artifacts. The denuder-filter system,
Chembcomb Model 3500 speciation sampling cartridge developed by Rupprecht & Patashnick Co, Inc.
could be used to collect N03, S042 . and NH4+ simultaneously. The sampling system contains a single-
nozzle size-selective inlet, two honeycomb denuders, the aerosol filter and two backup filters (Keck and
Wittmaack, 2005). The first denuder in the system is coated with 0.5% sodium carbonate and 1% glycerol
and collects acid gases such as HC1, S02, HN02, and HN03. The second denuder is coated with 0.5%
phosphoric acid in methanol for collecting NH3. Backup filters collect the gases behind denuded filters.
The backup filters are coated with the same solutions as the denuders. A similar system based on the same
principle was applied by Possanzini et al. (1999). The system contains two NaCl coated annular denuders
followed by two denuders coated with Na2C03+ glycerol and citric acid, respectively. This configuration
was adopted to remove HN03 quantitatively on the first NaCl denuder. The third and forth denuder
remove S02 and NH3, respectively. A polyethylene cyclone and a two-stage filter holder containing three
filters are placed downstream of the denuders. Aerosol fine particles are collected on a Teflon membrane.
A backup nylon filter and a subsequent citric acid impregnated filter paper collect dissociation products,
NH4N03, HN03 and NH3, evaporated from the filtered particulate matter.
2-58
-------
2.8. Methods to Compute NOx and SOx Concentrations,
Chemical Interactions, and Deposition
2.8.1. Chemical Transport Models
CTMs are the prime tools used to compute the interactions among NOx, SOx, other pollutants and
their precursors, the transport and transformation of air toxics, the production of secondary aerosols, the
evolution of the particle size distribution, and deposition of pollutants. CTMs are driven by emissions
inventories for primary species such as NOx, SOx, NH3, and primary PM, and by meteorological fields
produced by other numerical prediction models. Meteorological quantities such as winds and
temperatures are taken from operational analyses, reanalyses, or weather circulation models. In most
cases, these are off-line meteorological analyses, meaning that they are not modified by radiatively active
species generated by the air quality model (AQM).
Emissions of precursor compounds can be divided into anthropogenic and biogenic source
categories, and biogenic sources can be further divided into biotic (vegetation, microbes, animals) and
abiotic (geogenic biomass burning, lightning) categories as presented in Section 2.2 above. However, the
distinction between biogenic sources and anthropogenic sources is often difficult to make, as human
activities affect directly or indirectly emissions from what would have been considered biogenic sources
during the preindustrial era. Thus, emissions from plants and animals used in agriculture have been
referred to as anthropogenic or biogenic in different applications. Wildfire emissions may be considered
to be biogenic, except that forest management practices may have led to the buildup of fuels on the forest
floor, thereby altering the frequency and severity of forest fires.
The initial conditions, or starting concentration fields of all species computed by a model, and the
boundary conditions, or concentrations of species along the horizontal and upper boundaries of the model
domain throughout the simulation, must be specified at the beginning of the simulation. Both initial and
boundary conditions can be estimated from models or data or, more generally, model + data hybrids.
Because data for vertical profiles of most species of interest are sparse, results of model simulations over
larger, usually global, domains are often used. As might be expected, the influence of boundary conditions
depends on the x of the species under consideration and the time scales for transport from the boundaries
to the interior of the model.
Each of the model components described above has associated uncertainties and the relative
importance of these uncertainties varies with the modeling application. The largest errors in
photochemical modeling are still thought to arise from the meteorological and emissions inputs to the
model (Russell and Dennis, 2000). Within the model itself, horizontal advection algorithms are still
thought to be significant sources of uncertainty (see, e.g., Chock and Winkler, 1994), though more
recently, those errors are thought to have been reduced (see, e.g., Odman and Ingram, 1996). There are
also indications that problems with mass conservation continue to be present in photochemical and
meteorological models (see, e.g., Odman and Russell, 1999) and can result in significant simulation
errors. The effects of errors in initial conditions can be minimized by including several days spin-up time
in a simulation to allow the model to be driven by emitted species before the simulation of the period of
interest begins.
While the effects of poorly specified boundary conditions propagate through the model's domain,
the effects of these errors remain undetermined. Because many meteorological processes occur on spatial
scales smaller than the model grid spacing (either horizontally or vertically) and thus are not calculated
explicitly, parameterizations of these processes must be used and these introduce additional uncertainty.
Specific uncertainty also arises in modeling the chemistry of NOx transformations because they are
strongly nonlinear. Thus, the volume of the grid cell into which emissions are injected is important
because, for example, 03 production or loss depends in a complicated way on the concentrations of NOx
2-59
-------
and OH as explained in Section 2.6.2.1 above. Use of ever-finer grid spacing allows more valid separation
of regions of high NOx concentrations from low NOx regions and from regions where NOx
concentrations are optimal for P(03).
The use of grid spacing fine enough to resolve the chemistry in individual power-plant plumes is
too demanding of computer resources for this to be attempted in most regional air quality simulations.
Instead, parameterizations of the effects of sub-grid-scale processes such as these must be developed, else
serious errors can result if emissions are allowed to mix through an excessively large grid volume before
the chemistry step in a model calculation is performed. In light of the significant differences between
atmospheric chemistry taking place within and outside of a power plant plume identified by Ryerson et al.
(1998), inclusion of a separate module for treating large, tight plumes can be useful. Because the
photochemistry of NOx transformation is nonlinear, emissions correctly modeled in a tight plume may be
incorrectly modeled in a more dilute plume. Fortunately, it appears that the chemical mechanism used to
follow a plume's development need not be as detailed as that used to simulate the rest of the domain, as
the inorganic reactions are most important in the plume; see e.g., Kumar and Russell (1996).
Because the chemical production and loss terms in the continuity equations for individual species
are coupled, the chemical calculations must be performed iteratively until calculated concentrations
converge to within some preset criterion. The number of iterations and the convergence criteria chosen
also can introduce error.
CTMs have been developed for application over a wide range of spatial scales ranging up from
neighborhood to global. Regional scale CTMs are used to: obtain better understanding of the processes
controlling the formation, transport, and destruction of gas- and particle-phase criteria and hazardous air
pollutants; understand the relations between concentrations of secondary pollutant products and
concentrations of their precursors such as NOx and VOCs and the factors leading to acid deposition and
possible damage to biota and ecosystems; understand relations among the concentration patterns of
various pollutants that may exert adverse effects; and evaluate how changes in emissions propagate
through the atmospheric system to secondary products and deposition.
CTMs in current use mostly have one of two forms. The first, grid-based or Eulerian air quality
models subdivide the region to be modeled or the modeling domain, into a three-dimensional array of grid
cells. Spatial derivatives in the species continuity equations are cast in finite-difference form over this
grid and a system of equations for the concentrations of all the chemical species in the model are solved
numerically at each grid point. Finite element Eulerian models also exist and have been exercised, but less
frequently. Time-dependent continuity or mass conservation equations are solved for each species
including terms for transport, chemical production and destruction, and emissions and deposition (if
relevant), in each cell. Chemical processes are simulated with ordinary differential equations, and
transport processes are simulated with partial differential equations. Because of a number of factors such
as the different time scales inherent in different processes, the coupled, nonlinear nature of the chemical
process terms, and computer storage limitations, not all of the terms in the equations are solved
simultaneously in three dimensions. Instead, operator splitting, in which terms in the continuity equation
involving individual processes are solved sequentially, is used.
In the second common CTM formulation, trajectory or Lagrangian models, a number of
hypothetical air parcels are specified as though following wind trajectories. In these models, the original
system of partial differential equations is transformed into a system of ordinary differential equations.
A less common approach is to use a hybrid Lagrangian-Eulerian model, in which certain aspects of
atmospheric chemistry and transport are treated with a Lagrangian approach and others are treaded in an
Eulerian manner; see e.g., Stein et al., 2000.
Each approach has advantages and disadvantages. The Eulerian approach is more general in that it
includes processes that mix air parcels and allows integrations to be carried out for long periods during
which individual air parcels lose their identity. There are, however, techniques for including the effects of
mixing in Lagrangian models such as FLEXPART (Zanis et al., 2003), ATTILA (Reithmeier and Sausen,
2002), and CLaMS (McKenna et al., 2002).
2-60
-------
2.8.1.1. Global Scale
Global-scale CTMs are used to address issues associated with climate change and stratospheric 03
depletion, and to provide boundary conditions for the regional-scale models. The CTMs include
simplified mathematical descriptions of atmospheric transport, the transfer of solar radiation through the
atmosphere, chemical reactions, and removal to the surface by turbulent motions and precipitation for
pollutants emitted into the model domain. The upper boundaries of the CTMs extend anywhere from the
top of the mixed layer to the mesopause at -80 km to obtain more realistic boundary conditions for
problems involving stratospheric dynamics.
The importance of global transport of 03 and 03 precursors and their contribution to regional 03
levels in the U.S. is now apparent. There are at present on the order of 20 three-dimensional global
models developed by various groups to address problems in tropospheric chemistry. These models resolve
synoptic meteorology, 03-N0x-C0-HC photochemistry, have parameterizations for wet and dry
deposition, and parameterize sub-grid scale vertical mixing processes such as convection. Global models
have proven useful for testing and advancing scientific understanding beyond what is possible with
observations alone. For example, they can calculate quantities of interest that cannot be measured directly,
such as export of pollution from one continent to the global atmosphere, or the response of the
atmosphere to future perturbations in anthropogenic emissions.
Global simulations are typically conducted at a horizontal resolution of 200 km2 or more.
Simulations of the effects of transport from long-range transport link multiple horizontal resolutions from
the global to the local scale. Finer resolution will only improve scientific understanding to the extent that
the governing processes are more accurately described at that scale. Consequently, there is a critical need
for observations at the appropriate scales to evaluate the scientific understanding represented by the
models.
During the recent IPCC-AR4 tropospheric chemistry study coordinated by the European Union
Atmospheric Composition Change: the European Network of excellence (ACCENT), 26 atmospheric
CTMs were used to estimate the impacts of three emissions scenarios on global atmospheric composition,
climate, and air quality in 2030 (Dentener et al., 2006a, b). All models were required to use anthropogenic
emissions developed at IIASA (Dentener et al., 2005) and GFED version 1 biomass burning emissions
(Van der Werf et al., 2003) as described in Stevenson et al. (2006). The base simulations from these
models were evaluated against a suite of present-day observations. Most relevant to this assessment report
are the evaluations with N02 and N deposition (Dentener et al., 2006b; Stevenson et al., 2006), which are
summarized briefly below.
A subset of 17 of the 26 models used in the Stevenson et al. (2006) study was used to compare with
three retrievals of N02 columns from the GOME instrument (van Noije et al., 2006) for the year 2000.
The higher resolution models reproduced the observed patterns better, and the correlation among
simulated and retrieved columns improved for all models when simulated values were smoothed to a 5° *
5° grid, implying that the models did not accurately reproduce the small-scale features of N02 (van Noije
et al., 2006). Van Noije et al. (2006) also suggested that variability in simulated N02 columns may reflect
model differences in OH distributions and the resulting NOx lifetimes, as well as differences in vertical
mixing which strongly affected partitioning between NO and N02. Overall, the models tended to
underestimate concentrations in the retrievals in industrial regions including the eastern U.S. and to
overestimate them in biomass burning regions (van Noije et al., 2006).
Over the eastern U.S. and in industrial regions more generally, the spread in absolute column
abundances is generally larger among the retrievals than among the models, with the discrepancy among
the retrievals particularly pronounced in winter (van Noije et al., 2006), suggesting that the models were
biased low, or that the European retrievals may be biased high as the Dalhousie SAO retrieval is closer to
the model estimates. The lack of seasonal variability in fossil fuel combustion emissions may have
contributed to a wintertime model underestimate (van Noije et al., 2006) manifested most strongly over
Asia. In biomass burning regions, the models generally reproduce the timing of the seasonal cycle of the
retrievals, but tend to overestimate the seasonal cycle amplitude, partly due to lower values in the wet
2-61
-------
season, which may reflect an underestimate in wet season soil NO emissions (Jaegle et al., 2004; van
Noije et al., 2006).
2.8.1.2. Regional Scale
Most major modeling efforts within the U.S. EPA use the Community Multiscale Air Quality
modeling system (CMAQ) (Byun and Schere, 2006; Byun and Ching, 1999). A number of other modeling
platforms using Lagrangian and Eulerian frameworks were reviewed in the 2006 03 AQCD (U.S. EPA,
2006b) and in Russell and Dennis (2000). Evaluations of the performance of CMAQ are given in Appel
et al. (2005), Arnold et al. (2003), Eder and Yu (2006), and Fuentes and Raftery (2005). CMAQ's domain
can extend from several hundred kilometers to the entire hemisphere. CMAQ is most often driven by the
MM5 mesoscale meteorological model (Seaman, 2000), though it may be driven by other meteorological
models, including the Regional Atmosphere Modeling System (RAMS); see http://atmet.com.
Simulations of pollution episodes over regional domains have been performed with a horizontal
resolution as low as 1 km, and smaller calculations over limited domains have been performed at even
finer scales. However, simulations at such high resolutions require better parameterizations of
meteorological processes such as boundary layer fluxes, deep convection, and clouds (Seaman, 2000), as
well as finer-scale emissions data than are generally available. Finer spatial resolution is necessary to
resolve features such as urban heat island circulation; sea, bay, and land breezes; mountain and valley
breezes; and the nocturnal low-level jet; all of which can affect pollutant concentrations.
The most common approach to setting up the horizontal domain is to nest a finer grid within a
larger domain of coarser resolution. However, there are other strategies such as the stretched grid (see
e.g., Fox-Rabinovitz et al., 2002) and the adaptive grid. In a stretched grid, the grid's resolution
continuously varies throughout the domain, thereby eliminating any potential problems with the sudden
change from one resolution to another at the boundary. Caution should be exercised in using such a
formulation because certain parameterizations like those for convection valid on a relatively coarse grid
scale may not be valid on finer scales. Adaptive grids are not fixed at the start of the simulation, but
instead adapt to the needs of the simulation as it evolves (see e.g., Hansen et al., 1994). They have the
advantage that they can resolve processes at relevant spatial scales. However, they can be very slow if the
situation to be modeled is complex. Additionally, if adaptive grids are used for separate meteorological,
emissions, and photochemical models, there is no reason a priori why the resolution of each grid should
match, and the gains realized from increased resolution in one model will be wasted in the transition to
another model. The use of finer horizontal resolution in CTMs will necessitate finer-scale inventories of
land use and better knowledge of the exact paths of roads, locations of factories, and, in general, better
methods for locating sources and estimating their emissions.
The vertical resolution of these CTMs is variable, and usually configured to have more layers in the
PBL, and fewer higher up. Because the height of the boundary layer is of critical importance in
simulations of air quality, improved resolution of the boundary layer height would likely improve air
quality simulations. Additionally, current CTMs do not adequately resolve fine scale features such as the
nocturnal low-level jet in part because little is known about the nighttime boundary layer.
CTMs require time-dependent, three-dimensional wind fields for the period of simulation. The
winds may be generated either by a model using initial fields alone or with four-dimensional data
assimilation to improve the model's performance; i.e., model equations can be updated periodically or
nudged to bring results into agreement with observations. Modeling series durations can range from
simulations of several days' duration to several months or multiple seasons of the year.
Chemical kinetics mechanisms (sets of chemical reactions) representing the important reactions
occurring in the atmosphere are used in CTMs to estimate the rates of chemical formation and destruction
of each pollutant simulated as a function of time. Because of different approaches to the lumping of
organic compounds into surrogate groups for computational efficiency, chemical mechanisms can
produce somewhat different results under similar conditions. The CB-IV chemical mechanism (Gery
2-62
-------
et al., 1989), the RADMII mechanism (Stockwell et al., 1990), the SAPRC (e.g., Carter, 1990; Wang
et al., 2000a, b), and the RACM mechanisms can be used in CMAQ. Jimenez et al. (2003) provided brief
descriptions of the features of the main mechanisms in use and compared concentrations of several key
species predicted by seven chemical mechanisms in a box model simulation over 24 h. The average
deviation from the average of all mechanism predictions for 03 and NO over the daylight period was
<20%, and was 10% for N02 for all mechanisms. However, much larger deviations were found for HN03,
PAN, H02, H202, ethylene, and isoprene. The large deviations shown for most species imply differences
between the x of atmospheric species and the assignment of model simulations to either NOx-limited or
radical-limited regimes between mechanisms. Gross and Stockwell (2003) found small differences
between mechanisms for clean conditions, with differences becoming more significant for polluted
conditions, especially for N02 and R02 radicals.
CMAQ and other state-of-the-science CTMs incorporate processes and interactions of aerosol-
phase chemistry (Mebust et al., 2003). There have also been several attempts to study the feedbacks of
chemistry on atmospheric dynamics using meteorological models like MM5 (Grell et al., 2000; Liu et al.,
2001; Lu et al., 1997; Park et al., 2001). This coupling is necessary to simulate accurately feedbacks such
as may be caused by the heavy aerosol loading found in forest fire plumes (Lu et al., 1997; Park et al.,
2001), or in heavily polluted areas. Photolysis rates in CMAQ can now be calculated interactively with
model-produced 03, N02, and aerosol fields (Binkowski et al., 2007).
Spatial and temporal characterizations of anthropogenic and biogenic precursor emissions must be
specified as inputs to a CTM. Emissions inventories have been compiled on grids of varying resolution
for many HCs, aldehydes, ketones, CO, NH3, and NOx. Emissions inventories for many species require
the application of algorithms for calculating the dependence of emissions on physical variables such as
temperature and to convert the inventories into formatted emission files that can be used by a CTM. For
example, emissions data preprocessing for CMAQ is done by the Spare-Matrix Operator Kernel
Emissions (SMOKE) system; see http://smoke-model.om. For many species, information concerning the
temporal variability of emissions is lacking, so annual or 03_season averages are used in short-term,
episodic simulations. Annual emissions estimates are often modified by the emissions model to produce
emissions more characteristic of the time of day and season. Significant errors in emissions can occur if
inappropriate time dependence or a default profile is used.
2.8.1.3. Sub-Regional Scale
The grid spacing in regional CTMs of between 1 and 12 km2 is usually too coarse to resolve spatial
variations on the neighborhood scale. The interface between regional scale models and models of smaller
exposure scales described is provided by smaller scale dispersion models. AERMOD (http://www.epa.gov
/scram001/dispersion_prefrec.htm) is one example of a steady-state plume model formulated as a
replacement to the ISC3 dispersion model. In the stable boundary layer (SBL), it assumes the
concentration distribution to be Gaussian in both the vertical and horizontal dimensions. In the convective
boundary layer, the horizontal distribution is also assumed to be Gaussian, but the vertical distribution is
described with a bi-Gaussian probability density function. AERMOD has provisions to be applied to flat
and complex terrain, and multiple source types (including, point, area and volume sources) in both urban
and rural areas. It incorporates air dispersion based on PBL turbulence structure and scaling concepts, and
is meant to treat both surface and elevated sources and simple and complex terrain in rural and urban
areas. The dispersion of emissions from line sources like highways is treated as the sum of emissions
from a number of point sources placed side by side. However, emissions are usually not in steady state
and there are different functional relationships between buoyant plume rise in point and line sources. It
should be remembered that N02 is largely secondary in nature. However, AERMOD does not have
provision for including secondary sources. The more appropriate use of AERMOD would be to simulate
the total of NO and N02, or NOx.
2-63
-------
There are non-steady state models that incorporate plume rise explicitly from different types of
sources. For example, CALPUFF (http://www.src.com/calpuff/calpuff 1 .htm) is a non-steady-state puff
dispersion model that simulates the effects of time- and space-varying meteorological conditions on
pollution transport, transformation, and removal and has provisions for calculating dispersion from
surface sources. However, it should be noted that neither model was designed to treat the dispersion of
emissions from roads or to include secondary sources.
2.8.1.4. Modeling Effects of Convection for Chemical Transport
The effects of deep convection can be simulated using cloud-resolving models or in regional or
global models in which the convection is parameterized. The Goddard Cumulus Ensemble (GCE) model
(Tao and Simpson, 1993) has been used by Pickering et al. (1991, 1992a, b, 1993, 1996), Scala et al.
(1990) and Stenchikov et al. (1996) in the analysis of convective transport of trace gases. The cloud
model is nonhydrostatic and contains a detailed representation of cloud microphysical processes. Two-
and three-dimensional versions of the model have been applied in transport analyses. The initial
conditions for the model are usually from a sounding of temperature, water vapor, and winds
representative of the region of storm development. Model-generated wind fields can be used to perform
air parcel trajectory analyses and tracer advection calculations.
Such methods were used by Pickering et al. (1992a) to examine transport of urban plumes by deep
convection. Transport of an Oklahoma City, OK, plume by the 10-11 June 1985 PRE-STORM squall line
was simulated with the 2-D GCE model. This major squall line passed over the Oklahoma City
metropolitan area, as well as more rural areas to the north. Chemical observations ahead of the squall line
were conducted by the PRE-STORM aircraft. In this event, forward trajectories from the boundary layer
at the leading edge of the storm showed that almost 75% of the low-level inflow was transported to
altitudes exceeding 8 km. Over 35% of the air parcels reached altitudes over 12 km. Tracer transport
calculations were performed for CO, NOx, 03, and HCs. Rural boundary layer NOx was only 0.9 ppb,
whereas the urban plume contained about 3 ppb. In the rural case, mixing ratios of 0.6 ppb were
transported up to 11 km. Cleaner air descended at the rear of the storm lowering NOx at the surface from
0.9 to 0.5 ppb. In the urban plume, mixing ratios in the updraft core reached 1 ppb between 14 and 15 km.
At the surface, the main downdraft lowered the NOx mixing ratios from 3 to 0.7 ppb.
Regional CTMs have been used for applications such as simulations of photochemical 03
production, acid deposition, and fine PM. Walcek et al. (1990) included a parameterization of cloud-scale
aqueous chemistry, scavenging, and vertical mixing in the chemistry model of Change et al. (1987). The
vertical distribution of cloud microphysical properties and the amount of sub-cloud layer air lifted to each
cloud layer are determined using a simple entrainment hypothesis (Walcek and Taylor, 1986). Vertically
integrated 03 formation rates over the northeast U. S. were enhanced by -50% when the in-cloud vertical
motions were included in the model.
Global models with parameterized convection and lightning have been run to examine the roles of
these processes over North America. Lightning contributed 23% of upper tropospheric NOY over the
SONEX region according to the UMD-CTM modeling analysis of Allen et al. (2000). During the summer
of 2004 the NASA Intercontinental Chemical Transport Experiment - North America (INTEX-NA) was
conducted primarily over the eastern two-thirds of the U.S., as a part of the International Consortium for
Atmospheric Research on Transport and Transformation. Deep convection was prevalent over this region
during the experimental period. Cooper et al. (2006) used a particle dispersion model simulation for NOx
to show that 69-84% of the upper tropospheric 03 enhancement over the region in summer 2004 was due
to lightning NOx. The remainder of the enhancement was due to convective transport of 03 from the
boundary layer or other sources of NOx. Hudman et al. (2007) used a GEOS-Chem model simulation to
show that lightning was the dominant source of upper tropospheric NOx over this region during this
period. Approximately 15% of North American boundary layer NOx emissions were shown to have been
vented to the free troposphere over this region based on both the observations and the model.
2-64
-------
2.8.2. Computed Deposition
Wet and dry deposition are important removal processes for pollutants on urban, regional, and
global scales and so are included in CTMs. The general approach used to estimate deposition velocity
(Vd) in most models is the resistance-in-series method described above and represented in Equation 5.
This approach works for a range of substances, although it is inappropriate for species with substantial re-
emissions from the surface or for species where deposition to the surface depends on concentrations at the
surface itself. The approach is also modified somewhat for aerosols in that the terms Rj, and Rc are
replaced with a surface Vd to account for gravitational settling.
Vd=J/(Rm+Rb+RJ
Equation 5
aerodynamic
Atmospheric
Resistances
laminar' sub-layer
stomatal
cuticuiar
'"A
chemistry
Canopy ^
Resistances
soil
c3
,
Resistance analogy for the deposition of atmospheric pollutants
Source: Courtesy of T. Pierce, USEPA/ORD / NERL / Atmospheric Modeling Division.
Figure 2-25. Schematic of the resistance-in-series analogy for atmospheric deposition. Function of wind
speed, solar radiation, plant characteristics, precipitation/moisture, and soil/air temperature.
where l(ct. Rb, and Rc represent the resistance due to atmospheric turbulence, transport in the fluid
sublayer very near the elements of surface such as leaves or soil, and the resistance to uptake of the
surface itself, respectively, as shown in Figure 2-25.
2-65
-------
Wesley and Hicks (2000) listed several shortcomings of the then-current knowledge of dry
deposition. Among those shortcomings were difficulties in representing dry deposition over varying
terrain where horizontal advection plays a significant role in determining the magnitude of Ra and
difficulties in adequately determining Vd for extremely stable conditions such as those occurring at night;
see the discussion by Mahrt (1998), for example. Under optimal conditions, when a model is exercised
over a relatively small area where dry deposition measurements have been made, models still generally
showed uncertainties on the order of ± 30% (see, e.g., Brook et al., 1996; Massman et al., 1994; Padro,
1996; Wesely and Hicks, 2000). Wesely and Hicks (2000) concluded that an important result of those
comparisons was that the level of sophistication of most dry deposition models was relatively low, and
that deposition estimates, therefore, must rely heavily on empirical data. Still larger uncertainties exist
when the surface features in the built environment are not well known or when the surface comprises a
patchwork of different surface types, as is common in the eastern U.S.
2.8.2.1. Deposition Forms
Wet Deposition
Wet deposition results from the incorporation of atmospheric particles and gases into cloud droplets
and their subsequent precipitation as rain or snow, or from the scavenging of particles and gases by
raindrops or snowflakes as they fall (Lovett, 1994). Wet deposition depends on precipitation amount and
ambient pollutant concentrations. Receptor (i.e., vegetation) surface properties have little effect on wet
deposition, although leaves can retain liquid and solubilized PM. In terrain containing extensive
vegetative canopies, any material deposited via precipitation to the upper stratum of foliage is likely to be
intercepted by several foliar surfaces before reaching the soil. This allows such processes as foliar uptake,
chemical transformation, and re-suspension into the atmosphere to occur.
Landscape characteristics can affect wet deposition via orographic effects and by the closer
aerodynamic coupling to the atmosphere of tall forest canopies as compared to the shorter shrub and
herbaceous canopies. Following wet deposition, humidity and temperature conditions further affect the
extent of drying versus concentrating of solutions on foliar surfaces, which influence the rate of metabolic
uptake of surface solutes (Swietlik and Faust, 1984). The net consequence of these factors on direct
physical effects of wet deposited PM on leaves is not known (U.S. EPA, 2004).
Rainfall introduces new wet deposition and also redistributes throughout the canopy previously
dry-deposited particles (Peters and Eiden, 1992). The concentrations of suspended and dissolved materials
are typically highest at the onset of precipitation and decline with duration of individual precipitation
events (Hansen et al., 1994). Sustained rainfall removes much of the accumulation of dry-deposited
particles from foliar surfaces, reducing direct foliar effects and combining the associated chemical burden
with the wet-deposited material (Lovett, 1994) for transfer to the soil. Intense rainfall may contribute
substantial total particulate inputs to the soil, but it also removes bioavailable or injurious pollutants from
foliar surfaces. This washing effect, combined with differential foliar uptake and foliar leaching of
different chemical constituents from particles, alters the composition of the rainwater that reaches the soil
and the pollutant burden that is taken-up by plants. Once in the soil, these particle constituents may affect
biogeochemical cycles of major, minor, and trace elements. Low intensity precipitation events, in
contrast, may deposit significantly more particulate pollutants to foliar-surfaces than high intensity
precipitation events. Additionally, low-intensity events may enhance foliar uptake through the hydrating
of some previously dry-deposited particles (U.S. EPA, 2004).
Dry Deposition
Dry particulate deposition, especially of heavy metals, base cations, and organic contaminants, is a
complex and poorly characterized process. It appears to be controlled primarily by such variables as
atmospheric stability, macro- and micro-surface roughness, particle diameter, and surface characteristics
2-66
-------
(Hosker and Lindberg, 1982). The range of particle sizes, the diversity of canopy surfaces, and the variety
of chemical constituents in airborne particulates have made it difficult to predict and to estimate dry
particulate deposition (U.S. EPA, 2004).
Dry deposition of atmospheric particles to plant and soil surfaces affects all exposed surfaces.
Larger particles >5 |im diameter are dry deposited mainly by gravitational sedimentation and inertial
impaction. Smaller particles, especially those with diameters between 0.2 and 2 (jm, are not readily dry-
deposited and may travel long distances in the atmosphere until their eventual deposition, most often via
precipitation. Plant parts of all types, along with exposed soil and water surfaces, receive steady deposits
of dry dusts, EC, and heterogeneous secondary particles formed from gaseous precursors (U.S. EPA,
1982b).
Estimates of regional particulate dry deposition infer fluxes from the product of variable and
uncertain measured or modeled particulate concentrations in the atmosphere and even more variable and
uncertain estimates of Vd parameterized for a variety of specific surfaces (see e.g., Brook et al., 1996).
Even for specific sites and well-defined particles, uncertainties are large. Modeling the dry deposition of
particles to vegetation is at a relatively early stage of development, and it is not currently possible to
identify a best or most generally applicable modeling approach (U.S. EPA, 2004).
Occult Deposition
The occurrence of occult deposition tends to be geographically restricted to coastal and high
mountain areas. Several factors make occult deposition particularly effective for the delivery of dissolved
and suspended particulates to vegetation. Concentrations of particulate-derived materials are often many-
fold higher in cloud or fog water than in precipitation or ambient air due to orographic effects and gas-
liquid partitioning. In addition, fog and cloud water deliver particulate chemical species in a bioavailable-
hydrated form to foliar surfaces. This enhances deposition by sedimentation and impaction of submicron
aerosol particles that exhibit low Vd before fog droplet formation (Fowler et al., 1989). Deposition to
vegetation in fog droplets is proportional to wind speed, droplet size, concentration, and fog density. In
some areas, typically along foggy coastlines or at high elevations, occult deposition represents a
substantial fraction of total deposition to foliar surfaces (Fowler et al., 1991).
2.8.2.2. Methods for Estimating Dry Deposition
Methods for estimating dry deposition of particles are more restricted than for gaseous species and
fall into two major categories: surface analysis methods, which include all types of estimates of
contaminant accumulation on surfaces of interest; and atmospheric deposition rate methods, which use
measurements of contaminant concentrations in the atmosphere and descriptions of surrounding surface
elements to estimate deposition rates (Davidson and Wu, 1990). Surface extraction or washing methods
characterize the accumulation of particles on natural receptor surfaces of interest or on experimental
surrogate surfaces. These techniques rely on methods designed specifically to remove only surface-
deposited material. Total surface rinsate may be equated to accumulated deposition or to the difference in
concentrations in rinsate between exposed and control (sheltered) surfaces and may be used to refine
estimates of deposition. Foliar extraction techniques may underestimate deposition to leaves because of
uptake and translocation processes that remove pollutants from the leaf surface (Garten and Hanson,
1990; Taylor et al., 1988). Foliar extraction methods also cannot distinguish gas from particle-phase
sources (Bytnerowicz et al., 1987a, 1987b; Dasch, 1987; Lindberg and Lovett, 1985; Van Aalst, 1982).
The National Dry Deposition Network was established in 1986 to document the magnitude, spatial
variability, and trends in dry deposition across the U.S.. Currently, the network operates as a component
of the CASTNet (Clarke et al., 1997). A significant limitation on current capacity to estimate regional
effects ofNOx and SOx deposition is inadequate knowledge of the mechanisms and factors governing
particle dry deposition to diverse surfaces (U.S. EPA, 2004).
2-67
-------
Dry deposition cannot be directly measured. Deposition rates and totals are often calculated as the
product of measured ambient concentration and a modeled Vd. This method is widely used because
atmospheric concentrations are easier to measure than are dry deposition rates, and models have been
developed to estimate Vd. Ambient pollutant concentrations and meteorological conditions required for
application of inferential models are routinely collected at CASTNet dry deposition sites. Monitored
chemical species are limited to 03, S042 . N03 . NH/, S02, and HN03. The temporal resolution for the
ambient concentration measurements and dry deposition flux calculations is hourly for 03 and weekly for
the other species (Clarke et al., 1997).
Collection and analysis of stem flow and throughfall can also provide useful estimates of
particulate deposition when compared to directly sampled precipitation. The method is most precise for
particle deposition when gaseous deposition is a small component of the total dry deposition and when
leaching or uptake of compounds of interest out of or into the foliage is not a significant fraction of the
deposition because these lead to positive and negative artifacts in the calculated totals.
Foliar washing, whether using precipitation or experimental lavage, is one of the best available
methods to determine dry deposition to vegetated ecosystems. Major limitations include the site
specificity of the measurements and the restriction to elements that are largely conserved within the
vegetative system. Surrogate surfaces have not been found that can adequately replicate essential features
of natural surfaces; and therefore do not produce reliable estimates of particle deposition to the landscape.
Micrometeorological methods employ eddy covariance, eddy accumulation, or flux gradient
protocols for quantifying dry deposition. These techniques require measurements of particulate
concentrations and of atmospheric transport processes. They are currently well developed for ideal
conditions of flat, homogeneous, and extensive landscapes and for chemical species for which accurate
and rapid sensors are available. Additional studies are needed to extend these techniques to more complex
terrain and more chemical species.
2.8.2.3. Factors Affecting Dry Deposition Rates and Totals
In the size range of -0.1 to 1.0 |im where Vd is relatively independent of particle diameter as shown
in Figure 2-26, particulate deposition is controlled by roughness of the surface and by the stability and
turbulence of the atmospheric surface layer. Impaction and interception dominate over diffusion as dry
deposition processes, and the Vd is considerably lower than for particles that are either smaller or larger
than this size range (Shinn, 1978).
Deposition of particles between 1 and 10 |im diameter is strongly dependent on particle size
(Shinn, 1978). Larger particles within this size range are collected more efficiently at typical wind speeds
than are smaller particles (Clough, 1975), suggesting the importance of impaction. Impaction is related to
wind speed, the square of particle diameter, and the inverse of receptor diameter as a depositing particle
fails to follow the streamlines of the air in which it is suspended around the receptor. When particle
trajectory favors a collision, increasing either wind speed or the ratio of particle size to receptor cross-
section increases the probability of collision.
2-68
-------
0)
E
o
o
5
c
o
5
tf>
o
CL
D
1,000 -
100 —
10 —
1 —
0.1 —
0.01 —
0.001 —
0.000 -
Stokes Law
¦¦¦¦¦¦ Brownian Diffusion
¦ ^ i ™™ Pelers and Erten (1 $82)
i Little and Wlffen (1977)
\ \ \ \
0.001 0.01 0.1 1 10
Particle Diameter (pm)
100
Figure 2-26. The relationship between particle diameter and deposition velocity for particles. Values
measured in wind tunnels by Little and Wiffen (1977) over short grass with wind speed of 2.5
m/s closely approximate the theoretical distribution determined by Peters and Eiden (1992) for
a tall spruce forest. These distributions reflect the interaction of Brownian diffusivity
(descending dashed line), which decreases with particle size and sedimentation velocity
(ascending dotted line from Stokes Law), which increases with particle size. Intermediate-
sized particles (.0.1 to 1.0 |jm) are influenced strongly by both particle size and sedimentation
velocity, and deposition is independent of size. Source: U.S. EPA (2004).
Empirical estimates of Vd for fine particles under wind tunnel and field conditions are often
several-fold greater than predicted by theory (Unsworth and Wilshaw, 1989). A large number of transport
phenomena, including streamlining of foliar obstacles, turbulence structure near surfaces, and various
phoretic transport mechanisms are not well characterized (U.S. EPA, 2004). The discrepancy between
estimated and predicted values of Vd may reflect model limitations or experimental limitations in the
specification of the effective size and number of receptor obstacles. Available reviews (e.g., U.S. EPA,
1996a, 2004) suggest the following generalizations: particles >10 |im exhibit variable Vd between 0.5 and
1.1 cm/s depending on friction velocities, whereas a minimum particle Vd of 0.03 cm/s exists for particles
in the size range 0.1 to 1.0 |im: the Vd of particles is approximately a linear function of friction velocity;
and deposition of particles from the atmosphere to a forest canopy is from 2 to 16 times greater than
deposition in adjacent open terrain like grasslands or other low vegetation.
Leaf Surface Effects on Vd
The chemical composition of a particle is not usually considered to be a primary determinant of its
Vd. Rather, the plant leaf surface has an important influence on the Vd of particles, and therefore on the
flux of dry deposition to the terrestrial environment. Relevant leaf surface properties include stickiness,
microscale roughness, and cross-sectional area. These properties affect the probability of impaction and
particle bounce. The efficiency of deposition to vegetation also varies with leaf shape. Particles impact
2-69
-------
more frequently on the adaxial (upper surface) surface than on the abaxial (lower surface). Most particles
accumulate in the midvein, central portion of leaves. The greatest particle loading on dicotyledonous
leaves is frequently on the adaxial surface at the base of the blade, just above the petiole junction.
Precipitation washing probably plays an important role in this distribution pattern (U.S. EPA, 2004).
Lead particles have been shown to accumulate to a greater extent on older as compared with
younger needles and twigs of white pine, suggesting that wind and rain may be insufficient to fully wash
the foliage. Fungal mycelia (derived from windborne spores) were frequently observed in intimate contact
with other particles on leaves, which may reflect minimal re-entrainment of the spore due to shelter by the
particles, mycelia development near sources of soluble nutrients provided by the particles, or simply co-
deposition (Smith and Staskawicz, 1977).
Leaves with complex shapes tend to collect more particles than do those with shapes that are more
regular. Conifer needles are more efficient than broad leaves in collecting particles by impaction,
reflecting the small cross-section of the needles relative to the larger leaf laminae of broadleaves and the
greater penetration of wind into conifer canopies than broadleaf ones (U.S. EPA, 2004).
Canopy Surface Effects on Vd
Surface roughness increases particulate deposition, and Vd is usually greater for a forest than for a
nonforested area and greater for a field than for a water surface. Different size particles have different
transport properties and Vd. The upwind leading edges of forests, hedgerows, and individual plants are
primary sites of coarse particle deposition. Impaction at high wind speed and the sedimentation that
follows the reduction in wind speed and carrying capacity of the air in these areas lead to preferential
deposition of larger particles (U.S. EPA, 2004).
Air movement is slowed in proximity to vegetated surfaces. Canopies of uneven age or with a
diversity of species are typically aerodynamically rougher and receive larger inputs of dry-deposited
pollutants than do smooth, low, or monoculture vegetation (Garner et al., 1989; U.S. EPA, 2004).
Canopies on slopes facing the prevailing winds receive larger inputs of pollutants than more sheltered,
interior canopy regions.
All foliar surfaces within a forest canopy are not equally exposed to particle deposition. Upper
canopy foliage tends to receive maximum exposure to coarse and fine particles, but foliage within the
canopy tends to receive primarily fine particles.
2.8.2.4. Nitrogen Deposition and Flux with Biota
Several Nr are species are deposited to vegetation, among them, HN03, N02, and PAN and other
rono2.
Field observations based on concentration gradients of HN03 and using eddy covariance
techniques demonstrate rapid deposition that approaches the aerodynamic limit (as constrained by
atmospheric turbulence) in the Weseley and Lesht (1989) formulation based on analogy to resistance; see
Figure 2-25 and Equation 5. Surface resistance to HN03 uptake by vegetation is negligible and its
deposition rates are independent of leaf area or stomatal conductance, implying that deposition occurs to
branches, soil, and the leaf cuticle as well as leaf surfaces. The HN03 Vd typically exceeds 1 cm/s and
exhibits a diel pattern controlled by turbulence characteristics of midday maxima and lower values at
night in the more stable boundary layer.
Compared with HN03, N02 interaction with vegetation is more difficult to understand in part
because very fast measurements of N02 flux are confounded by the rapid interconversion of NO and N02
with 03 (Gao et al., 1991). Application of 15N-labeled N02 has demonstrated that N02 is absorbed and
metabolized by foliage (Mocker et al., 1998; Segschneider et al., 1995; Siegwolf et al., 2001; Weber et al.,
1995). Exposure to N02 induces activation of N03 reductase (Weber et al., 1995, 1998), a necessary
enzyme for assimilating oxidized N. Current understanding of N02 interactions with foliage is largely
based on leaf cuvette and growth chamber studies, which expose foliage or whole plants to controlled
2-70
-------
N02 concentrations, and measure the fraction of N02 removed from the chamber air. A key finding is that
the fit of N02 flux to N02 concentratin has a non-zero intercept, implying a compensation point or
internal concentration. In studies at very low N02 concentrations, emission from foliage is observed
(Teklemariam and Sparks, 2006). Evidence for a compensation point is not solely based on the fitted
intercept. The N02 uptake rate to foliage is clearly related to stomatal conductance. Internal resistance is
variable, and may be associated with concentrations of reactive species such as ascorbate in the plant
tissue that react with N02 (Teklemariam and Sparks, 2006). Foliar N02 emissions show some dependence
on N content (Teklemariam and Sparks, 2006). Internal N02 appears to derive from plant N metabolism.
Two approaches to modeling N02 uptake by vegetation are the resistance-in-series analogy which
considers flux (F) as the product of concentration (Q and Vd, related to the sum of aerodynamic,
boundary layer, and internal resistances (Ra, Rb, and Rc, respectively, from Equation 5 and Figure 2-25);
by convention, positive fluxes are in the direction from atmosphere to foliage. Note that this approach is
the method most often used to predict deposition in AQMs, that of Wesely and Lesht (1989), as described
above. Typically, the N02 Vd is less than that for 03 due to the surface's generally higher resistance to
N02 uptake, consistent with N02's lower reactivity.
Alternatively, N02 exchange with foliage can be modeled from a physiological standpoint where
the flux from the leaf (J) is related to the stomatal conductance (gs) and a concentration gradient between
the ambient air N02 concentration (Ca) and interstitial air N02 concentration (C,) in the leaf (Ca - Ct) as
depicted in Equation 6.
J = g.(Ca-Ct)
Equation 6
This approach best describes results for exchange with individual foliage elements, and is expressed per
unit leaf or needle area. While this approach provides linkage to leaf physiology, it is not straightforward
to scale up from the leaf to the ecosystem. This model implicitly associates the compensation point with a
finite internal concentration. Typically observed compensation points are ~1 ppb; values of internal N02
concentrations are consistent with metabolic pathways that include NOx. In this formulation, the uptake
will be linear with N02, which is typically measured in foliar chamber studies.
Several studies have shown the UV dependence of N02 emissions, which implies some photo-
induced surface reactions to release N02. Rather than model this as a UV-dependent internal
concentration, it would be more realistic to add an additional term to account for emission that is
dependent on light levels and other surface characteristics
J = gs(Ca-C,)=Js(UV)
Equation 1
PAN is phytotoxic and absorbed at the leaf. Observations based on inference from concentration
gradients and rates of loss at night (Schrimpf et al., 1996; Shepson et al., 1992) and from leaf chamber
studies (Teklemariam and Sparks, 2004) have indicated that uptake of PAN is slower than that of 03.
However, recent work in coniferous canopies with direct eddy covariance PAN flux measurements
indicated a Vd more similar to that of 03. Uptake of PAN is under stomatal control, has non-zero
deposition at night, and is influenced by leaf wetness (Turnipseed et al., 2006). On the other hand, flux
measurements determined by gradient methods over a grass surface showed a Vd closer to 0.1 cm/s, with
uncertainty on the order of a factor of 10 (Doskey et al., 2004). Whether the discrepancies are
methodological or indicate intrinsic differences between different vegetation is unknown. Uptake of PAN
is a smaller loss process than its thermal decomposition in all cases.
The biosphere also interacts with NOx through HC emissions and their subsequent reactions to
form multi-functional RON02. Formation of the hydroxyalkyl nitrates occurs after OH attack on VOCs.
In one sense, this mechanism is simply an alternate pathway for OH to react with NOx to form a rapidly
2-71
-------
depositing species. If VOC were not present, OH would be available to react with N02 when it is present
to form HN03.
Isoprene nitrates are an important class of R0N02. Isoprene reacts with OH to form a radical that
adds N02 to form the hydroxyalkyl nitrate. The combination of hydroxyl and N03 functional group
makes these compounds especially soluble with low vapor pressures, so they likely deposit rapidly
(Shepson et al., 1996; Treves et al., 2000). Many other unsaturated HCs react by analogous routes.
Observations at Harvard Forest show a substantial fraction of the total of all gas-phase forms of oxidized
N is not accounted for by NO, N02 and PAN, and the remainder is attributed to the missing fraction of
RONO2 (Horii et al., 2006; Munger et al., 1998). Furthermore, the total NOY flux exceeds the sum of
HNO3, NOx, and PAN, which implies that the RON02 are a substantial fraction of the total N deposition.
Other observations showing evidence of hydoxyalkyl nitrates include those of Grossenbacher et al. (2001)
and Day et al. (2003).
HN02 formation on vegetative surfaces at night has long been observed based on measurements of
positive gradients (Harrison and Kitto, 1994). Surface reactions of N02 enhanced by moisture were
proposed to explain these results. Production was evident at sites with high ambient N02; at low
concentration, uptake of HN02 exceeded the source. Daytime observations of HN02 when rapid
photolysis is expected to deplete ambient concentrations to very low levels implies a substantial source of
photo-induced HN02 formation at a variety of forested sites where measurements have been made.
Estimated source strengths are 200 to 1800 ppt/h in the surface layer (Zhou et al., 2002a, 2003b), which is
-20 times faster than all nighttime sources.
HN02 sources could be important to HOx budgets as HN02 is rapidly photolyzed by sunlight to
OH and NO. Additional evidence of light-dependent reactions to produce HN02 comes from discovery of
a HN02 artifact in Pyrex sample inlet lines exposed to ambient light. Either covering the inlet or washing
it eliminated the HN02 formation (Zhou et al., 2002b). Similar reactions might serve to explain
observations of UV-dependent production ofNOx in empty foliar cuvettes that had been exposed to
ambient air (Hari et al., 2003; Raivonen et al., 2003).
Production of HN02 in the dark is currently believed to occur via a heterogeneous reaction
involving N02 on wet surfaces (He et al., 2006; Jenkin et al., 1988; Pitts et al., 1984; Sakamaki et al.,
1983). It has been proposed that the mechanism has first-order dependence in both N02 and H20
(Kleffmann et al., 1998; Svensson et al., 1987) despite the stoichiometry. However, the molecular
pathway of the mechanism is still under debate. Jenkin et al. (1988) postulated a H20»N02 water complex
reacting with gas phase N02 to produce HN02, which is inconsistent with the formation of an N204
intermediate leading to HN02 as proposed by Finlayson-Pitts et al. (2003). Another uncertainty is whether
the reaction forming HN02 is dependent on water vapor (Stutz et al., 2004; Svensson et al., 1987) or
water adsorbed on surfaces (Kleffmann et al., 1998). Furthermore, the composition of the surface and the
available amount of surface or surface-to-volume ratio can significantly influence the HN02 production
rates (Kaiser and Wu, 1977; Kleffmann et al., 1998; Svensson et al., 1987), which may explain the
difference in the rates observed between laboratory and atmospheric measurements.
There is no consensus on a chemical mechanism for photo-induced HN02 production. Photolysis of
HNO3 or N03 absorbed on ice or in surface water films has been proposed (Honrath et al., 2002;
Ramazan et al., 2004; Zhou et al., 2001, 2003b). Alternative pathways include N02 interaction with
organic surfaces such as humic substances (George et al., 2005; Stemmler et al., 2006). Note that either
N03 photolysis or heterogeneous reaction of N02 are routes for recycling deposited NOx back to the
atmosphere in an active form. N03 photolysis would return N that heretofore was considered irreversibly
deposited, while surface reactions between N02 and water films or organic molecules would decrease the
effectiveness of observed N02 deposition if the HN02 were re-emitted.
2-72
-------
2.8.3. Air Quality Model Evaluation
Urban and regional air quality is determined by a complex system of coupled chemical and
physical processes including emissions of pollutants and pollutant precursors, complex chemical
reactions, physical transport and diffusion, and wet and dry deposition. NOx in these systems has long
been known to act nonlinearly in P(03) and other secondary pollutants (Dodge, 1977); to extend over
multiple spatial and temporal scales; and to involve complicated cross-media environmental issues such
as acidic or nutrient deposition to sensitive biota and degradation of visibility.
NOy species emitted and transformed from emissions control the production and fate of 03 and
aerosols by sustaining or suppressing OH cycling. Correctly characterizing the interrelated NOY and OH
dynamics for 03 formation and fate in the polluted troposphere depends on new techniques using
combinations of several NOY species for diagnostically probing the complex atmospheric dynamics in
typical urban and regional airsheds.
Arnold et al. (1998) described a model evaluation methodology that distinguished several types of
AQM testing. Two components of that methodology were: operational testing to judge the performance
and overall behavior of a model over specific attributes; and diagnostic testing to help reveal potential
compensating error in model inputs or processing. Diagnostic testing is in situ testing of model
components using data that emphasize atmospheric processes, often with mass balance techniques, special
species ratios, and process rate and reaction rate information not typically stored by the model for output.
Some of these probes have been developed through process-oriented studies using theoretical
assumptions, model-derived explanations, and results from instrumented models ranging from one-
dimensional box models to the full 4-dimensional photochemical modeling system (Tonnesen and Dennis,
2000). Additional information on instrumenting AQMs for diagnostic analysis with model process and
reaction rate information is found in Dennis et al. (2002); information pertaining to the specific
implementation of these techniques in CMAQ is found in (Gipson and Young, 1999); and results from
application of diagnostic probes to modeling experiments are found in Arnold and Dennis (2003, 2006).
Evaluation results from a recent U.S. EPA exercise of CMAQ in the Tampa Bay, FL, airshed are
presented here as an example of the present level of skill of state-of-the-science AQMs for predicting
atmospheric concentrations of the relevant NOx, SOx, and NHX species for this NAAQS assessment. This
modeling series exercised CMAQ version 4.4 and with the University of California at Davis (UCD)
sectional aerosol module in place of the standard CMAQ modal module and was driven by meteorology
from MM5 v3.6 and with NEI emissions as augmented by continuous emissions monitoring data where
available. (The UCD size-segregated module was preferred for this application because of the importance
of sea salt particles in the bay airshed. Testing of this new engineering extension revealed that its
performance was very similar to CMAQ's standard modal; hence, model behavior and performance
reported here can stand as a general indication of CMAQ's skill.)
The CTM was run with 21 vertical layers for the month of May 2002. For this evaluation, CMAQ-
UCD was run in a one-way nested series of three domains with 32 km, 8 km, and 2 km horizontal grid
spacings from the CONUS (32 km) to central Florida and the eastern Gulf of Mexico (2 km).
Depictions of the 8 km and 2 km domains used here zoomed over the central Tampa area are shown
in Figure 2-27 and Figure 2-28.
2-73
-------
8KM RESOLUTION; ZOOMED2
. v.
rainpa
Sydney
Gaudy
aipcicrtibuio
Gulf of Mexico
him i- i>Hi
Figure 2-27. 8 km southeast U.S. CMAQ domain zoomed over Tampa Bay, FL.
2KWI RESOLUTION; ZOOMED2
SIPOttfflmHB
Guff of Mexico
Figure 2-28. 2 km southeast U.S. CMAQ domain zoomed over Tampa Bay, FL.
2.8.3.1. Ground-based Comparisons of Photochemical Dynamics
Errors in the NOv concentrations in the model most likely from on-road emissions (Figure 2-29)
affected NOx predictions, but CMAQ-UCD's general responses were reasonable. The model also
replicated well anthropogenic and biogenic VOC emissions; see Figure 2-30 and Figure 2-31,
2-74
-------
respectively. After initial errors leading to underprediction in the first 21 days, CMAQ's predictions of
hourly PM2 5 concentrations and trends over the whole month also replicated the observed concentrations
well; see Figure 2-32.
Z2
Hours (EST)
20
Hours (EST)
20
Hours (EST)
Figure 2-29. Hourly averages for May 1-31,2002. CMAQ 8 km and 2 km results and measured
concentrations of NO (top), NO2 (middle), and total NOx (bottom). Legend: blue circles =
Observations (OBS); green squares = 8 km; red diamonds = 2 km CMAQ solutions.
2-75
-------
4 T
3 - -
¦Q
Q.
a.
LU
1 -May 2-May 3-May 4-May 5-May 6-May 7-May 8-May 9-May 10-
May
Hours (date mark = 0000 EST)
4 r
3 - -
2 - -
11- 12- 13- 14- 15- 16- 17- 18- 19- 20-
May May May May May May May May May May
Hours (date mark = 0000 EST)
4 r
3 - -
v 2 --
21- 22- 23- 24- 25- 26- 27- 28- 29- 30- 31-
May May May May May May May May May May May
Hours (date mark = 0000 EST)
Figure 2-30. May 2002 daily concentrations and 8 km CMAQ predictions for ethene at Sydney, FL. Legend:
blue circles = OBS; green squares = 8 km CMAQ solutions.
2-76
-------
8 T
6 - -
¦Q
Q.
Q.
I 4--
c
2
Q.
O
t/)
2 - -
1 -May 2-May 3-May 4-May 5-May 6-May 7-May 8-May 9-May 10-
May
Hours (date mark = 0000 EST)
8
6
4
2
o ¦ I ¦ I ¦ I ¦ I ¦ I ¦ I ¦ I ¦ \ ¦ I ¦ I ¦ I ¦ I ¦ I iffi ¦ I .g|> ¦ I ¦*! . i
11 -May 12-May 13-May 14-May 15-May 16-May 17-May 18-May 19-May 20-May
Hours (date mark = 0000 EST)
4
3
2
1
0
21- 22- 23- 24- 25- 26- 27- 28- 29- 30- 31-
May May May May May May May May May May May
Hours (date mark = 0000 EST)
Figure 2-31. May 2002 daily concentrations and 8 km CMAQ predictions for isoprene at Sydney, FL.
Legend: blue circles = OBS; green squares = 8 km CMAQ solutions.
2-77
-------
40 -•
1 -May 2-May 3-May 4-May 5-May 6-May 7-May 8-May 9-May 10-
May
Hours (date mark = 0000 EST)
w 30
11 -May 12-May 13-May 14-May 15-May 16-May 17-May 18-May 19-May 20-May
Hours (date mark = 0000 EST)
w 30
21- 22- 23- 24- 25- 26- 27- 28- 29- 30- 31-
May May May May May May May May May May May
Hours (date mark = 0000 EST)
Figure 2-32. Observed hourly PM2.5 concentrations at St. Petersburg, FL and results from CMAQ 8 km.
Legend: blue circles = 0BS; green squares = 8 km CMAQ solutions.
2-78
-------
140 j
>120--
(0
£100 -¦
I" 80 - -
3
o
.= 60 +
o
P3
d)
40 --
O
o ° 0 8
§gi 8
o o
o ^ ^ q i o
o ©
iiiiiiiiiiiiitlf i
0.00 2.00 4.00 6.00 8.00 10.0 12.0 14.0 16.0 18.0 20.0 22.0
0 0 0 0 0 0 0
Diurnal Time (EST)
200 j
co
180 ¦¦
160 --
0
CT)
140 --
120 ¦¦
O
O
100 --
t
80 --
TO
60 --
O
40 ¦¦
O
O
20 -•
a
LL
0 -¦
.a , ^ M .
0.0-0.5 0.5-1.0 1.0-3.0 3.0-5.0 5.0-10 10-15 15-25 >25
O-yNOv ratio
Figure 2-33. Observed and modeled ratios of O3 to N0X. Diel curves from hourly averages over May 1-31,
2002 (top), and distribution of O3 to NOx ratio values binned to show fractions of total daylight
hours, May 1-31, 2002 (bottom panel). Legend: blue circles or bars = OBS; green squares or
bars = 8 km; red triangles or bars = 2 km.
The P(03) efficiency curves in the model agreed well with those observed at Sydney, FL; see
Figure 2-33. However, tests of CH20 (Figure 2-34) and H202 (Figure 2-35) seemed to indicate an error in
the model's OH chemistry related to these radical reservoir species since both were substantially and
systematically different from the observations at the ground-based Sydney site. These species have
historically been very difficult to model well, however, and the overall excellent agreement of CMAQ-
UCD to production curves in relation to NOx processing mean that this error was likely restricted to these
species and of limited influence in the overall model solutions and for this evaluation.
2-79
-------
8 T
o o
0.00 2.00 4.00 6.00 8.00 10.00 12.00 14.00 16.00 18.00 20.00 22.00
Diurnal Time (EST)
8 T
o O
0 5 10 15
NOz (ppb)
Figure 2-34. Observed and CMAQ 8 km and 2 km predicted formaldehyde concentrations. Hourly averages
from each day, May 1-31, 2002 (top), and formaldehyde concentrations as a function of NOz
concentrations (bottom). Legend: blue circles = OBS; green squares = 8 km; red triangles or
diamonds = 2 km.
2-80
-------
0
1-May 2-May 3-May 4-May 5-May 6-May 7-May 8-May 9-May 10-
May
Hours (Date Mark = 0000 EST)
o
11-May 12-May 13-May 14-May 15-May 16-May 17-May 18-May 19-May 20-May
Hours (Date Mark = 0000 EST)
21-May22-May23-May24-May25-May26-May27-May28-May29-May30-May31 -May
Hours (Date Mark = 0000 EST)
Figure 2-35. Hourly concentrations of hydrogen peroxide, observed and predicted by CMAQ 8 km and 2
km, May 1-31, 2002 at Sydney, FL. Legend: blue circles = OBS; green squares = 8 km; red
diamonds = 2 km.
2-81
-------
2.8.3.2. Predicted Chemistry for Nitrates and Related Compounds
pN03 plays a crucial and complex role in the health of aquatic and estuarine ecosystems. Gas-phase
N03 replacement of CI on sea salt particles is often favored thermodynamically and the Vd of the coarse
N03 formed through this replacement is more than an order of magnitude greater than for fine N03. Over
open bodies of salt water such as the Gulf of Mexico and Tampa Bay, FL, N03 from this reaction
dominates dry deposition and is estimated to be of the same order as N03 wet deposition.
However, N03 concentrations are driven, buffered, and altered by a wide range of photochemical
gas-phase reactions, heterogeneous reactions, and aerosol dynamics, making them especially difficult to
model well. Because pN03 is derived mostly from gas-phase HN03 and will interact with Na+, NH4+, Cl~,
and S042 , all these species and the physical parameters governing their creation, transport,
transformation, and fate must be accurately replicated to predict pN03 with high fidelity. This has
historically been a difficult problem for numerical process models, owing not least to the pervasive dearth
of reliable ambient measurements of N03~in its various forms. Normalized mean error (NME) for the
large-scale Eulerian CTM-predicted pN03 has typically been on the order of a factor of 3 greater than the
NME for pS04 (Odman et al., 2002; Pun et al., 2003).
S042 . NH4+, Na+, and CP were all predicted to within a factor of 2 and with no significant bias
during the photochemical day in the 8 km CMAQ-UCD solution, although a significant bias in Na+ and
Cl was evident in the 2 km solution for two near water sites. This grid-size dependent bias is still being
explored. Size segregation maxima were correct to within two size bins every day for which there were
observations for both S042 and NH4+ (0.2 to 1.0 |im). and Na+ and Cr(2.0 to 10.0 |im). Cl~
concentrations were greatly overpredicted during dark hours, but were nearer to observed values during
the photochemical day. CMAQ performance for HN03 and NH3 are shown in Figure 2-37 and Figure
2-40, respectively.
Figure 2-36 shows that CMAQ-UCD systematically underpredicted the hourly time series of
measured pN03 concentrations at the Sydney supersite, the only location with discrete pN03 data. These
time series data establish that CMAQ-UCD's largest errors were on four days in the first 2 weeks of the
month, but that the total peak pN03 concentrations were nearly all underpredicted.
Since pN03 is derived in large part from gas-phase HN03, its underprediction may be due to an
underprediction of HN03 concentrations or an underrepresentation of the gas- to aerosol-phase change. At
Sydney, FL, in fact, both these conditions held. Figure 2-37 depicts the model's bias for HN03
underprediction in both the 8 km and 2 km solutions, excepting four days of very large peak
overpredictions. This trend was especially true overnight; on eight other days the model overpredicted the
one hour peak concentration as well, though not so substantially, but the chief effect was still one of an
artificial and inappropriate N limitation in the model.
A time series molar equivalent ratio of HN03 to total N03 depicts which phase stores the N03 and
how that storage ratio changes over time. Figure 2-38 shows that at Sydney, FL, CMAQ-UCD stored too
much N03 in the gas phase as HN03 (and recall that the daytime HN03 concentrations were sometimes
overpredicted by the model) and too little in the gas phase overnight, when the model was regularly low
against the measurements; compare Figures 2-37 and 2-38. Note again here the self-similarity of the 8 km
and 2 km solutions in this comparison.
Interestingly, the 23 hour integrated data did not reveal this important difference in nitrate form
between the model and measurements as Figure 2-39 shows in the stacked bar percentage plots of fine
and coarse pN03 together with gas-phase HN03. Both the 8 km (Figure 2-39 (middle panel) and the 2 km
(Figure 2-39 (bottom panel) solutions predicted distributions between the two general ranges of aerosol
size, and between gas and aerosol phases, with good fidelity to the daily observations (Figure 2-39 (top
panel)) at Sydney, FL. This result illustrates that while discrete time series data are crucial for diagnosing
model behavior, on the integrated total daily and longer basis used for computing total annual N loads,
CMAQ-UCD predicted approximately the correct distributions for pN03, even though the total N03
concentration prediction was biased low.
2-82
-------
6 T
' 2-May 3-May 4-May 5-May 6-May 7-May 8-May 9-May 10-
May
Hours (date mark = 0000 EST)
11-May 12-May 13-May 14-May 15-May 16-May 17-May 18-May 19-May 20-May
Hours (date mark = 0000 EST)
21-May22-May23-May24-May25-May26-May27-May28-May29-May30-May31-May
Hours (date mark = 0000 EST)
Figure 2-36. Hourly and CMAQ-UCD-predicted total pN03 concentrations at Tampa Bay, FL, and
observations at Sydney, FL May 1-31, 2002. Legend: blue circles = OBS; green x's = 8 km; red
diamonds = 2 km.
2-83
-------
0
1-May 2-May 3-May 4-May 5-May 6-May 7-May 8-May 9-May 10-
May
Hours (Date Mark = 0000 EST)
21-May22-May23-May24-May25-May26-May27-May28-May29-May30-May31-IVlay
Hours (Date Mark = 0000 EST)
11-May 12-May 13-May 14-May 15-May 16-May 17-May 18-May 19-May 20-May
Hours (Date Mark = 0000 EST)
Figure 2-37. Hourly and CMAQ-predicted HNO3 concentrations at Sydney, FL, May 1-31, 2002. Legend:
blue circles = OBS; green squares = 8 km; red diamonds = 2 km.
2-84
-------
o
Eo.6
0.0
1-May 2-May 3-May 4-May 5-May 6-May 7-May 8-May 9-May 10-May
Hours (date mark = 0000 EST)
0.8
0.0 I T I ¦ I ¦ I ¦ I ¦ I ¦ I ¦ I T I ¦ I ¦ I T I ¦ I ¦ I T I T I ¦ I ¦ I T I ¦ I
11-May 12-May 13-May 14-May 15-May 16-May 17-May 18-May 19-May 20-May
Hours (date mark = 0000 EST)
0.8 -¦
0.2 --
0.0 I i I ¦ I ¦ I ¦ I ¦ I ¦ I i I i I ¦ I ¦ I ¦ I ¦ I ¦ I i I ¦ I ¦ I ¦ I ¦ I ¦ I i I ¦ I
21-May22-May23-May24-May25-May26-May27-May28-May29-May30-May31~May
Hours (date mark = 0000 EST)
Figure 2-38. Hourly and CMAQ-UCD-predicted ratio of HNO3 to total NO3 at Tampa Bay, FL and
observations at Sydney, FL, May 1-31, 2002. Legend: blue circles = OBS; green x's = 8 km; red
diamonds = 2 km.
2-85
-------
MOUDI aerosol N03 and ARA HN03 (23 h daily meai)
1
5/1 5/3 5/5 5/7 5/9 5/11 5/13 5/15 5/17 5/19 5/21 5/23 5/25 5/27 5/29 5/31
Date, 2002
¦ <25um 125-10 irn ~ HN03
CMAQ-UCD 8 km solution (23 h daily meai)
1
5/1 5/3 5/5 5/7 5/9 5/11 5/13 5/15 5/17 5/19 5/21 5/23 5/25 5/27 5/29 5/31
¦ <25 im 125-10 irn DHN03
CMAQ-UCD 2 km solution (23hotivteans)
I
5/1 5/3 5/5 5/7 5/9 5/11 5/13 5/15 5/17 5/19 5/21 5/23 5/25 5/27 5/29 5/31
Date. 2002
¦ <2.5 um 12.5-10 um ~ HN03
Figure 2-39. CMAQ-UCD predicted fractions and totals of total NO3 for days in May 2002 with
measurements in Tampa Bay, FL. Legend: red bars = <2.5 |jm; blue bars = 2.5-10 |jm; green
bars = HNO3. Measured concentrations (top panel), CMAQ-UCD 8 km solution (middle panel)
and CMAQ-UCD 2 km solution (bottom panel).
2-86
-------
-May 2-May 3-May 4-May 5-May 6-May 7-May 8-May 9-May 10-
May
Hours (date mark = 0000 EST)
11- 12- 13- 14- 15- 16- 17- 18- 19- 20-
May May May May May May May May May May
Hours (date mark = 0000 EST)
-May22-May23-May24-May25-May26-May27-May28-May29-May30-May31-May
Hours (date mark = 0000 EST)
Figure 2-40. Hourly and CMAQ-predicted NH3 concentrations at Sydney, FL, May 1-31, 2002. Legend: blue
circles = OBS; green squares = 8 km; red diamonds = 2 km.
While the coarse fraction inorganic aerosol anion totals were dominated by NO;, . and S042
dominated the fine fraction aerosol inorganic anions, there was sufficient NHX (NHX = NH3 + NFL) at
Sydney, FL, to form fine aerosol NH4NO3 in some circumstances. Figure 2-40 depicts the hourly mass
concentration of NH3 at Sydney, FL, showing again the strong self-similarity of the 8 km and 2 km
2-87
-------
solutions. Each solution, however, underpredicted the measured NH3 concentrations consistently, and
especially for the nine very large excursions of 10-20 |ig/m3 during the month.
Overall, CMAQ-UCD was found to be operationally sound in this evaluation of its 8 km and 2 km
solutions for the Tampa Bay airshed using the ground-based and aloft data (not shown here) from the May
2002 field intensive. Moreover, results from diagnostic tests of the model's photochemical dynamics were
generally in excellent agreement with results from the ambient atmosphere. However, CMAQ-UCD was
biased low in this application for total N03 and for N03 present as gas-phase HN03. In addition, the
model was biased low for the HOx radical reservoir species CH20 and H202, though this bias appeared to
have been limited to these species. Performance of the new UCD aerosol module was judged to be
entirely adequate, allocating aerosols by chemical makeup to the appropriate size fractions. Model
performance for fine-mode aerosols was also judged to be fully adequate.
2.8.3.3. Evaluating Deposition with CTMs
Global CTM Performance
Both wet and dry deposition are of necessity highly parameterized in all CTMs. While all current
models implement resistance schemes for dry deposition, the Vd generated from different models can vary
highly across terrain types (Stevenson et al., 2006). The accuracy of wet deposition in global CTMs is tied
to spatial and temporal distribution of model precipitation and the treatment of chemical scavenging.
Dentener et al. (2006b) compared wet deposition across 23 models with available measurements around
the globe. Figure 2-41 and Figure 2-42 extract results of a comparison of the 23-model mean versus
observations over the eastern U.S. for pN03 and pS04 deposition, respectively. The mean model results
were strongly correlated with the observations (r >0.8), and usually captured the magnitude of wet
deposition to within a factor of 2 over the eastern U.S. Dentener et al. (2006b) concluded that 60 to 70%
of the participating models captured the measurements to within 50% in regions with quality controlled
observations.
600
400 -
!
x
200
0
200
400
600
Measurement
Source: Dentener et al. (2006b). Reprinted with permission.
Figure 2-41. Scatter plot of total nitrate (HNO3 plus PNO3) wet deposition (mg N/m2/yr) of the model mean
versus measurements for the North American Deposition Program (NADP) network. Dashed
lines indicate factor of 2. The gray line is the result of a linear regression fitting through zero.
2-88
-------
1000
800
600
400
200
2 param lit y ¦ 114.0 ~ 0.77x
1 pararn fit J -1.00x
| pfltaiBflt within : HHi: 66-0
0
200
400
600
800
1000
Measurement
Source: Dentener et al. (2006b). Reprinted with permission.
Figure 2-42. Scatter plot of total SO42" wet deposition (mg S/m2/yr) of the model mean versus
measurements for the National Atmospheric Deposition Program (NADP) network. Dashed
lines indicate factor of 2. The gray line is the result of a linear regression fitting through zero.
2.8.3.4. Regional CTM Performance
Regional CTM performance for concentration and deposition of relevant NOx and SOx species is
illustrated here with examples from CMAQ version 4.6.1 as configured and run for exposure and risk
assessments reported in the Draft Risk and Exposure Assessment for the Review of the Secondary
National Ambient Air Quality Standards for Oxides of Nitrogen and Oxides of Sulfur (U.S. EPA 2008c);
additional details on the model configuration and application are found there. A map of the 36 km parent
domain and two 12 km (east and west) progeny domains appears in Figure 2-43.
Figure 2-43. CMAQ modeling domains for the OAQPS risk and exposure assessments: 36 km outer parent
domain in black; 12 km western U.S. (WUS) domain in red; 12 km eastern U.S. (EUS) domain
in blue.
2-89
-------
Comparisons from the 2002 annual run of CMAQ for the exposure assessment are shown here
against measured concentrations and deposition totals from nodes in three networks: IMPROVE, CSN
(labeled STN in the plots) and CASTNet; see the full description of these networks in Section 2.9 below.
Comparisons were made as model-observation pairs at all sites having sufficient data for the seasonal or
the 2002 annual time period ill the two 12 km east and west domains and were evaluated with these
descriptive statistics: correlation, root mean square error, normalized mean bias, and normalized mean
error.
Summertime pS04 concentrations are well predicted by CMAQ, to within a factor of 2 at nearly
even point, and with R2 >0.8 across all three networks; see Figure 2-44. This result tracks the generally
well-predicted pS04 concentrations found in earlier CMAQ evaluations: see Eder and Yu, 2006; Mebust
et aL 2003; and Tesche et aL 2006. Since pS04 concentrations are strongly a function of precipitation,
care must be taken to ensure that the meteorological solution driving individual CMAQ chemical
applications produces precipitation fields with low bias as discussed by Appel et aL (2008).
2002ac_m«l2v33_l2kmE S04 for June to August 2002
~ IMPROVE {20G2ac me12v33 12kmE| *
A STN(20Q2ac met2v33 l2kmE) a* /
CASTNet {2002ac met2v33 12kmE) /
o
00
CM
O
2
12
0
4
8
10
14
6
Figure 2-44, 12-km EUS Summer sulfate PM, each data point represents a paired monthly averaged
(June/July/August) observation and CMAQ prediction at a particular IMPROVE, STN, and
CASTNet site. Solid lines indicate the factor of 2 around the 1:1 line shown between them.
Wintertime pN03 (Figure 2-45) and total NO.< (HNO3+PNO3) (Figure 2-46) concentrations are
predicted less well by CMAQ; but N03~ is a pervasively difficult species to measure and model for the
reasons described in detail in Sections 2.6 and 2.7 above. Still, at the CASTNet nodes where the total NO;
concentrations are higher than they are at all but a few of the remote IMPROVE sites, CMAQ predicts
concentrations for nearly every node to within a factor of 2 and with an R2 >0.8.
2-90
-------
2002ac_m.et2v33_12l
CO
Monthly Average
TN03 { ug/m3}
2002ac mel2v33 12kmE
o
0
1
2
3
5
6
7
4
Observation
Figure 2-46. 12-km EUS Winter total nitrate (HNO3 +total pNCh), each data point represents a paired
monthly averaged (December/January/February) observation and CMAQ prediction at a
particular CASTNet site. Solid lines indicate the factor of 2 around the 1:1 line shown between
them.
These CMAQ-predicted concentrations, coupled with modeled cloud and precipitation fields
produce wet deposition fields for SO f and N03 in the east domain as shown in Figures 2-47 and 2-48,
respectively.
2-91
-------
Z002ac_met2v33_12kmESO4 lor 20020101 |o 20021231
n NADP^dep {2002ac met2v33_ 12kmE)
in
o
o
0
5
10
15
20
25
30
35
Observation
Figure 2-47. 12-km EUS annual sulfate wet deposition, each data point represents an annual average
paired observation and CMAQ prediction at a particular NADP site. Solid lines indicate the
factor of 2 around the 1:1 line shown between them.
2Q02ac_met2v33_12kmE N03 tor 20020101 to 20021231
~ NADP
-------
2.8.4. Computing Atmospheric Deposition to Specific Locations
Inputs of new N, i.e., non-recycled, exogenous N mostly anthropogenic in origin, are often key
factors controlling primary productivity in N-sensitive estuarine and coastal waters (Paerl et al., 2000).
Increasing trends in urbanization, agricultural intensity, and industrial expansion have led to increases in
N deposited from the atmosphere on the order of a factor of 10 in the previous 100 years (Swackhamer
et al., 2004). Direct loadings of atmospheric N to ocean and gulf waters along the northeast and southeast
U.S. are now roughly equal to or exceed the load of new N from riverine inputs (Duce et al., 1991) at 11,
5.6, and 5.6 kg N/ha for the northeast Atlantic coast of the U.S., the southeast Atlantic coast of the U.S.,
and the U.S. eastern Gulf of Mexico, respectively (Paerl, 2002).
This N deposition takes different forms physically and chemically. Physically, deposition can be
direct, with the loads resulting from air pollutants depositing directly to the surface of a body of water,
usually a large body of water like an estuary or lake. In addition, there is indirect deposition component
derived from deposition to the rest of the watershed, both land and water, of which some fraction is
transported through runoff, rivers, streams, and groundwater to the waterbody of concern.
Airshed extents developed and used courtesy of R. Dennis, U.S. EPA/ORD/NERL/Atmospheric Modeling Division.
Figure 2-49. Principal airsheds and watersheds NOx for these estuaries: Hudson/Raritan Bay; Chesapeake
Bay; Pamlico Sound; and Altamaha Sound.
2-93
-------
Direct and indirect depositions to watersheds depend on air pollutant concentrations in the airshed
above the watershed. The shape and extent of the airshed is quite different from that of the watershed. In a
watershed, everything that falls in its area, by definition, flows into a single body of water. An airshed, by
contrast, is a theoretical concept that defines the area containing the emissions contributing a given level,
often 75%, to the deposition in a particular watershed or to a given waterbody. Hence, airsheds are
modeled domains containing the sources estimated to contribute a given level of deposition from each
pollutant of concern. Four U.S. East Coast watersheds and their corresponding NOx airsheds are shown in
Figure 2-49.
Table 2-10. Atmospheric N loads relative to total N loads in selected great waters *
Waterbody
Total N Load
(million kg/yr)
Atmospheric N Load
(million kg/yr)
Percent Load from the Atmosphere
Albemarle-Pamlico Sounds
23
9
38
Chesapeake Bay
170
36
21
Delaware Bay
54
8
15
Long Island Sound
60
12
20
Narragansett Bay
5
0.6
12
New York Bight
164
62
38
BASED ON ADN LOADS FROM THE WATERSHED ONLY (EXCLUDING DIRECT N DEPOSITION TO THE BAY SURFACE):
Waquoit Bay, MA
0.022
0.0065
29
BASED ON ADN DIRECTLY TO THE WATERBODY (EXCLUDING ADN LOADS FROM THE WATERSHED):
Delaware Inland Bays
1.3
0.28
21
Flanders Bay, NY
0.36
0.027
7
Guadalupe Estuary, TX
4.2-15.9
0.31
2-8
Massachusetts Bays
22-30
1.6-6
5-27
Narragansett Bay
9
0.4
4
Newport River Coastal Waters, NC
0.27-0.85
0.095-0.68
>35
Potomac River, MD
35.5
1.9
5
Sarasota Bay, FL
0.6
0.16
26
Tampa Bay, FL
3.8
1.1
28
ADN = atmospheric deposition of N
Source: U.S. EPA, 2000d
N inputs have been studied in several East and Gulf Coast estuaries owing to eutrophication
concerns there. N from atmospheric deposition in these locations is estimated to be 10 to 40% of the total
input of N to many of these estuaries, and could be higher for some. Estimates of total N loadings to
estuaries or to other large-scale elements in the landscape are then computed using measurements of wet
and dry N deposition where these are available and interpolated with or without a set of air quality model
predictions such as the Extended Regional Acid Deposition Model (Ext-RADM) (Dennis et al., 2001;
2-94
-------
Mathur and Dennis, 2003). Ext-RADM has been shown to capture spatial and seasonal variations in N
deposition; to predict the constituent deposition species correctly; and to simulate the chemistry and
physics relating reduced and oxidized forms of N with high validity.
Extensive evaluation by Mathur and Dennis of the performance of Ext-RADM showed that model-
predicted ambient levels, gas-to-particle partitioning ratios, and deposition totals were in good agreement
with available measurements, having R2 for both annual and seasonal totals in the range of 0.4 to 0.7 for
most species. Ext-RADM correctly predicted that most particles in the eastern U.S. are fully neutralized,
further demonstrating that most modeled chemistry is correct as judged against measurements.
Experiments with Ext-RADM to characterize atmospheric conditions over the eastern U.S. in the period
at the end of the 1980s and early 1990s showed that the model predicted that reduced N species were
contributed 47 ± 8% of total N wet deposition, in excellent agreement with the number inferred from
measurements, 43 ± 9%.
Table 2-10 lists several waterbodies for which atmospheric N inputs have been computed and
ratioed to total N loads. The contribution from the atmosphere ranges from a low of 2-8% for the
Guadalupe Estuary in South Texas to highs of -38% in the New York Bight and the Albemarle-Pamlico
Sound in North Carolina.
The North Carolina case is a particularly rich example of computing defined airsheds and
characterizing the contributions from oxidized and reduced forms of N to underlying water bodies; see the
summary reported by Dennis and Mathur (2001). The Albemarle-Pamlico principal N airsheds were
computed to be 665,600 km2 for oxidized N, and 406,400 km2 for reduced N; these are factors of 25 and
15, respectively, larger than the watershed drainage area. NO emissions from within the oxidized N
principal airshed was estimated to explain 63% of all oxidized N deposition to the Albemarle-Pamlico
system, very similar to the total of 60% of all reduced N deposition accounted for by NH3 emissions in
the reduced N principal airshed. The regional component to these computed N deposition totals varied
with the form of N such that local NH3 emissions inside North Carolina were estimated to account for
45% (hence, 55% left from the regional component) of the total reduced N deposition, while local NO
and N02 emissions accounted for only 20% of the oxidized N deposition total (hence, leaving a regional
component of 80%).
Chemically, N deposited from the atmosphere directly or indirectly can be present as an oxide or in
reduced form as NH3 and NH/ or as dissolved or particulate organic N; see the listing in Table 2-11 for a
division of these and an approximate ranking of source strengths. NO and N02, chiefly from fossil fuel
combustion, dominate total N pollution in the U.S. at -50 to 75% of the total; see the descriptions of this
chemistry in Section 2.6.2 and of sources in Section 2.2. above.
CAFOs and other intensified agricultural production methods have resulted in greatly increased
volumes of animal wastes, of which 30 to 70% may be emitted as NH3 (Whitall and Paerl, 2001). The
increase in reduced N deposition in the U.S. measured as increased NH44" deposition correlates well with
the local and regional increases in this agricultural intensity (Whitall and Paerl, 2001). Moreover, the
increases in NH4+ deposition in the U.S. track the effects in Europe where animal operations have
dominated agricultural production for much of the previous 100 years and where NH4+ is the dominant
form of N deposited from the atmosphere (Holland et al., 1999). Tables 2-12 and 2-13 list several
important watersheds and their respective oxidized (Table 2-12) and reduced (Table 2-13) airsheds.
Airsheds for oxidized N tend to be larger than those for reduced N owing to differences in the transport
and deposition of NOx and NHX described above in Section 2.6.
2-95
-------
Table 2-11. Natural and anthropogenic sources of atmospheric N compounds.
Chemical Form
Sources (in approximate order of importance)
Reduced N
Agricultural
NH3NH4
Livestock waste (volatilized NH3)
Chemical fertilizers (volatilized NH3)
Biomass burning
Dust from deforestation and land clearing
Urban and Rural (non-agricultural)
Wastewater treatment (volatilized NH3)
Fossil fuel combustion (from automobile catalytic converters)
Natural
Biomass burning (forest and grass fire)
Decomposition of organic matter
Dust and aerosols
Volcanism
Oxidized Nitrogen
Urban and Rural (non-agricultural)
NO, N02, NO3-
Fossil fuel combustion
Mobile and stationary engines
Powerplants and industrial
Natural
Biomass burning
Lightning
Photolysis of N2O (air, land, water)
Dust and aerosols generated by storms
Microbially mediated volatilization
Organic Nitrogen
Agricultural
(Dissolved and Particulate)
Dust and volatilization of wastes*
Urban and Rural (non-agricultural)
Dust or aerosols*
Natural
Atmospheric photochemical and lighting
Biological production in oceans*
* = possible, but little known about sources (the major chemical forms of atmospheric N compounds are the reduced, oxidized, and organic forms)
Source: Swackhamer et al. (2004). Reprinted with permission.
Considerable uncertainty attaches to estimates of the third form of atmospherically derived N,
organic N, in part because convenient methods for measurement and analysis are not widely available; see
Table 2-11. Intensive studies at individual sites have shown, however, that for the North Carolina coast,
for example, 30% of rain water N and deposition consisted of organic N, 20-30% of which was then
available to primary producers on time scales of hours to days (Peierls and Paerl, 1997).
2-96
-------
Table 2-12.
Characteristics of oxidized-nitrogen airsheds.
Watershed
Size
(km2)
Size Factor
over
Watershed
Area
% Oxidized-N
Deposition Explained
Airshed NOx
Emissions as % of
Eastern North
America
Efficiency of
Deposition:
% Deposition /
% Emission
Casco Bay
624,000
244
47
10
4.7
Great Bay
547,000
214
60
13
4.6
Narragansett
Bay
595,200
138
73
18
4.1
Long Island
Sound
905,600
22
70
23
3.0
Hudson/Raritan
Bay
912,000
22
62
25
2.5
Barnegat Bay
505,600
361
67
16
4.2
Delaware Bay
729,600
22
75
26
2.9
Delaware Inland
Bays
326,400
584
52
12
4.3
Chesapeake
Bay
1,081,600
6.5
76
34
2.2
Pamlico Sound
665,600
25
63
18
3.5
Winyah Bay
886,400
19
69
24
2.9
Charleston
Harbor
806,400
20
56
18
3.1
St. Helena
Sound
588,800
48
59
11
5.4
Altamaha
678,400
18
68
13
5.2
Tampa Bay
256,000
45
76
5
15.2
Apalachee Bay
441,600
31
50
9
5.6
Apalachicola
Bay
812,800
16
69
17
4.1
Mobile Bay
992,000
8.7
68
17
4.0
Lake
Pontchartrain
659,200
17
63
11
5.7
Barataria-
Terrebonne
409,600
55
63
8
7.9
Source: http://www.eDa.xv/AMD/Multimedia/characteristicstable.html. Table generated by and used courtesy of R. Dennis, U.S. EPA/ORD/NERL/Atmospheric Modeling Division
2-97
-------
Table 2-13. Characteristics of principal airsheds for reduced-N (Red-N) deposition.
Watershed
Principal Red-N
Airshed Area
(km2)
Red-N Area as %
of Ox-N Area
% Red-N Deposition
Explained by Airshed
Emissions
Airshed NH3 Emission as %
of Eastern North American
Emissions
Chesapeake Bay
668,000
64%
55%
11%
Pamlico Sound
406,000
61%
60%
6.8%
Apalachee Bay
310,000
70%
45-50% est.
4.3%
Source: http://www.epa.aov/AMD/Multimedia/rBducedTable.html. Table generated by and used courtesy of R. Dennis, U.S. EPA/ORD/NERL/Atmospheric Modeling Division
Although deposition of atmospheric N to most locations is dominated at present by oxidized N
produced from fuel combustion, reduced N also contributes, perhaps becoming the dominate contributor
in areas near source regions. However, not all of the reduced N in the form of NH3 is deposited locally
near to the source, as is often assumed.
AQMs like CMAQ can be instrumented to provide analytical output on modeled processes
including emissions, chemical production, and horizontal and vertical transport; see Gipson, 1999. This
process analysis output makes possible experiments with modeled emissions and transport to explore the
fraction of species like total NHX deposited locally and the size of the remaining fraction available for
transport to locations remote from the high concentration sources of NH3. An example of this work for
areas on the east coast of the U.S. near areas of extremely high NH3 emissions in North Carolina was
performed by Dennis and Mathur in the U.S. EPA Atmospheric Modeling Division using CMAQ version
4.5 with 12 km grid spacings. This example is provided here to illustrate the importance of transport to
calculated local and remote deposition totals.
CMAQ includes all chemical and physical processes now known to be relevant for building an
analysis of the vertical and horizontal budget of NH3 for either surface cells in the model's first layer,
nominally 38 m deep, or the total column of air above the surface, generally extending through the mixed
layer into the free troposphere. The diagram in Figure 2-50 illustrates these processes and the surface cell
and total column as present in the model.
I op
1 (Him
Free Troposphere
Mixed Layer (-2 km)
Vertical [Redistribution
(mixing, jadvectioij, clouds)
Gas to Particl
Conversion
38m GastoParicle r'
Surface Conversioi A
-t
Horizontal
Advection
c
nh3
Emissions
Layer 1 Analysis
Dry Deposition
FT
Horizontal
Advection
Wet Deposition
NH3 j)ry Deposition
Emissions
Total Column Analysis
Figure 2-50. Typical surface layer cell and total column structure and processes represented in CMAQ.
2-98
-------
Importantly, CMAQ captures the chief spatial patterns and magnitudes of air concentrations and
wet deposition relevant to computing these budgets, as shown in Figures 2-51 for concentrations and 2-52
for deposition.
J4c_emisv17 S04 for June to August 2002
eg
o
CO
CM
J4c_emisv17 NH4 tor June to August 2002
J4c emisv17
CORP NM0 % NME %
IMPROVE
0 81 -23 67 38 61
STN
0 76 -8 24 35 6
CASTNet
0 88 -24 34 27 13
J4c emtev17
com MMB % NME %
IMPROVE
0.53 -2.72 51 09
STN
0 71 -004 41 66
CASTN«H
0 76 -25.86 32 73
~ IMPROVE (J4c emisv 17)
a STN {J4c_emlsvt7J
CASTNet (J4c emisv17)
~ IMPROVE (J4c emisv 17)
A STN(J4c„emlsv17)
CASTNel {J4c emisv17) A
monthly average
NH4 ( ug/m3 )
RPO - None
State - All
monthly average
S04 ( ug/m3)
RPO ¦ None
State - All
Observation Observation
Figure 2-51. CMAQ vs. measured air concentrations from east-coast sites in the IMPROVE, CSN (labeled
STN), and CASTNet networks in the summer of 2002: (left) sulfate and (right) ammonium.
Solid lines indicate the factor of 2 around the 1:1 line shown between them.
More specifically, CMAQ's predictions of NH3 and S042" for both high and low concentration sites
are well-within the range of measurements there; see Figure 2-53.
Deposition velocities are difficult to estimate for reasons described in Sections 2.8.2 and 2.8.3
above Recent work in the U.S. EPA Atmospheric Modeling Division with CMAQ showed that the
original Vd for NH3 was very likely too high and should be nearer to the values for S02 deposition, or
even lower over some land use surface types. A sensitivity study with the model was performed to test the
effects of changing NFI3 Vd on the fraction of NFI3 available for transport away from cells with high
emissions concentrations. Comparisons were made for the surface cells and total column NFI3
concentrations similar to the analysis represented in Figure 2-54.
In the highest emissions cells during June 2002, the surface NHX budget was dominated by
turbulent transport or vertical mixing moving a majority of the surface NH3 emissions up and away from
the surface into the mixed layer. Figure 2-54 depicts the NHX budget under the base case (Base Vd) and
the sensitivity case (S02 Vd) for which the NH3 Vd was set equal to the S02 Vd- Lower NH3 Vd decreased
NHX deposition to the surface from 15 to 8 %, leaving more NHX for transport horizontally, 22% up from
20% in the base case, and vertically, 69% up from 64% in the base case.
2-99
-------
Deposition
CMAQ Ammonium Ion Wet
NADP Ammonium Ion Wet Deposition
Ammonium as
NH; (kg/ha)
s o 5 MB
05-10 i
1.0-15
1 5-2.0
2.0-2.5
2 5 - 3.0
30-35
3 5-40
4 0-4 5 |
>4 5 ¦
WET NH* DEPOSITION (KSfl-WJ
CMAQ pq
VS. NADP pCOl—2003 AVBttGO)
LMTH? TO ares N H-E EASIEFN UE
ANNUAL
C123 + &67»9IJ
WOP
SUM MBXMi SMOOTH UK
fUDPSnNOeHFSKW *
200 2003 200
Figure 2-52. Comparison of CMAQ-predicted and NADP-measured NH/ wet deposition, (top left) CMAQ
prediction; (bottom left) NADP-measurements; (right) regression and smoothed median line
through CMAQ predictions and NADP measurements with sites in the Chesapeake Bay
watershed highlighted.
Typically, -67% of surface emissions were moved aloft where most was advected away from the
high emissions cell, with a small fraction converted to pNH4 and an even smaller fraction wet-deposited
to the surface. The total column analyses for NH3 and NHX are shown in Figure 2-55.
Local total deposition (wet + dry) is a significant but not dominant loss pathway for surface NH3
emissions. In these simulations, CMAQ deposited -25% of the NH, emissions from the single high
concentration cell in Sampson County, NC, back into that cell. By far, the largest contribution to the local
deposition total was dry deposition: dry-to-wet deposition ratios for the Sampson County high emissions
cell and surrounding surface cells ranged from 2 to 10.
Deposition to cells farther away from the high concentration, immediately surrounding cells was
significantly affected by the change in NH3 Vd tested in this case. Figure 2-56 depicts the range of
influence of the high concentration cell, where that range is defined to be the distance by which 50% of
the emissions attributable to that cell have deposited. The range of influence of the high concentration
Sampson County cell was extended in the Vd sensitivity tested here from -180 km in the base case to
-400 km in the case using the lower, more realistic Vd for NH,. The areal extent of this difference in
range of influence is mapped in Figure 2-57.
2-100
-------
Kenansville Ammonia July 2004
12-hour Averages: 6am-6pm
184 186 188 190 192 194 196 198 200 202 204 206 208 210 212 214
Julian Day: TickMark at Midnight (July 2004)
Kenansville NH3 -B-CMAQ-J4c NH3 I
Kenansville Sulfate July 2004
12-hour Averages: 6am-6pm
Site
184 186 188 190 192 194 196 198 200 202 204 206 208 210 212 214 |
Julian Day: TickMark at Midnight (July 2004)
-Kenansville S04 -»-CMAQ-J4c AS04
Millbrook (Raleigh) Sulfate July 2004
12-hour Averages: 6am-6pm
Millbrook (Raleigh) Ammonia July 2004
12-hour Averages: 6am-6pm
3 10
184 186 188 190 192 194 196 198 200 202 204 206 208 210 212 214 |
Julian Day: TickMark at Midnight (July 2004)
P*- Millbrook S04 CMAQ-J4c AS04 I
184 186 188 190 192 194 196 198 200 202 204 206 208 210 212 214
Julian Day: TickMark at Midnight (July 2004)
-Millbrook NH3 CMAQ-J4c NH3
Figure 2-53. CMAQ-predicted (red symbols and lines) and 12-h measured (blue symbols and lines) NH3 and
SO42' surface concentrations at high and low concentration cells in North Carolina in July
2004. (top left) High concentration NH3 in Kenansville; (top right) high concentration S042~ in
Kenansville; (bottom left) low concentration NH3 in Raleigh; (bottom right) low concentration
SO42" in Raleigh.
o^eT- ITHim*
Free Troposphere
:69-64%:
¦Vertical j
SDiffusiort
_±_ 0.04%
38m Gas to
Surface Particle
tt
c
Mixed Layer (-2 km)
22-20%
Horizontal
Advection
NH:
Emissions
8-15%
Dry Deposition
Right Number: BaseVd
Left Number: S02Vd
NH3 Layer 1 Analysis
Figure 2-54. Surface cell (layer 1) analysis of the sensitivity of NHx deposition and transport to the change
in NH3 Vd in CMAQ for June 2002.
2-101
-------
2.1-2.2%
Gas to Particle
Conversion
(unusually
low)
n
TfHcm
Free Troposphere
89-82%
Horizontal
Advection Mixed Layer (-2 km)
"T 1.55-0.54%
Wet Deposition
N l'< Dry Deposition
Emissions 8-15%
NH3 Column Analysis
91-84%
Horizontal
Bight Number BaseVd
Left Number: SOjVd
"T" 0.58-0.57%
Wet Deposition
Dry Deposition
Emissions 8-15%
NHX Column Analysis
Figure 2-55. Total column analysis for NH3 (left) and NHx (right) showing modeled NH3 emissions,
transformation, and transport throughout the mixed layer and up to the free troposphere.
-180km
Distance from Center (km)
June 2002 NHx Range of Influence: BaseVd vs. S02Vd
Sampson County (single cell)
S02Vd
-400km
BaseVd
- - Wet+Dry Dep BaseVd
—0—Wet+Dry Dep S02Vd
Advection BaseVd
Advection S02Vd
Figure 2-56. Range of influence (where 50% of emitted NH3 deposits) from the high concentration
Sampson County, NC, cell in the June 2002 CMAQ simulation of Vd sensitivities. Base case
and sensitivity case total deposition (blue symbols and lines); base case and sensitivity case
advection totals (red symbols and lines).
2-102
-------
Figure 2-57. Areal extent of the change in NHx range of influence as predicted by CMAQ for the Sampson
County high concentration cell (center of range circles) in June 2002 using the base case and
sensitivity case Vd.
2.8.5. Policy Relevant Background Concentrations of NOx and SOx
Background concentrations of NOx and SOx used for purposes of informing decisions about
NAAQS are referred to as PRB concentrations. PRB concentrations are those concentrations that would
occur in the U.S. in the absence of anthropogenic emissions in continental North America, defined here as
the U.S., Canada, and Mexico. PRB concentrations include contributions from natural sources
everywhere in the world and from anthropogenic sources outside those three countries. Biogenic
emissions from agricultural activities are not considered in the formation of PRB concentrations.
Background levels so defined facilitate separation of pollution levels that can be controlled by U.S.
regulations (or through international agreements with neighboring countries) from levels that are
generally uncontrollable by the U.S. EPA assesses risks to human health and environmental effects from
N02 and S02 levels in excess of these PRB concentrations.
The MOZART-2 global model of tropospheric chemistry (Horowitz et al., 2003) is used to
diagnose the PRB contribution to NOx and SOx levels and to total (wet + dry) deposition. The model
setup for the present-day simulation has been published in a series of papers from a recent model
intercomparison (Dentener et al., 2006a, b; Shindell et al., 2006; Stevenson et al., 2006; van Noije et al.,
2006).
2-103
-------
Total
50°N
4^N
40°N
35°N
30°N
Z5°N
1 ao°w
iao°w
80°W
< d.StT^^ZsIT 5?l 0 7.40 ppb
Background
30°N
Percent Background Contribution
30°N
Figure 2-58. Annual mean concentrations of NO2 (ppb) in surface air over the U.S. in the present-day
(upper panel) and policy relevant background (middle panel) MOZART-2 simulations. The
bottom panel shows the percentage contribution of the background to the present-day
concentrations.
12CAV ioo°w eo°w
iao°W iaa°W bo°W
2-104
-------
i uiai
50°N
45°W
40°N
35°N
30DN
25°N
120°W
100°W
80°W
< O.Ol"^^ 2.41 3.60 ppb
Background
30°N
< O.DO?^ 0.011 0.015 ppb
Percent Background Contribution
3CJ°N
Figure 2-59. Annual mean concentrations of SO2 (ppb) in surface air over the U.S. in the present-day
(upper panel) and policy relevant background (middle panel) MOZART-2 simulations. The
bottom panel shows the percentage contribution of the background to the present-day
concentrations.
120°W lOtAV flO°W
12D°W 100°W SO°W
2-105
-------
First, the role of PRB in contributing to N02 and S02 concentrations in surface air is considered.
Figure 2-58 shows the annual mean predicted N02 concentration in surface air in the base case simulation
(top) and from the PRB simulation (middle panel), along with the percentage contribution of the predicted
background to the total base case N02 concentrations (bottom). Maximum concentrations in the base case
simulation occurred along the Ohio River Valley and in the Los Angeles basin just as they do in reported
measurements; see Section 2.2 above. While present-day concentrations are often >5 ppb, predicted PRB
was <300 ppt over most of the CONUS and <100 ppt in the eastern U.S. The distribution of PRB largely
reflects the distribution of soil N02 emissions, with some local enhancements due to biomass burning
such as is seen in western Montana. In the northeastern U.S., where present-day N02 concentrations are
highest, PRB was predicted to contribute <1% to the total concentrations.
Background S02 concentrations computed with MOZART-2 are orders of magnitude smaller than
measured ambient concentrations, <10 ppt over much of the CONUS as shown in the middle panel of
Figure 2-59. Maximum PRB S02 concentrations are 30 ppt. In the Northwest where there are geothermal
sources of S02, the contribution of PRB to total S02 is 70 to 80%. However, excepting this, PRB
contributes <1% to present-day S02 concentrations in surface air as shown in the bottom panel.
The spatial pattern of NOY (defined in the model as HN03 + NH4NO3 + NOx + H02N02 +
RON02) in wet and dry deposition is shown in Figure 2-60. The upper panel of this figure shows that
highest values are found in the eastern U.S. in and downwind of the Ohio River Valley. The pattern of N
deposition in the PRB simulation shown in the Figure 2-60 middle panel, however, shows maximum
deposition centered over Texas and in the Gulf Coast region, reflecting a combination of N emissions
from lightning in the Gulf region, biomass burning in the Southeast, and microbial activity in soils with
maxima in central Texas and Oklahoma. The bottom panel of Figure 2-60 shows that the PRB
contribution to N deposition is <20% over the eastern U.S., and typically <50% in the western U.S. where
NOy deposition is already lower.
Present-day deposition of SOx (S02 and pS04 is largest in the Ohio River Valley, due to coal-
burning power plants in that region, while background deposition is typically at least an order of
magnitude smaller; see Figure 2-61. Over the eastern U.S., the predicted background contribution to SOx
deposition was <10%, and even smaller, <1%, where present-day SOx deposition was highest. The
predicted contribution of PRB to S deposition was highest in the western U.S. at >20% because of the
geothermal sources of S02 and oxidation of DMS at the water surface of the eastern Pacific.
Figure 2-62 shows results from MOZART-2 discussed above as compared with those from another
tropospheric chemistry model, GEOS-Chem (Bey et al., 2001), which was previously used to diagnose
PRB 03 concentrations (Fiore et al., 2003). In both models, the predicted surface PRB NOx
concentrations tended to mirror the distribution of soil NO emissions, which were highest in the Midwest.
The NO emissions in GEOS-Chem were greater than those in MOZART-2 by nearly a factor of 2. This is
largely explained by the different assumptions regarding the contribution to soil NO emissions through
fertilizer since GEOS-Chem total soil NO emissions were actually higher than MOZART-2 at 0.07 versus
0.11 Tg N over the U.S. in July. Even with the larger PRB soil NO emissions, however, surface NOx
concentrations in GEOS-Chem were typically <500 ppt.
It is also instructive to consider measurements of S02 at relatively remote monitoring sites, i.e.,
ones located in sparsely populated areas not subject to obvious local sources of pollution. Berresheim
et al. (1993) used a type of atmospheric pressure ionization mass spectrometer (APIMS) at Cheeka Peak,
WA (48°30'N, 124°62'W, 480 m asl), in April 1991 during a field study for DMS oxidation products: S02
concentrations there ranged from 20 to 40 ppt. Thornton et al. (2002a) have also used an APIMS with an
isotopically labeled internal standard to determine background S02 levels and found 25 to 40 ppt in
northwestern Nebraska in October 1999 at 150 m above ground using the NCAR C-130. These values are
comparable to remote central south Pacific convective boundary layer S02 (Thornton, 1999).
In summary, the PRB contribution to NOx and SOx concentrations and deposition over the
CONUS is very small, except for S02 in areas with volcanic activity.
2-106
-------
Total
12D°W
100°W
Background
i2j°W iaa°W ao°w
< 25 45 65 85 105 125
Percent Background Contribution
laAV ioa°W bo°w
< ^ 23 32
Figure 2-60. Annual mean concentrations of wet and dry deposition of NHNO3, NH4NO3, NOx, HO2NO2, and
organic nitrates (mg N/m2/yr) in surface air over the U.S. in the present-day (upper panel) and
policy relevant background (middle panel) MOZART-2 simulations. The bottom panel shows
the percentage contribution of the background to the present-day concentrations.
2-107
-------
Total
Figure 2-61,
50°N
45°N
40°N
35°N
30°N
25°N
I20°W 100°W SO°W
1 DO 800 1500 2200 2900 3600
Background
50°N
45°N
40 °N
35°N
30°N
25°N
I20°Vi/ 100°W SO°W
10 16 22 28 34 40
Percent Background Contribution
30°N
Annua! mean concentrations of SOx deposition (S02 + pS04 (mg S/m2/yr) in surface air over
the U.S. in the present-day (upper panel) and policy relevant background (middle panel)
MOZART-2 simulations. The bottom panel shows the percentage contribution of the
background to the present-day concentrations.
120°W 100°W 00 °W
2-108
-------
GEOS-C/iern SOIL MO
MOZARTV2 SOIL NO,
tfloTW
5 11 IS 22 28
i«p°w ecCw
< If
GEOS-Cn inri S'jrfilCfr HO
M0ZART-2SuTBCe-MO JUL
i
-------
2.9.1. Routine Air Monitoring Networks in North America
Routine ambient air concentration and deposition monitoring networks in North America provide
more than 3000 fixed platforms measuring numerous species and chemical and physical properties. Many
of these long-standing network systems were initialized after the 1970 CAA, subsequent CAAA, NAAQS
reviews and National Academy of Sciences (NAS) recommendations resulting in periodic step
enhancements to these routine networks. Examples include CASTNet and NADP addressing
acidification; the Photochemical Assessment Monitoring Stations (PAMS) in response to persistent 03
pollution and to monitor 03 precursors including NOx; and the PM2 5 network. Table 2-14 lists the
networks, sponsoring agencies, monitoring site densities, and their dates of operation, locations, and
measurement parameters.
Table 2-14. Major routine operating air monitoring networks.
Network
Lead
Agency
ft of
Sites
Initiated
Measurement Parameters
Location of Information and/or Data
STATE/LOCAL/FEDERAL NETWORKS
NCore—National Core
Monitoring Network
U.S. EPA
75
2008
03, NO/NO2/NOY, S02, CO,
PM2.5/ PM10-2.5, PM2.5
speciation, NH3, HNO3, surface
meteorology
htto://www.eDa.aov/ttn/
amtic/monstratdoc.htm
SLAMS—State and Local
Ambient Monitoring Stations
U.S. EPA
-3000
1978
03, N0x/N02, S02, PM2.5/PM10,
CO, Pb
http://www.epa.aov/ttn/airs/
airsaqs/aqsweb/aqswebhome.htm
STN—PM2.5 Speciation
Trends Network
U.S. EPA
300
1999
PM2.5, PM2.5 speciation, major
ions, metals
htto://www.eDa.aov/ttn/airs/
airsaas/aasweb/aaswebhome.html
PAMS—Photochemical
Assessment Monitoring
Network
U.S. EPA
75
1994
O3, NOx/NOy, CO, speciated
VOCs, carbonyls, surface
meteorology and upper air
htto://www.eDa.aov/ttn/airs/
airsaqs/aqsweb/aqswebhome.htm
IMPROVE—Interagency
Monitoring of Protected
Visual Environments
NPS
110 plus
67 proto-
col sites
1988
PM2.5/PM10, major ions, metals,
light extinction, scattering
coefficient
htto://vista. cira.colostate.edu/IMPROVE/
CASTNet—Clean Air Status
and Trends Network
U.S. EPA
80+
1987
O3, SO2, major ions, calculated
dry deposition, wet deposition,
total deposition for S/N, surface
meteorology
http://www.epa.aov/castnet/
GPMP—Gaseous Pollutant
Monitoring Network
NPS
33
1987
03, N0x/N0/N02, S02, CO,
surface meteorology, (plus
enhanced monitoring of CO,
NO, NOx, NOy, and SO2 plus
canister samples for VOC at
three sites)
htto://www2. nature. nDs.aov/air/
Monitorina/network.cfm#data
POMS—Portable Ozone
Monitoring Stations
NPS
14
2002
O3, surface meteorology, with
CASTNet-protocol filter pack
(optional) S042", N03, NH4+,
HNOs, SO2
http://www2.nature.nDS.aov/
air/studies/port03.cfm
Passive Ozone Sampler
Monitoring Program
NPS
43
1995
O3 dose (weekly)
http://www2.nature.nps.aov/
air/Studies/Passives.cfm
2-110
-------
Network
Lead
Agency
ft of
Sites
Initiated
Measurement Parameters
Location of Information and/or Data
NADP/NTN—National
Atmospheric Deposition
Program / National Trends
Network
USGS
200+
1978
Major ions from precipitation
chemistry
http://nadD. sws.uiuc.edu/
NADP/MDN—National
Atmospheric Deposition
Program / Mercury
Deposition Network
None
90+
1996
Mercury from precipitation
chemistry
http://nadD.sws.uiuc.edu/mdn/
AIRMoN—National Atmos-
pheric Deposition Program /
Atmospheric Integrated Re-
search Monitoring Network
NOAA
8
1984
Major ions from precipitation
chemistry
httD://nadD.sws.uiuc.edu/AIRMoN/
IADN—Integrated
Atmospheric Deposition
Network
U.S. EPA
20
1990
PAHs, PCBs, and organochlo-
rine compounds are measured
in air and precipitation samples
http://www.epa.aov/alnoo/monitorina/air
NAPS—National Air Pollution Canada
Surveillance Network
152+
1969
S02, CO, 03, NO, NO2, NOx,
VOCs, SVOCs, PM10, PM2.5,
TSP, metals
http://www.etcentre.ora/NAPS/
CAPMoN—Canadian Air and Canada
Precipitation Monitoring
Network
29
2002
O3, NO, NO2, NOymass, PM2.5
speciation, major ions for
particles and trace gases,
precipitation chemistry for major
ions
http://www.msc.ec.ac.ca/ cap
mon/index e.cfm
Mexican Metropolitan Air
Quality Network
Mexico
93
???
03, NOx, CO, S02, PM10, TSP
See (Ceccotti and Messick, 1997)
http://www.cec.ora/oubs docs/-
publications/index.cfm?varlan=enalish
AIR TOXICS MONITORING NETWORKS
NATTS—National Air Toxics
Trends Station
U.S. EPA
23
2005
VOCs. Carbonvls. PM10 metals. http://www.eDa.aov/ttn/airs/airsaas/
Ha aasweb/aaswebhome.htm
State/Local Air Toxics
Monitoring
U.S. EPA
250+
1987
VOCs. Carbonvls. PMm metals. htto://www.eDa.aov/ttn/airs/airsaas/
Ha aasweb/aaswebhome.htm
NDAMN—National Dioxin Air U.S. EPA
Monitoring Network
34
1998-
2005
CDDs, CDFs, dioxin-like PCBs
http://cfpub2.epa.aov/ncea/cfm/
recordisplav.cfm?deid=22423
TRIBAL MONITORING NETWORKS
Tribal Monitoring
U.S. EPA
120+
1995
O3. NOx/N02. SO2. PM2.5/ PM10. http://www.eDa.aov/ttn/airs/
PAN. NH3. PM2.5. PM10 and airsaas/aasweb/aaswebhome.htm
coarse fraction CO, Pb
INDUSTRY/RESEARCH NETWORKS
New Source Permit
Monitoring
None
variable
variable
03, NOx/N02, SO2, PM2.5/PM10,
CO Pb
Contact specific industrial facilities
HRM Network—Houston
Regional Monitoring Network
None
9
1980
O3. NOx. PM? 5/PM10. CO. SO2. httD://hrm.radian.com/ houston/
Pb. VOCs. surface meteoroloav how/index.htm
2-111
-------
Network
Lead
Agency
it of
Sites
Initiated Measurement Parameters Location of Information and/or Data
ARIES/ SEARCH—Aerosol None
Research Inhalation
Epidemiology Study /
Southeastern Aerosol
Research and Characteriza-
tion Study experiment
1992 O3, NO/NO2/NOY, SO2, CO,
PM2.5/PM10, PM2.5 speciation,
major ions, NH3, HNO3,
scattering coefficient, surface
meteorology
httB://wvvw.atmospheric-
research.com/studies/SEARCH/index.html
SOS - SERON-Southern U.S. EPA
Oxidant Study - Southeast-
ern Regional Oxidant
Networks
-40 1990
03, NO, NOy, VOCs, CO,
surface meteorology
htto://ww¥.ncsu.edu/sos/ix!hs/
sos3/State of SOS 3.pdf
Two important ambient air networks focused on environmental welfare effects were established in
the mid-1980s. IMPROVE with >100 sites in National Parks and other remote locations is used primarily
to assess visibility impairment, but has provided a reliable long-term record of particulate mass and major
speciation components and served as a model for the later deployment of STN; see Figure 2-63. STN
(now part of CSN) has provided an urban complement to characterize aerosol composition; see Figure
2-64. Additional, minor networks identified in Figure 2-63 include those of the state and local air agencies
deployed since the mid-1980s measuring a variety of aerosol- and gas-phase, hazardous air pollutants
(HAPs) at -200 locations, and a modest National Air Toxics Trends (NATTS) network of 23 sites; see the
list and brief descriptions in Table 2-14 above.
Ambient Air Monitoring Stations in the United States
/o c »
' o *: ^ 1
rSK
¦
nat rs
~
PM % Speciabon
0
PAMS
•
CASING
ir
Improve
•
O,
O
pm14
O
SO
•
NOx
•
PM„
•
NOx
•
Lead
•
CO
Figure 2-63. Aggregate map of most routine U.S. monitoring stations.
2-112
-------
CASTNet was established in the early 1990s to track changes in dry deposition of major inorganic
ions and gaseous precursors associated with the CAA Title IV reductions in S02 and NOx or S and N.
Complementing ongoing precipitation measurements from NADP, CASTNet (see Figure 2-63) has
provided a valuable source of model evaluation data for many of the large regional scale applications
since the 1990s.
Upper Midwest
Industrial Midwest
4-
20-i Northwest
E 16-
§12-
2Q_. Northeast
E 16-
5 i2-i—1|—|_r
02 03 04 05 06
02 03 04 05 06
02 03 04 05 06
02 03 04 05 06
Southern CA
Southeast
rp2 500
sites each for O,
and PMf5 and
related spatial
^Aapgin^Sugpor^
NCore Measurements
Level 2i ~ 75 Multi-
pollutant (MP)
Sites."Core Species"
Plus Leveraqinq From
PAMS,
Speciation Program,
Air Toxics
Level 1. 3-10 Master
Sites Comprehensive
Measurements,
Advance Methods
Serving Science and
Technology Transfer
Needs
Minimum "Core" Level 2 Measurements
Continuous NO,NO¥,SOZ,CO.PM2 5,
PMI0/PMc, 03 Meteorology (T,RH.WS.WD>;
Integrated PM2 5 FRM, HN03. NH3,
\ ' NCore Level 2
w. <
STATUS
• Agreed
to • Prapawu
Figure 2-65. Original 3—tiered NCore design (left) and proposed site locations for Level 2 multiple pollutant
sites.
2-113
-------
Deployment of the PAMS and the PM2 5 networks from the early 1990s through 2002 markedly
enhanced the spatial, temporal, and compositional attributes of gases and aerosols.
A new multiple pollutant-monitoring network, NCore was begun in 2006. When implemented in
2009, NCore will provide a minimum of 75 Level-2 sites (Figure 2-65) in most major urban areas and
important transport corridor and background locations. NCore will include a variety of trace gas and
aerosol mass and speciation measurements which are intended to support multiple data user needs
including air quality model evaluation and long-term epidemiological studies.
2.9.1.1. Pollutant Categories
Inorganic Gas-phase Species
Most U.S. sites measuring NOx and S02 are incorporated within the State and Local Air
Monitoring Systems (SLAMS) networks. Most of the SLAMS sites are located in populated urban
locations with a variety of monitor siting requirements typically intended to site for high concentration
locations resulting in an emphasis on center-city locations for NOx and proximity to major power
generating facilities for S02.
Measurements of NOY, HN03, and NH3 are useful in a variety of ways important for assessing
NOx and SOx environmental pollution effects; some of these ways include: evaluation of emissions
inventories; inputs to and sources of evaluation for numerical and observation-based models; and
establishing baseline N budgets for watershed and field accountability assessments. NOY, HN03, and
NH3, together with true N02 and pNLL, are significant components of the total N budget but remain
poorly characterized at the national scale. In largest part, a lack of reliable, cost effective continuous
measurement methods has hindered deployment of instruments for HNO3 and NH3 as described in
Section 2.7. In the U.S., the Southeastern Aerosol Research and Characterization Study (SEARCH)
network of eight monitoring sites is the only source of routine, continuous ambient air measurements of
NOy together with NH3 and HN03; see discussions in Blanchard and Hidy (2003) and Zhang et al.
(2006). CASTNet recently deployed a network of inexpensive passive NH3 samplers which have promise
for characterizing broad spatial patterns, with extended averaging times beyond 24 h.
Particulate Matter Mass
Nearly 1500 PM2 5 gravimetric sites were established before 2000 to determine nonattainment
status of counties throughout the U.S. following the 1997 promulgation of the PM25 particulate matter
standard. The network has evolved to add over 500 continuous PM2 5 monitors and a reduction of 24-h
gravimetric samplers below 1000 sites (see Figure 2-66) that support air quality forecasting and public
notification of adverse air quality using the Air Quality Index (AQI), a generalized indicator of exposure
concern linked to the NAAQS (http://www.epa.gov/airnow/). While this expansion of continuous PM2 5
sites adds spatial coverage of highly temporally resolved information, the mix of instrument types
compromises data harmonization across sites and geographic areas with different operational
characteristics.
2-114
-------
m,
PM,, FRM/FEM Sites
a Correlated Radtence Nephelomeier
Figure 2-66. Maps illustrating coverage of PM2.5 FRM and FEM and O3 network (left); and PM2.5 continuous
samplers (right).
Particulate Matter Speciation
Hie speciation networks typically collect a 24-h sample once every 3 or 6 days. CASTNet provides
weekly averaged measurements of major ions including SO42 . NO-, . calcium (Ca ). Na+, potassium
(K+), NH4+, and magnesium (Mg2") integrated over all aerosol sizes by means of open-face filter packs.
Daily, 24-h speciated samples is limited to fewer than five sites in the U.S. and Canada. Similarly, a small
and variable number of sites, fewer than 10 in most years, provide near-continuous speciation data,
usually limited to some combination of pS04, pN03, and EC and OC. In addition, the 22 NATTS sites
include aetholometers measuring semi-continuous light absorption, often used as a surrogate for EC.
The U.S. EPA PM Supersites Program (Wittig and Solomon, 2006) provided highly time-resolved
aerosol measurements at eight cities in the U.S. for a mix of time periods between 1999 and 2004.
Depending on location and time period, a number of different instrument configurations were deployed
ranging from additional spatial coverage of standard speciation sites to systems capturing near-continuous
size-distributed chemical composition profiles.
Hie SEARCH program, funded since 1998 by EPRI and Southern Company, has provided
continuous, semi-continuous and integrated data on a wide variety of species from eight highly
instrumented paired research sites in the southeastern U.S. in the states of AL, FL, GA, and MS; see
descriptions in Hansen et al. (2003) for additional details. At present, the suite of measurements made at
all sites includes: 24-h PM2 5 filter samples, analyzed for mass, ions (SO " . N03 , NH_. ). OC, EC, BC,
and elements as measured by XRF; 24-h PM coarse mass, ions, and XRF elements; 24-h gaseous NH3;
continuous (minute-to-hourly) PMjj mass, OC, EC, NH4 . NO-, . and SO+~; light-scattering and light
absorption; continuous gaseous 03, N02, NOy, N03 , CO, and S02; and continuous, 10-m meteorological
parameters: wind speed, wind direction, temperature, RH, solar radiation, barometric pressure and
precipitation.
Precipitation-based Networks
Precipitation chemistry is the primary measured link between atmospheric, terrestrial, and aquatic
systems. NADP oversees a network of more than 250 sites (see Figure 2-67) where most of the major ions
key to aquatic chemistry addressing acidification and eutrophication effects are measured. The NADP
includes the Mercury Deposition Network (MDN) of -90 sites and seven Atmospheric Integrated
Research Monitoring Network (AIRMoN) sites providing greater temporal resolution.
2-115
-------
2.9.2. Intensive Field Campaigns
Intensive field campaigns of relatively short duration supplement routine, longer-term monitoring
networks by enhancing spatial, temporal, and compositional distribution of atmospheric species to better
elucidate physical and chemical processes relevant to the fate, transport, and removal of secondarily
formed gases and aerosols. Typically, these campaigns utilize some combination of aircraft studies, high
time resolved instrumentation and advanced analytical methods (in-situ and laboratory) to complement
routine ground-based measurements.
4
* c E ? **
The Canadian Air
and Precipitation
Monitoring Network
Reseau oanadien
d'£cha nt 111 onn age ties
precipitations et de I'air
<2>
Q«turna
Pinkie I akc»
LG4
O Mingan
rhsnai<;
0Ous« Bdy B
CI>
Ba^O'Espoir B
E L'% AEdQljard Harc01J„ B
v X
• . .
Loke pp,ra pare
N Ontario
V
t •
% •
*4
&
Long-Tom Monitoring
o NADP/NTN (wet deposition)
• NADP/MDN (mercury deposition)
° TIMEATM (surfacewater acidity)
• CASTNET (dry depositioiVcccne)
Ecological Resources
Acidic Surface Waters
^ N-Saturated Forest
•* Class I Areas
03 Highly Eutrophic Estuaries
NaT
Figure 2-67. Routinely operating North American precipitation and surface water networks. Upper left,
Canadian Air and Precipitation Monitoring Network (CAPMoN); Upper right, Integrated
Atmospheric Monitoring Deposition Network (IADN); Bottom, National Atmospheric
Deposition Program (NADP) with Time/LTM surface chemistry sites.
2-116
-------
Table 2-15. Air monitoring networks/campaigns for non-routine special intensive studies conducted since
the mid-1990s.
Lead
Agency1
§ of
Sites
Initiated
Measurement
Parameters
Location of Information
and/or Data
Notes
Texas
17 2006
O3, NOx, NOy, SO2,
haze, visibility, CO,
VOC, solar radia-
tion, surface mete-
orology, upper air
http://www.utexas.edu/research/
ceer/texaasll/PDF/12-12-04
Projected SurfaceSites tbl.pdf
Researchers from universities, state and federal agen-
cies, private industry, and local governments are join-
ing forces to conduct a major field study to address air
quality issues in the eastern half of Texas. The study,
planned for a period extending from Apr 2005 through
Oct 2006, will examine regional O3 formation, transport
of O3 and O3 precursors, meteorological and chemical
modeling, issues related to O3 formation by highly
reactive emissions, and PM formation. It is anticipated
that the information from the study will be the scientific
basis used for developing State Implementation Plans
(SIPs) for O3 (with concentrations averaged over 8 h),
regional haze, and, if necessary, for fine PM (PM < 2.5
|jm in diameter, PM2.5).
NOAA
1 ship, 2006
2
aircraft
03, NO, NO2, NOy,
VOCs, CO2, CO,
SO2, HNOs, NH3,
other reactive
pollutants, aerosols,
meteorological
parameters and
upper air
http://esrl.noaa.gov/csd/2006/
For TexAQS 2006, the NOAA air quality component will
investigate, through airborne and sea-based measure-
ments, the sources, and processes that are responsi-
ble for photochemical pollution and regional haze dur-
ing the summertime in Texas. The focus of the study
will be the transport of O3 and O3 precursors within the
state and the impact of the long-range transport of O3
or its precursors.
NOAA 3 2006 O3, NO, NO2, NOy, http://cloud1 .arc.nasa.gov/intex-b/ The export of air pollutants from urban to regional and
aircraft VOCs, CO2, CO, global environments is a major concern because of
SO2, HNO3, NH3, wide-ranging potential consequences for human
other reactive health, cultivated and natural ecosystems, visibility
pollutants, aerosols, degradation, weather modification, changes in radia-
meteorological tive forcing, and tropospheric oxidizing capacity. During
parameters, altitude the spring of 2006, a highly integrated atmospheric
— NOAA aircraft field experiment was performed over and around North
America. The Megacity Initiative: Local and Global
Research Observations (MILAGRO),
http://www.eol.ucar.edu/projects/milagro/, resulted
through a highly coordinated collaboration between
NSF (through MIRAGE-Mex), DOE (through MAX-
Mex), NASA (through INTEX-B) and a variety of re-
search institution in the U.S. and Mexico and involved
ground and air borne activities centered on Mexico
City, Mexico during March 2006. MILAGRO goals were
greatly facilitated and enhanced by a number of
concurrent and coordinated national and international
field campaigns and global satellite observations.
1 EPA— Environmental Protection Agency; NASA — National Aeronautics and Space Administration; NOAA — National Oceanic and Atmospheric Administration; NPS — National Park
Service\NSF — National Science Foundation; UCSD — University of California San Diego (Scripts Institution of Oceanography)
2This study is part of the Central California Air Quality Studies (CCAQS) which comprise the California Regional Particulate Air Quality Study (CRPAQS) and the Central California
Ozone Study (CCOS). CCAQS is a multi-year effort of meteorological and air quality monitoring, emission inventory development, data analysis, and air quality simulation modeling.
Prior studies in California included: Southern California Ozone Study (SCOS97) — 1997; Integrated Monitoring Study (IMS95) — 1995; San Joaquin Valley Air Quality Study (SJVAQS)
— 1990; SARMAP Ozone Study — 1990; Southern California Air Quality Study (SCAQS)— 1987.
historically, there have been many other field studies in the 1960s - 1990s that are not reflected in this table that involve both fixed monitoring sites and aircraft; well known examples
include Regional Air Pollution Study (RAPS), Large Power Plant Effluent Study (LAPPES), Northeast Corridor Regional Modeling Program (NECRMP), Northeast Regional Oxidant
Study (NEROS), Persistent Elevated Pollutant Episode (PEPE), and Lake Michigan Ozone Study (LMOS).
2-117
-------
There has been a long history of intensive field campaigns starting with the RAPS in the 1970s,
which formed the basis for evaluating the early photochemical gridded Eulerian airshed models used in
acid deposition and 03 assessments. Landmark campaigns in the U.S. through the 1980s and 1990s such
as the SCAQS (Lawson, 1990), the SJAQS/Atmospheric Utility Signatures, Predictions, and Experiments
(AUSPEX) (Roth et al. 1988) and the Southern Oxidant Study (SOS) (Cowling and Furiness, 2001). Over
the last decade there have been a series of field campaigns focusing on characterization of surface level
aerosols through the PM Supersites program (Solomon and Hopke, 2008). While the early campaigns
focused on urban environments, the Eulerian Model Evaluation Field Study (EMEFS) and SOS during the
early 1990s shifted focus toward regional spatial scales. In addition to addressing urban areas of concern
such as Houston, TX, and Los Angeles, CA, more recent campaigns have extended spatial scales beyond
regional studies to address oceanic transport and a variety of air pollution issues across the Northern
Hemisphere, recognizing the importance of far-ranging source regions and continental-scale atmospheric
processes. Some of these campaigns included local and regional studies for the northeast and southeast
U.S., portions of Texas, and central and southern California. A variety of federal and state entities have
served as lead agencies for these studies. Table 2-15 provides a listing of studies conducted since the mid-
1990s.
A synthesis of key findings from major field campaigns conducted over the last two decades would
elevate exposure of these programs to a wider audience potentially generating support to enhance and
sustain atmospheric process and model evaluation studies, which are important complements to routine
ground-based and satellite observation platforms. While the NARSTO database
(http://eosweb.larc.nasa.gov/PRODQCS/narsto/table narsto.html) provides access to raw data for various
field campaigns, coverage of campaigns beyond North America must be acquired from other sources. The
National Aeronautics and Space Agency (NASA)'s Atmospheric Data Science Center
(http://eosweb.larc.nasa. gov/) also provides access to some of the more recent field campaigns.
2.9.3. Satellite-Based Air Quality Observing Systems
An extensive array of satellite-based systems (see Table 2-16 and Table 2-17) with the capability of
measuring atmospheric column total species has been established by U.S. and European Satellite
programs lead by NASA and the National Oceanic and Atmospheric Administration (NOAA) in the U.S.
and the European Space Agency (ESA). A suite of satellites including Aqua, Aura, CALIPSO, OCO,
Glory, as well as NOAA-17, NOAA-18 and NPOESS, have either been launched since about the year
2000 or have other near-term proposed launch dates. Collectively, the remote sensing platforms include
techniques for measuring columns and/or profiles of aerosol optical depth (AOD), 03, CO, C02, CH4,
S02, N02, chlorinated fluorocarbon compounds (CFCs), other pollutants, and atmospheric parameters
such as temperature and H20 content. Most of these satellites have a near-polar orbit allowing for two
passes per day over a given location. When taken together, a group of six satellites (Aqua, Aura,
CALIPSO, OCO, as well as CloudSat and PARASOL), coined the A-Train, is being configured to fly in a
formation that crosses the equator a few minutes apart at around 1330 local time to give a comprehensive
picture of earth weather, climate, and atmospheric conditions.
Satellite imagery offers the potential to cover broad spatial areas; however, an understanding of
their spatial, temporal and measurement limitations is necessary to determine how these systems
complement ground based networks and support air quality management assessments.
2-118
-------
Table 2-16.
Satellite-based air quality observing systems.14
Instrument
Satellite
Platform3
Lead
Agency
Initiated
Measurement Parameters
Orbit & Hor.
Resolution
Location of Information
and/or Data
OLS (Operational
Linescan System)
DMSP satellites
DOD
1962?
Identify fires and smoke plume
Polar Imagery
only
http://www.af.mil/factsheets/-
factsheet.asp?fslD=94
BUV (Backscatter
Ultraviolet Spectrometer)
Nimbus 4
NASA
1970-1980
O3, CO2, SO2
Sun
synchronous
http://nssdc.asfc.nasa.aov/da
tabase/MasterCataloa?sc=19
70-025A
SBUV (Solar Backscatter
Ultraviolet Spectrometer)
Nimbus 7
NASA
1978-1993
O3, SO2
Polar
http://iwockv.asfc.nasa.aov/-
n7toms/nimbus7tech.html
TOMS (Total Ozone
Mapping Spectrometer)
Nimbus 7 Meteor
3 Earth-Probe
NASA
1978-1993
1991-1994
1996
O3, SO2, Aerosols
Polar-100
km
http://toms.asfc.nasa.aov/-
fltmodel/spacecr.html
LIMS (Limb Infrared
Monitor of the
Stratosphere
Nimbus 7
NASA
1978-1979
03, HNOs, NO2,
Polar
http://lims.aats-inc.com/-
about lims.html
ATM OS (Atmospheric
Trace Molecule
Spectroscopy)
Spacelab 3
ATLAS-1,2,3
NASA
1985,1992,
1993,1994
03, CFCb, CF2CI2, CIONO2,
HCI, HF, CO, CH4, HCN, HNOs,
NO, NO2, N2O5, Aerosols
http://remus.ipl.nasa.aov/-
atmos/sl3.html
CLAES (Cryogenic Limb
Array Etalon
Spectrometer)
UARS
NASA
1991-1993
03, CFCb, CF2CI2CIONO2, CH4,
HNOs, NO, NO2, N2O, N2O5,
Aerosols
http://umpaal.asfc.nasa.aov/
HALOE (Halogen
Occultation Experiment)
UARS
NASA
1991-2005
03, HCI, HF, CH4, NO, NO2,
Aerosols
http://umpaal.asfc.nasa.aov/
ISAMS (Improved
Stratospheric and
Mesospheric Sounder)
UARS
NASA
1991-1992
03, CO, CH4, NO2, N2O, N2O5
Aerosols
http://umpaal.asfc.nasa.aov/
MLS (Microwave Limb
Sounder)
UARS
NASA
1991-1999
03, CIO, CHsCN, HNOs, SO2
http://umpaal.asfc.nasa.aov/
GOES Imager (Geosta-
tionary Operational
Environmental Satellites)
GOES-10 GOES-
12
NOAA
1994
Fire products for WF_ABBA
(imagery) and GASP (aerosol
optical depth)
Geostationary
http://www.nesdis.noaa.aov/
GOES Sounder
(Geostationary
Operational
Environmental Satellites)
GOES-10 GOES-
12
NOAA
1994
Total column O3
Geostationary
http://cimss.ssec.wisc.edu/-
aoes/aoesmain. html#sndrinfo
AVHRR (Advanced Very
High Resolution
Radiometer)
NOAA-15 NOAA-
16 NOAA-172
NOAA
1998
Aerosol optical depth, particle
size information and vegeta-
tion/drought index products
related to air quality through
fires
Polar
4 km
http://noaasis.noaa.aov/-
NOAASIS/ml/avhrr.html
SBUV/2 (Solar
Backscattered Ultraviolet
Radiometer Model 2)
NOAA-16
NOAA-172
NOAA
2000
Total and profile O3 from surface Polar
to top of atmosphere in ~5 km
thick Umkehr layers
http://www2.ncdc.noaa.aov/-
docs/podua/html/c4/sec4-
4.htm
2-119
-------
Instrument
Satellite
Platform3
Lead
Agency
Initiated
Measurement Parameters
Orbit & Hor.
Resolution
Location of Information
and/or Data
MOPITT (Measurement of EOS Terra
Pollution in the Tropo-
sphere)
NASA
1999
co,ch4
Polar
22 x 22 km2
http://www.eos.ucar.edu/mopi
fi/
MISR (Multi-angle
Imaging
SpectroRadiomenter)
EOS Terra
NASA
1999
Aerosol properties and plume
height information near the
vicinity of fires
Polar ~1 km
htto://www-
misr.ipl.nasa.aov/mission/-
introduction/welcome.html
MODIS (Moderate
Resolution Imaging
Spectroradiometer)
EOS Terra
EOS Aqua
NASA
1999
2002
O3, aerosol optical depth, par-
ticle size information, fine
particle fraction, and forest fires
Polar
1 km
http://modarch.asfc.nasa.aov/
index.php
AIRS (Atmospheric
Infrared Sounder)
EOS Aqua
NASA
2002
Total column O3, surface
temperature, temperature and
moisture vertical profiles, (plus
under development are CO and
CO2 total column, O3 vertical
distribution, and CH4
distribution)
Polar
50 km
http://www-airs. ipl. nasa.aov/
HIRDLS (High Resolution
Dynamics Limb Sounder)
EOS Aura
NASA
2004
03, CFCb, CF2CI2, CIONO2,
CH4, HNOs, NO2, N2O, N2O5,
Aerosols
Polar: http://aura.gsfc.nasa.
gov/index.html
MLS (Microwave Limb
Sounder)
EOS Aura
NASA
2004
03, BrO, CIO, HOCI, HCI, CO,
HCN, CH3CN, HNO3, N20, OH,
HO2, SO2
Polar
http://aura.asfc.nasa.aov/inde
x.html
OMI (Ozone Monitoring
Instrument)
EOS Aura
NASA
2004
03, BrO, OCIO, HCHO, NO2,
SO2 and aerosol
Polar
12x24 km2
http://aura.asfc.nasa.aov/inde
x.html
TES (Total Emission
Spectrometer)
EOS Aura
NASA
2004
03, NOy, CO, SO2, CH4
Polar 26 x 42
km
http://aura.asfc.nasa.aov/inde
x.html
CALIPSO (Cloud-Aerosol
Lidar & Infrared Pathfind-
er Satellite Observations)
CALIPSO
NASA
2005
AOD, backscatter, extinction
Polar 0.3 x
0.3 km2
http://www-
calipso.larc.nasa.aov/about/
OMPS
Ozone Mapping
and Profiling
Suite NPOESS
Preparatory
Project
NOAA
2006
Total column and vertical profile
O3 data
Polar
http://www.ipo.noaa.aov/-
Proiects/npp.html
VIIRS (Visible Infrared
Imaging Radiometer
Suite)
NPOESS-
Preparatory
Project
NOAA
2006
AOD
Polar
http://www.ipo.noaa.aov/-
Proiects/npp.html
Orbiting Carbon
Observatory
OCO
NASA
2008
CO2
Polar
http://oco.iol.nasa.aov/
APS & TIM (Aerosol
Polarimetry Sensor &
Total Irradiance Monitor)
Glory
NASA
2008
BC soot, other aerosols, total
solar irradiance, cloud images
Sun- synchronous, circular
Low Earth
Orbit
http://alorv.asfc.nasa.aov/
1 Non-U.S. satellite systems are not included in table at this time.
2As of 3/15/06 the operational satellite platforms may need to include NOAA-18.
3CALIPSO — Cloud-Aerosol Lidar & Infrared Pathfinder Satellite Observations
DMSP— Defense Meteorological Satellite Program. EOS — Earth Observing System. GOES — Geostationary Operational Environmental Satellites
NOAA — National Oceanic and Atmospheric Administration!. NPOESS — National Polar-orbiting Operational Environmental Satellite System. OCO — Orbiting Carbon Observatory
UARS — Upper Atmosphere Research Satellite. 4See the following table for additional information on NASA satellites, instrument systems, pollutants measured, and data availability:
2-120
-------
Table 2-17. Key atmospheric chemistry and dynamics data sets at the NASA Goddard DAAC.
T.S
Missions Nimbus 4 Nimbus 7 /\DEOS 1 Nimbus 7 ATLAS 1, UARS ERS-2 Terra Aqua Aqua Aura
Earth-Probe ^
Instruments BUV SBUV TOMS LIMS ATMOS CLAES HALOE SAMS MLS GOME MODIS AIRS OmI HIRDLS MLS TES*
Data Period
Apr
'70-
May
'77
Nov
'78-
May
'93
Nov
'78-
Present
Oct
'78-
May
'79
'85, '92,
'93,'94
Oct
'91-
May '93
Oct '91-
Present
Sep '91 —
Jul '92
Sep
'91-
Jul
'99
April
'95-
Present
Mar
'00-
Present
Sep
'02-
Present
Jul '04-
Present
Jul '04-
Present
Jul '04-
Present
Jul '04-
Present
Spectral
Region
255-
380 nm
255-
340 nm
309-
360
312-
380 nm
6.2-
15 pm
2.98-15 pm
3.5-
12.7 pm
2.43-
10.25 pm
4.6-
16.6 pm
63,
183,
205
GHz
240-
790 nm
0.4-
14 pm
0.4-1.1,
3.74-
15.4 pm
270-
500 nm
6.12-
17.76 pm
118,190,
240, 640
GHz, 2.5
THz
3.2-
15.4 pm
Bands
13
13
6
6
16
9
8
8
3
3072
36
2382
1560
22
5
12
03
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
BrO
•
•
•
CFCIs
•
•
•
CF2CI2
•
•
•
CIO
•
•
OCIO
•
•
CIONO2
•
•
•
HOCI
•
HCI
•
•
•
HF
•
•
HCHO
•
•
CO
•
•
•
•
•
ch4
•
•
•
•
•
•
•
CHsCN
•
HCN
•
•
HNOs
•
•
•
•
•
•
•
NO
•
•
•
NO2
•
•
•
•
•
•
•
•
•
N2O
•
•
•
•
•
N2O5
•
•
•
•
OH
•
H20/
Humidity
•
•
•
•
•
•
•
•
•
•
•
•
S02
•
•
•
•
•
•
•
Aerosols
•
•
•
•
•
•
•
•
Cloud
•
•
•
•
•
•
•
Temperature
•
•
•
•
•
•
•
•
•
•
Geopotential
Height
•
•
•
•
•
Reflectance
•
•
•
•
•
•
Please note that the table above does not contain parameters from all sensors and products. Also available from the GES DAAC are many more Atmospheric and Earth Sciences data
products from AIRS, AMSU-A, HSB, MODIS, SeaWiFS, OCTS, CZCS, TRMM (PR, TMI, VIRS), TOVS Pathfinder, Data Assimilation Model (GEOS-1, GEOS-DAS, CPC/ACDB), UARS
(HRDI, WINDII, SOLSTICE, SUSIM, PEM), SORCE, several Field Campaigns, and Interdisciplinary data sets consisting of 70 geophysical Earth Sciences parameters. TOMS & SBUV
reprocessed data (version-8) are now available on DVD-ROM. The MLS and OMI-Aura products and Visualization tools are now available from GES DIC.
Source: Aura instrument 'TES' is archived at the NASA Langley Atmospheric Sciences Data Center (http://eosweb.larc.nasa.gov/) http://disc.asfc.nasa.gov/.
2-121
-------
2.9.3.1. Satellite Coverages
The near polar orbiting tracks of most satellites performing trace gas measurements provides wide
spatial coverage of horizontal resolution on the order of 10 to 50 km, but delivers only twice-daily
snapshots of a particular species. Consequently, temporal patterns of pollutants as well as a time-
integrated measure of pollutant concentrations cannot be delineated explicitly through satellite
measurements alone. The geostationary satellite platforms such as the GOES systems of NOAA do
provide near-continuous coverage of physical parameters for weather tracking and forecasting purposes.
Polar orbiting satellites typically provide horizontal spatial resolution between 10 and 100 km, depending
on the angle of a particular swath segment. Spatial resolution less than 10 km is possible with
geostationary platforms.
Characterization of elevated pollutants delivered by satellite systems complements of our ground
based in-situ measurement networks - especially considering that a considerable fraction of pollutant
mass resides well above Earth's surface. With few exceptions, satellite data typically represents a total
atmospheric column estimate. For certain important trace gases (e.g., N02, S02, CH20) and aerosols,
most mass resides in the boundary layer of the lower troposphere, enabling associations linking column
data to surface concentrations or emissions fields. For example, reasonable correlations, especially in the
eastern U.S., have been developed between concentrations from ground level PM2.5 stations and AOD
from NASA's Moderate Resolution Imaging Spectroradiometer (MODIS) aboard the Aqua and Terra
satellites (Engel-Cox et al., 2004); see the example in Figure 2-68. The Infusing Satellite Data into
Environmental Applications (IDEA; http://idea.ssec.w isc.edu/) site provides daily displays and
interpretations of MODIS and surface air quality data. The Cloud-Aerosol Lidar and Infrared Pathfinder
Satellite Observation (CALIPSO) satellite (discussed below) provides some ability to resolve aerosol
vertical gradients.
2.9.3.2. Measurement Issues
Most satellite air quality observations are based on spectroscopic techniques typically using
reflected solar radiation as a broad source of UV-through-IR electromagnetic radiation (LIDAR aboard
CALIPSO does utilize an active laser as the radiation source). While the science of satellite based
measurements of trace gases and aerosols is relatively mature, interferences related to surface reflections,
cloud attenuation and overlapping spectra of nearby species require adequate filtering and accounting for
in processing remote signals. For example, aerosol events episodes associated with clouds often are
screened out in developing in applications involving AOD characterizations through MODIS.
Correlations between AOD and surface aerosols generally are better in the eastern U.S. relative to the
West due to excessive surface light scattering from relatively barren land surfaces.
2.9.4. European Air Monitoring Networks
Extensive air monitoring networks have also been implemented in Europe. In addition to the
programs discussed above, many European-based programs are served by centralized organization
structures linked to international efforts such as Convention on Long Range Transport of Air Pollution
(LRTAP) (http://www.unece.org/env/lrtap/) and the underlying technical assessment body, the Co-
operative Programme for Monitoring and Evaluation of the Long-range Transmission of Air Pollutants in
Europe (EMEP). The Global Atmospheric Watch (GAW) program (http://www.wmo.int/pages/prog/-
arep/gaw/gaw home en.html) under the World Meteorological Organization (WMO) provides quality
assurance guidelines and data access to an important body of air quality measurements relevant to
assessing intercontinental pollution transport and climate forcing phenomena. The Norwegian Institute for
2-122
-------
Air Research (NILU) (http://www.nilu.iio/index.cfm?ac=topics&folder xd=4572&lan id=3). maintains a
database for much of the European based networks.
10 ~
to
-
o
^r
m _
CO
o _
CO
-120 -110 -100 -90 -80 -70
-0.8 -0.6 -0.4 -0.2 0 0.2 0.4 0.6 0.8 1
Source: Engel-Cox et al. (2004).
Figure 2-68. Correlation surfaces between MODIS AOD and hourly PM2 5 surface sites from April-
September 2002.
Table 2-18 includes combined contributions from all countries ranging from a few sites to tens of
sites per country. Measurements for a variety of air pollutants are addressed including 03, heavy metals,
persistent organic pollutants (POPs). PM, VOCs, and deposition from acidifying and eutrophying
compounds.
2-123
-------
Table 2-18. International and European air monitoring programs.
Network
. . . Number ......
Lead Agency ^ g.^eg Initiated
Measurement Parameters
Location of
Information / Data
EMEP-
Co-operative Programmed for
Monitoring and Evaluation of
the Long-range Transmission
of Air Pollutants in Europe
(encompasses networks for
-37 European countries and
organizations)
UNECE
270 1977 Acidifying / Eutrophying Compounds
(precipitation): SO42", NO3, NH4, trace
elements, pH, acidity
(A): S02, N02, HNOs, NH3, PM10, PM2.5,
major ions
O3 Heavy Metals precipitation, major
ions, PM2.5, PM10, Hg, wet deposition
POPs precipitation, air, deposition PM,
PM2.5, PM10, EC, OC, TC, BC VOC HCs,
Carbonyls
http://www.nilu.no/proiects/
cccfemepdata.html
EUROTRAC—The European International
Experiment on the Transport Executive
and Transformation of Committee
Environmentally Relevant (European
Trace Constituents over Countries)
Europe
??? 1986 EUROTRAC programs performed
analyses utilizing data from existing or
specially designed monitoring networks
to:
1. elucidate the chemistry and transport
of O3 and other photo-oxidants in the
troposphere, e.g., TOR—30 O3 stations
and ALPTRAC—15 snow-monitoring
sites
2. identify processes leading to the
formation of acidity in the atmosphere,
particularly those involving aerosols and
clouds.
3. understand uptake and release of
atmospheric trace substances by the
biosphere.
http://www.asf.de/eurotrac/
index what is.html
EUROTRAC-2 —The International ??? 1996
EU REKA project on the Scientific
transport and chemical Secretariat
transformation of trace (European
constituents in the Countries and
troposphere over Europe; EU)
second phase. Subprojects:
-AEROSOL
- BIATEX-2
-CAPMAN
-CMD
- EXPORT-E2
-GENEMIS
- GLOREAM
-LOOP
-MEPOP
-PROCLOUD
- SATURN
-TOR-2
-TRAP45
-TROPOSAT
EUROTRAC-2 programs performed http://www.asf.de/eurotrac/
analyses utilizing data from existing index what is.html
monitoring networks to: support the
further development of abatement
strategies within Europe by providing an
improved scientific basis for the
quantification of source-receptor
relationships for photo-oxidants and
acidifying substances.
2-124
-------
2.9.5. Ambient Concentrations of Relevant N Compounds
2.9.5.1. NO arid NO2
Species concentrations described in this section were taken from two types of networks described
in Section 2.9.1 above: the mostly urban networks designed and maintained for NAAQS attainment
demonstrations; and the mostly rural and remote networks designed and operated to comply with a range
of requirements for protection of landscapes and views.
Figure 2-69 shows the distribution of moni toring sites for ambient-level, continuous N02 across the
U.S. Data for ambient N02 are missing or are collected at very few sites over large areas of the U.S. Few
cities have more than two monitors and several large cities, including Seattle, WA, have none. Note that
the number of N02 monitors has been decreasing in the U.S. as ambient average concentrations have
fallen to far below the level of the NAAQS. There were, for example, 375 NQ2 monitors identified in
mid-2006, but only 280 in November 2007.
Figure 2-69. Location of ambient-level NO? monitors for NAAQS compliance in 2007. Shaded states have
NO2 monitors; unshaded states have none.
Criteria for siting ambient monitors for NAAQS pollutants are given in the SLAMS /NAMS /
PAMS Network Review Guidance (U.S. EPA, 1998). As might be expected, criteria for siting monitors
differ by pollutant. N02 monitors are meant to be representative of several scales: middle, or several city
blocks, 300 to 500 m; neighborhood, or 0.5 to 4 km; and urban, or 4 to 50 km. Middle- and
neighborhood-scale monitors are used to determine highest concentrations and source effects, while
neighborhood- and urban-scale monitors are used for monitoring population exposures. As can be seen,
there is considerable overlap between monitoring objectives and scales of representativeness. The
distance of neighborhood- and urban-scale monitor inlets from roadways increases with traffic volume
and can vary from 10 to 250 m away from roadways as traffic volume increases. Where the distance of an
inlet to a road is shorter than the value in this range for the indicated traffic volume on that road, that
monitor is classified as middle scale. Vertically, the inlets to N02 monitors can be set at a height from 2 to
15 m.
Figure 2-70 shows box plots of ambient concentrations of N02 measured at all monitoring sites
located within MSAs or urbanized areas in the U.S. from 2003 through 2005. As can be seen, mean N02
2-125
-------
concentrations are -15 ppb for averaging periods ranging from a day to a year, with an interquartile range
(IQR) of 10 to 25 ppb. However, the average of the daily 1-h maximum N02 concentration over this 3
year period is -30 ppb. These values are about twice as high as the 24-h average. The highest maximum
hourly concentration, -200 ppb, found during the period of 2003 to 2005 was more than a factor of 10
greater than the overall mean 24-h concentrations. The ratio of the 99th percentile concentration to the
mean ranges from 2.1 for the 1-year averages, to 3.5 for the 1-h averages.
100
90
80
70
80
50
40
30
20
10
0
f ^201 ^ —^201 jg —^129
1- h ma*
1-h
•
•
24-h
*
"X
2 week
* MAX
50 • Mean
1-year
Figure 2-70. Ambient concentrations of NO2 measured at all monitoring sites located within Metropolitan
Statistical Areas (MSAs) in the U.S. from 2003 through 2005. * max; • mean
Because ambient N02 monitoring data are so sparse across the U.S., and are particularly so in rural
areas, it would not be appropriate to use these data for constructing a complete coverage map of N02
concentrations. The short x of N02 with respect to conversion to NOz species and the concentrated nature
of N02 emissions result in steep gradients and low concentrations away from major sources that are not
adequately captured by the existing monitoring networks. For this reason, model predictions might be
more useful for showing large-scale features in the distribution of N02 and could be used in conjunction
with the values shown in Figure 2-71 to provide a more complete picture of the variability of N02 across
the U.S. Monthly average N02 concentrations for January and July 2002 calculated using U.S. EPA's
CMAQ model are shown in Figure 2-72. (A description of the capabilities of CMAQ and other three-
dimensional CTMs is given in Section 2.8) The high variation in N02 concentrations of at least a factor of
10 is apparent in these model estimates. As expected, the highest N02 concentrations are seen in large
urban regions, such as the northeast corridor, and lowest values are found in sparsely populated regions
located mainly in the West. N02 concentrations tend to be higher in January than in July.
2-126
-------
¦ 20.000 112
15.000
10.000
5.000
0.000 1
Figure 2-71.
Trends in N02 concentrations across the U.S. from 1990 to 2006 are shown in Figure 2-72. The
white line shows the mean values and the upper and lower borders of the shaded areas represent the 10th
and 90th percentile values. Information on trends at individual local air monitoring sites can be found at
www.epa.gov/airtrends/nitrogen.html.
0.06
— 0.05
Q.
& 0.04
C
o
"S 0.03
§ 0.02
C
o
° 0.01
o
90 92 94 96 98 00 02 04 06
1990 to 2006: 30% decrease
S ou rc e: www.e pa .g ov/a irtre nd s/nitro gen. htm I.
170 sites
National Standard
90 percent of sites are below this line.
I
Average
10 percent of sites are below this line.
Figure 2-72. Nationwide trend in NO2 concentrations. The white line shows the mean values, and the
upper and lower borders of the green (shaded) areas represent the 10th and 90th percentile
values. The current NAAQS of 0.53 ppm is shown with the dotted line.
Concentrations were substantially higher during earlier years in selected locations and contributed
in those years to the brown clouds observed in many cities. Residents in Chattanooga, TN, for example,
were exposed more than 30 years ago to high levels of N02 from a munitions plant (Shy and Love, 1980)
Annual mean N02 concentrations there declined from -102 ppb in 1968 to ~5 I ppb in 1972. There was a
strike at the munitions plant in 1973 and levels declined to -32 ppb. With the implementation of control
January 2002
Min = 0.019 at (1,1), Max = 45.966 at (23,46)
0.000
July 2002
Min = 0.012 at (6,4), Max = 40.802 at (23,46)
Monthly average NO2 concentrations (ppb) for January 2002 (left panel) and July 2002 (right
panel) calculated by CMAQ (36 X 36 km horizontal resolution).
20.000 112
15.000
10.000
5.000
2-127
-------
strategies, values dropped further. In 1988, the annual mean N02 concentration varied from -20 ppb in
Dallas, TX and Minneapolis, MN to 61 ppb in Los Angeles, CA. However, New York City, with the
second-highest annual mean concentration in the U.S. in 1988, the mean N02 concentration was 41 ppb.
The month-to-month variability in 24-h average N02 concentrations at two sites in Atlanta, GA, is
shown in Figure 2-73. (Similar plots of variability at other individual sites in selected urban areas are
shown in Figure 2-74 through Figure 2-81; these cities were chosen to represent regions with large
populations and, hence, large emissions from on-road vehicles and combustion for energy production, the
two largest sources of NO and N02.
a, Atlanta, GA. SUBURBAN
0.09
0.08
I 0.0?
ta06
'3 0-05
¦§ 0.04
| 0.03
O 0.02
0.01
0.00
01/01/2003 07/01/2003 01/01/2004 07/01/2004 01/01/2005 07/01/2005 01/01/2006
Sample Date (mm/dd/yyyy)
site id=130890002 poc«1
* Natural Spline Fit w/9df
b. Atlanta, GA,
0.09
__ 0.08
1. 0.07
f 0.06
XS 0.05
"g 0.04
g 0.03
O 0.02
URBAN and CENTER CITY
site 10=131210048 poe=1
Figure 2-73.
0.01
0.00
01/01/2003 07/01/2003 01/01/2004 07/01/2004 01/01/2005 07/01/2005 01/01/2006
Sample Date (mm/dd/yyyy)
Time series of 24-h average NO2 concentrations at individual monitoring sites in Atlanta, GA
from 2003 through 2005. A natural spline function (with 9 degrees of freedom) was fit and
overlaid to the data (dark solid line).
2-128
-------
a, N»w York, NY, SUBURBAN
b. New York, NY.
URBAN and CENTER CITY
0074
OOH
I !...
i I j j j
otfownro ctcm/mb owisoq* 87*>t/2ow mmmm
Sample Date (mmMd/yyyy)
0.57
0.08
mmtzm, Hoiaae
owi/ww oroiaae quowzom btoisskm ot/oi/20B& ww/sow owiooob
Sample Date (mm/dd/yyyy)
e. n«w York, ny. URBAN and CENTER CITY
I
«s«3"0d00SCHC9oC-1
GO?
006
&Q5
0.04
003
002
Cfr»
&0S
I
&mmn
P"M
d. New York, ny. URBAN and CENTER CITY
9.09
—
a.
o.o®
006.
oicw<
aoa.
002
001
0.00
OMttfflDO3 MS?®® CWOVMCH 01*51/2005 CM)toOC6 OlfllfflMB
Sample Oat© (mm/dd/yyyy)
t ft
i "t i
oi/dioou o&oiaxQ oiavatxw qtosom
? 1 - f
ow/zoos t>7mm& mmmm
Sample Date (mm/dd/yyyy)
•. New York, ny. URBAN and CENTER CITY
%m *}• jw»ioosspcx. * \
I • I
"t 'HTPfl^T
»T .
0011 «
o.eo^
OtfcVMOJ OWH/JOQJ 0UO1/20W OfllMW 0MM*0» 07«1/200S. 0^/2006
SampI® Date {mm/dd/yyyy)
Source: U.S. EPA AQS, 2007
Figure 2-74. Time series of 24-h average NO2 concentrations at individual monitoring sites in New York
City from 2003 through 2005. A natural spline function (with 9 degrees of freedom) was fit and
overlaid to the data (dark solid line).
2-129
-------
a, Chicago, IL.
RURAL
b. Chicago, IL.
SUBURBAN
0.00'
0.08'
0.0?'
0.05-
0.05'
0.04'
0.03-
002'
0.01'
0.00'
='teto3^SpfeeFa'^'9^|
01/0-1/2003 07/01/2003 01/01/2004 07/01,C2Q04 01/01,3005 07/01/2005 01/01/2006
Sample Date (mm/dd/yyyy)
1/03100f'6poc
01/01/2003 07/01/2003 01/01/200* 07/01.^2004 01/01/200& 07/01/2003 01/01/200S
Sample Date (mm/dd/yyyy)
c. Chicago, IL.
0.09'
0.06'
0-0?'
006.
0.D5'
0.04'
003'
0.02'
0.01'
0.00'
SUBURBAN
d. Chicago, IL.
SUBURBAN
site W» 170314002 0oc
it
01/01/2003 07/01(2003 01/01/2004 07/01/2004 01/01/2005 07/01/2005
Sample Date (mm/dd/yyyy)
01/01*2003 07/01/2003 01/01/2034 G7M2G04 01/01/200S
Sample Date (mm/dd/yyyy)
07/01/2005 01/01/2006
e. Chicago, IL.
SUBURBAN
0.00-.
*i s 1 r
01/01/2003 07/01/2003 01/01/2004 07/01/2004
f. Chicago, IL.
URBAN and CENTER CITY
!70310063
m
07/01/2005 01/01/2006
i 1 r~
01/01*2003 07/01/2003 01/01/2004 07/01 ,>2004 01/01/2.035 07/01/2005 01/01/2006
Sam pie Date (mm/dd/yyyy)
Sample Date (mm/dd/yyyy)
g. Chicago, IL.
0.09'
0.08-
E 007'
B 0.06-
URBAN and CENTER CITY
i703>0072 poc " 1
1 1 s 1 1 r
01/01/2003 Q7/01/2QQ3 01/01/2004 07/01/2004 01/01/2005 07/01/2005 01/01/2006
Sample Date (mm/dd/yyyy)
Figure 2-75. Time series of 24-h average NO2 concentrations at individual monitoring sites in Chicago, IL
from 2003 through 2005. A natural spline function (with 9 degrees of freedom) was fit and
overlaid to the data (dark solid line).
2-130
-------
a. Baton Rouge, LA,
SUBURBAN
sft»kt«221210001 poc=1
= Natural Spline fit wl 9 df
0.09:
0.0B:
0.07'
0.06
0.05-
004
0.03.
0.02'
0.01.
0.00
01/01/2003 07/01/2003 01/01/2004 07/01/2004 01/01/2005 07/01/2005 01/01/2008
Sample Date (mm/dd/yyyy)
P|^i|r|
b. Baton Rouge, LA,
0.03-
01/01/2003
URBAN and CENTER CITY
site id-220330009 poc=1
r* TfPCi
f Is if
n
~r
07/01/2003 01/01/2004 07/01/2004 01/01/2005 07/01/2005 01/01/2006
Sample Date (mm/dd/yyyy)
Figure 2-76. Time series of 24-h average N02 concentrations at individual monitoring sites in Baton Rouge,
LA from 2003 through 2005. A natural spline function (with 9 degrees of freedom) was fit and
overlaid to the data (dark solid line).
2-131
-------
a. Houston, TX.
SUBURBAN
s0*4820*1035 pec *
0.07-
0,06'
PIP
'
Iff?
ao»'
0.08'
0.0?'
j ¦ - wnj - •'v, iTrr v*3 - -n
'r1""1" 'i i ——t i n* i
0V01/2CC3 07W/2003 01/01/2004 07/01/2004 01/01/2005 07/01/2005
Sample Date (mm/dd/yyyy)
01/01/2003 07/01/2003
01/01/2004 07/01/2004 01/01/2005 07/01/2006
Sample Date (mm/dd/yyyy)
e, Houston, TX.
URBAN and CENTER CITY
f. Houston, TX. URBAN and CENTER CITY
sire w " 48ZCH20?9 poc
009'
aos-
007'
01/01/2003 07/01/2003 01/01/3004 07/010004 01/05,-2005 07/01/2005 01/01/20C6
005-
0.00-
t 1 r 1 1 1 r
01/01/2003 07/01/2033 01/01 >2004 07/01/2004 01/01/2005 07/01/2005 01/01^2006
Sample Date (mm/dd/yyyy)
Sample Date (mm/dd/yyyy)
g. Houston, TX. URBAN and CENTER CITY
Si« id = 463390078 poc* 1
01A51/2G03 07/01'2003 01/01/2004 07/01/2004 01»1fflX» 07/01/2005 01/Q1/2006
Sample Date (mm/dd/yyyy)
Figure 2-77. Time series of 24-h average NO2 concentrations at individual monitoring sites in Houston, TX
from 2003 through 2005. A natural spline function (with 9 degrees of freedom) was fit and
overlaid to the data (dark solid line).
2-132
-------
a. Los Angeles, CA.
SUBURBAN
b. Los Angeles, CA.
SUBURBAN
io ~06Q3?!X)02 ooc *¦ 2
Uiill
.,a
01/01 izm 07/01/2003 01/01/2004 07/01/2004 W/OlfflOQS 07/01/2005 01/01/2006
=d->0603?t2C1 poc»?
mm
Sample Date (mm/dd/yyyy)
07/01/2003 01/01/2004 07/01,2004 01»1S006 07/011-2005
Sample Date (mm/dd/yyyy)
c. Los Angeles, CA.
SUBURBAN
d. Los Angeles, CA.
SUBURBAN
060371'01 poc = 2
01/01^003 07/01/2003 01'OV?OCH CTO1/2004 01/01/2005 07/01/2005 01/01/2006
Sample Date (mm/dd/yyyy)
01,-01,2003 07/01/2003 0W1/2Q04 07/D1/2004 01/01,-2006 07/01/2005 01/01/2006
Sample Date (mm/dd/yyyy)
e. Los Angeles, CA,
SUBURBAN
f. Los Angeles, CA.
SUBURBAN
0603T4002 poc « 2
SitS *J = 060375055BOC
-i 1 1 1 5 r
01/01/2003 0MM/2Q03 01/01/2004 07/01/2004 01/01/2005 07/01/2005 01/01-
01/01/2003 07/01/2003 01/01/2004 07/01/2004 01/01,-2005 07/01/2005 01/01/2006
Sample Date (mm/dd/yyyy)
Sample Date (mm/dd/yyyy)
g. Los Angeles, CA.
SUBURBAN
h. Los Angeles, ca. URBAN and CENTER CITY
0 094
o.oe|
0.07-3
AkMk
01/01/2003 07/01/2003 01/01/2004 07/01/2004 01/01/2005 07/01/2005 01/01/2006
Figure 2-78.
Sample Date (mm/dd/yyyy)
0101/2003 S7&1SK3 01/012004 07B1/2004 01/O1MS QTffilffiOOS 01i£s1fiM6
Sample Date (mm/dd/yyyy)
Time series of 24-h average NO2 concentrations at individual monitoring sites in Los Angeles,
CA from 2003 through 2005. A natural spline function (with 9 degrees of freedom) was fit and
overlaid to the data (dark solid line).
2-133
-------
i. Lo. Anjefes, ca, URBAN and CENTER CITY
s
! . 1
EE
miff ]'i ¦
j. ios Angeles, ca, URBAN and CENTER CITY
©i«ii2C03 owwawj amm* mtrnm mmmm mmom omam
Sample Date (mm/dd/yyyy)
OlOB
0.07
m
I. I
h*Y\T ' ?!
• i'
m.
f I i ' i 1 f
vmrAm omom otoicow ouotooos mmvoo* mmmm
Sample Dat# (mm/dd/yyyy)
k. tos Angeles, ca. URBAN and CENTER CITY
am-
§ as?-
I nJ 1
»d»c*2
6014
©»3
I r-r
,!
i. Los Anjeies, ca. URBAN and CENTER CITY
m
o.<
00T
506
COS.]
0©H
wIW.l
&«| Ifffm
B»l
I rw'vtii^¥ii%
**r
T
T*
wmtssm o7/oi,m\rm owiaow mmmm mm txm mmmm y?®im ot«i/2oo4 wmrm
Sample Date (mm/dd/yyyy) Sample Date (mm/dd/yyyy)
m. Lo® Angers, ca, URBAN arid CENTER CITY
ft us Angeles, ca. URBAN and CENTER CITY
tita 4 •- 0W375W1 Poc »
Figure 2-79.
T"
T"
T"
ooe
008
007
008
»«iC*0603:«5JJ!>oe»t
m-
ei— r"— 1 """"T"
dimmm mmm mum •iw. owvaws mmm mtow mmm mm» mum mm* mmm mnm
Sample Date (mm/dd/yyyy) Sample Date (rmn/dd/yyyy)
Time series of 24-h average NO2 concentrations at individual monitoring sites in Los Angeles,
CA from 2003 through 2005. A natural spline function (with 9 degrees of freedom) was fit and
overlaid to the data (dark solid line).
2-134
-------
a. Riverside, CA. RURAL b. Riverside, CA. SUBURBAN
01/01/2003 07/01/2003 01/01/2004 07AJ1/2004 01/01/2006 07/01/2006 01/01/2006 01/01/2003 07/01/2003 01/01/2004 07/01/200* 01/01/200$ 07/01/2005 01/01/2006
Sample Date (mm/dd/yyyy) Sample Date (mm/dd/yyyy)
c. Riverside, CA. SUBURBAN d. Riverside, CA. SUBURBAN
01/01/2003 07/01/2003 01/01/2004 07/01/2004 01/01/2006 07/01/2005 01/01/2006 01/01/2003 07*51/2003 01/01/2004 07/01/2004 01*11/2005 07/01/2006 01/01/2006
Sample Date (mm/dd/yyyy) Sample Date (mm/dd/yyyy)
Figure 2-80. Time series of 24-h average NO2 concentrations at individual monitoring sites in Riverside, CA
from 2003 through 2005. A natural spline function (with 9 degrees of freedom) was fit and
overlaid to the data (dark solid line).
Strong seasonal variability exists in N02 concentrations in the data shown above. Higher
concentrations are found during winter, consistent with the generally lower PBL depths in winter. Lower
concentrations are found during summer, consistent with deeper PBLs and increased rates of
photochemical oxidation of N02 to NOz. Note also the day-to-day variability in N02 concentration, which
also tends to be larger during the winter. There appears to be a somewhat regular pattern for the other
southern cities examined with their winter maxima and summer minima.
Monthly maxima tend to be found from late winter to early spring m Chicago, IL, and New York,
N Y, with minima occurring from summer through fall. However, in Los Angeles and Riverside, CA,
monthly maxima tend to occur from fall through early winter, w ith minima occurring from spring through
early summer. Mean and peak N02 concentrations during winter can be up to a factor of 2 greater than
those during the summer at sites in Los Angeles.
The diel variability in N02 concentrations at the same two sites in the Atlanta metropolitan area
shown in Figure 2-73 is illustrated in Figure 2-82. As can be seen from these figures, N02 typically
exhibits daily maxima during the morning rush hours, although they can occur at other times of day. In
addition, there are differences between weekdays and weekends. At both Atlanta sites, N02
concentrations are generally lower and the diel cycles more compressed on weekends than on weekdays.
The diel variability of N02 at these sites is typical of that observed at other urban sites. Monitor siting
plays a role in determining diel variability in the sense that monitors located farther from traffic will
sometimes measure lower concentrations and show a flatter overall distribution of data compared to
monitors located closer to traffic.
2-135
-------
a. RivartM*, CA.
SUBURBAN
1 Riverside, CA.
SUBURBAN
I f I .
r * i * t*' •*
»x*mi*o«oy»:-i>3JcH>£»
03?|
QQW
,1 .! 1
1
qwoom o7/oi'"2ocs3 oi/owtox mmmm mmtmm mm .®g§ a ignore
OWiM© 0WKZOQ3 01JQU30Q* @M«l!3SM 01JQ1QQQG OHWZOOfr 01i&faa&
Sample Oat© (mm/dd/yyyy)
Sample Date (mm/dd/yyyy)
g. Riverside, CA.
SUBURBAN
oo? i
0.9
0-054
taerf-mrtSW-'iK*;' '
t I - T I
h, RiversWe, ca, URBAN and CENTER CITY
Pff
0W- '
303
ocs|
0Q1-
chdmob duotscx* omtam mmmm mmmm mmmm mmtxm
Sample Date (mm/dd/yyyy)
m mmm wmm* omrns otmam mmmm
Sample Date {mm/dd/yyyy}
i, Riverside, CA,
URBAN and CENTER CITY
v*
*«? f
1 I
»1L , is
it
"i•
Ml-
005 .
i 1 1 f 1 s r1
mmtms musm mmsm mam* mimm mmmm oiwomb
Sample Date (mm/dd/yyyy)
Source: U.S. EPA (2006a)
Figure 2-81. Time series of 24-h average NO2 concentrations at individual monitoring sites in Riverside, CA
from 2003 through 2005. A natural spline function (with 9 degrees of freedom) was fit and
overlaid to the data (dark solid line).
2-136
-------
A. Atlanta, GA Suburban
E
Q-
Q»
C
o
H3
-------
HHln h|ihO I lu| 11 h|im j i ml h h|
-------
0.20
0 20 40 60 80 100
[NOx] (ppb)
Source: U.K. AQEG (2004).
Figure 2-84. Ratios of PAN to NO2 observed at Silwood Park, Ascot, Berkshire, U.K. from July 24 to August
12,1999. Each data point represents a measurement averaged over 30 min.
HNO2
Measurements of HN02 in urban areas are extremely limited; however, data from Stutz et al.
(2004) and Wang and Lu (2006) indicate that levels of HN02 are <1 ppb even under heavily polluted
conditions, with the highest levels found during the night and just after dawn and the lowest values found
in the afternoon. However, data collected in the U.K. (Lammel and Cape, 1996; U.K. AQEG, 2004) and in
the U.S. (Kirchstetter and Harley, 1996) indicate that HN02-to-NOx ratios could be of the order of -5% in
motor vehicle emissions. These results indicate that HN02 levels in traffic could be comparable to those
of N02. Several field studies conducted at ground level (Hayden et al., 2003, near Boulder, CO) and
aircraft flights (Singh et al., 2007, over eastern North America), have found much higher NOz
concentrations than NOx concentrations in relatively unpolluted rural air, some of which could be
attributed to PAN.
The ratio of HN02 to N02 as a function of NOx measured at a curbside site in a street canyon in
London, UK is shown in Figure 2-85, where HN02 is labeled HONO. The ratio is highly variable, ranging
from about 0.01 to 0.1, with a mean -0.05. As N02 constitutes several percent of motor vehicle emissions
of NOx, the above implies that emissions of HN02 represent a few tenths of a percent of mobile NOx
emissions. A similar range of ratios has been observed at other urban sites in the United Kingdom
(Lammel and Cape, 1996).
2-139
-------
0,3
O
z
S
o
Z
o
£
0.2
0.1
o 0o
«o°©cDo
o o
200
o
o
o
o5>
400 600
[NOxJ (ppb)
800
1000
Source: UK AQEG (2004).
Figure 2-85. Ratios of HNO2 to NO2 observed in a street canyon (Maryiebone Road) in London, U.K. from 11
a.m. to midnight during October 1999. Data points reflect 15-min average concentrations of
HONO and N02.
HNOs and NO3-
Data for concentrations of HN03 and N03 in urban areas in the U.S. are sparse. The most
geospatially intensive set of data for any HN03 were taken as part of the Children's Health Study for
which gas-phase HN03 was measured at 12 sites in southern California from 1994 through 2001 (Alcorn
and Lurmann, 2004). Two week average concentrations ranged from <1 ppb to >10 ppb, with the highest
HNO3 concentrations and highest ratio of HN03 to N02, -0.2, was found downwind from central Los
Angeles in the San Bernardino area during summer, as one would expect for this more oxidized N
product.
HNO3 data have also been reported from the SEARCH network of four pairs (eight total sites) of
urban and rural sites in the southeastern U.S using increasingly sophisticated methods since 1998; see
Zhang et al. (2006). Concentrations of HN03 in this area have ranged from <1 ppb to >10 ppb.
Maps of ambient concentrations from CASTNet data for rural and remote areas are available
below. The CASTNet ambient concentration maps were produced with Arclnfo using an inverse distance
weighting (IDW) interpolation technique. Using IDW, the surface is most influenced by the nearest point
values and less so by more distant points. CASTNet sites within 400 km of each grid point were used in
this calculation. As noted above, thin data coverage complicates interpretation of these maps and renders
them most useful as heuristic guides to large areas of possible differences. Strict quantitative values
should not be imputed to areas away from the measurement sites.
Ranges of years in the chart represent 3 year averages. For example, 2004-2006 is the average
concentration of 2004, 2005 and 2006, as calculated from gridded output for each of the years. The three
annual grids from the 3 year period were averaged to derive the mean concentration of the 3 year period.
Only sites meeting completeness criteria for at least two of the three years of the averaging period were
included.
Figure 2-86 shows annual average concentrations for gas-phase HN03 from CASTNet for the years
2004 through 2006. Because HNO3 is produced mostly as a secondary product from emitted NO, the
regions of higher concentrations HN03 are geographically similar to those of high concentration NO and
N02: the northeast corridor and southern California.
2-140
-------
Source: CASTNet, USEPA/CAMD 7/30/07
Figure 2-86. Annual average gas-phase HN03 concentrations, 2004-2006. White areas on the map are
areas where monitoring sites are absent and no information is available.
HNO3
(^g/m3)
-0.0
-0.5
-1.5
Because they have the same precursor reactants in NO and NO> elevated 03 concentrations are
often associated with elevated HN03 concentrations. However, HN03 can be produced in significant
quantities in winter, even when 03 concentrations are low. The ratio between 03 and HNO-. also shows
great variation in air pollution events, with NOx-saturated environments having much lower ratios of 03
to HN03 (Ryerson et al., 2001). pN03 is formed primarily by combining N03 supplied by HN03 and NH3
and may be limited by the availability of either reactant. pN03 is expected to correlate loosely with 03,
whereas NH3 is not expected to correlate with 03.
Thus, annual average pN03 can account for several ppb of NOVr with higher values in the West.
There is strong seasonal variation, which is especially pronounced in western areas where there is
extensive wood burning in the winter resulting in a larger fractional contribution of local sources. Areas in
the East where there are topographic barriers might be expected to show higher fractional contributions
from local sources than other eastern areas that are influenced by regionally dispersed sources. Figure
2-87 shows a map of annual average NO; concentrations in the years 2004 to 2006 produced from
CASTNet measurements of ambient concentrations. This maps indicates at least qualitatively that
maximum NO-, concentrations are found in areas of maximum NO and N02 emissions.
2-141
-------
N03-
(pg/m3)
,— 0.0
Source: CASTNet, USEPA/CAMD 7/30/07
Figure 2-87. Annual average gas-phase NO3 concentrations, 2004-2006. White areas on the map are areas
where monitoring sites are absent and no information is available.
2.9.5.3. Nitro-PAHs
Nitro-PAHs are widespread and found even in high altitude, relatively unpolluted environments
(Schauer et ai., 2004). However, there are differences in composition and concentration profiles both
within and between monitoring sites (rural vs. urban) as well as between and within urban areas (Albinet
et al., 2006; Naumova et al., 2003; Soderstrdm et al., 2005), with some differences in relative abundances
of nitro- and oxo-PAHs also reported. Source attribution has remained largely qualitative with respect to
concentrations or mutagenicity (Eide et al., 2002). The spatial and temporal concentration pattern for the
nitro-PAHs may differ from that of the parent compounds because concentrations of the latter are
dominated by direct emission from local combustion sources. These emissions results in higher
concentrations during atmospheric conditions more typical of wintertime when mixing heights tend to be
low. The concentrations of secondary nitro-PAHs are elevated under conditions that favor hydroxyl and
NO; radical formation, i.e., during conditions more typical of summertime, and are enhanced downwind
of areas of high emission density of parent PAHs and show diurnal variation (Fraser et al., 1998; Kameda
et al., 2004; Reisen and Arey, 2005). Nitro-napthalene concentrations in Los Angeles, CA varied between
about 0.15 to almost 0.30 ng/nr compared to 760 to 1500 ng/m for napthalene. Corresponding values for
Riverside, CA were 0.012 to more than 0.30 ng/nr for nitro-napthalene; and 100 to 500 ng/m 5 for
napthalene. Nitro-pyrene concentrations in LA varied between approximately 0.020 to 0.060 ng/m'
compared to 3.3 to 6.9 ng/m' pyrene, whereas corresponding values for Riverside were 0.012 to
0.025 ng/nr and 0.9 to 2.7 ng/nr'.
2.9.5.4. Ammonia
Section 2.7.1.5 above established that a successful real-time continuous monitoring technique for
ambient NH3 has not been identified; and, of equal importance, Section 2.4 above described the severely
limiting unknowns related to NH3 emissions on national and local scales. With these important
2-142
-------
information gaps, estimates ofNH3 concentrations at any scale for the U.S. must be constructed and
interpreted with caution. It is possible, for example, to rank NH3 concentrations by land use types from
the few special field campaigns where it was measured as Walker et al. (2004) did for agricultural, non-
agricultural, and urban types; see Table 2-19. This table shows the enormous range in NH3 concentrations
by season and land use type, from a low of 0.02 (ig/m3 in summer over alpine tundra on Niwot Ridge,
CO, (Rattray and Sievering, 2001) to 11.0 (ig/m3 over fertilized lands in Wekerom in the central
Netherlands (Buijsman et al., 1998).
Table 2-19. Ambient NH3 concentrations summarized by study.
Concentration (ngm3)
Land Use
Comment
3.0
Agricultural
Low NH3 emissions
11.0
Agricultural
Moderate NH3 emissionsa
10.48
Agricultural
Fall"
0.65-1.2
Agricultural
Springc
0.26
Agricultural
Winterc
0.34
Non-agricultural
High elevation, summer and falld
0.02
Non-agricultural
High elevation, summere
0.62-1.47
Non-agricultural
High elevation, summer'
0.29
Non-agricultural
High elevation, summer s
0.22
Non-agricultural
Coastal, summer s
0.21
Non-agricultural
Forest, summer s
0.16
Non-agricultural
Foresth
0.21
Non-agricultural
Wetland, summer s
0.23
Non-agricultural
Wetland, summer s
0.63
Non-agricultural
Desert, summer s
4.75
Non-agricultural
Grassland, summer s
0.38-1.49
Urban
Pittsburgh, PA; summer'
0.63-0.72
Urban
Research Triangle Park, NC; fall i
1.18
Urban
Vinton, VA; summerk
8 Buijsman et al. (1998). e Rattray and Sievering (2001) 1 McCurdy et al. (1999).
bMcCulloch et al. (1998). f Aneja et al. (2000). i Sickles et al. (1990).
c Pryor et al. (2001). sLangford et al. (1992). k Leaderer et al. (1999)
dTarnay et al. (2001). hLeferetal. (1999)
A preliminary draft U.S. national-scale, county-level NH3 map was created by the U.S. EPA using
emissions data from the widely-used Carnegie Mellon University (CMU) NH3 emissions model,
version 3. See Pinder et al., (2007) for a description of the model and application information. The map
included an empirical relationship between emissions and ambient concentration derived from field
measurements over five different land use types in North Carolina, covering the range of expected NH3
2-143
-------
emissions densities. In the CMU emissions model used for this analysis, fertilized soils were included but
NH3 emissions from natural (non-fertilized soils) were not because of overwhelmingly large uncertainties
in their emissions factors. Emissions from mobile sources are included in this analysis, although
uncertainties in those values are also large; see Section 2.4. above. Emissions from a small number of
counties having very low emissions rates were set to missing because their extremely low values
invalidated determination of densities in those counties.
The CMU-derived county-level NH3 emissions map is shown in Figure 2-88. Differences in the
techniques for estimating NH3 emissions and the very large uncertainties in NH3 emissions factors and
totals complicate direct, fine-scale comparisons to other NH3 emissions maps. However, both the
magnitude and areal extent of Walker's emissions map compare very favorably to the U.S. EPANEI
database NH3 emissions map shown in Section 2.4 above.
Ammonia Emission Density
2,001 - 2,500
I I 2,801 -S,000
I I 5.001 -7,500
I I 7,501 - 12,000
Source: J. Walker, USEPA/ORD/NRMRL
Figure 2-88 County-scale NH3 emissions densities from the CMU inventory model.
The empirical relationship between NH3 emissions and ambient NH3 concentration is shown in
Figure 2-89.
2-144
-------
7
y = 0.63 + 5.562e-4x
R2 = 0.99
6
Kenansville
Clinton ••
5
4
3
Kinston
2
• Lewiston
Morehead City
1
0
0
2000 4000 6000 8000 10000 12000
County-scale NH3 emission density (kg NH3/km2)
Source : J Walker, U.S. EPA/ORD/NRMRL
Figure 2-89. Ambient NH3 concentration as a function of county scale NH3 emissions.
The map of estimated ambient NH3 concentration from the CMU emissions model and empirical
relationship illustrated above is shown in Figure 2-90.
Ambient Ammonia Concentration
Figure 2-90. Estimated county-scale ambient NH3 concentrations.
H 1.23-1.36
1 37 -1.56
1.57 - 2 25
[ I 2 26-350
I I 3.51 - 5 00
I I 5 01 - 6 60
Source: J. Walker, USEPA/ORD/NRMRL
2-145
-------
2.9.5.5. NH4N03
The IMPROVE network is the premier source of data about the spatial and temporal patterns of
rural and remote pN03 in the U.S. Data about urban pN03 and other particles comes primarily from the
CSN, which includes STN and others (Jayanty, 2003). Although IMPROVE began and still functions
primarily to characterize visibility impairments in protected areas, the network has been expanded several
times since its inception to include coverage of remote areas in central and western states of the U.S. to
better understand regional components of particulate pollution. Except where noted, information in this
section was derived from the IMPROVE IV Report (DeBell, 2006) with data displays created with data
and tools available at http://vista.cira.colostate.edii/views/. Much of these data and conclusions are also to
appear in Pitchford et al. (2008).
IMPROVE monitors the major fine particle components pS04, pN03, crustal EC and OC, and
coarse mass concentration computed as the concentration difference PMi0 minus PM2 5. An implicit
assumption is that most of the pN03 is present as NH4N03 in the PM2 5 size range. One component of the
Big Bend Regional Aerosol and Visibility Observational (BRAVO) Study, conducted at Big Bend
National Park, TX, in the summer and fall of 1999, entailed use of detailed measurements of aerosol
chemical composition, size distribution, water growth, and optical properties to characterize the aerosol
and assess the relationships among aerosol physical, chemical, and optical properties (Schichtel, et al.,
2004). Fine pN03 accounted for <5% of the mass concentration in these samples and was present mostly
as NaN03. Approximately 67% of the pN03, again inferred to be NaN03, was found in the coarse mode
where it comprised -8% of the mass concentration.
However, depending on the acidity of the particles, which in turn depends strongly on their S042
and NH4+ contents, higher pN03 concentrations could be found in coarse mode particles (PM10-2.5) than in
PM2 5 samples. The average pN03 content of PM25 and PMi0 is typically -1% in the eastern U.S. and
15.7% and 4.5%, respectively, in the western U.S. (U.S. EPA, 1996a). These values suggest that most of
the pN03 was in the PM2 5 size fraction in the studies conducted in the western U.S., but pN03 in the
studies in the eastern U.S. was mainly in the PMi0.2.5 size fraction.
A year-long special study of PMi0 speciation was conducted at nine IMPROVE remote area
monitoring sites during 2003 and 2004 to provide additional information about the geographic and
seasonal variations in coarse particle composition; see Malm et al. (2007). The same sampling and
analytical procedures were used for the PMi0 samples as are routinely used on the IMPROVE PM2 5
samples. IMPROVE PM10 speciation study did not include NH44" analysis, so pN03 was assumed again to
be NH4N03. As expected, crustal minerals were the largest contributors to PMi0 mass overall at -60%,
and organic particles contributed -25%. On average, N03 was the third largest contributor to PMi0 mass
at -8% on average for the nine monitoring sites. The sites with the highest coarse pN03 concentrations
were the two in California, San Gorgonio, 0.74 |ig/nr' and Sequoia, 0.69 (ig/m3, where PM2 5 pN03 were
also high on average, 2.66 |ig/m3 and 2.14 |ig/m3 respectively. Brigantine, a coastal site in New Jersey had
the highest fraction of total PMi0 pN03 at 36%. The authors speculated that Brigantine's pN03 was likely
NaN03, the result of HN03 reactions with sea-salt NaCl. The nine-site average fraction of total coarse
pN03 was 26%.
Figure 2-91 shows maps of remote NH4N03 for two years selected to demonstrate the additional
information available after expansion of IMPROVE into the central U.S. The locations of monitoring sites
supplying the data shown as color contours are shown as dot on the maps. As above, concentration
contour maps carry the caution that their isolines are provided merely to guide visual similarities among
sites reporting similar values, and should not be read to suggest any quantitative spatial interpolation
where sites do not exist. These plots, showing the so-called Midwest pN03 bulge illustrate why that
caution is always warranted.
2-146
-------
&OUJ1
Wj*Dl
2GG4 Annual
ammN03f
ammN03f
Source: http:/A/ista .cira.colostate.edu/views
Figure 2-91. IMPROVE network measured annual averaged NH4NQ3 concentration for 2000 (left) and for
2004 (right).
Before 2001, no IMPROVE or any other regular remote-area aerosol speciation monitoring sites
existed in the central states between northern Minnesota and Michigan to the north, and Arkansas and
Kentucky to the south. The lack of monitoring over such a large region in the center of the country hid the
largest regional distribution of particulate matter dominated by NH4NO3, previously thought to be a
phenomenon isolated to California. However, with fewer than 6 years of complete data for this region,
insufficient information exists to test for long-term trends.
Combining IMPROVE and CSN data makes possible comparison of urban pN03 to surrounding
remote-area regional values. These are shown as paired color contour maps for IMPROVE and
IMPROVE plus CSN in Figure 2-92. U.S. EPA (2004) used the pairing of IMPROVE and CSN
monitoring sites at 13 selected areas to separate local and regional contributions to the major contributors
of PM25, as shown for pNO ; in Figure 2-93.
• IMPROVE Site
¦ IMPROVE Urban Site
• O
Puerto Rico /
Virgin Islands
Alaska
Hawaii
Source: http:/A/ ista.cira.colostate.edu/views
Figure 2-92. IMPROVE and CSN (labeled STN) monitored mean NH4NO3 concentrations for 2000 through
2004.
2-147
-------
Fresno
Missoula
Salt Lake City
Tulsa
St. Louis
Birmingham
Indianapolis
Atlanta
Cl&zeland
Charlotte
Richmond
Baltimore
New York City
0 2 4 6 8 10 12
Annual Average Concentration
of Nitrates, ug/m3
Source: U.S. EPA (2004).
Figure 2-93. Regional and local contributions to annual average PM2.5 by PNO3 for select urban areas
based on paired IMPROVE and CSN monitoring sites.
Urban pN03 concentrations in western states are, in general, more than a factor of 2 greater than
the remote-area regional concentrations. For the Central Valley of California and Los Angeles areas, the
urban excess NH4NO3 exceed regional concentrations by amounts ranging from 2 (.ig/nr1 to 12 |ig/nr\ In
the region of the recently identified Midwest pN03 bulge (see Figures 2-91 and 2-92), the urban
concentrations were less than twice the background concentrations for an annual urban excess of about
1 |ig/m \ Northeast and southeast of the Midwest N03 bulge, annual urban pN03 were ~1 |ig/nr or less
above the remote-area regional concentrations, with warmer southern locations tending to have smaller
concentrations of both regional and urban excess pN03.
Holland et al. (1999) developed NOx emissions trends from 1989 to 1995 and compared them to
corresponding trends in total N03 (pN03 + gas-phase HN03) for the states in the U.S. between Louisiana
and Minnesota and east of that line based on data from 34 rural CASTNet dry deposition monitoring sites.
They reported a decrease in total N03 median values of -8% associated with a decrease of 5.4% in non-
biogenic NOx emissions. Because of the form of total N03 assumed in this analysis, it is not possible to
determine whether this change is larger for HN03 or pN03. For situations with limited gas-phase NH3 or
with elevated temperatures, it may be assumed that the trend in total N03 is principally in HN03 with no
net change in pN03. Where NH3 concentrations are substantially in excess of those required to neutralize
pS04 and where temperatures are lower, this trend may be assumed to be reversed.
Potential causes of the Midwest pN03 bulge can be examined through comparison of the pN03
areal extent to that of NOx and NH3 emissions. Figure 2-94 shows a map of the annual average pN03
concentrations (top) with a map of NH3 emissions (bottom). The spatial extent of NH3 emissions in the
Nitrates
WEST
EAST
~ Regional
Contribution
¦ Local
Contribution
2-148
-------
Midwest is strikingly similar to that of pNCh concentrations, each having regional maxima centered on
Iowa. NO and N02 emissions are high over a broad region of the country associated with the larger
population densities and greater numbers of fossil fueled EGUs to the east of the Midwest pN03 bulge.
While both NH3 and HN03 are needed to form NH4N03, the maps suggest the Midwest N03 bulge is due
primarily to the abundance of free NH3, defined as the amount beyond that required to neutralize the
acidic pS04. By contrast, the region east of the Midwest N03 bulge might be expected to have excess
HN03 given greater emissions of NO and N02, but apparently has a deficiency of free NH3. The few
eastern monitoring sites with locally high pN03 (near southern Pennsylvania) are located within a small
region of high density animal agricultural sites identified as a high NH3 emissions region in Figure 2-94.
Note that California's South Coast and Central Valley have both high NH3 and NO and NO; emissions,
explaining its own high pN03 concentrations.
|2004 Annual
I ammNOSf
Figure 2-94. Maps of spatiai patterns for average annual pNOs measurements (top), and for NH3 emissions
for April 2002 from the WRAP emissions inventory (bottom).
Importantly, although the Midwest pN03 bulge was not apparent before measurements were made
in its region, air quality modeling using national-scale emissions data predicted it in 1996 in CMAQ
applications for the U.S. EPA Western Regional Air Partnership (http://www.wrapair.org). Figure 2-95
shows the CMAQ-predicted average pN03 concentration for the month of January 1996 (left panel) and
the model-predicted sensitivities of pN03 to a 50% decrease in NH3 emissions (right panel). Decreases of
~3 |ig/m in the heart of the Midwest N03 bulge and in its areal extent were predicted as a result of the
NH3 emissions decrease.
2-149
-------
Figure 2-95. CMAQ simulation of January monthly averaged pN03 concentration using 1996 emissions
(left), and for a 50% decrease in NH3 emissions (right).
CMAQ Jan Avg Nitrate
8 Layers
k=CCTM_a2b1_36.combine.01 .avg
pi 12.000 90
* 10.500
9.000
7.500
6.000
4.500
3.000
1.500
0.000
ug/m3
CMAQ Jan Avg Nitrate- 50% NH3 Red.
8 Layers
o=CCTM_a2b9_36.combine.01 .avg
January 1,1996 1:00:00
0.000 at (5,90), Max= 10.577 at (115,40)
Source: U.S. EPA/WRAP.
4
\
/
2002 2002 2002
Nitrate 1.5 ua/rri3 Nitrate ~ A uo/mS Nitrate D .5 ua/m3
WRAP
Pacific Offshore
2002 2002 2002 CENRAP
Nitrate 1.6 uo/m3 Nitrate 0.6 ua/m3 Nitrate 1 2 uo/rn3
Eastern U.S.
Canade
Mexico
Outside Domain
Figure 2-96. Particulate NOs source attribution by region using CAMx modeling for six western remote area
monitoring sites Top left to right Olympic NP, WA; Yellowstone NP, WY; Badlands NP, SD;
bottom left to right San Gorgonio (W), CA; Grand Canyon NP, AZ; and Salt Creek (W), NM.
WRAP includes North Dakota, South Dakota, Wyoming, Colorado, New Mexico and all states
further west. CENRAP includes all states east of WRAP and west of the Mississippi River
including Minnesota. Eastern U.S. includes ail states east of CENRAP. The Pacific Offshore
extends 300km to the west of California, Oregon, and Washington. Outside Domain refers to
the modeling domain, which extend hundreds of kilometers into the Pacific and Atlantic
Oceans and from Hudson Bay Canada to just north of Mexico City.
Several example monitoring locations distributed across the northern and southern portions of the
eastern U.S. have been selected to illustrate the attribution results from air quality simulation modeling by
source region and source type for pN(X The modeling for this work was done with the Comprehensive
Air Model with extensions (CAMx); see http://camx.com for model code and descriptions. They include
Olympic National Park (NP), WA; Yellowstone NP, WY; and Badlands NP, SD across the north, and San
1
January 1,1996 1:00:00
Min= 0.000 at (5,90), Max= 14.025 at (13,44)
12.000 90
10.500
9.000
7.500
6.000
4.500
3.000
1.500
0.000 1
132 ug/m3
2-150
-------
Gorgonio Wilderness (W), CA; Grand Canyon NP, AZ and Salt Creek W, NM across the south. Pie
diagrams of pN03 attribution results by source region for each of these sites are shown in Figure 2-96.
Based on these sites, 25% or less of the pN03 in remote areas of the Pacific coastal states is from outside
of the U.S. (Pacific Offshore and Outside of the Domain). The Outside of the Domain values are derived
by simulating the fate of the boundary condition concentrations, which for the WRAP air quality
modeling were obtained using output from the GEOS-CHEM global air quality model (Fiore et al., 2003).
By comparison, pN03 at these western sites is much more from domestic regional emission
sources, with -60 to -80% being from emissions within the WRAP region. For the west coast sites, -25%
of pN03 is from a combination of Pacific Offshore emissions (i.e. marine shipping) and Outside Domain
regions. Canadian emissions are responsible for -10 to 30% of pN03, but Mexican emissions do not
contribute appreciably to pN03 for the three southern sites. Motor vehicles are the largest contributing
NO + N02 source category responsible for pN03 for these six WRAP sites, with a combination of point,
area, and wildfire source categories also contributing from -10 to 50% of the WRAP regional emissions.
2.9.6. Ambient Concentrations of Relevant Sulfur Compounds
2.9.6.1. SO2 and SO? Near Urban Areas
S02 data collected from the SLAMS and NAMS networks show that the decline in S02 emissions
from EGUs has improved air quality. There has not been a single monitored exceedance of the S02 annual
ambient air quality standard in the U.S. since 2000 (U.S. EPA, 2006a). EPA's trends data
(www, epa. siov/airt rends) reveal that the national composite average S02 annual mean ambient
concentration decreased by 48% from 1990 to 2005, with the largest single-year reduction coming in
1994-1995 (U.S. EPA, 2006a).
In 2007, there were -500 S02 monitors reporting values to the U.S. EPA Air Quality System
database (AQS). Trace level S02 monitoring is currently required at the approximately 75 proposed
NCore sites, as noted in CFR 40 Part 58 Appendices C and D. Continued operation of existing SLAMS
for S02 using FRM or FEM is required until discontinuation is approved by the U.S. EPA Regional
Administrator. Where SLAMS S02 monitoring is required, at least one of the sites must be a maximum
concentration site for that area.
Figures 2-97 through 2-102 illustrate geospatial locations of monitors for S02, N02, CO, PMi0,
PM2 5, and 03in 2005. These locations, sited in several cities in six states, were selected as relevant for
S02 environmental effects to complement measurements from rural and remote CASTNet sites and to be
near large sources of S02. For each state, map A of each figure shows locations of each monitor for all six
pollutants; map B of each figure shows only the S02 monitor locations. Totals for each monitor type are
included. These figures demonstrate the important point that not all S02 monitors in any Consolidated
Metropolitan Statistical Area (CMSA) are co-located with monitors for other pollutants. Two examples
are given below.
Table 2-20 lists the totals for all criteria air pollutant monitors (except Pb) in California, as well as
the subset of these monitors in San Diego County. At each of the four sites where S02 was measured,
N02, CO, PM10, PM2 5, and 03 were also measured, with the exception of PM2 5 at one site (AQS ID
060732007) in Otay Mesa, CA. Table 2-21 lists the totals for all criteria air pollutant monitors (except Pb)
in Ohio, as well as the subset of in Cuyahoga County.
In Cuyahoga County, OH, PMi0 and PM2 5 were measured at all four sites where S02 was also
measured in 2005, but 03 and CO were not measured at any of those four sites; N02 was only measured at
one site (AQS ID 39050060) near Cleveland's city center and -0.5 km from the intersection of Interstate
Highways 77 and 90.
2-151
-------
Table 2-20.
Number of monitors in California and San Diego County, 2005.
S02
NO2
03
CO
PM10
PM2.5
California (all)
35
105
176
86
177
97
San Diego County 4 9 10 6 7 7
Table 2-21. Number of monitors in Ohio and Cuyahoga County, 2005.
SO2
NO2
03
CO
PM10
PM2.5
Ohio (all)
31
4
49
15
49
49
Cuyahoga County 4 2 3 4 6 7
The regional distribution of S02 and S042 concentrations through the CONUS is shown in Table
2-22. As for the country on the whole, in and around most individual CMSAs, the trends are also toward
lower S02 levels. Table 2-23 shows that many annual and even 1-h mean concentrations for the years
2003 through 2005 were consistently at or below the operating LOD of ~3 ppb for the standard sensitivity
UV fluorescence S02 monitors deployed in the regulatory networks. The aggregate mean value over all 3
years and all monitoring sites in and around the CMSAs was just above the LOD at ~4 ppb, and was
identical to the 1-h and 24-h means.
The maximum 1-h concentration observed at some sites in and around some CMSAs exceeded the
mean by a large margin, with maximum 1-h values of >600 ppb. However, the 50th percentile maximum
value outside CMSAs, 5 ppb, was only slightly greater than the 1-h, 24-h, and annual mean value, 4 ppb.
The 50th percentile maximum value inside CMSAs, 7 ppb, was 75% greater than these longer-term
averages, reflecting heterogeneity in source strength and location. In addition, even with 1-h max values
of >600 ppb, the maximum annualized mean value for all CMSAs was still <16 ppb, which is below the
current annual primary S02 NAAQS.
The strong west-to-east increasing gradient in S02 emissions described above is well replicated in
the observed concentrations in individual CMSAs. For example, Table 2-24 shows the mean annual
concentrations from 2003-2005 for the 12 CMSAs with four or more S02 regulatory monitors. Values
ranged from a reported low of ~1 ppb in Riverside, CA, and San Francisco, CA, to a high of-12 ppb in
Pittsburgh, PA, and Steubenville, OH, in the highest S02 source region.
2-152
-------
Sacramento ^
San FrancisCqi
San Jos
San Francii;
San Diego
Highway
Monitor Location
Interstate
Federal
State
Source US EPA Office of Air and Radtafeor AOS Database
Figure 2-97. Criteria pollutant monitor locations (A) and SO2 monitor locations (B), California, 2005.
Shaded counties have at least one monitor. Map A shows locations of each monitor for all six
criteria pollutants. Map B shows only the SO2 monitor locations.
2-153
-------
Toledo
Cleveland
Toledo
Clevelan
incinnati
Highway
Monitor Location
Interstate
+
CO (15)
Federal
A
N02 (4)
State
V
03 (49)
O
PM10(49)
~
PM2S(49)
£
S02 (31)
Source: US EPA Office of Air and Radiation AOS Database
Figure 2-98. Criteria pollutant monitor locations (A) and S02 monitor locations (B), Ohio, 2005. Shaded
counties have at least one monitor. Map A shows locations of each monitor for all six criteria
pollutants. Map B shows only the SO2 monitor locations.
2-154
-------
Lima
Tucson
Highway
Monitor Location
Interstate
+
CO (20)
Federal
A
N02 (13)
State
V
03 (45)
O
PM10(67)
~
PM25(16)
¦fr
S02 (7)
Source: US EPA Office of Air and Radiation AOS Database
Figure 2-99. Criteria pollutant monitor locations (A) and S02 monitor locations (B), Arizona, 2005. Shaded
counties have at least one monitor. Map A shows locations of each monitor for all six criteria
pollutants. Map B shows only the SO? monitor locations.
Tucson
2-155
-------
Aiteijt&wn
' *
isburg
Alierit^vyri
ir/ ^ Pittsburgh
Highway
Monitor Location
Interstate
+
CO (25)
Federal
A
NOj (29)
State
V
03 (47)
O
PM10 (46)
~
PM2 5 (49)
~
S02 (42)
Sourc®: US EPA Office of Air and Radiation AQS Database
Figure 2-100. Criteria pollutant monitor locations (A) and SO2 monitor locations (B), Pennsylvania, 2005.
Shaded counties have at least one monitor. Map A shows locations of each monitor for all six
criteria pollutants. Map B shows only the SO2 monitor locations.
2-156
-------
Albany'
Albany
Highway
Monitor Location
Interstate
+
CO (11)
Federal
A
N02 (9)
State
V
03 (34)
O
PM10<10)
n
PM25(28)
~
S02 (25)
Source: US EPA Office of Air and Radiation AQS Database
Buffalo
Buffalo
Figure 2-101. Criteria pollutant monitor locations (A) and SO2 monitor locations (B), New York, 2005.
Shaded counties have at least one monitor. Map A shows locations of each monitor for all six
criteria pollutants. Map B shows only the SO2 monitor locations.
2-157
-------
afn bridge
pringfie!<
Boston
Highway
Monitor Location
Interstate
+
CO (5)
Federal
A
N02 (13)
State
V
03 (16)
O
PM10 (11)
~
PM25 (22)
~
SO2(10)
A
Source: US EPA Office of Air and Radiation AQS Database
New
Boston
Figure 2-102. Criteria pollutant monitor locations (A) and SO2 monitor locations (B), Massachusetts, 2005.
Shaded counties have at least one monitor. Map A shows locations of each monitor for all six
criteria pollutants. Map B shows only the SO2 monitor locations.
2-158
-------
Table 2-22.
Regional distribution of SO2 and SCV" ambient concentrations, averaged for 2003-2005.
Region
Concentration
SO2 (ppb)
S042" (Mg/m3)
Mid-Atlantic
3.3
4.5
Midwest
2.3
3.8
Northeast
1.2
2.5
Southeast 1.3 4.1
Table 2-23. Distributions of temporal averaging of S02 concentrations inside and outside CMSAs.
Averaging Time
Monitor Locations
Percentiles
Mean
10 25 30 50 70 75 90 95
99
Max
1-h MAX CONCENTRATION
Inside CMSAs
332405
13 1 1
1 3
4
7
13
16
30
45
92
714
Outside CMSAs
53417
13 1 1
1 1
2
5
10
13
31
51
116
636
1-h AVG CONCENTRATION
Inside CMSAs
7408145
4 1 1
1 1
1
2
4
5
10
15
34
714
Outside CMSAs
1197179
4 1 1
1 1
1
2
3
3
7
13
36
636
24-hAVG CONCENTRATION
Inside CMSAs
327918
4 1 1
1 1
2
3
5
6
10
13
23
148
Outside CMSAs
52871
4 1 1
1 1
1
2
3
4
8
12
25
123
ANNUAL AVG CONCENTRATION
Inside CMSAs
898
4 1 1
1 1
2
4
5
6
8
10
12
15
Outside CMSAs
143
4 1 1
1 1
2
3
4
5
8
9
13
14
AGGREGATE 3-YR AVG CONCENTRATION, 2003-2005
Inside CMSAs
283
4 1 1
1 2
3
3
5
5
8
10
12
14
Outside CMSAs
42
4 1 1
1 2
2
3
4
5
8
9
13
13
* Values are ppb. ** CMSA = Consolidate.d Metropolitan Statistical Area
The Pearson correlation coefficients (r) for multiple monitors in these CMSAs were generally very
low for all cities, especially at the lower end of the observed concentration ranges, and even negative at
the very lowest levels on the West Coast; see Table 2-24. This reflects strong heterogeneity in S02
2-159
-------
ambient concentrations even within any one CMSA. At higher concentrations, the r-values were also
higher. In some CMSAs, this heterogeneity may result from meteorological effects, whereby a generally
well-mixed subsiding air mass containing one or more S02 plumes with relatively high concentration
would be more uniformly spread than faster-moving plumes with lower concentrations. However,
instrument error may also play a role, because the highest r-values, i.e., those >0.7, correspond to the
highest S02 concentrations, i.e., >6 and >10 ppb. Since the lowest S02 concentrations are at or below the
operating LOD, and demonstrate the lowest /correlation across monitors that share at least some air mass
characteristics most of the year, the unbiased instrument error in this range may be confounding
interpretation of any possible correlation. This could be because the same actual ambient value would be
reported by different monitors (with different error profiles) in the CMSA as different values in this
lowest concentration range.
To better characterize the extent and spatiotemporal variance of S02 concentrations within each of
the CMSAs having four or more S02 monitors, the means, minima, and maxima were computed from
daily mean data across all available monitors for each month for the years 2003 through 2005. Because
many of these CMSAs with S02 monitors also reported S042 . it is possible to compute the degree of
correlation between S02, the emitted species, and S042 , the most prominent oxidized product from S02.
S042 values, however, while averaged over all available data at each site are generally available at their
monitoring sites on a schedule of only 1 in 3 days or 1 in 6 days. Furthermore, S02 and S042 monitors
are not all collocated throughout the CMSAs. For each of the five example CMSAs in Figures 2-103
through 2-107, monthly aggregated values are depicted from daily means of: (a) the monthly mean,
minimum, and maximum S02 concentrations; (b) the monthly mean, minimum and maximum S042
concentrations; and (c) a scatterplot of S02 versus S042 concentrations.
Table 2-24. Range of mean annual SO2 concentrations and Pearson correlation coefficients in urban areas
having at least four regulatory monitors, 2003-2005.
CMSA (# Monitors)
Mean SO2 Concentration (ppb)
Pearson Correlation Coefficient
Philadelphia, PA (10)
3.6-5.9
0.37-0.84
Washington, DC (5)
3.2-6.5
0.30-0.68
Jacksonville, FL (5)
1.7-3.4
-0.03-0.51
Tampa, FL (8)
2.0-4.6
-0.02-0.18
Pittsburgh, PA (10)
6.8-12
0.07-0.77
Steubenville, OH (13)
8.6-14
0.11 -0.88
Chicago, IL (9)
2.4-6.7
0.04-0.45
Salt Lake City, UT (5)
2.2-4.1
0.01 -0.25
Phoenix, AZ (4)
1.6-2.8
-0.01 -0.48
San Francisco, CA (7)
1.4-2.8
-0.03-0.60
Riverside, CA (4)
1.3-3.2
-0.06-0.15
Los Angeles, CA (5)
1.4-4.9
-0.16-0.31
2-160
-------
Moving across the CONUS from highest to lowest S02 concentrations, first consider Steubenville,
OH (Figure 2-103), where the area of highest S02 concentrations of all 12 CMSAs with more than four
monitors, all monthly mean S02 concentrations (a) were substantially <30 ppb, though max daily means
in some months were often >60 ppb, or even >90 ppb. S042 data (b) at Steubenville were insufficient to
make meaningful comparisons, though the 12 months of available S042 data suggest no correlation with
S02 (c).
Next, consider Philadelphia, PA (Figure 2-104). S02 in Philadelphia, PA (a) is present at roughly
one-half the monthly mean concentrations in Steubenville, OH, and demonstrates a strong seasonality
with S02 concentrations peaking in winter. By contrast, S042 concentrations in Philadelphia peak in the
three summer seasons, with pronounced wintertime minima. This seasonal anticorrelation still contains
considerable monthly scatter, however.
Los Angeles, CA (Figure 2-105) presents a special case, since its size and power requirements
place a larger number of S02 emitters near it than would otherwise be expected on the West Coast.
Concentrations of S02 demonstrate weak seasonality in these 3 years, with summertime means of ~3 to 4
ppb, and maxima generally higher than wintertime ones, though the highest means and maxima occur
during the winter of 2004-2005. S042 at Los Angeles shows stronger seasonality, most likely because the
longer summer days of sunny weather allow for additional oxidation of S02 to S042 than would be
available in winter. Weak seasonal effects in S02 likely explain the complete lack of correlation between
S02 and S042 here.
2-161
-------
(a >120
100
80
60
40
20
0
111
5§> ,5?" J? xS? S? J?" sJ" & »
# # ^ ^ # -# # # ^ # ^p # # ^ ^ c/f <£>
(b) 40
36
32
_ 28
*E 24
1 20
«* 1S
o
m 12
§ f '
•*> ^
(c) 24
20
— 16
J
a
3 12
O
« 8
4
0
>? »»*• .«v -jS-
5? .5? 5? J? & ,# P8"
& & ^
f ^ •#' #" ^ #' ,/ ^ # o/ ^
o
o
%
10
15
so,
-------
(a) 50
45
40
„ 35
a. 30
25
d5
CO
20
15
10 +
5 ¦¦
-Illlltllltllllllll-
//////////////////
(fa) 70
80
50
2 40
30
20
10
0
8
iliii
\l tffi"
f
f ± i
J? tS? .
-------
(a) 16
14 •
43
a
&
4 --
2
I
S? sS3 $?> 5? jjs j5> (J» 5J* yj* c?* y}" 5I3 5^ 5^ 5$ $ fp
# ¦#' #" ^ •/' #' ^ I?' ej# # #" ^ e#>' ^
(c> 20
18
£ «2
10
O
e» e
o ^
~
«- o 0 #
O
-------
Figure 2-106. Riverside, CA, 2003-2005. (a) Monthly mean, minimum, and maximum SO2 concentrations, (b)
Monthly mean, minimum, and maximum SO42" concentrations, (c) Monthly mean SO42"
concentrations as a function of SO2 concentrations.
2-165
-------
14
J2
&
0.
II
lii
p> J? 6^ 5? jS> s§> 5^ $j* # ,5*" S^5 ,& j£
# # ^ ^ ^ ^ # # #
gS$?
4"
<£> J? rv>
jy1 S5^ ,# <1* yf" ,«* ^ jf5
^ ^ ^ c$* ^ if* -y^ ^ ^ ^°A ^ ^ ^* ^ ^p'
o
m
*o
~
V
o 0
S02 (ppb)
Figure 2-107. Phoenix, 2003-2005. (a) Monthly mean, minimum, and maximum SO2 concentrations, (b)
Monthly mean, minimum, and maximum SO42" concentrations, (c) Monthly mean SO42"
concentrations as a function of SO2 concentrations.
The Riverside, CA CMSA (Figure 2-106) presents the strongest example among the 12 examined
for this study of correlation between S02 and S042 . though even here the R2 value is merely 0.3.
Seasonal peaks are obvious in summertime for S02 and S042 , both at roughly one-half the ambient
concentrations seen in Los Angeles. This is very likely due to Riverside's geographic location just
downwind of the regionally large electric generating utility sources near Los Angeles and the prevailing
westerly winds in summer. Again, as with Los Angeles, the summertime peaks in S042 are most likely
due to the combination of peaking S02 and favorable meteorological conditions allowing more complete
oxidation.
2-166
-------
Phoenix, AZ was the CMSA with the lowest monthly mean S02 and S042 concentrations
examined here (Figure 2-107). In Phoenix, nearly all monthly mean S02 values were at or below the
regulatory monitors' operating LOD of ~3 ppb. S042 concentrations were equivalently low, roughly one-
half the concentrations seen in Riverside, CA, for example. The monthly mean data show strong
summertime peaks for even these very low-level S042 observations, which, at ~1 to 3 (ig/m3, were
generally one-half of those in Philadelphia. This suggests some seasonality in S02, though anticorrelated
with S042 ; however, the trend is very weak, as the correlation scatterplot shows.
2.9.6.2. SO2 and SO4 in Rural and Remote Areas
The mean annual concentrations of S02 and S042 from CASTNet's long-term monitoring sites can
be compared using two 3-year periods, 1989-1991 and 2003-2005, shown in Figure 2-108 for S02 and
Figure 2-109 for SO/
From 1989 through 1991 the highest ambient mean concentrations of S02 and S042 were observed
in western Pennsylvania and along the Ohio River Valley: >20 (ig/m3 (~8 ppb) S02 and >15 |_ig/m3 S042 .
As with S02, in the years since the Acid Rain Program controls were enacted, both the magnitude of
S042 concentrations and their areal extent have significantly decreased, with the largest decreases again
along the Ohio River Valley.
IMPROVE monitors the major fine particle components including S042 , N03, crustal, elemental,
and organic carbon plus coarse mass concentration defined as PM10 minus PM2 5. An implicit assumption
is that most of the pS04 is present as (NH4)2S04. Much of the information contained below is based on
particulate elemental S used to infer the (NH4)2S04 levels. A discussion of IMPROVE S-to-pS04 history
and trends is available (Eldred, 2001). As with data from IMPROVE used above for pN03, and except
where noted, information in this section was derived form the IMPROVE IV Report (DeBell, 2006) with
data displays created with data and tools available at http://vista.cira.colostate.edii/views/. Much of these
data and conclusions are also to appear in Pitchford et al., 2008.
In contrast to results seen for pN03, additional monitoring sites in the U.S. Midwestern states
produced no surprises for (NH4)2S04 concentration or areal extent. Figure 2-111 shows highest
(NH4)2S04 concentrations in the mid-Atlantic and upper southern U.S. states where annual concentrations
ranged from ~3 (.ig/ni3 to ~6 |ig/nr\ The (NH4)2S04 concentrations in most western U.S. states was <~1
(ig/m3.
Combining IMPROVE and CSN data makes possible the comparison of urban pS04 to surrounding
remote-area regional values. These are shown as paired color contours maps for IMPROVE and
IMPROVE plus CSN in Figure 2-112. U.S. EPA (2004) used the pairing of IMPROVE and CSN
monitoring sites at 13 selected areas to separate local and regional contributions to the major contributors
of PM2 5, as shown for pS04 in Figure 2-113.
As shown in Figures 2-112 and 2-113, annual-averaged urban pS04 concentrations were in general
not significantly greater than the regional values, with urban excess pS04 generally <~0.5 (ig/m3.
Exceptions to the general case appear in Texas and Louisiana where urban excess pS04 concentrations
were > 1 |ig/nr\ Urban contributions were a larger fraction of the total pS04 in western U.S. states because
the regional levels were much lower there than in eastern ones. The small area decrease in pS04 evident
on these maps between the two eastern high concentrations regions may not be real, but cannot be verified
without speciation monitoring sites in southern Ohio, Kentucky, West Virginia, and Virginia. U.S. EPA
(2004) estimates of the contribution of local sources to pS04 were made for the same cities included for
pN03 shown above; see Figure 2-93. The east-west divide in both concentration and the regional
contribution to pS04 is strongly apparent here.
2-167
-------
S02
(ftg/m 3)
SO2
(jig/m3)
Source: U.S. EPA CASTNet
Figure 2-108. Annual mean ambient S02 concentration, 1989 through 1991 (top), and for 2003 through 2005
(bottom). White areas on the maps are areas where monitoring sites are absent and no
information is available.
2-168
-------
S042-
Wm3)
S042"
W"3)
_-0.0
Source: CAST Net
Figure 2-109. Annual mean ambient SO42" concentration, 1989 through 1991 (top), and 2003 through 2005
(bottom). White areas on the maps are areas where monitoring sites are absent and no
information is available.
2-169
-------
AC AD 1
2000 Annual
ammS04f
"acadi
2004 Annual
arnmS04f
Source: http:/A/ista.cira.colostate.edu/vi ews
Figure 2-110. IMPROVE network measured annual averaged pS04 concentration for 2000 (top) and for 2004
(bottom). Note difference in scale.
2-170
-------
• IMPROVE Site
Alaska
|jg/m3
Hawaii
Puerto Rico /
Virgin Islands
¦ IMPROVE Urban Site
Source: http:/A/ista.cira.colostate.edu/vi ews
Figure 2-111. IMPROVE mean (NH^SO,!2" concentrations for 2000 through 2004.
A STN Site
• IMPROVE Site
¦ IMPROVE Urban Site
Puerto Rico /
Virgin Islands
° Alaska
Hawaii
Source: http:/A/ista.cira.colostate.edu/vi ews
Figure 2-112. IMPROVE and CSN (labeled STN) monitored mean (NH^SO^- concentrations for 2000
through 2004.
2-171
-------
Sulfates
Fresno
Missoula
Salt Lake City
Tulsa
St Louis
Birmingham
Indiana pol is
Atlanta
Cleveland
Charlotte
Richmond
Baltimore
New York City
0 2 4 6 8 10 12
Annual Average Concentration
of Sulfates, ug/m3
Source: U.S. EPA (2004).
Figure 2-113. Regional and local contributions to annual average PM2.5 by pS04 for select urban areas
based on paired IMPROVE and CSN monitoring sites.
Source attribution of the pS04 contribution to haze at Big Bend NP, TX was a primary motivation
for the BRAVO Study. Schichtel, et al. (2005) showed that during the four-month field monitoring study
(July through October, 1999) S02 emissions sources in the U.S. and Mexico were responsible for -55%
and -38% of the pS04 respectively. Among U.S. source regions, Texas was responsible for -16%, eastern
U.S. -30%, and the western U.S. -9%. A large coal fired power plant, the Carbon facility in Mexico just
south of Eagle Pass, TX was responsible for about -19%, making it the largest single contributor.
Modeling for a three-day pollution episode in September 1996 in the California South Coast Air
Basin (SCAB) and for another episode in January 1996 in the San Joaquin Valley (SJV) by Ying and
Kleeman (2006) has shown that -80% of pS04 for both regions is derived from upwind sources, with
most of the remaining local contributions associated with diesel and high-S fuel combustion. Kleeman,
et al. (1999) used a combination of measurements and modeling to show that the upwind pS04 source
region for the SCAB was over the Pacific Ocean, and this was confirmed by measurements on Santa
Catalina Island. Moreover, these particles subsequently grew with accumulation of additional secondary
aerosol material, principally NH3NO3, as they traversed the SCAB. Most the HN03 that forms pN03 in
the SCAB is from diesel and gasoline combustion (-63%), while much of the NH3 is from agricultural
sources (-40%) and catalyst equipped gasoline combustion (-16%) and upwind sources (-18%). In the
SJV most of the HNO3 that forms pN03 is from upwind sources (-57%) with diesel and gasoline
combustion contributing most of the rest (30%), while much of the NH3 is from upwind sources (-39%)
and a combination of area, soil and fertilizer sources (-52%).
WEST
EAST
~ Regional
Contribution
~ Local
Contribution
2-172
-------
¦ • H—•—
• V—\
(
¦
* •
*
T- [ •—1
»
• •
•
*: * • . .
* *
• * "|
|(a)
Contribution of the Pat Hk Goarst 1c Carat*nlT»ttcn
-0,1
0.1-0.2
¦ CL2-G.4
¦ 04-0,$
¦Cut'OJ
¦ OS- 1.4
Source: Xu et al. (2006). Reprinted with permission.
Figure 2-114. Contributions of the Pacific Coast area to the (NH4)2S04 (pg/m3) at 84 remote-area monitoring
sites in western U.S. based on trajectory regression. Dots denote locations of the IMPROVE
aerosol monitoring sites.
Using a regression analysis to find the dependence of pS04 concentration measured over a 3 year
period (2000-2002) at 84 western IMPROVE monitoring sites on the modeled transport trajectories to the
sites for each sample period, Xu et al. (2006) were able to infer the source regions that supplied pS04 in
the western U.S. Among the source regions included in this analysis is the near-coastal Pacific Ocean (i.e.
a 300 km zone off the coast of California, Oregon, and Washington states). Up to 50% of the pS04
measured at southern California monitoring sites is associated with this source region. As shown in Figure
2-114 the zone of impact from this source region includes large regions of California, Arizona, and
Nevada, possibly owing to the high S-content fuel used in marine shipping and port emissions.
The S04? attribution results of the WRAP air quality modeling (available from http://wrapair.org)
are largely in line with these empirical results, finding that the Pacific Offshore source region contributed
somewhat smaller amounts than reported by Xu et al., (2006) with concentrations at the highest affected
site in California of -45% compared to 50% by the regression analyses and even greater differences for
more distant monitoring sites.
2-173
-------
2002
Sulfate 1.1 ua/m3
A
I tv
2002
Sulfate 0.6 ua/rn3
2002
Sulfate 0.8 ua/m3
\ T
2002
Sulfate 0.5 ua/m3
2002
Sulfate 1.2 ua/m3
(a)
WRAP
Pacific Offshore
CENRAP
Eastern U.S.
Canada
Mexico
Outside Domain
2002
Sulfate 1.4 ua/m3
Figure 2-115. pS04source attribution by region using CAMx modeling for six western remote area
monitoring sites. Top left to right Olympic NP, WA; Yellowstone NP, WY; Badlands NP, SD;
bottom left to right San Gorgonio (W), CA; Grand Canyon NP, AZ; and Salt Creek (W), NM.
WRAP includes North Dakota, South Dakota, Wyoming, Colorado, New Mexico and all states
further west. CENRAP includes ail states east of WRAP and west of the Mississippi River
including Minnesota. Eastern U.S includes all states east of CENRAP. The Pacific Offshore
extends 300 km to the west of California, Oregon, and Washington. Outside Domain refers to
the modeling domain, which extend hundreds of kilometers into the Pacific and Atlantic
Oceans and from Hudson Bay Canada to just north of Mexico City.
Several example monitoring locations distributed across the northern and southern portions of the
western U.S. have been selected to illustrate the attribution results from the WRAP air quality simulation
modeling by source region and source type for pS04. They include Olympic NP, WA; Yellowstone NP,
WY; and Badlands NP, SD across the north, and San Gorgonio Wilderness (W), CA; Grand Canyon NP,
AZ and Salt Creek W, NM across the south. Pie diagrams of the pS04 attribution results by source region
for each of these sites are shown in Figure 2-115. Based on these sites, >50% of the pS04 in remote areas
of the Pacific coastal states is from outside the U.S., Pacific Offshore and Outside of the Domain. The
Outside of the Domain values are derived by simulating the fate of the boundary condition concentrations,
which for the WRAP air quality modeling were obtained using output from the GEOS-CHEM global air
quality model (Fiore et al., 2003). The pS04 fraction from the region labeled Outside of Domain is
approximately uniform throughout the western U.S. with site-to-site variation in the fraction mostly
caused by the variations in the total S042 concentration. The more northerly sites have effects from
Canadian emissions, while the southern sites have effects from Mexican emissions. Half of the Salt
Creek, NM, pS04 is from the domestic source emissions further to the east (CENRAP and eastern U.S.),
which also contribute -20% to Badland pS04 concentrations. A breakout of the emission sources from
within the WRA P region by source type (not shown) has most of the emissions from point sources, with
the combination of motor vehicle, area and wildfire emissions contributing from a few percent at the
furthest eastern sites to roughly half at San Gorgonio.
2-174
-------
2.10. Deposition of Nitrogen and Sulfur Species Across the
Landscape
As established in Sections 2.2 and 2.3 above, total emissions of NOx and SOx have decreased
substantially in the last 35 years. Between 1970 and 2005, NOx emissions fell from 26.9 million tons per
year (Mt/yr) to 19 Mt/yr, and SO? emissions fell from 31.2 to 15 Mt/yr. These decreases m emissions led
to correlative decreases in N and S concentrations, described just above in Section 2.9, and in atmospheric
deposition ofN and S species across the landscape; see the trends summarized in Figure 2-116.
Importantly, however, these very recent decreases in deposition still leave current deposition amounts,
which are a factor of 2 greater than in pre-industrial times for NO;, and NH4 . and a factor of 5 greater for
S042 . according to modeling experiments by Luo et al. (2007).
90% of sites have annual sulfur deposition
below this line
Average
*
Median
10% of sites have annual
sulfur deposition below this line
'90 '91 '92 '93 '94 '95 '96 '97 '98
Year
'99 '00 '01 '02 '03 '04 '05
Coverage: 34 monitoring sites in the eastern United States.
Data source: MACTEC Engineering and Consulting, inc2006
90% of sites have annual nitrogen deposition
below this line
cn jS
1 %
cz —
Median
Average
10% of sites have annual nitrogen deposition
below this line
'90 '91 '92 '93 '94 '95 '96 '97 '98 '99 '00 '01 '02 '03 '04 '05
Year
Coverage: 34 monitoring sites in the eastern United States.
Data source: MACTEC Engineering and Consulting, Inc., 2006
Figure 2-116. Trends, 1990-2005 in S (left) and N (right) deposition for 34 sites in the eastern U.S.
Deposition maps were developed by CASTNet to show the composition of dry deposition for N
and S. The maps are labeled with inferred total deposition at each site, and a pie chart showing the
relative proportion of wet and dry deposition or the chemical components of the deposition. Wet
deposition is estimated from the interpolated concentration as measured by NADP multiplied by the
measured rainfall at the site. Dry deposition is inferred from the measured ambient air concentrations of
the chemical multiplied by the dry deposition rate obtained from an inferential model of linked resistances
to derive species and location-dependent Vd. (See the CASTNet QAPP for more information on methods
of computing their deposition totals). Note that NH3 is not included m these total N estimates because it is
not currently measured in these networks.
Data in this section are presented to show deposition across the landscape; finer-scale data and
maps of sensitive and vulnerable regions and ecosystems are presented in other sections. Data presented
in the maps and charts represent 3 year averages. For example, "89-91" is the average deposition of 1989,
1990, and 1991 for a given site. Only sites having valid total deposition for at least two of the three years
are shown and in some instances sites only met this criterion for one of the two reporting periods.
Because of differences like these, direct site-by-site comparisons are not possible everywhere.
2-175
-------
-pt
X) ¥ I I l J, m
14
a* e.o
Ml VO
12
3 t:
V3
¥
13 ^ " 24) V?
15
€.7 g g 5,9
81 bo®"®
80 7,9 as
3.0 '-• -.r f . i^_
*2 J 9 t» »•»
u\ v ,.» l v —¦: MS
• on 1.5 -r s. — -^HE.. ¦
»
a.i 9.1
2.4 / w6, I w *6 ^
* 2J > —-rX
*» »J! Jo 6 6 7 ' « lnOr0 N
'4J ) I j | f* y W {3S^Bf ' (kg/ha)
__ 15
J I /• M \/ t—=10
® —J t—I I r f V s
/ , -''¦ DOtyN
g \ Bw«n
\\ XvM. {V v*
SoiifCtr USBPA/CASTNETNADPJNJN \ I f f USBPA/CAMD 08/10/07
68 8.5
\ * JTfc, «.
V —J W'4 10 2
>».e w |r ipj
as 103 o, ll.D
" *8 •¦! ,,D
cS
M *¦*
£ M«Br80
Inorg. N
M 6-= (kg/ha)
15
b=n10
V 5
~ Dry N
BwetN
Source USEPA/CASTNET NADP/NTN USEPA/CAMD 10W04
Figure 2-117. Total average yearly wet and dry inorganic N deposition, excepting NHs, for 2004-2006 (top)
and 1989-1991 (bottom).
2-176
-------
V?
1.1
v
12
"16
»-4 8.0 V-
^ 7* 4.7 ^ CO
W \ W % «3 Ag,
T~ J I v_f" ' MT7 &
3 0 m ! J^V.9 8«®
e ' *:-v—i %wm
13 v Z0 \f
4,8
a.i
76
7.0 u ~ —^*s3s.
¦fi S « T.I +-6 Inorg. N
% I A V ^ ^ -*F » (kg/ha)
6,7 5,7
¦%- « _j —--t- * V #7
£—, I r- w —f\ ~DryHNOs-
„ £ ^Viw-MFT V ~ cryNH*-
" i ® Bl*y NOS-
ED Wet NH4T
IS H Wat NO 3'
Sonic*: USEPA/CASTNETNADPWTW \ 1, £" 59 f I USEPAJCAMD MOTTO
£
Cr ea
6 ft B.t»
i# JZ
tr-r-—/ I . «SL
/ ¥1 J _ \ / V£K :3w»%
/ J— _1:—( ttTkgJWr7
f.1^1.7 W*W
p/ |3 —f— - i v/ _ v ™ - ^
^ Inorg.N
j \ / (kg^a)
T-i I \ V _, 15
110
~ cry HNO3-
~ DryNHi*
¦ Dry NOs-
\J ~ Wet NH4-
¦ wetNOs-
Source: U SEFA/CASTNET NADF/NTN USEPA/CAMD IftWH
Figure 2-118. Total average yearly inorganic N deposition by species, excepting MH3, for 2004-2006 (top)
and 1989-1991 (bottom).
2.10.1. Nitrogen
For the years 2004-2006, mean N deposition was greatest in the Ohio River Valley, specifically in
the states of Indiana and Ohio, with values as high as 9.2 and 9.6 kg N/ha/yr, respectively; see Figure
2-117. Recent work by Elliott et al. (2007) using Sl5N to trace deposition totals and isolate them to point
to mobile source type shows that for 33 NADP/NTN sites in the East and Upper Midwest, spatial
distributions of 5lDN concentrations were strongly correlated with NQX emissions from point sources, and
2-177
-------
that wet NO; deposition at the 33 sites considered was strongly associated with NOx emissions from the
surrounding point sources. N deposition was lower in other parts of the East, including the Southeast and
in northern New England. In the central U.S., Kansas and Oklahoma reported the highest deposition, 7.0
and 6.5 kg N/ha/yr, respectively. N deposition was generally much lower in the western U.S., where it
was highest in urban areas in southern California and Denver, CO, 4.8 and 3.3 kg N/ha/yr, respectively. It
should be noted, however, that large portions of the U.S. west of the Mississippi River are poorly covered
by the current deposition monitoring networks as the location icons on these maps make clear. Hence, the
actual degree of heterogeneity and magnitude of real deposition in much of the West is largely unknown.
Because NOx emissions decreased by -25% between 1990 and 2005, recent N deposition is lower
compared with average deposition for the years 1989 to 1991. For 1989 to 1991, several recording
stations in the Ohio River basin reported average annual deposition rates in excess of 10 kg N/ha/yr. Data
are lacking, however, for much of the central and western U.S. and little can be said for changes between
the two reporting periods in these areas for the reasons given above. The greatest mass of N deposition
primarily occurred as wet N03 and NH4 . followed in importance by dry HNO,. dry NH4 . and dry NO; ;
see Figure 2-118. Although most deposition for both reporting periods occurred as wet deposition, there
were some exceptions, including parts of California where N deposition was primarily dry.
Figure 2-119 and Figure 2-120 show maps of wet deposition from NADP's data and IDW
interpolation technique. pN03 concentration and wet deposition amounts track dry deposition in locations
where the two monitoring networks overlap. Thus pN03 ambient concentration and deposition are highest
in the upper Ohio River Valley, excepting the large reported pNO, deposition, in excess of 6 kg N/ha/yr
where NADP has a substantial number of monitors but CASTNet does not; see the description of the
Midwest NO, bulge in Section 2.9.6.2.
-
Wet NO3
(kg/ha)
MJ
Source: NADP, US EPA, CAMD 7/19/07
Figure 2-119. N03~ concentration in NADP wet deposition samples, 2004-2006.
2-178
-------
t- j Wet NH4
(kg/ha)
Source: NADP, US EPA, CAMD 7/19/07
Figure 2-120. Average NO3" concentration in NADP wet deposition samples, 2004-2006.
2.10.1.1. Example of NO2 and HNO3 Deposition and Flux Data from Harvard Forest
Harvard Forest is a rural site in central Massachusetts, where ambient NOx, NO?, and other
pollutant concentrations and fluxes of total NOY have been measured since 1990 (Munger et al., 1996).
An intensive study in 2000 used a TDLAS to measure N02 and HN03. Absolute concentrations of
UNO; were measured; and the flux inferred was based on the dry deposition inferential method that uses
momentum flux measurements to compute a Vd and derives an inferred flux (Hicks et al., 1987; Wesely
and Hicks, 1977). Direct eddy covariance calculations for HN03 were not possible because the
atmospheric variations were attenuated by interaction with the inlet walls despite very short residence
time and use of fluorinated silane coatings to make the inlet walls more hydrophobic. NO response was
adequate to allow both concentration and eddy covariance flux determination. Simultaneously, NO and
NO eddy covariance fluxes were determined with two separate 03 CL detectors, one equipped with a H2
gold catalyst at the inlet to convert all oxidized N compounds to NO. Additionally, the measurements
include concentration gradients for NO, N02, and 03 over several annual cycles to examine their vertical
profiles in the forest canopy.
Overall, the results showed typical N02 concentrations of 1 ppb under clean-air conditions and
mean concentrations up to 3 ppb at night and 1 ppb during daytime for polluted conditions. Net positive
fluxes (emission) of N02 were evident in the daytime and negative fluxes (deposition) were observed at
night (Figure 2-121). NO fluxes were negative during the daytime and near zero at night.
2-179
-------
NW
SW
o
E
o
E
£
6
c
o
o
o
E
x
3
4 -
-2
[NO]
[NOJ
[01/10
F NO
. fno2 ¦
FO3/10
1 1
12
Hours
18
••
Hours
Source: Horii et al. (2004). Reprinted with permission.
Figure 2-121. Diel cycles of median concentrations (upper panels) and fluxes (lower panels) for the
northwest clean sector, left panels) and southwest (polluted sector, right panels) wind sectors
at Harvard Forest, April-November, 2000, for NO, NO2, and O3/IO. NO and O3 were sampled at
a height of 29 m, and NO2 at 22 m. Vertical bars indicate 25th and 27th quartiles for NO and
N02 measurements. N02 concentration and nighttime deposition are enhanced under
southwesterly conditions, as are O3 and the morning NO maximum.
2-180
-------
Simple Model
100
O)
ffl
NO
N02
0.0 0,2 0.4 0.6 0.8 1.0 -2
1
0
1
2
Concentration (nmol mol"1) Flux (nmol mol"1 cm s"1)
Source: Horii (2002).
Figure 2-122. Simple NOx photochemical canopy model outputs. Left panel, concentrations of NO (dashed)
and NO2 (solid); right, fluxes of NO (dashed) and NO2 (solid). Symbols indicate measurement
heights for NO (29 m) and N02 (22 m) at Harvard Forest. The model solves the continuity
equation for NO concentration at 200 levels, d/dz(-Kc[dNO/dz]) = PNO'LNO, where PNO =
[NO]/t1, LNO = [NO]/t2, and zero net deposition or emission of NOx is allowed. NOx (NO + NO2)
is normalized to 1 ppb. t1 = 70 s in this example. Due to the measurement height difference,
observed upward NO2 flux due to photochemical cycling alone should be substantially larger
than observed downward NO flux attributable to the same process.
In part, the opposite NO and N02 fluxes are simply consequences of variable NO-to-N02 ratio
distributions responding to vertical gradients in light intensity and 03 concentration, which resulted in no
net flux of NOx (Gao et al., 1993). In the Harvard Forest situation, the NO and N02 measurements were
not at the same height above the canopy, and the resulting differences derive at least in part from the
gradient in flux magnitude between the two inlets (Figure 2-122).
2-181
-------
FN02 (night) = F0 + V0 [N02] + a [N02]2
Hourly Data (fitted)
Nightly Medians
5
0
¦5
10
15
V0= -0.08 ±0.03 (cm s'1)
a = -0.013 ± 0.001 (nmol1 mol cm s1)
R2= 0 63
20
0
5
10
15
20
25
30
[N02] (nmol mol1)
Source: Horii et al. (2004). Reprinted with permission.
Figure 2-123. Hourly (dots) and median nightly (pluses) NO2 flux vs. concentration, with results of least
squares fit on the hourly data (curve).
At night, when NO concentrations are near zero due to titration by ambient 03 there is not a flux of
NO to offset N02 fluxes. Nighttime data consistently show N02 deposition (Figure 2-123), which
increases with increasing N02 concentrations. Concentrations above 10 ppb were rare at this site, but the
few high N02 observations suggest a nonlinear dependence on concentration. The data fit a model with
Vd of-0.08 plus an enhancement term that was second order in N02 concentration. The presence of a
second order term implies that N02 deposition rates to vegetation in polluted urban sites nearer to the
prominent sources of N02 could be considerably larger than what was observed at this rural site.
After accounting for the time of the NO-N02 null cycle during the measurement-sampling period,
the net NOx flux can be derived. Overall, there was a net deposition of NOx during the night and
essentially zero flux in the day, with large variability in the magnitude and sign of individual flux
observations. For the periods with N02 concentrations >2 ppb, deposition was always observed. These
canopy-scale field observations are consistent with a finite compensation point for N02 in the canopy that
offsets foliar uptake or even reverses it when concentrations are especially low. At concentrations above
the compensation point, NOx is absorbed by the canopy. Examination of concentration profiles
corroborates the flux measurements (Figure 2-124). During daytime for low-NOx conditions, there was a
local maximum in the concentration profile near the top of the canopy where 03 has a local minimum,
which is consistent with foliar emission or light-dependent production of NOx in the upper canopy.
Depletion was evident for both NOx and O , near the forest floor. Air reaching the ground has passed
2-182
-------
through the canopy where uptake is efficient and the vertical exchange rates near the ground are slow. At
night, the profiles generally decreased with decreasing height above the ground, showing only uptake. At
higher concentrations, the daytime NOx concentrations were nearly constant through the canopy; no
emission was evident from the sunlit leaves.
NOx PROFILES
Canopy
"fop **"$* 25
0.70 075 0.80 0.85 15 20 25 30 3,4 3.8 4.2 4.6 14 16 18 20 22 24
Canopy
Top "
0.60 0.65 0.70 0.75 0.80 28 30 32 34 36 4,2 4.4 4.6 4,8 5.0 22 2 3 24 25 26 27 28
Concentration (nmol mol'1) Concentration (nmol mol"1)
Source: Horii et al. (2004). Reprinted with permission.
Figure 2-124. Averaged profiles at Harvard Forest give some evidence of some NO2 input near the canopy
top from light-mediated ambient reactions, or emission from open stomates.
Figure 2-125 compares observed fluxes of all the observed species. The measured NOx and
estimated PAN fluxes were small relative to the observed total NOY flux. In clean air, HN03 accounted
for nearly all the NOY flux and the sum of all measured species ws about equal to the NOY concentration.
However, in polluted conditions, unmeasured species were up to 25% of the NOY, and HN03 fluxes
cannot account for all the total NOY flux observed. These unmeasured NOY species likely are
hydroxyalkyl nitrates and similar compounds rapidly deposited to surfaces but not routinely measured;
see the descriptions of measurement techniques and challenges in Section 2.3. The deposition ofHN03
and multifunctional RON02 were the largest elements of the measured N dry deposition budget.
2-183
-------
Two key areas of remaining uncertainty were the production of HN02 over vegetation and the role
of very reactive biogenic VOCs. HN02 is important because its photolysis is a source of OH radicals, and
its formation may represent an unrecognized mechanism to regenerate photochemically active NOx from
N03 that had been considered terminally removed from the atmosphere; see the discussion in Section 2.3
above on the atmospheric chemistry of NOx and the role of oxidized N compounds in atmospheric N
transport.
Summer 2000
NW
SW
o
c
o
O
12
10
8
6
4
2
0
1.2 --
"5 0.6 +
c
o
tj 0.4
re
0.0 --
O
E -15
-25
NC5
NO +HNO,+PAN
no*+hno5
NO*= NO+N02 "
NO*
I ' '
_i 1 i i i 1 i
I | I I 1_
—^—
— /
1_A^_
v-., //
\^\'l
V
j—[—|—1——1—|—!—I I—I—I—I II I—I—I I I—L
frfrll-
\
FPAN (est.) X
FNO (param.) y
FN02 (param.) A
FHNOa(DDIM) o
FNOy (e.c.) u
I I I I I I I I
uj fli oi
%
1 I 1 1 1 1 I 1 1 1 1 I 1 1 1 -I- I 1 1 1 1
n • rr
T"r-p-|-r--.-T-^--r-T
~-i 1 I 1 1 1 1 1 1 1 1 1 I 1 1 1 1 1 r i
<)> I
1 „ I
"U "'"-[I |i I
f .A
J—L
C)
i [1'
,[] [iin]
1 1 1 1 I 1 1 1 1 I 1 1 1
10
15
20 0
Hour
10
15
20
Source: Horii et al. (2006).
Figure 2-125. Summer (June-August) 2000 median concentrations (upper panels), fractions of NOy (middle
panels), and fluxes (lower panels) of NOY and component species separated by wind direction
(northwest on the left and southwest on the right). Vertical lines in the flux panels show 25th
and 75th quartiles of F(NOY) and F(HN03); negative fluxes represent deposition; F(NOx) is
derived from eddy covariance F(NO) and F(N02) measurements (corrected for photochemical
cycling), F(HN02) is inferred, and F(NOy) was measured by eddy covariance. The sum of NOx,
HNO3, and PAN accounts for all of the NOY concentration and flux for northwesterly
(unpolluted background) flows, whereas up to 50% of NOy and F (NOv) under southwesterly
flows are in the form of Nr species whose fluxes are not measured or estimated here.
2-184
-------
2.10.2. Sulfur
For the most recent 3-year reporting period available (2004 to 2006), mean S deposition was
greatest in the eastern U.S. east of the Mississippi River, with the highest deposition of 21.3 kg S/ha/yr in
the Ohio River Valley; see Figure 2-126. Most recording stations throughout the Ohio River Valley report
3-year total S deposition averages >10 kg S/ha/yr and many other stations in the East report deposition >5
kg S/ha/yr. Data are sparse for the central and west U.S. However, where available, data indicate lower
values than in most of the East, ranging from 4.1 to 5.3 kg S/ha/yr. Total S deposition in the U.S. west of
the 100th meridian is lower, with all recording stations reporting <2 kg S/ha/yr and many reporting <1.0
kg S/ha/yr. These values can also be compared to S deposition totals from 1989-1991 in Figure 2-126.
Station-by-station comparisons between averaging periods are difficult because some stations do not have
sufficient data to report a mean for the sampling period. There are, however, clear regional decreases in S
deposition across the country. S deposition for 1989-1991 (the earliest 3-year reporting period available)
is almost uniformly greater than for the most recent 3-year average (2004-2006). Deposition since 1989—
1991 has declined throughout the Ohio River basin from a previous high of 25.4 kg S/ha/yr, and in New
England, and the Mid-Atlantic regions, consistent with the -48% decrease in S02 emissions nationwide
between 1990 and 2005. Very little coverage for western and central U.S. was available for the 1989—
2001 reporting period, but sites with data show a similar decrease. Figure 2-127 shows that for both the
1989-1991 and 2004-2006 reporting periods, S was primarily deposited as wet S042 followed by a
smaller proportion as dry S02 and a much smaller proportion as dry S042 .
2-185
-------
0.7
0.7
11
12
0.8
1.2
1.0
Soutee: USEPA/CASTNET NA DP/NTN
06
0.7
¥
Soiacc ITSEPA/CASTNET N ADP.-NTN
3.4
4.7
6.S
7.5
V
* I
-
B.9
W
10-2
ia*W4o i,8
41 - 15.9
¦W 21*1SL11JLB
l£ 12.313.2
7"7 1fi-1 m? '
11.5 i0-2 10,0
&4 *W
, ®-3 a,i
6.3
a.o
e s2
few 7 6
A x
Tota'S
(kg/Via)
=
6,3
O
if
SJ
~ wets
riDfyS
USEPA/CAMD 08/30/07
12,5
12,9
16.fi
185
ifl.o
2QM j£ t 21.e
20 2 199
»• *ij«« 1S4">» ,5'°
14.5 25.4
16.2
15.9
19.8
13.1
|08 1
J-B
€ "
Total S
(kgrtia)
10
LI Wets
I j Dry S
USEPA/CAMD 10/06/04
Figure 2-126. Total average yearly wet arid dry S deposition for 2004-2006 (top) and 1989-1991 (bottom).
2-186
-------
iv Tt*
L
1.3
3.4
47
(
0.6
0.7
0.7
3>5
V
12
6.4
5.5
A5
4£
*0.2
03
A
1.0
AJ
0.8 °-e
1.2
V
*
Hi1" .
,,, ^ 18.3 «.«
08 21 3 .18.1
»8 J2.913J!
7.7 -16.1
Same usfiPA/rA.fmBTNADPiNlK
1.3
0 6
0.7
14
Source USEPA/CASTNET NADP/NTN
¥
9.4
11.5 W'2 (0.0
-•
„ 03 a.i
S.2
6.9
5.2 5£
M
i-9
4.7
7i5 7i Totals
(kg/ha)
20
V^510
a
n Dry S02
I Dry SO42'
5-0 ~ Wot SO12-
USEPA/CAMD 0SV30/07
fi.a $.3
0,2
JSS
1?9
-16.8
*382 .
a0.« 2I.3 -1M
-j 13.3 ~16 S 20.2J"
*• -*•>« ]61'7,;so
J4.e 25.4
r.
(0.6 ri-3
£
Totals
(kgma)
.2Q
P*10
~ DrySOs
¦ Dry SO'2-
D Wet SO42-
USEPA/CAMD 10/06^)4
Figure 2-127. Total average yearly S deposition by species for 2004-2006 (top) and 1989-1991 (bottom)
2-187
-------
2.11. Summary
2.11.1. Emissions and Atmospheric Concentrations
Total anthropogenic NO and N02 emissions in the U.S. in 2002 were 23.19 Tg. Combustion
chemistry at EGUs contributed -22% of this total, and transportation-related sources contributed -56%.
Ambient annual NOx concentrations have decreased -35% in the period 1990-2005 to current annual
average concentrations of -15 ppb.
Biogenic NOx sources are substantially smaller than anthropogenic ones and include biomass
burning, lightning, and soils. The NO and N20 emitted from soils as intermediate products from
denitrification can evolve either naturally or as stimulated by addition of N-containing fertilizers to crops
and other soil management practices. N20, another member of the oxides of N family of compounds, is
also a contributor to total U.S. GHG emissions: -6.5% on a Tg C02e basis in 2005, and its U.S. emissions
decreased -3% in the period 1990-2005, though there remains considerable interannual variation in this
value.
Concentrations of N02 in the CONUS from biogenic sources in the U.S., Canada, and Mexico and
from all sources elsewhere in the world are defined as policy-relevant background concentrations. On an
annual average basis, these concentrations are calculated to be <300 ppt over most of the CONUS and
<100 ppt in the eastern U.S. where NO emissions are greatest. The 24-h ambient N02 levels in CMSAs
where most of the regulatory monitors are located and where most anthropogenic emissions originate
were, on average, <20 ppb with a 99% percentile value <50 ppb for the years 2003-2005. Annual-average
N02 concentrations over the CONUS are calculated to be <5 ppb for nearly all urban and rural and remote
sites.
On a national scale, energy production at EGUs accounted for -66% of total S02 emissions in the
U.S. in 2001-2002; -5% of total S02 is emitted by transportation-related sources, with on-road vehicles
accounting for -40% of the transportation fraction and off-road diesel and marine traffic together
accounting for the remainder. As with NOx, emissions of SOx have been significantly reduced in recent
years: ambient annual SOx concentrations have decreased -50% in the period 1990-2005 and now stand
at -4 ppb for both aggregate annual and 24-h average concentrations nationwide.
Annual-average policy-relevant background S02 concentrations in the U.S. are <10 ppt over most
of the CONUS, or <1% of observed S02 concentrations everywhere except areas in the Pacific Northwest
where geogenic S02 sources are particularly strong.
NH3 emissions are chiefly from livestock and from soils as stimulated by addition of N-containing
fertilizers to crops and other soil management practices. Confined animal feeding operations and other
intensified agricultural production methods over a period of many decades have resulted in greatly
increased volumes of animal wastes high in N; 30 to 70% of these wastes may be emitted as NH3. These
increases in NH3 emissions, and the consequent increases in ambient NH3 concentrations and NH4+
concentration and deposition, are highly correlated geospatially with the local and regional increases in
agricultural intensity. However, estimates of total NH3 emissions on national and sub-national scales
range widely owing to three complex issues: the high spatial and temporal variability in NH3 emissions;
the high uncertainty in the magnitude of those emissions; and the lack of real-time, reliable, ambient NH3
monitoring techniques. Nonetheless, U.S. national NH3 emissions totals have been calculated, taking into
account these three drivers of uncertainty; for 2001-2002 the national NH3 emissions total from the NEI
and as corrected by methods described in Section 2.5 was 4.08 Tg/yr.
2-188
-------
2.11.2. Field Sampling and Analysis
The coverage of the networks is very thin over large expanses of the interior U.S. and especially so
west of the 100th meridian. This lack of monitored sites increases the likelihood that significant exposure
from N and S deposition is now occurring at current atmospheric concentrations where no measurements
are available, as predicted in numerical experiments with large-scale, first-principles models of
atmospheric chemistry and physics and deposition, and as measured at some few special experimental
sites.
The instrumentation deployed at present in the routine monitoring networks for determination of
gas-phase N02 and S02 concentrations is likely adequate for determining compliance with the current
NAAQS. But in application for determining environmental effects, all these methods have important
limitations, which make them inadequate for fully characterizing the state of the atmosphere at present,
correctly representing the complex heterogeneity of N and S deposition across the landscape, and for
realistically apportioning the contributions of reduced and oxidized forms of atmospheric N and S in
driving observed biological effects at a national scale.
For example, routine N02 measurements by CL are contaminated by unknown and varying
concentrations of higher-order oxidized N species, including gas-phase HN03, important in itself for N
deposition to the biosphere and also as a precursor to pN03. Moreover, dry deposition of NO, N02, and
PANs is not at present estimated in dry deposition networks, but could account for as much as 30% of
total dry oxidized N deposition in areas near strong NOx sources. This would include estuaries and other
wetlands near large urban areas.
As concerns S02, the present-day ambient annual average S02 concentrations are very near or even
below the operating LOD of most of the FRM monitors in the largest regulatory network. This produces
irresolvable uncertainty in these data, which may be important for environmental effects from S
compounds since they result in some cases from exposure at these current low concentrations.
Routine field sampling techniques for NH3 are at present limited to integrated values from several
days to one week because higher frequency semi-continuous methods are not yet sufficiently robust to
deploy for routine operation in national networks although passive NH3 samplers show excellent
potential. Estimates for the contribution of NH3 to the total N deposition budget range as high as 30% of
total N, and are perhaps the dominant source of reduced N. Moreover, routine national-scale sampling and
analysis for particulate-phase N03 . S042 . and NH/ are subject to positive and negative errors, chiefly
from the loss or production of constituent species on the surface of the filter used for the long time-
integrated measurement.
The aggregate effect of these uncertainties and errors very likely is to underestimate total N and S
atmospheric deposition and subsequent biological exposures.
2.11.3. Nitrogen and Sulfur Deposition
Increasing trends in urbanization, agricultural intensity, and industrial expansion during the
previous 100 years have produced a nearly 10-fold increase in N deposited from the atmosphere. NOx,
chiefly from fossil fuel combustion, often dominates total N pollution in the U.S. and comprises -50 to
75% of the total N atmospheric deposition.
For the period 2004-2006, the routine monitoring networks report that the mean N deposition in
the U.S. was greatest in the Ohio River Valley, specifically in Indiana and Ohio, with values as high as 9.2
and 9.6 kg N/ha/yr, respectively. N deposition was lower in other parts of the east, including the southeast
and northern New England. In the central U.S., Kansas and Oklahoma reported the highest deposition: 7.0
and 6.5 kg N/ha/yr, respectively.
2-189
-------
Table 2-25. Regional changes in wet and dry N and S atmospheric concentrations and deposition, 1989-
1991 and 2003-2005.
Measurement
Unit
Region
Average
1989-
1991
Average
2003-2005
Percent
Change*
Wet SO42" deposition
kg/ha
Mid-Atlantic
Midwest
Northeast
Southeast
27
23
23
18
20
16
14
15
-24
-32
-36
-19
Wet SO42" concentration
mg/L
Mid-Atlantic
Midwest
Northeast
Southeast
2.4
2.3
1.9
1.3
1.6
1.6
1.1
1.1
-33
-30
-40
-21
Ambient SO2 concentration
|jg/m3
Mid-Atlantic
Midwest
Northeast
Southeast
13
10
6.8
5.2
8.4
5.8
3.1
3.4
-34
-44
-54
-35
Ambient SO42" concentration
|jg/m3
Mid-Atlantic
Midwest
Northeast
Southeast
6.4
5.6
3.9
5.4
4.5
3.8
2.5
4.1
O CO CO "=3"
CO CO CO CSI
Wet inorganic N deposition
kg/ha
Mid-Atlantic
Midwest
Northeast
Southeast
5.9
6.0
5.3
4.3
5.5
5.5
4.1
4.4
-8
-8
-23
+2
Wet NO3 concentration
mg/L
Mid-Atlantic
Midwest
Northeast
Southeast
1.5
1.4
1.3
0.8
1.0
1.2
0.9
0.7
-29
-14
-33
-9
Ambient NO3" concentration
|jg/m3
Mid-Atlantic
Midwest
Northeast
Southeast
0.9
2.1
0.4
0.6
1.0
1.8
0.5
0.7
+5
-14
+20
+17
Total ambient NO3" concentration (NO3" + HNO3)
|jg/m3
Mid-Atlantic
Midwest
Northeast
Southeast
3.5
4.0
2.0
2.2
3.0
3.5
1.7
2.1
-14
-12
-13
-5
* Percent change is estimated from raw measurement data, not rounded; some of the measurement data used to calculate percentages may be at or below detection limits.
Source: CASTNet and the National Atmospheric Deposition Program / National Trends Network (NADP/NTN)
N deposition estimated from measurements primarily occurred in the form of wet N03 and NH/
followed with decreasing amounts of dry HN03, dry NH/, and dry N03 . Although deposition in most
areas of the U.S. occurred in wet form, there were some exceptions, including parts of California where N
deposition was primarily dry. Data are very sparse for the central U.S. between the 100th meridian and the
2-190
-------
Mississippi River; but where available, N deposition values there were lower than in most of the eastern
U.S., ranging from 4.1 to 5.3 kg N/ha/yr.
Estimates of total N loadings to estuaries, or to other large-scale elements in the landscape, are
computed using the measurements of wet and dry N deposition (as reported above) where these are
available, and then can be interpolated with or without a set of air quality model predictions to determine
the relative contribution from the atmosphere of various species of reduced and oxidized N. Measurement
and modeling experiments like these have shown that atmospheric inputs of Nr directly to the surface of
some coastal waters are essentially equal to or greater than those contained in riverine flow in the absence
of deposition and may contribute from 20 to >50% of external N loadings to these systems. For example,
atmospheric N inputs to the northeast Atlantic coast of the U.S., the southeast Atlantic coast of the U.S.,
and the eastern Gulf of Mexico have been estimated to be 11, 5.6, and 5.6 kg N/ha/yr, respectively. More
specifically and at finer spatial scales, atmospheric N loads to great waters and estuaries in the U.S. have
been estimated to range from 2 to 8% for Guadalupe Bay, TX, on the lowest end, up to -72% for the St.
Catherine-Sapelo estuary at the highest end (Castro et al., 2003). At Chesapeake Bay, atmospheric N is
estimated to contribute up to 30% of total N and 14% of the NH4 loadings to the Bay.
For the period 2004-2006, mean S deposition in the U.S. was greatest east of the Mississippi River
with the highest deposition amount, 21.3 kg S/ha/yr, in the Ohio River Valley where most recording
stations reported 3 year averages >10 kg S/ha/yr. Numerous other stations in the East reported S
deposition >5 kg S/ha/yr. Total S deposition in the U.S. west of the 100th meridian was relatively low,
with all recording stations reporting <2 kg S/ha/yr and many reporting <1 kg S/ha/yr.
S was primarily deposited in the form of wet S042 followed in decreasing order by a smaller
proportion of dry S02 and a much smaller proportion of deposition as dry S042 . However, these annual S
data in the western U.S., like those for N deposition, are derived from measurements in networks with
many fewer nodes in the West than in the East and so cannot represent all subregions in the West.
Table 2-25 lists separate concentration and deposition totals for wet and dry N and S species in 4
sub-regions of the U.S. as annual averages for the years 1989-1991 and 2003-2005 as a summary of the
foregoing data. Note that the U.S. West region is not present in this table owing to the dearth of annual-
scale measured concentration and deposition data at a sufficient number of sites to compare to those in the
eastern and Midwestern regions. Data on concentrations and deposition for individual years are available
at some sites in the West, as described for those specific cases in Sections 2.9 and 2.10 above. In
summary, measured N deposition in the West ranges from lows in the Pacific Northwest of between 1 and
2 kg N/ha/yr (N) (http://nadp.sws.iiiuc.edu) to highs in the Sierra Nevada and San Bernardino Mountains
in California of 20 to 40 kg N/ha/yr (total N) (Fenn et al., 2002 and 2003; Bytnerowicz et al., 2002), and
may be higher in particular locations for some sub-annual seasons.
2-191
-------
Chapter 3. Ecological and
Other Welfare Effects
This chapter is organized into four sections. The introduction (Section 3.1) frames the organizing
principle of this chapter and several basic concepts of ecology. A discussion of acidification is presented
in Section 3.2. N enrichment is discussed in Section 3.3. Lastly, other welfare effects are presented in
Section 3.4, including interactions between S deposition and Hg methylation and direct gas-phase injury
to vegetation.
3.1. Introduction to Ecological Concepts
3.1.1. Critical Loads as an Organizing Principle for Ecological Effects
of Atmospheric Deposition
This chapter uses the critical loads concept as an organizing principle. The components that are
necessary to develop a critical load provide a conceptual framework for linking atmospheric pollutants to
ecological endpoints that indicate impairment. The generally accepted definition of a critical load of
atmospheric pollutant deposition emerged from a pair of international workshops held in the late 1980s .
The workshop participants defined a critical load as:
"A quantitative estimate of an exposure to one or more pollutants below
which significant harmful effects on specified sensitive elements of the
environment do not occur according to present knowledge."
The development of a quantitative critical load estimate requires a number of steps. An illustrative
example of the eight general steps is shown in Table 3-1. A more detailed description of these steps is
given in Annex D.
This chapter presents information with a focus on the following questions:
¦ What is the disturbance?
¦ What receptors are affected?
¦ What indicator organisms are (or previously were) present and observable?
¦ What chemical indicators are changing and can be measured?
¦ What atmospheric pollutant is driving the changes in the chemical indicators?
It is important to recognize that there is no single "definitive" critical load for a natural resource.
Critical loads estimates reflect the current state-of-knowledge and policy priorities. Changes in scientific
understanding may include, for example, new dose-response relationships; better resource maps and
inventories; larger survey datasets; continuing time-series monitoring; and improved numerical models.
Changes in the policy elements may include: new mandates for resource protection; focus on new
pollutants; and inclusion of perceived new threats that may exacerbate the pollutant effects (e.g., climate
change).
3-1
-------
Table 3-1. An example of the matrix of information that must be considered in the definition and
calculation of critical loads (see discussion in text). Note that multiple alternative biological
indicators, critical biological responses, chemical indicators, and critical chemical limits
could be used.
1) Disturbance
Acidification
Eutrophication
2) Receptor
Forest
Lake
Grassland
Lake
3) Biological
indicator
Sugar
Maple
Norway
Spruce
Brook trout
Fish species
richness
Species
diversity
Primary
productivity
4) Critical
biological
response
Failure to
reproduce
Seedling
death
Presence
absence
Species
loss
Species
loss
Excess
productivity
5) Chemical
indicator
Soil % Base
Saturation
Soil Ca/AI
ratio
Lakewater
ANC
Lakewater
ANC
Soil C/N
ratio
Lakewater
no3
6) Critical
chemical
limit
10%
1.0
0 peq/L
50 (jeq/L
20
10 peq/L
7) Atmospheric
pollutant
S04, no3,
nh4
S04, no3,
nh4
so4, no3,
nh4
S04, no3,
nh4
no3, nh4
no3, nh4
8) Critical
pollutant load
???
???
???
???
???
???
This procedure can result in calculation of multiple critical loads for a given pollutant at a single
location. The multiple solutions derive from the nested sequence of disturbances, receptors, and biological
indicators that must be considered for a given pollutant. Multiple critical load values may also arise from
an inability to agree on a single definition of "significant harm." Calculation of critical loads for multiple
definitions of "harm" may be deemed useful in subsequent discussions of the analysis and in the decision-
making steps that may follow critical load calculation.
Finally, there is the inescapable heterogeneity of all natural environments. Consider soils, for
instance. The high spatial variability of soils almost guarantees that for any reasonably sized soil-based
"receptor" that might be defined in a critical load analysis, there will be a continuum of critical load
values for any indicator chosen. The range of this continuum of values may be narrow enough to be
ignored; nevertheless, there is an a prion expectation in any cntical load analysis that multiple values (or
a range of values) will result from the analysis. Given the heterogeneity of ecosystems affected by N and
S deposition, examples of published critical load values for a variety of endpoints and locations in the
U.S. are presented here (see Section 3.3.7).
3.1.2. Ecosystem Scale, Function, and Structure
Information presented in this ISA was collected at multiple scales, ranging from the physiology of
a given species to population, community, and ecosystem-level investigations. For this assessment,
"ecosystem" is defined as a functional entity consisting of interacting groups of living organisms and their
3-2
-------
abiotic (chemical and physical) environment. Ecosystems cover a hierarchy of spatial scales and can
comprise the entire globe, biomes at the continental scale, or small, well-circumscribed systems such as a
small pond.
Ecosystems have both structure and function. Structure may refer to a variety of measurements
including the species richness, abundance, community composition and biodiversity as well as landscape
attributes. Competition among and within species and their tolerance to environmental stresses are key
elements of survivorship. When environmental conditions shift, for example, by the presence of
anthropogenic air pollution, these competitive relationships may change and tolerance to stress may be
exceeded. Function refers to the suite of processes and interactions among the ecosystem components and
their environment that involve nutrient and energy flow as well as other attributes including water
dynamics and the flux of trace gases. Plant processes including photosynthesis, nutrient uptake,
respiration, and C allocation, are directly related to functions of energy flow and nutrient cycling. The
energy accumulated and stored by vegetation (via photosynthetic C capture) is available to other
organisms. Energy moves from one organism to another through food webs, until it is ultimately released
as heat. Nutrients and water can be recycled. Air pollution alters the function of ecosystems when
elemental cycles or the energy flow is altered. This alteration can also be manifested in changes in the
biotic composition of ecosystems.
3.1.3. Ecosystem Services
Ecosystem structure and function may be translated into ecosystem services. Ecosystem services
identify the varied and numerous ways that ecosystems are important to human welfare. Ecosystems
provide many goods and services that are of vital importance for the functioning of the biosphere and
provide the basis for the delivery of tangible benefits to human society. Hassan et al. (2005) define these
to include supporting, provisioning, regulating, and cultural services:
¦ Supporting services are necessary for the production of all other ecosystem services. Some
examples include biomass production, production of atmospheric 02, soil formation and
retention, nutrient cycling, water cycling, and provisioning of habitat. Biodiversity is a supporting
service that is increasingly recognized to sustain many of the goods and services that humans
enjoy from ecosystems. These provide a basis for three higher-level categories of services.
¦ Provisioning services, such as products (Gitay et al., 2001), i.e., food (including game, roots,
seeds, nuts and other fruit, spices, fodder), fiber (including wood, textiles), and medicinal and
cosmetic products (including aromatic plants, pigments).
¦ Regulating services that are of paramount importance for human society such as (a) C
sequestration, (b) climate and water regulation, (c) protection from natural hazards such as floods,
avalanches, or rock-fall, (d) water and air purification, and (e) disease and pest regulation.
¦ Cultural services that satisfy human spiritual and aesthetic appreciation of ecosystems and their
components.
3.2. Ecological Effects of Acidification
This section describes the major effects of acidification on terrestrial and aquatic ecosystems
resulting primarily from acidifying deposition. In this document, acidifying deposition includes gases and
particles derived from SOx, NOx, and NHX.
3-3
-------
3.2.1. Effects on Major Biogeochemical Processes
Acidifying deposition has altered major biogeochemical processes in the U.S. by increasing the S
(Figure 3-1) and N content of soils, accelerating S042 and NO ; leaching from soil to drainage water,
depleting base cations (especially Ca and Mg) from soils, and increasing the mobility of Al. The extent of
soil acidification is a critical factor that regulates virtually all acidification-related ecosystem effects from
S and N deposition. Soil acidification occurs in response to both natural factors and acidifying deposition.
To best integrate the effects of acidifying deposition, this assessment starts with a description of the
effects on soils and major biogeochemical processes within ecosystems, then summarizes the chemical
and biological effects on terrestrial, transitional, and aquatic ecosystems. More detailed information 011
acidification effects is provided in Annex B.
Little or no S Deposition
Major Fluxes
BC
BC
BC
Vegetation
Bedrock
Atmospheric
Deposition
Soil
Drainage
Water
Simplified Description
Base cation supply to soil from
deposition and weathering is
stable.
Base cation losses to soil water
and surface water are replaced
by the external supply from
weathering and BC deposition.
Base cation uptake into
vegetation is recycled.
Inorganic monomeric Al is not
mobilized to drainage water.
High S Deposition to Base-Poor Ecosystem
Low S Adsorption on Soil Hi9h s Adsorption on Soil
BC
BC
BC
BC
BC
BC
BC- H' Al. so;
BC" H' Al, SO/
Drainage
Water
Drainage
Water
Soil
Vegetation
Bedrock
Bedrock
Atmospheric
Deposition
Soil
Vegetation
Atmospheric
Deposition
Simplified Description
Increased S deposition is often
accompanied by increased base
cation deposition.
Sulfate leaches through soil to
drainage water if it is not
adsorbed to soil.
Sulfate flux is partially
neutralized by flux of BC and Al;
from soil to drainage water.
Over time, soils can become
depleted of BC, and drainage
water enriched in Al, and H\ both
of which can be toxic to plant
roots and aquatic biota.
Figure 3-1. Illustration of major fluxes of ions associated with S-driven acidification of drainage water.
The upper diagram represents ion fluxes in the absence of S deposition. The two lower
diagrams illustrate changes to these fluxes in response to S deposition on two types of soils,
without (left) and with (right) substantial S adsorption on soil. Effects are most pronounced
under high deposition with little adsorption, and include increased leaching of base cations
(contributes to soil acidification), H (reduced pH), and Ah (toxic to many plant roots and
aquatic species). Larger fluxes are represented by thicker arrows.
3-4
-------
3.2.1.1. Soil Acidification
Soil acidification is the loss of base cations plus the accumulation of acidic cations such as
hydrogen ions (H ) and aluminum ions (Aln+) in the soil; this happens when a proton donor is added to the
soil. Soil acidification is a natural process. Such natural acidity is contributed by carbonic acid, organic
acids, and plant cation uptake (Charles, 1991; Turner et al., 1991). However, the donor can also be a
mineral acid, such as HN03 and H2S04, the common components of acid rain that result from NOx and
SOx air pollution. Decreases in soil pH attributable to acidifying deposition have been documented in the
U.S. (Bailey et al., 2005; Johnson et al., 1994a, b; Sullivan et al., 2006a). Effects in the eastern U.S.
appear to have been limited mainly to the Northeast and portions of the Appalachian Mountains in both
hardwood and coniferous forests. Soil acidification has also likely occurred in localized areas of mixed
conifer forest and chaparral vegetation in, and near, the Los Angeles Basin, in response to locally high
levels of atmospheric dry N deposition (Fenn et al., 2003a).
To evaluate soil acidification, the soil must be considered in terms of the surface organic layer (the
primary rooting zone), of which the Oa horizon (or in some studies the O horizon, which combines the Oe
and Oa horizons) is an important component (See Figure 3-2). In addition, the mineral soil including the
A and/or B horizon, which lie below the Oa horizon and are primarily comprised of mineral matter, must
be considered.
Acidifying deposition can have a direct effect on soil pH. However, net uptake of nutrient cations
by vegetation can also generate acidity within the soil, and a considerable amount of natural organic
acidity is produced in the Oa horizon through the partial decomposition of organic matter. This process
can decrease the pH of soil water in the Oa horizon well below the lowest pH values measured in
acidifying deposition (Krug et al., 1985; Lawrence et al., 1995). Oa-horizon soils under coniferous
vegetation are strongly acidified by organic acids and are unlikely to have experienced a lowering of pH
as a result of acidifying deposition (Johnson and Fernandez, 1992; Lawrence et al., 1995). Soils
influenced by the growth of hardwood species tend to have surface horizons that are less acidic naturally
and are, therefore, more susceptible to decreased pH in the Oa horizon from acidifying deposition. By
taking up larger amounts of Ca from the soil, hardwoods can acidify lower soil horizons more than
conifers even though they enrich surface horizons with Ca via litterfall (Alban, 1982).
3-5
-------
Horizons 0
Source: http ://so il s. us da. go v/edu cati on/res ourc es/
Figure 3-2. Diagram illustrates soil horizons commonly found. Each main horizon is denoted by a capital
letter 0) Organic matter: Litter layer of plant residues in relatively undecomposed form. A)
Surface soil: Layer of mineral soil with most organic matter accumulation and soil life. This
layer is depleted of iron, clay, aluminum, organic compounds, and other soluble constituents.
B) Subsoil: This layer accumulates iron, clay, aluminum and organic compounds.
C) Substratum: Layer of unconsolidated soil parent material. This layer may accumulate the
more soluble compounds that bypass the "B" horizon.
Several studies document declines in soil pH within the Oa/A horizons and the upper B horizon in
sensitive regions of the U.S. over the past several decades (Bailey et ai.. 2005; Drohan and Sharpe, 1997;
Johnson et aL 1994a, b; Warby et al. (2009). These declines have been attributed at least partly to
acidifying deposition (Bailey et al., 2005).
In summary, soil acidification is a natural process, which is often exacerbated by acidifying
deposition. Natural acidification is particularly pronounced in coniferous forests. Acidifying deposition
can contribute to soil acidification, with consequent effects on the availability of nutrient cations in soil
and the chemistry of drainage water that flows from soil into streams and lakes. Despite recent decreases
in acidifying deposition and some improvement in surface water acid-base status, there are widespread
observations of ongoing soil acidification such as decreases in soil exchangeable base cations (Bailey
et al., 2005; Sullivan et al., 2006a, b; Warby et al., 2009) (See Section 3.2.1.4). More detailed descriptions
of soil acidification are provided by Binkley and Richter (1987) and van Breemen et al. (1983).
3-6
-------
3.2.1.2. Sulfur Accumulation and SO42- Leaching
Most acidification-related consequences of atmospheric S and N deposition in the U.S. are caused
by S042 (Driscoll et al., 2001b; Sullivan, 2000). The mobility within the watershed of S042 . derived
from atmospheric S deposition, is a key factor governing many aspects of soil and water acidification at
locations in the U.S. that are affected by acidifying deposition.
Upon deposition to the Earth's surface, S may be assimilated by vegetation or microbes,
accumulate in the soil or act as a mobile ion and leach out of the soil. When a given area is affected by
acidifying deposition, S deposition levels are typically much higher than plant demand for S and
consequently almost all deposited S is transported to the soil where it may accumulate and is available for
leaching as S042 . S042 acts as a mobile anion at many locations in the U.S. that receive high levels of S
deposition, notably the glaciated Northeast and Upper Midwest, where much of the deposited S leaches
through soils into streams and lakes. S042 leaching leads to most of the ecological effects from
atmospheric S deposition because it is accompanied by leaching of cations, and this contributes to
acidification of soil, soil water, and surface water.
Over time, sustained S042 leaching and associated soil acidification contributes to pronounced
changes in soils in some areas. When S is transported from soils to surface waters in the form of S042 . an
equivalent amount of cations, or countercharge, is also transported. When the countercharge is provided
by base cations, the base saturation of the soil is reduced as the acidity of the soil water is neutralized.
However, this process acidifies the soil, thereby decreasing the soil's capacity to neutralize additional
acidity deposited from the atmosphere and prevent acidification of soil water, and by connection, surface
water. As the base cations become depleted, the countercharge provided by acidic cations (H and
inorganic Al) increases, sometimes resulting in toxic conditions for plant roots and aquatic organisms
(Charles, 1991; Turner et al., 1991).
In the U.S., there are some regional trends of soil accumulation, retention, and leaching of S that
are discussed below.
Southeast
Accumulation of atmospherically deposited S in soil has resulted from anion adsorption and
incorporation of S into organic matter through biological assimilation. Such retention of S can
temporarily reduce S042 leaching and cause a delay in ecosystem recovery in response to decreases in S
deposition, as some accumulated S is slowly released from the soil into drainage water. S adsorption on
soil is especially pronounced in the southeastern U.S. Under continued loading of S deposition, it is
expected that many southeastern watersheds will exhibit a gradual decrease in the extent of S adsorption
in the future. This will likely contribute to further acidification of some streams, even under substantially
reduced levels of S deposition (Elwood et al., 1991; Sullivan et al., 2004; Turner et al., 1991).
Northeast
In the Northeast, the accumulation of a portion of the historic legacy of atmospheric S deposition in
soil was demonstrated by a positive relationship between wet deposition of S042 and concentrations of
total S in the forest floor of 12 red spruce stands (Driscoll et al., 2001a). However, net loss of S from soils
now appears to be occurring in a number of northeastern watersheds in response to decreased levels of
atmospheric S deposition. The potential for net mineralization of stored S might affect recovery of
drainage waters (Gbondo-Tugbawa et al., 2002; Likens et al., 2002; Novak et al., 2007). Where leaching
of previously stored S occurs, it delays soil and surface water chemical recovery from acidification
(Driscoll et al., 2001b).
Weathering contributes substantial S in some watersheds (Shanley et al., 2005). Uncertainties in
estimates of ecosystem S fluxes, such as weathering and dry deposition, and the difficulty in discerning
the effects of net S042 desorption and net S mineralization make it difficult to predict when S outputs in
3-7
-------
the northeastern U.S. will no longer exceed inputs. Recent research results, based on experimental
reduction of S inputs, suggest that this process will occur on a decadal timescale (Martinson et al., 2005;
Morth et al., 2005). The long-term role of C-bonded S adds further uncertainty because enhancement of S
mineralization by a warming climate could also affect S retention and release from soil (Driscoll et al.,
2001a; Knights et al., 2000),
In summary, atmospheric S deposition alters soil chemistry through the following
mechanisms: sustained S042 leaching and associated changes in soil chemistry, and accumulation of S in
the soil through physical/chemical adsorption and biological assimilation. The recent evidence of net loss
of S from soils at a number of sites in the Northeast is a likely response to recent decreases in atmospheric
S inputs (Driscoll et al., 2001b). The gradual loss of previously accumulated S is further contributing to
continued S042 leaching and soil acidification.
3.2.1.3. Nitrogen Accumulation and NO3- Leaching
The scope of this section is the role of N deposition in the process of acidification. This assessment
divides the effects of N deposition into the two broad categories of acidification and N nutrient
enrichment effects. The latter is discussed in Section 3.3. N deposition may cause acidification of
ecosystems via three main mechanisms: excess accumulation in soils followed by increased rates of
nitrification by microbes; change in base cation status of soils caused by N03 leaching; and increased
growth of vegetation causing increased cation uptake.
Nitrification and Accumulation
Atmospherically deposited N accumulates in soil through incorporation of N into organic matter.
Accumulation is either documented or suggested to occur across large areas of the U.S. (Aber et al.,
2003). Direct evidence for such accumulation has been found in the northeastern U.S. and in Colorado.
Increased accumulation of N in soil is suggested, for example, by a positive correlation between
atmospheric deposition levels and total N concentration in the Oa soil horizon at red spruce sites in New
York, Vermont, New Hampshire, and Maine (Driscoll et al., 2001b). Mass balance studies also show soil
N retention (e.g., Campbell et al., 2004). Further evidence that atmospheric deposition has increased the
availability of N in soil is provided by the strong negative correlation between atmospheric N deposition
and the C:N ratio of the Oa soil horizon across the northeastern U.S. (Aber et al., 2003).
The nitrification process is mediated by autotrophic bacteria that derive energy by oxidizing NH/
to N03 . Nitrification produces acidity in the form of HN03 as a byproduct. The HN03 produced
contributes to the acidification of soils and surface waters. If the C:N ratio of soils falls below about 20 to
25, nitrification is stimulated and net nitrification and associated production of acidity occurs in soils
(Aber et al., 2003; Emmett et al., 1998). This process often results in elevated N03 concentration in soil
waters and surface waters (Aber et al., 2003; Ross et al., 2004). Thus, data collected from streams and
lakes can yield important information about processes that occur in the soil. N saturation refers to the
condition when N inputs from atmospheric deposition and other sources exceed the biological
requirements of the ecosystem. Excess N supply reduces competition between plants and heterotrophic
microbes for NH/ to the point that net nitrification occurs (Aber et al., 1998, 2003).
Leaching
In many upland forested areas in the U.S., a large fraction of the N received in atmospheric
deposition is retained in soil or in plant biomass. Nevertheless, elevated N03 concentration in surface
waters during the growing season is common and widespread in the U.S. (Charles, 1991). High
concentrations of N03 in lakes and streams, indicative of ecosystem N saturation in most natural
systems, have been found at a variety of locations throughout the U.S. (Stoddard, 1994; U.S. EPA, 2004).
In general, atmospheric deposition of 8 to 10 kg N/ha/yr or more, results in N03 leaching to surface
3-8
-------
waters in the eastern U.S. Lower N deposition levels (<5 kg N/ha/yr) may lead to N03 leaching in the
mountainous West because of colder temperatures, shorter growing season, little soil development,
extensive exposed bedrock, and rapid melting of large snowpacks (Baron et al., 1994; Williams et al.,
1996a).
N03 leaching usually contributes to the leaching of base cations from soils to surface waters.
Although concentrations of N03 are typically less than S042 in drainage waters in most ecosystems in
the U.S., concentrations of N03 in some streams are high enough to suggest a substantial role for N03 in
base cation loss from soil, particularly during periods of high soil-water N03 flux during the non-
growing season (Cook et al., 1994; Van Miegroet et al., 1992).
The relationship between atmospheric N deposition and N03 leaching from forest ecosystems is
often modified by land-use history, current land-use, land disturbance, tropospheric 03 levels, and
climate. The N retention capacity of soils is highly dependent on land-use history and its effects on N
cycling and pool sizes. For example, the removal of trees reduces the amount of N in the watershed and
enhances the demand of vegetative regrowth for added N. This effect results in little or no N03 leaching
and can last for decades to more than a century (Goodale et al., 2000). N03 leaching is also affected by
current land use (U.S. EPA, 2006e). In the northeastern U.S., concentrations of N in streams of upland
forested watersheds tend to be considerably lower than in streams draining watersheds with other land
uses (Aber and Driscoll, 1997; Aber et al., 2003). Perhaps the most noteworthy effect of urban land use on
processes of nutrient enrichment from N deposition concerns the transport of N03 to N-limited estuarine
and near-coastal waters. This topic is discussed in Section 3.3.2.4. In agricultural, and especially in
forested areas, it is generally expected that most atmospherically deposited N is taken up by terrestrial
vegetation. This is usually not the case in urban landscapes, although it is sometimes possible. Due to the
relatively large impervious surface area in the urban landscape (e.g., buildings, roads, parking lots), a
higher percentage of precipitation is routed directly to surface waters, with less opportunity for vegetative
uptake of deposited N.
Climatic factors also play an important role in determining the extent of N03 leaching. In
particular, temperature and moisture have large effects on N cycling and N03 leaching. Murdoch et al.
(1998) found that, for at least one site, annual mean N03 concentrations in stream water were not related
to annual wet N deposition, but rather, were positively correlated with mean annual air temperature. This
pattern was likely due partly to the fact that microbial processes responsible for N03 production are very
sensitive to temperature. Fluctuations in microbial immobilization and mineralization in response to
climatic variability affect N03 losses to drainage waters.
Long-term data sets also suggest that climate may affect patterns of N03 loss. Many of the original
(sampled periodically since the early 1980s) long-term monitoring lakes in the Adirondack Mountains
showed increased N03 leaching from terrestrial ecosystems throughout the 1980s, which was followed
by a decline during the 1990s (Driscoll et al., 2003b, d). Decreasing stream N03 concentrations during
the 1990s were also observed in the Catskill Mountains and in New Hampshire (Driscoll et al., 2003b).
There was not a substantial change in N emissions or deposition in the Northeast region over that period.
Climatic factors, increases in atmospheric C02, and interactions with increasing availability of DOC have
been proposed as possible contributing factors for regional decreases in N03 in drainage water during the
1990s, but the driver of this decadal scale pattern remains under investigation. Snowmelt and rain-on-
snow, along with soil freezing, can influence N cycling in cold climates (Campbell et al., 2005; Eimers
et al., 2007; Park et al., 2003).
3.2.1.4. Base-Cation Leaching
Acidifying deposition has been shown to be an important factor causing decreases in
concentrations of exchangeable base cations in soil. Loss of base cations from soil is a natural process.
Under conditions of low atmospheric deposition of S and N the limited mobility of anions associated with
naturally derived acidity (organic acids and carbonic acid) controls the rate of base cation leaching. Inputs
3-9
-------
of S and N in acidifying deposition enhance inputs of strong acid anions that can accelerate natural rates
of base-cation leaching (Cronan et al., 1978; Lawrence et al., 1999).
Leaching of base cations from watershed soils to surface waters is a mechanism that depletes
essential plant nutrients from soil and limits the extent of surface water acidification in response to
acidifying deposition. When S042 and N03 leaching occur in equal magnitude to base cation leaching,
the drainage water is not acidified. However, in the process of neutralizing the acidity of drainage water,
base cation release from soil causes depletion of the base saturation of the soil. Soil base saturation
expresses the concentration of exchangeable bases (Ca, Mg, potassium [K], sodium [Na]) as a percent of
the total cation exchange capacity (which includes exchangeable H+ and inorganic Al). Under conditions
of low soil base saturation (<20%) and elevated concentrations of strong acid anions, Al is mobilized
from soil to drainage water (Cronan and Schofield, 1990), with potentially harmful consequences for
sensitive terrestrial plants and aquatic organisms throughout the food web.
In the 1990s, data were published supporting the occurrence of base cation depletion from soils in
the U.S. (Lawrence and Huntington, 1999; Lawrence et al., 1995; 1997), although decreases in
exchangeable Ca concentrations had earlier been identified in European soils through repeated sampling.
Recent data revealed that decreases in concentrations of exchangeable base cations and base saturation in
the Oa and B soil horizons have occurred over the past several decades in the eastern U.S. and most
studies attribute this change to the effects of acidifying deposition. For example, a soil re-sampling study
in the U.S. was conducted in northwestern Pennsylvania by Bailey et al. (2005). This study showed that
between 1967 and 1997 pronounced decreases, attributed largely to acidifying deposition, were measured
in exchangeable Ca and Mg concentrations in Oa/A horizons and throughout the B horizon (See also
comments by Johnson [2005]). Data compiled by Sullivan et al. (2006a, b) suggested decreases in base
saturation of B-horizon soils in the Adirondack Mountains between the mid-1980s and 2003. In another
re-sampling study in the northeastern U.S. Warby et al. (2009) observed large decreases in exchangeable
Ca and base saturation in the Oa horizons between 1984 and 2001, attributed to acidifying deposition. The
greatest decreases in exchangeable Ca and base saturation were observed in the regions of central New
England and Maine, where depletion of base cations was less apparent in 1984. Depletion of base cations
contributes to soil acidification and influences the ability of watershed soils to support acid-sensitive
vegetation and to neutralize acidity in future acidifying deposition. Both plant uptake of cations, for
example via forest regrowth subsequent to logging or land use conversion, and acidifying deposition can
acidify soils (Johnson and Todd, 1990; Richter and Markewitz, 2001; Trettin et al., 1999).
Upslope decreases in exchangeable soil base cation concentrations were found to be positively
correlated with higher S deposition in the Catskill Mountains (Lawrence et al., 1999). Furthermore,
declines in soil exchangeable pools of base cations have been documented in New Hampshire (Likens
et al., 1996) and Norway (Kirchner and Lydersen, 1995). In summary, leaching of base cations associated
with acidifying deposition is occurring in sensitive regions in the U.S. Base cation loss increases the
sensitivity of the watershed to further acidifying deposition. Watersheds that were capable of fully
neutralizing a particular level of acidifying deposition in the past may no longer be capable of fully
neutralizing that level today or at some time in the future because of the cumulative effect of acidifying
deposition on soil base saturation. Where the availability of exchangeable base cations is limited, the
leaching of potentially toxic inorganic Al into soil and surface waters can result.
3.2.1.5. Aluminum Leaching
If soil base saturation is 20 to 25%, or lower, acidifying deposition can mobilize inorganic Al,
which can lead to the leaching of this potentially toxic form of Al into soil waters and surface waters
(Cronan and Schofield, 1990; Reuss and Johnson, 1985). This is an extremely important effect of
acidifying deposition because some forms of inorganic monomeric Al, including Al3+ and various
hydroxide species, are toxic to tree roots, fish, algae, and aquatic invertebrates (see Section 3.2.3). In fact,
fish mortality in response to surface water acidification is usually attributable to Al toxicity. Increased
3-10
-------
concentrations of exchangeable inorganic A1 in the mineral soil have been identified through repeated
sampling in the U.S. and Europe over periods ranging from 17 years to 41 years in studies by Billet et al.
(1990), Falkengren-Grerup and Eriksson (1990), Bailey et al. (2005), and Lawrence et al. (1995). In areas
of Europe with excessively high acidic deposition levels, evidence of Al depletion in the mineral soil has
also been found (Lapenis et al., 2004; Mulder et al., 1989), but Al depletion has not been documented in
the U.S.
Acidifying deposition is an important cause of increased mobilization of inorganic Al from soils to
streams and lakes (Turner et al., 1991). Acidifying deposition introduces mineral acidity associated with
anions that are more mobile than those from organic matter. If the release of base cations from the soil is
insufficient to neutralize the inputs of sulfuric and nitric acid, then Al that had previously been deposited
by normal soil development in the upper mineral soil is mobilized. Al may also be mobilized by organic
acids. However, acidifying deposition mobilizes Al in inorganic forms, and in doing so increases the
amount of exchangeable inorganic Al within the B horizon and results in transport of inorganic Al into
soil waters and surface waters (Driscoll and Bisogni, 1984; Driscoll et al., 1985). Inorganic Al is
minimally soluble at pH about 6.0, but solubility increases steeply at pH values below about 5.5. This
distinction between organic and inorganic forms of Al is important because organic Al is not toxic,
whereas inorganic Al is toxic to a variety of plants and aquatic organisms (Baker and Schofield, 1982;
Baldigo and Murdoch, 1997; Joslin and Wolfe, 1988, 1989)(Section 3.2.3.1). Discussions in this
document of inorganic Al in solution refer to dissolved, rather than particulate or colloidal, forms. These
dissolved inorganic Al species are often collectively called inorganic monomeric Al (Driscoll, 1984).
Recovery of soil chemistry will require a decrease in exchangeable Al concentrations and Al
leaching. Once acidified, it is unlikely that soil exchangeable Al concentrations will decline again unless
soils are limed. Therefore, it is unclear what length of time would be required to decrease soil
exchangeable Al concentrations to levels characteristic of unpolluted systems. Furthermore, in most cases
it is unclear whether exchangeable Al concentrations are continuing to increase, remaining stable, or
decreasing. Predictions of trends in exchangeable Al concentrations remain uncertain because of our
incomplete understanding of mechanisms through which mineral matter and organic matter interact to
control dissolved Al concentrations. Possible changes in the dynamics of soil organic matter that could be
expected from climate change add further uncertainty to predictions of future change in exchangeable Al
concentrations in soils.
In summary, the natural downward movement and deposition of Al within the upper soil profile is
altered by acidifying deposition if the release of base cations is insufficient to buffer atmospheric inputs
of acidity. Rather than be deposited as an alumino-organic complex, Al mobilized by acidifying
deposition tends to remain in solution in inorganic forms that can be transported out of the soil and into
surface waters. Depletion of exchangeable base cations generally precedes the mobilization of inorganic
Al; therefore, as base cation concentrations in drainage water decrease, inorganic Al concentrations may
increase. Increases in concentrations of inorganic Al have been documented at several locations in base-
cation depleted soils in the U.S. and Europe. In soils with base saturation values less than about 15 to
20%, the ratio of exchangeable Ca to Al is typically low in upper mineral soils (Lawrence et al., 2005).
3.2.1.6. Episodic Acidification
The status of surface water chemistry can be examined and reported as chronic condition or
episodic condition. Chronic condition refers to annual average conditions, which are often represented as
summer and fall chemistry for lakes and as spring baseflow chemistry for streams. Episodic condition
refers to conditions during rainstorms or snowmelt when proportionately more drainage water is routed
through upper soil horizons, which tend to provide less neutralization of atmospheric acidity as compared
with deeper soil horizons. Surface water chemistry exhibits lower pH and acid neutralizing capacity
(ANC) during episodes than during baseflow conditions.
3-11
-------
One of the most significant effects of acidifying deposition on surface water chemistry is the short-
term change in chemistry termed "episodic acidification." While natural processes contribute to seasonal
and short-term increases in the acidity of surface waters, research from several regions in the U.S.
indicates that acidifying deposition likely has substantially increased the magnitude, frequency, and
biological effects of episodic acidification events. Many streams that exhibit chemical conditions during
base flow (relatively stable flows that occur between storms) that is suitable for aquatic biota, are subject
to occasional episodic acidification with adverse consequences. During such episodes, both stream flow
and water chemistry can change markedly (Figure 3-3). Episodic acidification can cause declines in pH
and ANC, and most significantly, increases in inorganic A1 concentrations in stream waters of the
Northeast, Pennsylvania, and in the central Appalachian Mountain region (Charles, 1991). Episodic
decreases in pH and ANC have been documented throughout the country (Wigington et al., 1990).
Episodes are generally accompanied by changes in at least two or more of the following chemical
parameters: ANC, pH, base cations, S042 , N03 . Aln+, organic acid anions, and DOC (Sullivan, 2000).
The U.S. EPA's Episodic Response Project (ERP) confirmed the chemical and biological effects of
episodic pH depressions in lakes and streams in parts of the U.S. (Wigington et al., 1993). The ERP
illustrated that episodic processes are mostly natural, that S042 and especially N03 attributable to
atmospheric deposition play important roles in the episodic acidification of some surface waters, and that
the chemical response that has the greatest effect on biota is increased Al; concentration. Similar findings
had been reported elsewhere, especially in Europe, but the ERP helped to clarify the extent, causes, and
magnitude of episodic acidification in portions of the U.S. (Sullivan, 2000a).
Aquatic biota vary greatly in their sensitivity to episodic decreases in pH and increases in inorganic
Al in waters having low Ca concentration. Baker et al. (1990b) concluded that episodes are most likely to
affect biota if the episode occurs in waters with pre-episode pH above 5.5 and minimum pH during the
episode of less than 5.0. The most thorough characterization of episodic variation in stream chemistry in
the U.S. was conducted through the ERP, in which 13 low-order streams (watershed areas less than
24 km2) in the Adirondack and Catskill regions of New York and the Appalachian Plateau in Pennsylvania
were monitored from 1988 to 1990 (Wigington et al., 1996). About 10% of the acid episodes involved
decreases in ANC of up to 200 (ieq/L, decreases in pH of up to one unit, and increases in concentrations
of inorganic Al of up to 15 |_iM (Wigington et al., 1996). Results showed that acid episodes reduced the
size of fish populations and eliminated acid-sensitive species if median high-flow pH was less than 5.2
and inorganic Al concentration exceeded 3.7 (.iM. despite the relatively short duration of episodes (Baker
et al., 1996).
Results from the ERP demonstrated that episodic acidification can have long-term adverse effects
on fish populations. Streams with suitable chemistry during low flow, but low pH and high inorganic Al
levels during high flow, had substantially lower numbers and biomass of brook trout than were found in
non-acidic streams (Wigington et al., 1996).
In many regions, the most severe acidification of surface waters generally occurs during spring
snowmelt (Charles, 1991). Stoddard et al. (2003) found that, on average, the difference between spring
and summer ANC during baseflow in New England, the Adirondacks, and the Northern Appalachian
Plateau was about 30 j^ieq/L during the period 1990 to 2000 (see Figure 3-4). This implies that lakes and
streams in these regions would need to recover to chronic ANC values above 30 |_icq/L before they could,
on average, be expected to not experience acidic episodes (Stoddard et al., 2003). However, the estimate
of 30 (ieq/L is certain to be low because the comparison was made with non-episodic sampling in spring,
expressed as average spring ANC. ANC measured during episodic spring events would be expected to be
lower than average ANC during spring.
3-12
-------
6
s 5
1 8 4
E w 3
-------
200
0 New England Lakes
0 Adirondack Lakes
O Appaiachian Streams
150 -
o
< 100 -
E
c
I
O) 50 -
c
1
w
C
to
I
-50
-50
50
0
100
150
200
Mean Summer ANC (peq/L)
Source: Stoddard et al. (2003)
Figure 3-4. Relationship between mean summer acid neutralizing capacity (ANC) and the mean of
minimum spring ANC values at long-term monitoring lake and stream sites in New England,
the Adirondacks, and the Northern Appalachian Plateau.
The most important factor governing watershed sensitivity to episodic acidification is the pathway
followed by snowmelt water and storm-flow water through the watershed. The routing of water as it flows
through a watershed determines the degree of contact with acidifying or neutralizing materials and
therefore influences (along with soils and bedrock characteristics) the amount of episodic acidification
that occurs. In any given watershed, surface water ANC may vary in time depending upon the proportion
of the flow that has contact with ANC supplying substrate; in general, the more subsurface contact, the
higher the surface water ANC (Turner et al., 1991). This pattern can be attributed in part to higher base
saturation and (in some watersheds) greater S042 adsorption capacity in subsurface soils. It may also
relate to the accumulation in the upper soil horizons of acidic material derived from atmospheric
deposition and decay processes (Lynch and Corbett, 1989; Turner et al., 1991).
Streams having acidic episodes show significantly higher fish mortality and other aquatic
community changes as compared with streams in which ANC remains above 0 j^icq/L (Wigington et al.,
1993). Results from in situ bioassay studies from across the eastern U.S. show that acidic episodes (with
associated low pH and elevated inorganic Al concentrations, and high streamwater discharge) caused
rapid fish mortality under some conditions (Baker et al., 1996; Bulger et al., 1999; Driscoll et al., 2001b).
For example, streams with suitable conditions during low flow, but moderate-to-severe episodic
acidification during high flow, had higher fish mortality in bioassays, higher net downstream movement
of brook trout during events, and lower brook trout abundance and biomass compared to streams that did
not experience appreciable episodic acidification. These episodically affected streams lacked the more
acid-sensitive fish species (blacknose dace and sculpin). Movement of trout into refugia (areas with
3-14
-------
higher pH and lower inorganic Al) during episodes only partially mitigated the adverse effects of episodes
(Baker etal., 1996).
Consideration of episodic acidification greatly increases the extent and degree of estimated effects
for acidifying deposition on surface waters. In the Northeast, inclusion of episodically acidified water
bodies in regional assessments substantially increases estimates of the extent of surface water
acidification. For example, baseflow samples collected from 1991 to 1994 through the U.S. EPA
Temporally Integrated Monitoring of Ecosystems (TIME) Program indicated that 10% of the 1,812 lakes
larger than 1 ha surface area in the Adirondack region could be considered chronically acidic (fall index
ANC values less than 0 j^icq/L). However, an additional 31% of these lakes had fall index ANC values
less than 50 j^ieq/L and were, therefore, estimated to be susceptible to episodic acidification (Driscoll
etal., 2001b).
Lawrence (2002) estimated the extent of episodically acidified stream reaches in a Catskill, NY
watershed (area = 85 km2) using an index site at the base of the watershed that became episodically
acidified at high flows. Upstream sites with a lower base flow ANC than the index site at the same date
and time were found to have a high likelihood of becoming episodically acidified. Base flow sampling of
122 upstream sites indicated that approximately 16% of the total upstream reaches had chronic ANC less
than 10 (ieq/L, but that 66% of the stream reaches had episodic ANC less than 10 j^icq/L.
In the Southeast, a recent study by Deviney et al. (2006) within Shenandoah National Park,
Virginia used hourly ANC predictions over short time periods to compute recurrence intervals of annual
water-year minimum ANC values for periods of 6, 24, 72, and 168 h. They extrapolated the results to the
rest of the Shenandoah National Park catchments using catchment geology and topography to stratify
watershed response patterns. On the basis of the models, they concluded that a large number of
Shenandoah National Park streams had 6- to 168-h periods of low ANC values, which may stress resident
fish populations (Deviney et al., 2006). Specifically, on the basis of a 4 year recurrence interval,
approximately 23% of the land area (44% of the catchments) can be expected to have conditions that are
classified with respect to brook trout response categories (Bulger et al., 1999) as indeterminate (ANC 20
to 50), episodically acidic (ANC 0 to 20) or chronically acidic (ANC less than 0) for 72 continuous hours.
Many catchments were predicted to have successive years of ANC values sufficiently low as to
potentially extirpate some aquatic species (Deviney et al., 2006). The authors of the study reported that
smaller catchments are more vulnerable to episodic acidification than larger catchments underlain by the
same bedrock. Results from a study of six intensively monitored sites in the Park demonstrated a clear
pattern of larger episodic ANC depressions in streams having higher median ANC than in streams with
lower ANC. However, streams with low median ANC typically experienced decreases that resulted in
minimum ANC values associated with toxicity to biota. These low ANC conditions were more likely to
occur in streams underlain by siliclastic bedrock than in those with granitic or basaltic bedrock.
In the West, episodic acidification is an especially important issue for surface waters throughout
high-elevation areas. Where soils are sparse, as in alpine regions, most snowpack N is flushed to surface
waters early in the snowmelt period. Even though there is evidence through use of isotopic tracers that
much of the N was cycled microbially, snowpack N has been reported to cause temporary acidification of
alpine streams (Campbell et al., 2002; Williams and Tonnessen, 2000). Snowmelt-related temporary
acidification of alpine lakes and streams and associated effects have been reported in the Rocky
Mountains (Brooks et al., 1996; Williams et al., 1996b) and Sierra Nevada (Johannessen and Henriksen,
1978; Stoddard, 1995).
There have been no studies in the U.S. to determine if either the severity or frequency of episodic
acidification has lessened in response to recent decreases in acidifying deposition over the past three
decades. In a study of two streams in Nova Scotia (Laudon et al., 2002) noticeable trends in ANC during
different phases of storm hydrographs from 1983 to 1998 were generally not detected other than during
the peak-flow phase of one stream (an increase of 0.87 j^icq/L).
In summary, the vast majority of water chemistry data for acid-sensitive lakes and streams in the
U.S. were collected at low stream flow. It is well known, however, that water chemistry changes with
season and with weather. Altered water chemistry is most stressful to aquatic biota (lowest pH and ANC;
3-15
-------
highest inorganic A1 concentration) during high flow following snowmelt and rainstorms. During such
conditions, stream chemistry can be toxic to species that thrive under chemical conditions more typical of
base flow. The U.S. EPA ERP and other more localized studies have quantified the effects of episodes on
fish. Episodes are driven by hydrological processes, but the acidification that occurs is largely a result of
acidifying deposition, especially in cases where inorganic A1 has been mobilized. Consideration of such
variability in water chemistry is critical for accurate assessment of the extent, magnitude, and biological
effects of surface water acidification. The biological effects of changes in surface water chemistry are
discussed in greater detail in Section 3.2.3.3.
3.2.2. Terrestrial Ecosystems
The changes in major biogeochemical processes and soil conditions described above contribute to a
series of effects on terrestrial ecosystems. These changes are manifest in both chemical and biological
effects that can include reduced soil base saturation, altered key element ratios, changes in plant
productivity, reduced stress tolerance of sensitive plant species, and in some cases, increased mortality of
canopy trees. Specific chemical indicators of change can be used to assess sensitivity to, and effects from,
acidifying deposition. In the U.S., terrestrial effects of acidification are best described for forested
ecosystems, with supportive information on other plant communities, including shrubs and lichens.
3.2.2.1. Chemical Effects
There are several chemical indicators that provide useful information about the acid-base status of
soils and its influence on terrestrial vegetation. These include soil base saturation; Ca:Al ratio; and C:N
ratio (see Table 3-2). Each chemical indicator provides insight into the level to which the ecosystem has
acidified and may be susceptible to associated biological effects. These chemical indicators may also be
used to monitor the extent of acidification or recovery that occurs in forest ecosystems as deposition rates
of S and N change. As such, several chemical indicators and possible effect thresholds have been
developed and applied in conjunction with efforts to estimate critical loads. The critical loads approach is
discussed in more detail in Section 3.1.1.
Table 3-2. Examples of chemical indicators of effects from acidifying deposition to terrestrial
ecosystems.
Examples of Chemical
Indicators
Example Possible Effect n ,
T. u u References
Threshold
Soil base saturation
10-20% Lawrence et al. (2005); Driscoll et al. (2001b); Cronan et al.
(1990)
Soil solution Ca:AI ratio
1.0 Cronan and Grigal (1995)
Soil C:N ratio
20-25 Aberetal. (2003)
3-16
-------
Soil Base Saturation
In soils with a base saturation less than about 15 to 20%, exchange ion chemistry is dominated by
A1 (Reuss, 1983). Under this condition, responses to sulfuric and nitric acid inputs largely involve the
release and mobilization of inorganic A1 through cation exchange. This is the form of A1 that interferes
with uptake of Ca by plant roots and is also toxic to many forms of aquatic biota (Baker et al., 1990b;
Cronan and Grigal, 1995).
The soil O horizon tends to have a much higher base saturation than the underlying mineral soil,
despite having lower pH due to organic acidity. The base saturation of the B horizon is in a Spodosol can
be sensitive to base cation depletion from leaching by S042 and N03 . and is therefore useful for
assessing base status with regard to acidifying deposition. Little direct work has been done to relate soil
base saturation to forest health, but Cronan and Grigal (1995) determined that base saturation values
below about 15% in the B horizon of forests in the northeastern U.S. could lead to effects from Al stress.
Lawrence et al. (2005) also observed pronounced decreases in diameter growth of Norway spruce in
northwestern Russia, where base saturation decreased from 30% to 20% in the upper 10 cm of the B
horizon over a period of 37 years.
Base saturation values less than 10% predominate in the soil B horizon in the areas in the U.S.
where soil and surface water acidification from acidifying deposition have been most pronounced,
including conifer and hardwood forests in the Adirondack Mountains (Sullivan et al., 2006a), red spruce
forests throughout the Northeast (David and Lawrence, 1996), hardwood forests in the Allegheny Plateau
(Bailey et al., 2004), and conifer and hardwood forests in the southern Appalachian Mountains (Sullivan
et al., 2003). In a study of sugar maple decline throughout the Northeast, Bailey et al. (2004) found
threshold relationships between base cation availability in the upper B soil horizon and sugar maple
mortality at Ca saturation less than 2%, and Mg saturation less than 0.5% (Bailey et al., 2004). The
authors concluded that base saturation varied as a function of topography, geologic parent material, and
acidifying deposition.
Aluminum Concentration in Soil Solution: Calcium to Aluminum Ratio
Al may be toxic to tree roots. Plants affected by high Al concentration in soil solution often have
reduced root growth, which restricts the ability of the plant to take up water and nutrients, especially Ca
(Parker et al., 1989) (Figure 3-5). Ca is well known as an ameliorant for Al toxicity to roots in soil
solution, as well as to fish in a stream. However, because inorganic Al tends to be increasingly mobilized
as soil Ca is depleted, elevated concentrations of inorganic Al tend to occur with low levels of Ca in
surface waters. Mg, and to a lesser extent Na and K, have also been associated with reduced Al toxicity.
Dissolved Al concentrations in soil solution at spruce-fir study sites in the southern Appalachian
Mountains frequently exceed 50 (.iM and sometimes exceed 100 (.iM (Eagar et al., 1996; Johnson et al.,
1991; Joslin and Wolfe, 1992a). All studies reviewed by Eagar et al. (1996) showed a strong correlation
between Al concentrations and N03 concentrations in soil solution. They surmised that the occurrence of
periodic large pulses of N03 in solution were important in determining Al chemistry in the soils of
southern Appalachian Mountain spruce-fir forests.
The negative effect of Al mobilization on Ca uptake by tree roots was proposed by Shortle and
Smith (1988). Substantial evidence of this relationship has accumulated over the past two decades
through field studies (Kobe et al., 2002; McLaughlin and Tjoelker, 1992; Minocha et al., 1997; Schlegel
et al., 1992; Shortle et al., 1997) and laboratory studies (see review by Cronan and Grigal, 1995; Sverdrup
and WarfVinge, 1993). Based on these studies, it is clear that high inorganic Al concentration in soil water
can be toxic to plant roots. The toxic response is often related to the concentration of inorganic Al relative
to the concentration of Ca, expressed as the molar ratio of Ca to inorganic Al in soil solution. As a result,
considerable effort has been focused on determining a threshold value for the ratio of Ca to Al that could
be used to identify soil conditions that put trees under physiological stress.
From an exhaustive literature review, Cronan and Grigal (1995) estimated that there was a 50% risk
of adverse effects on tree growth if the molar ratio of Ca to Al in soil solution was as low as 1.0. They
3-17
-------
estimated that there was a 100% risk for adverse effects on growth at a molar ratio value below 0.2 in soil
solution.
Transpiration
FUNCTION
Membrane integrity
Stomatal regulation
Enzyme activation
Carbohydrate metabolism
Cold hardiness
Defense/chemical-physical
I
GROWTH
Cell division
Cell wall synthesis
Stress tolerance
1
STRUCTURE
Canopy integrity
Leaf form
Wood quality
Tree height
Root distribution
Fo iar
Ca
Uptake
Exchange
Forrest
Mineral
Pool
Deposition
Cytoplasm
Cell Wain
Cal
Leaching
Throughfall
s In
Root xylem and cortex
-------
years in the Oa and B horizons of a high-elevation red spruce stand experiencing high mortality. In the
3-yr study of De Witt et al. (2001), A1 additions lowered molar Ca to inorganic A1 ratios in soil solutions
of a Norway spruce stand below 0.5, but the authors found no response other than reduced Mg
concentrations in needles in the third year, which was a possible precursor to damage.
In summary, a molar ratio of Ca to Al in soil solution can be used as a general index that suggests
an increasing probability of stress to forest ecosystems as the ratio decreases. The ratio value of 1.0 is
proposed as a general damage threshold, but cannot be interpreted as a universally applicable threshold in
all natural systems. Tree species vary widely in their sensitivity to Al stress (See Annex Table B-20). In
addition, Al concentrations in soil solution often exhibit pronounced spatial and temporal variability that
is difficult to relate to root activity. Finally, the form of Al present in solution plays an important role in
determining toxicity. For example, organically complexed Al, which predominates in upper, organic-rich
soil horizons, is essentially nontoxic (Baker and Schofield, 1982; Cronan and Grigal, 1995).
Soil N: Carbon to N Ratio
Mechanisms of retention and release of N in forest ecosystems are not fully understood, but the
adverse effects of nitrification and associated acidification and cation leaching have been consistently
shown to occur only in soils with a C:N ratio below about 20 to 25 (Aber et al., 2003; Ross et al., 2004).
This observation makes the C:N ratio especially useful because N mineralization and nitrification rates
are difficult to measure directly under natural conditions. All available measurement approaches disturb
the soil and often cause artificially high rates. Therefore, field measurement provides a relative index
rather than a realistic quantitative rate (Ross et al., 2004). Approaches for measuring N mineralization and
nitrification also are subject to high degrees of variability, both temporally (hourly to seasonal) and
spatially (down to the sub meter level). Measurements of total OC and N however, are less variable in
space and time and are therefore more straightforward to document than N mineralization and nitrification
rates. Also, ratios of C to N in the forest floor are inversely related to acidifying deposition levels,
although the relationship is stronger for hardwood stands than conifer stands (Aber et al., 2003). In
summary, these factors make the C:N ratio a reliable and relatively straightforward measure for
identifying forest ecosystems that may be experiencing soil acidification and base leaching as a result of
N input and increased nitrification.
DOC leaching
Over the past two decades, increased DOC concentrations in soils and surface waters were widely
reported across North America and Europe (Findlay 2005; Evans et al., 2006; Roulet and Moore 2006;
Monteith et al., 2007). It is very likely DOC levels will continue to rise, leading to an increase in C export
from the relative stable pool (soil) to the labile pool (riverine and marine) (Roulet and Moore 2006;
Monteith et al., 2007). Although there have been arguments about the mechanism responsible for
producing the high DOC concentration, changes in S and N atmospheric deposition both appears to play
key roles in this trend (Driscoll et al., 2003c; Findlay 2005; Evans et al., 2006; Roulet and Moore 2006;
Monteith et al., 2007). Through an assessment of time series data from 522 remote lakes and streams in
North American and Europe, Monteith et al. (2007) found that DOC concentrations had increased in
proportion to the rates at which atmospheric S and sea salt deposition declined. Numerous laboratory and
field studies showed that the solubility of SOM increases with decreasing soil acidity (Kalbitz et al.,
2000). Declining S deposition may increase soil DOC concentration by increasing soil pH. Decreasing S
deposition may also result in lower ionic strength by decreasing the concentrations of a suite of
multivalent ions, such as S042" and aluminum. The coagulation of DOC decreases under low ionic
strength, leading to higher DOC export rates. (Monteith et al., 2007).
Chronically high N deposition has been proposed as another significant factor that influences DOC
production and transportation. Increases in DOC concentrations were often observed in both experimental
sites which received chronic N addition (Pregitzer et al., 2004; Sinsabaugh et al., 2004; Adams et al.,
2005), and natural ecosystems which experienced high rates of N deposition, such as Hudson River and
3-19
-------
several European peatlands (Findlay 2005; Bragazza et al., 2006). Although N deposition contributes to
the acidification of soils, it may increase DOC export by increasing litter input and decreasing DOC
degradability (Roulet and Moore 2006). However, the mechanism behind the association between N
deposition and DOC production is not yet well understood (Pregitzer et al., 2004; DeForest et al., 2005;
Findlay 2005).
3.2.2.2. Summary of Biogeochemistry and Chemical Effects
The evidence is sufficient to infer a causal relationship between acidifying deposition and changes in
biogeochemistry related to terrestrial ecosystems. The strongest evidence for a causal relationship comes
from studies of forested ecosystems, with supportive information on other plant communities, including
shrubs and lichens; grasslands are likely less sensitive to acidification than forests. Soil acidification
occurs in response to inputs of sulfuric acid (H2S04) and nitric acid (HN03); the effect can be neutralized
by weathering or base cation exchange. Soil acidification is a natural process, but is often accelerated by
acidifying deposition. Acidifying deposition is important in decreasing concentrations of exchangeable
base cations in soils. Despite recent decreases in acidifying deposition, there are widespread observations
of ongoing soil acidification such as decreases in soil exchangeable base cations (Bailey et al., 2005,
Sullivan et al. 2006a, 2006b, Warby et al., 2009). The limited mobility of anions associated with naturally
derived acidity (organic acids and carbonic acid) controls the rate of base cation leaching from soil under
conditions of low atmospheric deposition of S and N. Because inputs of S and N in acidifying deposition
provide anions that are more mobile in the soil environment than anions of naturally derived acids, these
mineral acid anions can accelerate natural rates of base-cation leaching.
Nitrification is mediated by autotrophic bacteria that derive energy by oxidizing NH/ to N03 .
Nitrification produces acidity in the form of HN03 as a byproduct. The HN03 produced contributes to the
acidification of soils and surface waters.
There are three useful indicators of chemical changes and acidification effects on terrestrial
ecosystems, showing consistency and coherence among multiple studies: soil base saturation, Al
concentration in soil water, and soil C:N ratio.
¦ Soil base saturation is the concentration of exchangeable bases as a percent of the total soil cation
exchange capacity. Once base saturation decreases to a critical level (approximately 15-20%),
inputs of H2S04 and HN03 result in exchange of inorganic Al.
¦ Inorganic Al is toxic to some tree roots. Plants affected by high inorganic Al concentrations in
soil solution often have reduced root growth, which restricts the ability of the plant to take up
water and nutrients, especially calcium (Ca) (Parker et al., 1989).
¦ The C:N ratio of soil is used to indicate alterations to the N biogeochemical cycle. If the C:N ratio
of soils falls below about 20 to 25, nitrification is stimulated resulting in net nitrification and
increased acidity.
3.2.2.3. Biological Effects
Acidifying deposition can affect terrestrial ecosystems via direct effects on plant foliage and
indirect effects associated with changes in soil chemistry. Biological effects of acidification on terrestrial
ecosystems are generally attributable to Al toxicity and decreased ability of plant roots to take up base
cations (especially Ca) and water from the soil (Cronan and Grigal, 1995). Acidifying deposition to acid-
sensitive soils can cause soil acidification, increased mobilization of inorganic Al from soil to drainage
water, and depletion of the pool of exchangeable base cations in the soil. Effects on the soil and direct
effects of acidifying deposition on foliage can influence the response of plants to climatic stresses such as
3-20
-------
drought and cold temperature. They can also influence the sensitivity of plants to other stresses, including
insect pests and disease (Joslin et al., 1992).
The combined effects of acidifying deposition and other stressors on terrestrial vegetation are
typically measured using indices such as percent dieback of canopy trees, dead tree basal area (as a
percent), crown vigor index, and fine twig dieback (see Table 3-3). Each of these variables has a rating
system used to quantify forest condition and relate the variables to foliar and soil nutrient concentrations.
Table 3-3.
Example biological effects indicators in terrestrial ecosystems.
Indicator Species
Example of Health Indices
References
Red spruce
Percent dieback of canopy trees
Shortle etal. (1997)
DeHayes etal. (1999)
Sugar maple
Basal area dead sugar maple (as %)
Crown vigor index
Fine twig dieback
Bailey et al. (1999)
Drohan and DeWalle (2002)
The effects of acidifying deposition on the health, vigor, and productivity of terrestrial ecosystems
in the U.S. are best documented in spruce-fir and northern hardwood forests of the eastern U.S. Some
information is also available for individual species such as red spruce, sugar maple, and some species of
lichen. In the western U.S., the health of Ponderosa pine and Jeffrey pine has been affected by air
pollution, but such effects have largely been attributed to ozone exposure, not acidifying deposition.
Health, Vigor, and Reproduction of Tree Species in Forests
Both coniferous and deciduous forests throughout the eastern U.S. are experiencing gradual losses
of base cation nutrients from the soil due to accelerated leaching from acidifying deposition. This change
in base cation nutrient availability may reduce the quality of forest nutrition over the long term. Evidence
suggests that red spruce and sugar maple in some areas in the eastern U.S. have experienced declining
health as a consequence of acidifying deposition. Existing information regarding the effects of acidifying
deposition on these two forest tree species is summarized below and reference is made to specific health
indicators where such information is available.
Red Spruce
Red spruce (Picea rubens) is a conifer that occurs mainly in the northeastern U.S. and at scattered
high-elevation sites in the Appalachian Mountains (see Figure 3-6). Red spruce dieback or decline has
been observed across high elevation landscapes of the northeastern, and to a lesser extent, southeastern
U.S. At high elevations in the Adirondack and Green Mountains, more than 50% of the canopy red spruce
trees died during the 1970s and 1980s. In the White Mountains, about 25% of the canopy spruce died
during that same period (DeHayes et al., 1999). Dieback of red spruce has also been observed in mixed
hardwood-conifer stands at relatively low elevations in the western Adirondack Mountains, an area that
receives high inputs of acidifying deposition (Shortle et al., 1997); acidifying deposition has been
implicated as a causal factor (DeHayes et al., 1999). The frequency of freezing injury to red spruce
needles has increased over the past 40 years, a period that coincided with increased emissions of S and N
oxides and increased acidifying deposition (DeHayes et al., 1999).
From the 1940s to 1970s, red spruce growth also declined at high elevation in the southeastern U.S.
(Cook and Zedaker, 1992; Eagar et al., 1996; McLaughlin et al., 1987} as emissions of both NOx and S02
3-21
-------
increased to maxima of about 25 and 30 million tons/yr, respectively. The growth decline in Great Smoky
Mountains National Park in North Carolina and Tennessee started earlier at higher elevations (around the
1940s and 1950s) and was steeper, while the growth decline developed at lower elevation sites 20 years
later. After the 1980s, red spruce growth increased substantially at both the higher- and lower-elevation
sites, corresponding to a decrease in S02 emissions in the U.S. (to about 20 million tons/yr by 2000),
while NO emissions held fairly stead}' (at about 25 million tons/yr). Annual emissions of S plus NOx
explained about 43% of the variability in red spruce tree ring growth between 1940 and 1998. Climatic
variability accounted for about 8% of the growth variation for that period. At low elevation, changes in
radial growth could be explained by climatic variables only, and there was no correlation with national S
plus nitrogen oxide emissions trends. Recent reductions in S oxide emissions may have changed growth
trajectories (Webster et al., 2004).
Source: Little (1971)
Figure 3-6. Distribution of red spruce (rose) and sugar maple (green) in the eastern U.S. These two tree
species have experienced adverse effects in portions of their ranges that have been attributed
to acidification from acidifying deposition. Tree distribution data were obtained from Little's
Atlas.
The observed dieback in red spruce has been linked, in part, to reduced cold tolerance of red spruce
needles, caused by acidifying deposition. Results of controlled exposure studies showed that acidic mist
or cloud water reduced the cold tolerance of current-year red spruce needles by 3 to 10 °C (DeHayes
et al., 1999). There is a significant positive association between cold tolerance and foliar Ca in trees that
exhibit foliar Ca deficiency. The membrane-associated pool of Ca, although a relatively small fraction of
the total foliar Ca pool, strongly influences the response of cells to changing environmental conditions.
The plant plasma membrane plays an important role in mediating cold acclimation and low-temperature
injury (U.S. EPA, 2004). The studies of DeHayes et al. (1999) suggested that direct acidifying deposition
on red spruce needles preferentially removes membrane-associated Ca. More recently, a link has been
3-22
-------
established between availability of soil Ca and winter in jury (Hawley et al., 2006) based on an
experimental addition of Ca at the Hubbard Brook Experimental Forest, New Hampshire. This study
demonstrated that Ca depletion from soil was associated with winter injury of red spruce foliage during
2003 when winter injury was unusually high throughout the region (see Figure 3-7).
(A)
¦ Reference
OSCa-addition
3
ns
I
Dominant and Intermediate, suppressed,
codominant and understory
Tree crown class
¦ Reference
01 Ca-addition
ns
L
Dominant and Intermediate, suppressed,
codominant and understory
Tree crown class
Source : Hawley et a I. (2006). Reprinted with permission.
Figure 3-7. Mean {+ standard error bars) of current-year red spruce needle winter injury in reference and
calcium-addition watersheds and among crown classes, expressed as foliar injury (A) and
bud mortality (B). Watershed means were either not significantly different (ns) or statistically
different at p <0.05 (*) or p <0.01 (**) based on nested analyses of variance.
In summary, the weight of evidence suggests that changes in soil chemistry have contributed to
high mortality rates and decreasing grow th trends of red spruce trees in some areas over the past three
decades (Sullivan et al., 2002). In forests where this has occurred, which are mainly located at high
elevation, changes in red spruce growth rates are attributable, at least in part, to base cation deficiencies
related to decreased availability of Ca and increased availability of Al as a result of acidifying deposition
3-23
-------
effects on soils. Important factors appear to include depletion of base cations in upper soil horizons by
acidifying deposition, A1 toxicity to tree roots, and accelerated leaching of base cations from foliage as a
consequence of acidifying deposition. Recent studies also show improvements in red spruce growth with
decreasing emissions of S02 in the U.S. (Webster et al., 2004).
Sugar Maple
Sugar maple (Acer saccharum) is the deciduous tree species of the northeastern U.S. that is most
commonly associated with adverse acidification-related effects of S and N deposition, though other base
cation accumulating hardwoods may also be at risk (Driscoll et al., 2001b). Sugar maple is distributed
throughout the northeastern U.S. and central Appalachian Mountain region as a component of the
northern hardwood forest Figure 3-6.
A conceptual view of the interactions of acidifying deposition and other stressors in sugar maple
decline is provided in Figure 3-8. Several studies, mainly in Pennsylvania, have hypothesized that sugar
maple decline is linked to the occurrence of relatively high levels of acidifying deposition and base-poor
soils (Bailey et al., 2004; Hallett et al., 2006; Horsley et al., 2000; Moore and Ouimet, 2006; St. Clair et
al., 2005).
Acidifying deposition may be contributing to episodic dieback of sugar maple in the Northeast
through depletion of nutrient cations from marginal soils (Figure 3-8). Horsley et al. (2000) found that
dieback at 19 sites in northwestern and north central Pennsylvania and southwestern New York was
correlated with combined stress from defoliation and soil deficiencies of Mg and Ca. Dieback occurred
predominately on ridgetops and on upper slopes, where soil base cation availability was much lower than
occurred in the deeper soils found on middle and lower slopes (Bailey et al., 2004). A long-term decrease
in soil pH since 1960 (0.78 pH unit decrease in the O horizon, and 0.23 pH unit decrease in the A horizon)
in Pennsylvania hardwood forests has been documented, along with decreases in soil Ca and Mg
concentrations. Declining sugar maples were shown to be deficient in foliar Ca and Mg (Drohan and
Sharpe, 1997). More recent research has strengthened understanding of the role of cation nutrition in
sugar maple health at a regional scale across a broad range of conditions (Hallett et al., 2006).
Drohan et al. (2002) investigated differences in soil conditions in declining versus non-declining
sugar maple plots in northern Pennsylvania from the U.S. Department of Agriculture (USDA) Forest
Service's Forest Inventory and Analysis (FIA) program. Soils in plots with declining sugar maple tended
to have lower base cation concentrations and pH, and Ca:Al ratio less than 1. Regressions between foliar
and soil chemistry showed that foliar nutrition was highly correlated with the chemistry of the upper 50
cm of soil (Drohan et al., 2002).
Juice et al. (2006) added Cato watershed 1 (Wl) at HBEF in October 1999 sufficient to raise the
pH of the Oie soil horizon from 3.8 to 5.0 and the Oa horizon from 3.9 to 4.2. Subsequently, they
measured the response of sugar maples to the Ca fertilization. Foliar Ca of canopy sugar maples increased
markedly and foliar Mn declined. By 2005, crown condition was much healthier then in the untreated
reference watershed (W6). The density of sugar maple seedlings increased significantly following high
seed production in 2000 and 2002. In addition, sugar maple germinants were 50% larger on Wl and
mycorrhizal colonization of seedlings was much higher in the treated watershed (22.47% of root length)
as compared with the reference watershed (4.4%) (Juice et al., 2006).
In general, evidence indicates that acidifying deposition in combination with other stressors is a
likely contributor to the decline of sugar maple trees that occur at higher elevation, on geologies
dominated by sandstone or other base-poor substrate, and that have base-poor soils having high
percentages of rock fragments (Drohan et al., 2002). Such site conditions are representative of the kinds
of conditions expected to be most susceptible to adverse effects of acidifying deposition because of
probable low initial base cation pools and high base cation leaching losses.
3-24
-------
Growth
and Health
Secondary
Stressors
Carbohydrate Supply
Acidification
Net Photosynthesis
Available Mn
Ca and Mg Uptake
Available Al
Available Ca arid Mg
Source: Hallett et al. (2006). Reprinted with permission.
Figure 3-8. Conceptual diagram outlining the current understanding of sugar maple decline. Positive and
negative signs indicate the nature of the correlative relationship between variables.
Other Forest Ecosystems
Loss of base cations, specifically Ca2+, has also been implicated in increased susceptibility of
flowering dogwood (Cornus florida) to its most destructive disease, dogwood anthracnose (Figure 3-9).
Flowering dogwood is a dominant understory species of hardwood forests in the eastern U.S.
(Holzmueller et al., 2006), with important ecosystem functions as a food source for numerous species of
animals, and as a large contributor to available Ca in forest litter. It is also recognized as a significant
cultural and aesthetic resource throughout its range. Since dogwood anthracnose, a mostly fatal disease,
was first reported in 1976 in New York State, it has spread over a large portion of the species' range,
generally resulting in mortality greater than 90% in affected stands. Pacific dogwood (Cornus nutallii) is
similarly affected, but because its abundance within its range was much lower before the disease first
appeared, the effect has received less notice. Susceptibility to the disease, and disease severity in stands,
appear dependent on several factors, including acid deposition and various edaphic characteristics and
meteorological conditions.
3-25
-------
Source: Holzmueller et al. (2006). Reprinted with permission.
Figure 3-9. Native range of flowering dogwood (Cornus florida) (dk. gray) and the documented range of
dogwood anthracnose in the eastern U.S. (red). 2002 data from the U.S. Forest Service.
In 1990 and 1991, Britton et al. (1996) exposed 200 potted dogwood plants to simulated acid rain
(SAR) at 4 levels of pH between 2.5 and 5.5. The plants were then placed among natural stands showing
symptoms of the disease. In both years, there was a fourfold increase in percentage of leaf area affected
from the plants treated with SAR at a pH of 5.5 to those treated with SAR at a pH of 2.5. In 1992 and
1993, four combinations of SAR with a pH of 2.5 or 5.5 were applied separately to the foliage and the soil
before inoculation. The percent of leaf area affected was approximately two to four times greater for
plants grown in soil treated with acidic SAR, regardless of foliar treatment, suggesting that the worsening
of anthracnose damage by acid deposition occurs mostly through soil effects.
In a study of the effects of Ca, K, and Mg on dogwood density in forest stands, and on resistance to
anthracnose in containerized dogwood plants, (Holzmueller et al., 2007) found a strong relation between
soil available cations, particularly Ca2+, and dogwood density in Great Smoky Mountains National Park,
where dogwood anthracnose has resulted in significant damage. The mortality of potted dogwood plants
fertilized with solutions varying in Ca2+, K+, or Mg2+ concentration, and exposed to anthracnose, was both
greatest and most rapid when Ca2+ was deficient, but not when K+ and Mg2+ were deficient.
Data on the possible effects of S and N deposition on the acid-base characteristics of forests in the
U.S., other than the spruce-fir and northern hardwood forest ecosystems, are limited. Ponderosa pine
(Finns ponderosa) seedlings exposed to acidic precipitation (pH 5.3, 4.4, 3.5 of 1:1 NH03:H2S04)
showed no significant changes in growth (Temple et al., 1992). Fenn et al. (2003a) reported that
deposition of 20 to 35 kg N/ha/yr contributed to increased NO, leaching and soil acidity and decreased
base saturation in southern California forest ecosystems, but they did not report quantitative measures of
growth. Baron et al. (2000) showed that small differences in the N deposition between the east (3 to 5 kg
N/ha/yr) and the west (1 to 2 kg N/ha/yr) side of the Rocky Mountains were associated with significant
declines in foliar Mg levels and increased foliar N:Mg and N:Ca ratios in old-growth stands of
Engelmann spruce (Piceci engelmanii). It is not known if such changes in nutrient ratios affect the health
or growth of these forests.
3-26
-------
Despite the evidence for effects of acidifying deposition on the health and vigor of some terrestrial
plant communities, few studies are available that have directly documented species loss, reduced
biodiversity, or adverse effects on threatened and endangered species. A notable exception is the effect of
acidifying deposition on lichen abundance and diversity within forest communities (See discussion
below). In eastern North America and central Europe, areas that receive relatively high levels of
acidifying deposition and high atmospheric concentrations of S02, N oxides, and reduced N have
experienced noticeable reductions in cyanolichen abundance on both coniferous and deciduous trees
(Richardson and Cameron, 2004). Effects on lichen species biodiversity are also likely (McCune, 1988;
van Haluwyn and van Herk, 2002). In London, epiphyte diversity, including a majority of the lichen taxa,
declined in areas where NO exceeded 40 (ig/m3 and total N oxides exceeded 70 |_ig/nr\
Health and Biodiversity of Other Plant Communities
Shrubs
Forest trees are not the only vascular plants that are potentially sensitive to acidifying deposition.
Available data suggest that it is possible, or perhaps likely, that a variety of shrub and herbaceous species
are sensitive to base cation depletion and/or A1 toxicity. However, conclusive evidence is generally
lacking.
Research in Europe has illustrated a shift from shrub to grass dominance in heathlands in response
to acidifying deposition. However, such effects are probably more related to the nutrient enrichment
effects of N deposition than to the acidification effects of S and N deposition. (See further discussion in
Section 3.3.3.1.) In summary, whereas some evidence suggests that effects on shrubs and perhaps
herbaceous plants are possible, data in the U.S. are insufficient to support the use of shrub or herbaceous
plant species as indicators of the acidification-related effects of acidifying deposition at this time.
Lichens
Typically, lichens and bryophytes are among the first components of the terrestrial ecosystem to be
affected by acidifying deposition. Vulnerability of lichens to increased N input is generally greater than
that of vascular plants (Fremstad et al., 2005). Even in the Pacific Northwest, which receives uniformly
low levels of N deposition, changes from acid-sensitive and N-sensitive to pollution-tolerant and
nitrophillic lichen taxa are occurring in some areas (Fenn et al., 2003a). Lichens remaining in areas
affected by acidifying deposition were found by Davies et al. (2007) to contain almost exclusively the
families Candelariaceae, Physciaceae, and Teloschistaceae.
Effects of S02 exposure on lichens includes reduced photosynthesis and respiration, damage to the
algal component of the lichen, leakage of electrolytes, inhibition of N fixation, reduced K absorption, and
structural changes (Farmer et al., 1992; Fields, 1988). In response to reductions after the 1970s in S02
exposure and acidifying deposition in London, lichen diversity increased dramatically (Hawksworth,
2002). However, the recovery of lichens in response to reduced S and N inputs is inconsistent.
Improvement for bryophytes has been reported to occur in 1 year by Power et al. (2006) and Mitchell
et al. (2004), 5 years by Gordon et al. (2001), and 49 years by Strengbom et al. (2001).
Scott (1989a, 1989b) concluded that the S:N exposure ratio was as important as pH in causing toxic
effects on lichens, based on experiments on Cladina rangiferina and C. stellaris. Thus, it is not clear to
what extent acidity may be the principal stressor under high levels of air pollution exposure. The toxicity
of S02 to several lichen species is greater under acidic conditions than under neutral conditions. The
effects of excess N deposition to lichen communities are discussed in Section 3.3.5.1.
3-27
-------
Grasslands
Due to structural differences and their lower canopy, grasslands are thought to be less sensitive to
acidification than woodlands (Blake et al., 1999; Kochy and Wilson, 2001). Among grasslands, those with
calcareous soils will be less sensitive than those with acidic soils (Bobbink et al., 1998). Most literature
on the effects of atmospheric S and N deposition on grasslands documents effects of fertilization from N
deposition, not acidification. Such fertilization effects are discussed in Section 3.3.5.1.
Arctic and Alpine Tundra
The possible effects of acidifying deposition on arctic and alpine plant communities are also of
concern. Especially important in this regard is the role of N deposition in regulating ecosystem N supply
and plant species composition. (See further discussion of such effects in Section 3.3.5.1.) Soil
acidification and base cation depletion in response to acidifying deposition have not been documented in
arctic or alpine terrestrial ecosystems in the U.S. Such ecosystems are rare and spatially limited in the
eastern U.S., where acidifying deposition levels have been high. These ecosystems are more widely
distributed in the western U.S. and throughout much of Alaska, but acidifying deposition levels are
generally low in these areas. Key concerns are for listed threatened or endangered species and species
diversity. However, for most rare, threatened, or endangered herbaceous plant species, little is known
about their relative sensitivities to acidification from atmospheric deposition inputs. Although plant
species diversity of arctic and alpine ecosystems is highly valued, it is difficult to document changes in
this parameter in response to acidifying deposition.
3.2.2.4. Summary of Biological Effects
The evidence is sufficient to infer a causal relationship between acidifying deposition and changes in
terrestrial biota. The strongest evidence for a causal relationship comes from studies of terrestrial systems
exposed to elevated levels of acidifying deposition that show reduced plant health, reduced plant vigor,
and loss of terrestrial biodiversity. Consistent and coherent evidence from multiple species and studies
shows that acidifying deposition can affect terrestrial ecosystems by causing direct effects on plant foliage
and indirect effects associated with changes in soil chemistry. Biological effects of acidification on
terrestrial ecosystems are generally attributable to Al toxicity and decreased ability of plant roots to take
up base cations. There are several indicators of stress to terrestrial vegetation (see Table 3-3) including
percent dieback of canopy trees, dead tree basal area (as a percent), crown vigor index, and fine twig
dieback.
Species Level
¦ Changes in soil chemistry (depletion of soil base cations, Al toxicity to tree roots, leaching of
base cations into drainage water) have contributed to high mortality rates and decreasing growth
trends of red spruce trees (Picea rubens) in some areas of the eastern U.S. over the past three
decades.
¦ Acidifying deposition, in combination with other stressors, is a likely contributor to the decline of
sugar maple (Acer saccharum) trees that occur at higher elevation, in some portions of the eastern
U.S., on geologies dominated by sandstone or other base-poor substrate, and that have base-poor
soils.
¦ Lichens and bryophytes are among the first species affected by acidifying deposition in the
terrestrial ecosystem. Effects of S02 on lichens include reduced photosynthesis and respiration,
damage to the algal component of lichen, leakage of electrolytes, inhibition ofN fixation, reduced
potassium (K) absorption and structural changes.
3-28
-------
¦ Data are insufficient to draw general conclusions for other species.
Community Level
¦ Species loss and reduced biodiversity of forests, shrubs, and meadow plant communities may
occur, but has not been clearly demonstrated in the U.S.
3.2.3. Aquatic Ecosystems
3.2.3.1. Chemical Effects
The changes in major biogeochemical processes and soil conditions caused by acidifying
deposition have significant ramifications for the water chemistry and biological functioning of associated
surface waters. Surface water chemistry indicates the adverse effects of acidification on the biotic
integrity of fresh water ecosystems. Because surface water chemistry integrates the sum of soil and water
processes that occur upstream within a watershed, it also reflects the results of watershed-scale terrestrial
effects, including N saturation, forest decline, and soil acidification (Stoddard et al., 2003). Thus, water
chemistry integrates and reflects changes in soil and vegetative properties and biogeochemical processes.
The effects on aquatic ecosystems can be described by changes in several chemical effects
indicators such as S042 concentration, N03 concentration, base cation concentration, pH, ANC, and
inorganic Al. All of these are of interest, and each can provide useful information regarding both
sensitivity to surface water acidification and the level of acidification that has occurred. Importantly, these
chemical changes can occur over both long- and short-term timescales. Short-term (hours or days)
episodic changes in water chemistry have perhaps the most significant biological effects. The
acidification effects on aquatic biota are most commonly evaluated using either Al or pH as the primary
chemical indicator (Table 3-4). ANC is also used because it integrates overall acid status and because
surface water acidification models do a better job projecting ANC than pH and inorganic Al
concentrations. However, ANC does not relate directly to the health of biota. The usefulness of ANC lies
in the association between ANC and the surface water constituents that directly contribute to or
ameliorate acidity-related stress, in particular pH, Ca, and inorganic Al. The base cation surplus
(Lawrence et al., 2007) is an alternate index that integrates acid-base status. It is based on a measurement
of ANC (calculated from the charge balance of ionic concentrations in water) and also accounts for the
influence of natural organic acidity.
A synoptic illustration of the national patterns of surface water alkalinity in the conterminous U.S.
is provided in Figure 3-10. Alkalinity is the most readily available measure of the sensitivity of lakes and
streams to acidifying deposition. Although the actual sensitivity of a water body depends on many
watershed characteristics and processes, the low-alkalinity areas on the map indicate where sensitive
surface waters are most likely to be found. The map is based on data from approximately 39,000 lake and
stream sites and the associations of the data values with factors such as land use, physiography, geology,
and soils.
3-29
-------
Table 3-4. Examples of chemical indicators of effects from acidifying deposition to aquatic ecosystems.
Chemical Indicator
Examples of Potential
Thresholds
Reference
Surface water pH
5.0-6.0
Baker et al. (1990b)
Surface water ANC
0-50 peq/L
Bulger et al. (1999)
Inorganic Al
2-4 pmol/L
Baldigo etal. (2007); Driscoll etal. (2001b);
Wigington et al. (1996)
- f
4*
0W r H"
SrJVV * —-—_ _ j»
r'f r * \
7 ""W'Vi
¦<
\
f-
¦r.%
y--'
Source: Omernik et al. (1983).
Figure 3-10. Surface water alkalinity in the conterminous U.S. Shading indicates the range of alkalinity
within which the mean annual values of most of the surface waters of the area fail.
Surface Water SO42"
Measurements of SO ; concentration in surface water provide important information on the extent
of cation leaching in soils and how SO_f concentrations relate to ambient levels of atmospheric S
deposition. Assessments of acidifying deposition effects dating from the 1980s to the present have shown
S042 to be the primary anion in most, but not all, acid-sensitive waters in the U.S. (Driscoll and Newton,
1985; Driscoll et al., 1988; 2001b; Webb et al., 2004). In an analysis representative of over 10,000 acid-
sensitive lakes in the Northeast, inorganic anions represented most negative (anionic) charge in 83% of
the lakes, and in this group of lakes, 82% of the total negative charge was due to S04 (Driscoll et al.,
2001b). In contrast, naturally derived organic anions represented an average of 71% of total negative
charge in the 17% of lakes in which organic anions predominated (Driscoll et al„ 2001b).
3-30
-------
Atmospheric deposition of S is widely acknowledged as causing changes in concentrations of
S042 in surface water. No long-term data sets exist to document changes in S042 in surface waters since
the onset of the Industrial Revolution. One of the longest-running monitoring programs exists at the
Hubbard Brook Experimental Forest in New Hampshire. Surface water data from this Long-Term
Ecological Research site have been used to develop historic estimates of S042 concentrations using the
Photosynthesis and EvapoTranspiration-BioGeoChemical (PnET-BGC) model (Gbondo-Tugbawa et al.,
2002). Results from Hubbard Brook suggest that acidifying deposition has contributed to a nearly four-
fold increase in stream S042 concentration between 1850 and 1970 (Driscoll et al., 2001b).
Long-term data in other regions suggest similar trends in some cases. For example, a study of
seven streams in the Catskill region of New York, Stoddard (1991) identified increasing trends in S042
concentrations from 1952-54 to 1970 in three streams and no trend in the four other streams.
As emissions and deposition of S have declined over approximately the last 30 years, surface water
concentrations of S042 have decreased in most regions in the eastern U.S. For example, Stoddard et al.
(2003) found that surface waters monitored in the U.S. EPA Long-Term Monitoring program showed
consistent decreases in S042 concentrations from 1990 to 2000 in New England lakes (1.77 (ieq/L/yr),
Adirondack lakes (2.26 (ieq/L/yr), Appalachian streams (2.27 (ieq/L/yr) and Upper Midwest lakes
(3.36 (ieq/L/yr). The only exception to the pattern of decreasing S042 concentration in surface waters
during this period was for streams in the Blue Ridge Mountain region of Virginia, which showed a
significant increase in S042 concentrations (0.29 (ieq/L/yr) during this period. The increasing trend in
Virginia streams is presumably the result of decreased S adsorption on soils and net desorption from the
soil in response to decreased S deposition.
In summary, available data indicate a pattern of increasing concentrations of S042 in surface
waters before the year of peak S emissions in the early 1970s, followed by widespread decreasing trends
in S042 concentrations after the peak (with the only exception being the Blue Ridge Mountain region in
Virginia). On this basis, continued decreases in S emissions would be expected to result in further
decreases in S042 concentrations in surface waters, although the rate of response is variable and some
model results suggest that recovery may be delayed as accumulated S leaches from watersheds, even as
emissions and deposition decline.
Surface Water NO3"
As described in the previous section, the acidification potential of atmospherically deposited S is
primarily a function of the extent of S042 anion mobility in watershed soils and drainage waters.
Similarly, acidification of soil water and surface water from atmospheric N deposition is largely governed
by the mobility of the N03 anion. Both oxidized and reduced N deposition can contribute to the N03
flux in drainage water. Once N is deposited, processes within the N cycle, including microbial
assimilation, plant uptake, and loss to denitrification act to limit the extent of N03 leaching. In contrast,
processes such as mineralization, nitrification, fixation, and atmospheric deposition contribute to the N03
flux and increase the likelihood that substantial leaching of N03 in drainage water will occur. Such
leaching of N 03 is required in order for N deposition to cause N saturation, surface water acidification,
or base cation leaching and depletion. Ultimately, the balance of these processes in the N cycle will
determine the extent to which such effects will be manifested.
Whereas S042 is generally considered the dominant agent of surface water acidification in most
affected regions of the U.S., N03 plays a large role in acidification of surface waters in some regions,
particularly during snowmelt and rainstorms. Before the mid-1980s, atmospheric deposition effects
research in the U.S. focused almost exclusively on S. Within the 1980 to 1990 (NAPAP) research
program, relatively little attention was paid to N research.
Release of N03 from soil to surface waters may affect nutrient relationships and biological
neutralization processes in aquatic ecosystems (Bukaveckas and Shaw, 1998; Kelly et al., 1987; Momen
et al., 2006). Driscoll and Newton (1985) found that N03 concentrations in 20 lakes in the early 1980s in
the Adirondack region of New York averaged 12% of S042 concentrations, whereas Lovett et al. (2000)
3-31
-------
found that baseflow N03 concentrations in 1994-97 were an average of 37% of S042 concentrations in
39 streams in the Catskill region of New York. Murdoch and Stoddard (1993) demonstrated the
importance of N03 during high-flow conditions in Catskill streams in which concentrations periodically
equaled or exceeded S042 concentrations. Concentrations of N03 in most southeastern streams tend to
be considerably less than S042 concentrations (Webb et al., 2004). However, Cook et al. (1994)
documented very high N03 concentrations in stream water at high elevation in the Great Smoky
Mountains in North Carolina.
Surface water N03 concentrations have changed overtime and these trends vary by region.
Several regions in the northeastern U.S. showed increased N03 concentrations during the 1980s. For
example, in the Catskill Mountains of New York all 16 streams for which data were available showed
increasing trends in N03 concentration during that period. A similar increase in N03 concentration was
reported for Adirondack lakes in the 1980s (Stoddard et al., 1999). These increasing trends in N03
concentration were initially attributed to N saturation in response to atmospheric deposition (Aber et al.,
1998).
More recent information on N03 trends during the 1990s, when atmospheric N deposition was
relatively stable, suggest that the relationship between atmospheric N deposition and surface water N03
concentrations is complex. During the 1990s, the only significant change occurred in the two regions with
the highest ambient surface water N03 concentrations: lakes in the Adirondack Mountains and streams in
the Northern Appalachian Plateau (Figure 3-11). Both exhibited small but significant downward trends in
N03 concentration during the 1990s. The long-term record of dissolved inorganic N (which is largely
N03 ) concentrations at the Hubbard Brook Experimental Forest showed a similar pattern: high
concentrations in the late 1960s and 1970s, followed by decreases to minimum values in the mid-1990s
(Aber et al., 2002). Across New England and the Upper Midwest, where ambient surface water
concentrations are much lower than in the Adirondack Mountains and Northern Appalachian Plateau
(Figure 3-11), N03 concentrations in surface waters were unchanged during the 1990s. The Ridge/Blue
Ridge province registered a small, but significant, decrease in N03 concentration during the 1990s, but
interpretation of trends for N03 in this region was complicated by gypsy moth defoliation, which caused
large increases in the concentration of N03 in soil water and stream water (Eshleman et al., 1998).
Efforts to explain the complex patterns in N03 concentrations under conditions of reasonably
stable atmospheric N deposition have focused on both terrestrial and aquatic N cycling. Goodale et al.
(2003) reported that lower N03 concentrations measured in the 1990s at streams in New Hampshire
could not be accounted for by differences in stream flow or forest succession, but inter-annual climate
variation was proposed as a possible cause. In the Adirondacks, Driscoll et al. (2007a) proposed that
increased concentrations of atmospheric C02 may have resulted in a fertilization effect that increased N
assimilation. Studies by Mitchell et al. (1996) and Murdoch et al. (1998) provide some evidence of
climate effects on trends in N03 concentrations in surface waters in the Northeast. In particular, a region-
wide spike in N03 concentrations followed an unusually cold December that may have disrupted soil
N cycling processes (Mitchell et al., 1996). Murdoch et al. (1998) also found that mean annual air
temperatures were strongly related to average annual N03 concentrations during most years in a Catskill
watershed with elevated N03 concentrations in stream water.
3-32
-------
Regional Trends, 1990-2000
(in lakes and streams)
Sulfate (peq/L/yr)
Nitrate (peq/L/yr)
ANC (peq/L/yr)
Hydrogen Ion (peq/L/yr)
Base Cations (peq/L/yr)
DOC (mq/L/yr)
Aluminum (peq/L/yr)
-1 o 1
Slope of Trend
New England Lakes
l < Adirondack Lakes
i i Northern Appalachian Streams
¦¦ Upper Midwest Lakes
Ridge and Blue Ridge Streams
Source: Stoddard et al. (2003)
Figure 3-11. Summary of regional trends in surface water chemistry from 1990 to 2000 in regions covered
by the Stoddard et al. (2003) report.
Processes within lakes may have also played a role in the measured trends in Adirondack lakes (Ito
et al., 2005; 2007). In a study of 30 of the 48 long-term monitoring lakes investigated by (Driscoll et al..
2003d; 2007a) and Momen et al. (2006) found that concentrations of dissolved NO, were inversely
correlated with concentrations of chlorophyll a (Chi a) in 11 lakes, and that Chi a was increasing in
concentration in 9 lakes. The increase in pH observed in most of these lakes may have stimulated
productivity so that N assimilation by plankton increased (Momen et al., 2006).
In summary, N03 contributes to the acidity of many lakes and streams in the eastern U.S. that have
been affected by acidifying deposition, especially during spring months and under high-flow conditions.
Nevertheless, there is little or no apparent relationship between recent temporal trends in N deposition and
trends in NO; concentrations m surface waters in the eastern U.S. This observation is in sharp contrast to
observed responses for S deposition and SOr concentrations. These results likely reflect the
complexities of N use within terrestrial and aquatic ecosystems. Uptake of atmospherically deposited N
by plants and microorganisms in the terrestrial environment precludes drainage water acidification and
base cation leaching that would be caused if excess N leached as N0; from the terrestrial to aquatic
ecosystems. While great uncertainty exists, and the timescales of N saturation may be longer than
previously considered (e.g., centuries rather than decades), the long-term retention of N deposited in
forested regions and consequent dampening of deposition effects on surface waters is unlikely to continue
indefinitely (Aber et al., 2003). Moreover, spatial patterns of N03 concentrations in surface water across
the northeastern U.S. are consistent with atmospheric N deposition values although there is considerable
variation in these concentrations based upon watershed attributes.
Surface Water Base Cations
The results from several studies in the eastern U.S. suggest that base cation concentrations in
surface waters increased during the initial phases of acidification into the 1970s. This trend reversed and
base cations decreased in response to decreasing S042 and NOT concentrations. For example, Likens
3-33
-------
et al. (1996) evaluated trends in base cation concentrations in stream water in relation to long-term trends
in S042 plus N03 for the Hubbard Brook Experimental Forest. This record showed an approximately
linear increasing relationship between concentrations of base cations and S042 plus N03 from 1964 to
1969, then a reversal in 1970 and a decreasing trend up to 1994. The slope of the phase with increasing
cation concentrations was steeper than the slope for the phase with decreasing cation concentrations.
Regional declines in base cation concentrations were measured in the Long-Term Monitoring project
from 1990 to 2000 for lakes in New England, the Adirondack Mountains, and the Upper Midwest (Figure
3-11). Lawrence et al. (1999) showed decreased concentrations of base cations at a rate that exceeded
decreases in (S042 plus N03 ) in Catskill Mountain streams from 1984 to 1997. Within western Virginia
and in Shenandoah National Park, concentrations of base cations in streams did not exhibit significant
temporal trends from 1988 to 2001, perhaps due to the influence of S adsorption to soil on stream water
S042 concentrations.
In some surface waters, interpretation of the effects of, and changes in, the concentration of base
cations and ANC is complicated by the influence of naturally occurring organic acidity. The base cation
surplus provides an approach for distinguishing between the effects of organic acidity and acidifying
deposition (Lawrence et al., 2007). Base cation surplus is defined as the difference between the summed
concentrations of base cations (Ca, Mg, Na, K) and strongly acidic inorganic anions (S042 , N03 .
chloride), plus an estimate of the strongly acidic organic anions estimated from dissolved organic C and
an assumed charge density. These strongly acidic organic anions are dissociated at low pH, and function
essentially as mineral acid anions in terms of their effect on ANC. The calculated base cation surplus is
similar to the calculated ANC, but explicitly accounts for strongly acidic organic acids. When the base
cation surplus is plotted against inorganic Al concentration, a distinct threshold for Al mobilization occurs
at a base cation surplus value that closely approximates 0, regardless of the DOC concentration (Figure 3-
12) (Lawrence et al., 2007). This threshold provides an unambiguous reference point for evaluating the
effects of acidifying deposition on mobilization of inorganic Al. To date, this calculated variable has only
been used in one large-scale assessment of acidifying deposition effects on surface waters (Lawrence
et al., 2007).
In summary, decreases in base cation concentrations in surface water in the eastern U.S. over the
past two to three decades are ubiquitous and are closely tied to trends in S042 concentrations. In most
regions, rates of decrease for base cations have been similar to those for S042 plus N03 . with the
exception of streams in western Virginia and in the Shenandoah National Park, which are affected by
decreases in S042 adsorption in soils. Decreasing trends of base cation concentrations do not necessarily
indicate further acidification or recovery of surface waters, but may indicate either lower base cation
leaching rates in soils or depletion of base cations from the soil system.
Surface Water pH
Surface water pH is a commonly used as an indicator of acidification. In addition, pH correlates
with other biologically important components of surface water acid-base chemistry, including ANC,
inorganic Al, Ca concentration, and organic acidity. Low pH can have direct toxic effects on aquatic
species (Driscoll et al., 2001b). Threshold pH levels for adverse biological effects have been summarized
for a variety of aquatic organisms (Baker et al., 1990b; Haines and Baker, 1986). Common reference
values for pH, below which adverse biological effects are anticipated, are 6.0, 5.5, and 5.0 (Driscoll et al.,
2001b). Only the most acid tolerant fish species can survive below pH 5.0, and Kretser et al. (1989) found
that half the total number of fish species that occur in the Adirondack region were present in lakes with
pH less than 6.0. A pH value of 6.0 is often considered the level below which biota are at risk from
acidification (Driscoll et al., 2001b). The effects of low pH are specific to the study organism and depend
also upon the concentrations of other chemicals in the water, notably inorganic Al and Ca. Species-
specific effects are discussed in more detail in Section 3.2.3.3.
Long-term past changes in surface water pH have been inferred for lakes in the Adirondacks
through paleolimnological studies (Charles et al., 1989; Cumming et al., 1992; 1994; Sullivan et al.,
3-34
-------
1990). These studies of algal remains in lake sediments for regionally representative Adirondack lakes
suggested that about 25 to 35% of the Adirondack lakes that are larger than 4 ha have acidified since
preindustrial time. An estimated 80% of the Adirondack lakes that had ambient pH less than 5.2 in the
mid-1980s were inferred to have experienced declines in pH and ANC since the previous century. About
30 to 45% of the lakes with ambient pH between 5.2 and 6.0 have also acidified. The results suggest that
the low-ANC lakes of the southwestern Adirondacks acidified more compared to other lakes in the
Adirondacks since preindustrial time.
Adirondack Survey
October 2003
Median DOC = 743 (jmol L
slope = -0.076
x intercept = 20
r2 = 0.68
Adirondack Survey
March 2004
O cP
Median DOC = 411 umol L
slope = -0.13
x intercept = 2.0
r2 = 0.71
Winnisook Stream
2001-2004
Median DOC = 211 pmol L"1
slope = -0.16
x intercept = 2.0
r2 = 0.69
-100 -50 0 50 100 150 200
Base Cation Surplus (|jeq L"1)
Source: Lawrence et al. (2007)
Figure 3-12. Concentration of inorganic Al in Adirondack streams as a function of the calculated base
cation surplus.
Additional information regarding long-term changes in surface water pH has been gained through
site-specific dynamic modeling. For example, by applying the PnET-BGC model to the long-term stream
chemistry record at the Hubbard Brook Experimental Forest, (Gbondo-Tugbawa et al., 2002) estimated
that past stream pH (circa 1850) was probably about 6.3, compared with values just above 5.0 in 2000
(Driscoll et al., 2007c).
3-35
-------
In recent decades, measurements of pH have been routinely collected in surface waters in the U.S.
where effects of acidifying deposition have been monitored, but there has been a long-standing reliance
on titrated ANC as the primary chemical measurement for evaluation of surface water acidification.
Overall, between 1980 and 2000 most studies reported slight increases in surface water pH, including
lakes in the Adirondack Mountains (rate variable) (Driscoll et al., 2007a) and southern New England
(0.002 pH units per year) (Warby et al., 2005), and streams in the Catskill/Poconos region of New York
and Pennsylvania (0.008 pH units per year) (Warby et al., 2005).
Through frequent monitoring from 1990 to 2000, Stoddard et al. (2003) found a decrease in
hydrogen ion (0.19 (ieq/L/yr) that was similar to the rate of change observed in the same Adirondack
lakes by Driscoll et al. (2007a) from 1992 to 2004 (0.18 (ieq/L/yr). Stoddard et al. (2003) also reported an
increase in the hydrogen ion concentration of Appalachian streams (0.08 (ieq/L/yr) and Upper Midwest
lakes (0.01 (ieq/L/yr); no trends were found in New England lakes in this study (see Figure 3-11).
In summary, increasing trends in pH (decreasing hydrogen ion concentration) in surface waters in
the northeastern U.S. were common through the 1990s up to 2004, but many exceptions occur, and
overall, the rates of change have been small. Driscoll et al. (2001a, 2001b; 2007c) attributed the limited
pH recovery of lakes in acid-sensitive regions to three factors: (1) the levels of acid-neutralizing base
cations in surface waters have decreased markedly because of the depletion of available base cations from
the soil, and to a lesser extent, a reduction in atmospheric inputs of base cations; (2) as forests mature,
their requirements for N decrease, and they are expected to increasingly lose N03 as forests develop; and
(3) sulfur has accumulated in the soil under previous conditions of high atmospheric S deposition and is
now being gradually released to surface water as S042 , even though S deposition has decreased.
Surface Water ANC
The most widely used measure of surface-water acidification is ANC, which is often determined by
Gran titration (titrated ANC). This measurement is the primary chemical indicator for assessing past
effects of acidifying deposition, and the recovery expected from decreasing atmospheric deposition
(Bulger et al., 2000; Stoddard et al., 2003). Titrated ANC is useful because it reflects the ANC of the
complete chemical system, which is typically decreased by acidic deposition in acid-sensitive landscapes.
Surface water pH is a common alternative to ANC as an indicator of acidification. However, at pH values
above about 6.0, pH is not a good indicator of either sensitivity to acidification or level of effect. In
addition, pH measurements (especially at these higher values) are sensitive to levels of dissolved C02 in
the water. In contrast, ANC is more stable and it reflects sensitivity and effects of acidification in a linear
fashion across the full range of ANC values. Therefore, ANC is the preferred indicator variable for
surface water acidification. Both titrated and calculated ANC values are commonly determined in studies
aimed at resource characterization or long-term monitoring.
Bulger et al. (1999) defined ANC response categories for brook trout in Virginia as less than zero
(chronic damage likely), 0 to 20 j^ieq/L (episodic damage likely), 20 to 50 j^ieq/L (likelihood of damage
not determined), and greater than 50 j^ieq/L (brook trout not sensitive). Baker et al. (1990b) used ANC
cutoffs of 0, 50, and 200 j^ieq/L for reporting on national lake and stream population estimates. ANC less
than 0 (ieq/L is of significance because waters at or below this level have limited capacity to neutralize
acid inputs. Surface waters with ANC <50 j^ieq/L have been termed "extremely acid sensitive" (Schindler,
1988), are prone to episodic acidification in some regions (DeWalle et al., 1987; Eshleman, 1988), and
may be susceptible to future chronic acidification at current or increased rates of acidifying deposition.
In assessing changes in surface water ANC, it is important to distinguish between acidic waters and
acidified waters. "Acidic" describes a condition that can be measured (i.e., Gran ANC less than or equal
to 0). It may be due either to the effects of acidifying deposition, or to other causes such as the presence
of organic acidity or the oxidation of chemically reduced S-containing minerals in the watershed.
"Acidified" refers to the consequences of the process of acidification (a decrease in ANC observed
through time). It does not require that the water body be acidic, and does not imply a particular cause for
3-36
-------
the change in chemistry. The term "anthropogenically acidified" implies that human activity was
responsible for the increase in acidity that occurred.
Some of the most detailed studies of ANC have been conducted in the Adirondack Mountains.
Model simulations suggested that none of the lakes in the Adirondack target lake population identified by
the U.S. EPA (EMAP) were chronically acidic or had ANC less than 20 j^ieq/L under preindustrial
conditions, but that by 1980 there were hundreds of such lakes (Table 3-5). Many lakes were estimated to
have had preindustrial ANC below 50 (ieq/L, but this estimate more than doubled by 1990. Based on
MAGIC model outputs extrapolated to the regional population of Adirondack lakes larger than 1 ha that
currently have ANC below 200 (ieq/L, maximum past acidification occurred by about 1980 or 1990, with
median ANC of the lake population of about 61 |_icq/L (reduced from a median of 92 j^ieq/L estimated for
the preindustrial period). Changes in ANC produced an increase in not only the percentage of lakes that
were chronically acidic, but also in those that were deemed likely to experience episodic acidification and
its associated short-term changes in water chemistry (Sullivan et al., 2008).
Table 3-5. Estimates of change in number and proportion of acidic surface waters in acid-sensitive
regions of the North and East, based on applying current rates of change in Gran ANC to past
estimates of population characteristics from probability surveys.
Results of Regional Survey
Results of Monitoring during 1990s
Region
Population
Size
Number
Acidic1
%
Acidic2
Time Period
of Estimate
Rate of
ANC
change3
Estimated
Number Acidic in
2000
% Acidic in
2000
% Change in
Number of
Acidic
Systems
New England
6,834 lakes
386 lakes
5.6%
1991-94
+0.3
374 lakes
5.5%
-2%
Adirondacks
1830 lakes
238 lakes
13.0%
1991-94
+0.8
149 lakes
8.1%
-38%
N.Appalachians
42,426 km
5,014 km
11.8%
1993-94
+0.7
3,600 km
8.5%
-28%
Ridge/Blue Ridge 32,687 km
1,634 km
5.0%
1987
-0.0
1,634 km
5.0%
0%
Upper Midwest
8,574 lakes
251 lakes
2.9%
1984
+1.0
80 lakes
0.9%
-68%
1 Number of lakes/length of streams with Gran ANC <0 in past probability survey by the U.S. EPA (data collected at "Time Period of Estimate," in column 5).
2 Percent of population (from Column 2) with Gran ANC <0 in past probability survey (data collected at "Time Period of Estimate," in column 5).
3 Based on regional trends presented in the Stoddard et al. (2003) report, in peq/L/yr, for the 1990s.
In other regions, responses to reduced levels of acidifying deposition required by the CAA and
other emissions control legislation were reported by Stoddard et al. (2003). They found tendencies during
the 1990s toward increasing surface water Gran ANC in all of the glaciated regions of the eastern U.S.
(i.e., New England, Adirondacks, and Northern Appalachian Plateau) and Upper Midwest, and decreasing
Gran ANC in the Ridge/Blue Ridge province. Changes in ANC were relatively modest compared with
observed reductions in S042 concentrations in surface waters. The regional increases in the Adirondacks,
Northern Appalachian Plateau, and Upper Midwest were statistically significant (Table 3-6). Median
increases of about +1 (ieq/L/yr in the Northern Appalachian Plateau, Adirondacks, and Upper Midwest
represent significant trends towards ecological recovery from acidification (Stoddard et al., 2003).
Estimated change in the number of acidic surface waters decreased during the 1990s in all regions
investigated by Stoddard et al. (2003), except the Ridge and Blue Ridge Provinces in the mid-
Appalachian Mountains (Table 3-5). For other regions, the change in number of acidic systems ranged
from -2% in New England to -68% in the Upper Midwest.
3-37
-------
Table 3-6. Regional trend results for long-term monitoring lakes and streams for the period 1990 through
2000.
Region
S042"
(jjeq/L/yr)
no3-
(jjeq/L/yr)
Base
Cations
[Ca + Mg]
(jjeq/L/yr)
Gran ANC
(jjeq/L/yr)
Hydrogen
(jjeq/L/yr)
DOC
(mg/L/yr)
Aluminum
(Mg/L/yr)
New England Lakes
-1.77**
+0.01 ns
-1.48**
+0.11ns
-0.01ns
+0.03*
+0.09ns
Adirondack Lakes
-2.26**
-0.47**
-2.29**
+1.03**
-0.19**
+0.06**
Appalachian Streams
-2.27*
-1.37**
-3.40**
+0.79*
-0.08*
+0.03ns
+0.56ns
Upper Midwest Lakes
-3.36**
+0.02ns
-1.42**
+1.07**
-0.01*
+0.06**
-0.06ns
Ridge/Blue Ridge Streams
+0.29**
-0.07**
-0.01ns
-0.07ns
+0.01 ns
NA
NA
Values are median slopes for the group of sites in each region. Regional trend not significant (p >0.05)* p <0.05 ** p <0.01. NA = insufficient data.
In summary, ANC is the most widely used measure of acid sensitivity, acidification, and chemical
recovery of surface waters in response to changes in acidifying deposition. Lake and stream ANC values
decreased throughout much of the 20th century in a large number of acid-sensitive lakes and streams
throughout the eastern U.S. This effect has been well documented in monitoring programs, paleolim-
nological studies, and model simulations. Since about 1990 the ANC of many affected lakes and streams
have shown some increase, but such increases have been relatively modest.
Surface Water Aluminum
The concentration of inorganic A1 in surface waters is an especially useful indicator of acidifying
deposition effects because (1) it is widely toxic, and (2) it generally does not leach from the terrestrial
soils to surface waters in the absence of acidifying deposition (Driscoll et al., 1988; Lawrence et al.,
2007) with exceptions such as acid mine drainage and relatively rare geologic deposits. Lawrence et al.
(2005) showed that strong organic acid anions can contribute to the mobilization of inorganic Al in
combination with S042 and N03 . but in the absence of geologic S, the presence of inorganic Al in
surface waters is an ambiguous indication of acidifying deposition effects.
Considerable work was done to define pH sensitivity ranges for a wide variety of aquatic
organisms, but when pH values fall below approximately 5.5, inorganic Al generally becomes the greater
health risk to biota. Although organically complexed Al (organic Al) can occur in surface waters as a
result of natural soil and hydrologic processes, this form of Al is not harmful to aquatic life (Gensemer
and Playle, 1999). Inorganic Al, however, has been found to be toxic to plant and animal species
throughout the food web (Gensemer and Playle, 1999).
Earlier studies demonstrated reduced growth and survival of various species of fish (Baker and
Schofield, 1982; Baker et al., 1996) at inorganic Al concentrations between approximately 2 and
7.5 (imol/L. Most recently, 20% mortality of young-of-the year brook trout was documented in situ during
a 30-day period with a median inorganic Al concentration of 2 (imol/L (Baldigo et al., 2007). This study
estimated that 90% mortality would occur over 30 days with a median inorganic Al concentration of
4.0 |_imol/L.
The development of methods to fractionate Al into organic and inorganic forms (Driscoll, 1984;
Sullivan et al., 1986) resulted in collection of a considerable amount of data on Al concentrations in
surface waters in the 1980s. However, most of this sampling was done either once or for a limited period
of time (Cronan et al., 1990; Driscoll and Newton, 1985; Driscoll et al., 1987b; Lawrence et al., 1987).
Available long-term trend information for inorganic Al is limited. In Adirondack lakes, inorganic Al
3-38
-------
concentrations decreased slightly (e.g., by 0.02 (iM/yrto 0.18 (iM/yr) (Driscoll etal., 2007a) or remained
unchanged between 1982 and 2004 (Driscoll et al., 2007a; Stoddard et al., 2003). There was no trend in
inorganic Al for this period in New England lakes, Appalachian streams, or Midwest lakes. Monthly
stream chemistry monitoring at the Hubbard Brook Experimental Forest showed decreases in inorganic Al
concentrations at four locations along the reference stream for the experimental forest from 1982 to 2000,
but no trends at two other locations along this stream (Palmer et al., 2004).
Most recently, Lawrence et al. (in press) found that 49 of 195 streams (25%) in the western
Adirondack region had inorganic Al concentrations above 2.0 (.iM during August base flow. Although
there is not a clear benchmark value above which inorganic Al is toxic to aquatic biota, 2 (.iM is generally
recognized as a reasonable threshold for biological effects at a variety of trophic levels (Baldigo et al.,
2007; Driscoll et al., 2001b).
In summary, inorganic Al is an important chemical indicator of the effects of acidifying deposition
on surface water. It has well-documented effects on aquatic biota at specific thresholds. Limited data
suggest that acid-sensitive regions of the northeastern U.S. have elevated inorganic Al concentrations
which have been induced by years of acidifying deposition and which pose a threat to aquatic life.
Concentrations have decreased slightly in some surface waters in the northeastern U.S. during the last two
decades in response to decreased levels of acidifying deposition.
Quantification of Acidification
Changes in the acid-base status of soils and drainage waters operate on different time scales. Most
temperate forest soils have high exchangeable acidity, and relatively small changes in the acidity of
precipitation input would not be expected to have a large effect on soil acidity (Krug and Frink, 1983;
Turner et al., 1990). Therefore, projected recovery of soil acid-base chemistry in response to future
decreases in acidifying deposition is expected to be limited (Gbondo-Tugbawa and Driscoll, 2002;
Sullivan et al., 2006b). In contrast, changes in the chemistry of drainage water in response to changes in
acidifying deposition can occur more rapidly. This is because drainage water can become acidified by the
leaching of a mobile acid anion such as S042 or N03 even if the acidity of the soil is not measurably
affected (Reuss and Johnson, 1986; Seip, 1980; Turner et al., 1990). In areas (including the northeastern
U.S.) where S adsorption on soils is minimal, and therefore where S042 is highly mobile, changes in S
deposition input have been shown to cause changes in the ANC and base cation concentrations in lakes
and streams over a time period of years to decades (c.f., Driscoll et al., 2003).
One way to quantify acidification dose-response relationships is to calculate the changes in various
ionic constituents in solution that occur in response to changes in mineral acid anion (S042 and N03 )
concentrations due to changes in acidifying deposition input. As S042 + N03 concentration increases or
decreases in solution, equivalent changes must also occur in the concentrations of other anions (i.e.,
bicarbonate |HC03 |. organic acid anion |RCOO |) or cations (i.e., H. Al", sum of base cations [SBC])
to maintain the charge balance. Typically, the largest counteracting change is in SBC. It is generally
assumed that most of the base cation change is due to Ca2+ and Mg2+. In acid-sensitive waters, additional
changes can occur in ANC (which can be expressed as |HC03 -H |). Aln+, and/or RCOO ) (Husar and
Sullivan, 1991). Henriksen (1984) presented evidence for Norwegian lakes, suggesting that base cation
release accounted for up to a maximum of 40% of the added mineral acid anions. This proportional
change in base cations relative to S042 or | S042 + N03 | is called the F-factor:
_ ASBC
~ A/SO] + NO3 ]
Subsequently, diatom reconstructions for Adirondack lakes suggested higher F-factors, generally
ranging from 0.4 to greater than 1.0 (Sullivan et al., 1990).
Sullivan and Eilers (1994) compiled available data on proportional changes in SBC, ANC, and
Aln+, relative to the observed or estimated change in | S042 + N03 |. Their analysis included: 1) measured
3-39
-------
short-term ( <20 year) changes in drainage water chemistry in response to ambient or experimental
increases or decreases in S deposition loading, 2) results of space-for-time substitution analyses, 3) results
of diatom inferences of past lake chemistry, and 4) MAGIC model hindcast and forecast simulations.
Results indicated a wide range in the estimated proportional changes in SBC as a percent of change in
| S042 + N03 |. Estimated F-factors generally ranged from about 0.5 to 1.0, although some watersheds in
Norway and in the western U.S. showed F-factors as low as about 0.25. The estimated proportional
change in ANC was typically less than 0.3, and change in Aln+ was smaller. Quantitative data were not
available for the organic acid anion response, but this response is also expected to generally be relatively
small. More recent PnET-BGC simulations for Adirondack lakes were in close agreement with the diatom
results in Figure 3-13 (Zhai et al., 2008), ranging from about 0.4 to 1.0.
1.5
1.0 -
o
.CO
0.5 -
0.0
-50
F= -10"5* ANC2+0.005*ANC+0.52
R2=0.62
T
50
100 150
ANC (neq/L)
200
250
Source: Zhai et al. (2008)
Figure 3-13. F-factors calculated from PnET-BGC model results for the period 1850 to 1980 as a function of
simulated ANC in 1980 for 44 EMAP lakes in the Adirondack region of New York.
Measured, modeled, and inferred changes in surface water chemistry in areas that have experienced
relatively short-term (less than three decades) changes in acid deposition loading are available from many
sources. Process model hindcasts and paleolimnological reconstructions of pre-industrial surface water
chemistry provide insight into the extent to which individual lakes and streams have acidified over the
longer term (Sullivan et al., 2006b). Evaluation of acidification effects from assessment of current
conditions is generally not helpful. This is because lakes and streams vary with respect to their expected
chemistry (i.e., ANC or pH) in the absence of acidifying deposition. In many regions, pre-disturbance lake
and stream ANC values below 50 (j,eq/L and pH below 6.0 appear to have been common; in some cases,
ANC below 0 and pH below 5.0 also occurred (Sullivan et al., 2006b). The common occurrence of lakes
and streams that were naturally low in ANC and pH before the advent of acidifying deposition is further
complicated by the fact that disturbances other than air pollution have contributed to further changes in
acid-base status, including both ANC and pH increases and decreases. Such disturbances have included
3-40
-------
logging, fire, forest regrowth, erosion, road-building, forest insect infestation and disease, and other land-
disturbing activities and events. Current surface water chemistry is a complex function of inherent
sensitivity (which to some degree was reflected in pre-disturbance chemistry), levels of acidifying
deposition (historic and current), and the effects of other disturbances.
Dose-response functions may in some cases be similar from water body to water body within a
defined region. For example, model simulations conducted for the NAPAP Integrated Assessment
(NAPAP 1991) reported by Sullivan et al. (1992) found that, although substantial variability was found in
projected future change in ANC among the modeled Adirondack watersheds, there was a highly
consistent relationship between median change in acidifying deposition and projected median change in
ANC over 50 years. Each 1 kg S/ha/yr change in future S deposition caused approximately a 3.5 (j,eq/L
change in simulated lakewater ANC (see Figure 3-14). Results of MAGIC model hindcast simulations
suggested that all of the Adirondack lakes modeled for NAPAP (1991) had acidified (decreased in pH or
ANC) since pre-industrial times. The median and range of estimated changes in ANC were -46 |icq/L and
-31 to -84 respectively. None of the lakes were inferred to have been acidic (ANC < 0) in pre-industrial
times. The minimum simulated pre-industrial values were pH 5.4 and ANC = 30 (j,eq/L (Sullivan and
Eilers, 1994).
Historical changes in Adirondack lakewater chemistry inferred from measurements in diatoms
suggested somewhat more conservative estimates of historical acidification. From these data, Sullivan
(1990) concluded:
¦ the "median" Adirondack lake had not acidified;
¦ acidification was generally limited to lakes that had ambient ANC during the 1980s less than
about 50 |icq/L (or pH less than about 6.0);
¦ approximately 15% of the Adirondack lakes were inferred to have acidified by more than 0.28 pH
units;
¦ the median historical acidification (expressed as A ANC - A Al,) of lakes that were acidic
(ANC <0) at the time of sampling was -37 |icq/L:
¦ approximately 3% of the Adirondack lakes were acidic in pre-industrial times, compared to 14%
in the 1980s.
3-41
-------
20 -
§
6 10 -
O
5
£ 0 —
6 ~10
I
S -20 -
I -
§ -30 -
-40
0
2
4
6
•2
-e
-4
-a
Median Change In Sulfur Deposition (kg/ha/yr)
Source: Sullivan et al. (1992)
Figure 3-14. Median and range of projected change in ANC (|jeq/L) of Adirondack lakes for 50-year MAGIC
simulations versus median future change in sulfur deposition (kg S/ha/yr) for each deposition
scenario. Points on each line correspond to -50%, -30%, -20%, 0%, +20%, +30% change from
current deposition.
These paleolimnological estimates suggested that Adirondack lakes that were acidic in the 1980s
had decreased a median of about 4 (j,eq/L in [ANCG - Al , | for each kg S/ha/yr change in S deposition. This
was slightly more than one-half of the median historical rate of acidification projected by MAGIC
(7 (ieq/L of calculated ANC for each kg S/ha/yr) for acidic Adirondack lakes (Sullivan et al., 1992).
In comparing estimates, derived from different approaches, of past and future changes in
Adirondack lakewater chemistry in response to acidifying deposition, it is important to consider several
factors (Sullivan et al., 1992):
¦ Chemical changes estimated from paleolimnology or monitoring data incorporate all influences
on the acid-base chemistry of lakewater, including land use, disturbances, and climatic
differences. Model estimates commonly include only postulated or estimated changes in
acidifying deposition as having influenced the lakewater chemistry. Because watershed
disturbances generally cause an increase in surface water ANC, they may partially explain the
more conservative estimates of diatom-inferred acidification compared with MAGIC model
estimates of acidification.
¦ The use of a process-based model for hindcasting requires assumptions regarding historical
deposition of all major ions. In addition to uncertainties regarding historical sulfur deposition
levels, base cation deposition has also likely changed by an unknown amount, and the degree to
which sulfur and base cation deposition have been coupled is unclear (Chen and Gomez, 1989;
Driscoll et al., 1989).
3-42
-------
¦ Organic acids may have exerted a greater influence on lakewater pH during pre-industrial times
than they do currently because DOC and organic acid anion concentrations may have decreased
in response to increased organic acid protonation and increased concentrations of A1 (Aimer
et al., 1974; Davis et al., 1985; Kingston and Birks, 1990; Krug and Frink, 1983). Data sets from
different points in time are often not directly comparable because of differences in ANC
definition or pH measurement. For example, the calculated ANC used by MAGIC differs from
titrated ANC (ANCG) used to calibrate paleolimnological transfer functions and reported in
surveys. The differences are due to the partially counteracting influences of Al and organic acids
on ANCg and their omission from calculated ANC. These differences can be appreciable for
acidic and low-ANC waters (Sullivan et al., 1989).
There is not a clear definable relationship between atmospheric S deposition and ecological effects.
A given amount of S deposition can cause a wide range of ecological responses, from no ecological effect
to varying levels of adverse ecological effect. These can include changes in community composition, loss
of sensitive species, reduced biological diversity, and altered ecosystem functions. Such effects can occur
in both aquatic and terrestrial ecosystems. The observed wide range of responses within and among
regions is attributable to varying ecosystem sensitivity. Some soils (notably in many watersheds in the
southeastern U.S.) have the capacity to adsorb substantial quantities of S, with essentially no acidification
of drainage water. Nevertheless, there is a finite limit to this S adsorption capacity, and under continual
high S deposition loading, the adsorptive capacity of soil will eventually become depleted.
In addition to differences in S adsorption capacity of soils, watersheds also differ in their sensitivity
to acidification effects as a consequence of differing sensitivity of the species that make up the local
biological community. Some species of fish, aquatic insects, and mollusks, for example, are highly
sensitive to adverse effects from low pH and high inorganic Al concentrations; others are less sensitive.
Finally, watersheds differ in the size of the soil base cation pool available to neutralize deposited mineral
acidity. Some watersheds have sufficient quantities of base cations in their soils such that drainage waters
will remain well buffered, even under relatively high S deposition loads, for many decades or longer.
Other watersheds had relatively low base cation supply during preindustrial times due to low weathering
rates of the underlying geology, and the base cation supply may have been further depleted by past
acidifying deposition.
As a consequence of these, and other, differences in sensitivity to S inputs, watersheds differ in the
extent to which they acidify in response to a given amount of S deposition and they also differ in the
extent to which that acidification translates to biological effects. Thus, one cannot specify a level of S
deposition that would be likely to cause adverse effects across the landscape. Sensitivity differs from
watershed to watershed.
Despite these differences in watershed sensitivity to acidification, it is possible to place bounds on
the amount of acidification that has occurred in response to a given change in S deposition. Such
quantitative estimates of acidification have been derived using watershed models of acidification
response. Modeling results summarized in Table 3-7 illustrate a wide range in the model estimates of past
acidification of acid-sensitive lakes and streams in the eastern U.S.
3-43
-------
Table 3-7. Model projections of surface water SO42" and associated ANC, shown as changes between
dates, for Adirondack and Shenandoah streams.
Region
Water
Bodies
Dates
Model
Pollution
Scenario
Change in
Median Surface
Water S042-
(Heq/L)
Change in
Median Surface
Water ANC
(Heq/L)
Reference
HINDCASTS
Adirondacks,
NY
38 lakes
1850 to
2003
PnET-
BGC
+72.9
-39.9
Zhai et al.
(2008)
Adirondacks,
NY
37 lakes
1850 to
1984
PnET-
BGC
+107
-77.8
Chen et al.
(2005)
Adirondacks,
NY
44 potentially
acid-sensitive
lakes
1850 to
1990
MAGIC
+77.8
-38.3
Sullivan etal.
¦ (2006b)
PnET-
BGC
+57.3
-29.5
FORWARD PROJECTIONS
Shenandoah
NP.VA
5 streams on
siliciclastic
bedrock
1990 to
2040
MAGIC
constant
deposition
+13
-11.6
Sullivan et al.
(2008)
mild reduction
-21
+6.2
medium
reduction
-23
+7.2
strong
reduction
-40
+24.2
very strong
reduction
-44
+27.2
Shenandoah
NP.VA
4 streams on
granitic
bedrock
1990 to
2040
MAGIC
constant
deposition
+22
-8
Sullivan et al.
(2008)
mild reduction
+11
-5
medium
reduction
+11
-5
strong
reduction
+3
-2
very strong
reduction
+2
-2
Shenandoah
NP.VA
5 streams on
basaltic
bedrock
1990 to
2040
MAGIC
constant
deposition
+33
-5
Sullivan et al.
(2008)
mild reduction
+12
0
medium
reduction
+11
+1
strong
reduction
-4
+5
3-44
-------
Region
Water
Bodies
Dates
Model
Pollution
Scenario
Change in
Median Surface
Water SO42-
(Heq/L)
Change in
Median Surface
Water ANC
(Heq/L)
Reference
very strong
reduction
-9
+6
Adirondacks,
NY
44 potentially
acid-sensitive
lakes
1990 to
2050
MAGIC
current and
expected
controls
-42.4
+5.89
Sullivan et al.
(2006b)
moderate
emission
controls
-58.9
+18.6
aggressive
emission
controls
-64.6
+22.6
Adirondacks,
NY
44 potentially
acid-sensitive
lakes
1990 to
2050
PnET-
BCG
current and
expected
controls
-18
-3.7
Sullivan et al.
(2006b)
moderate
emission
controls
-32.2
+1.8
aggressive
emission
controls
-38.3
+9.3
3.2.3.2. Summary of Biogeochemistry and Chemical Effects
The evidence is sufficient to infer a causal relationship between acidifying deposition and changes in
biogeochemistry related to aquatic ecosystems. The strongest evidence for a causal relationship comes
from studies of changes in surface water chemistry including concentrations of S042 . N03 . inorganic Al,
and Ca, surface water pH, sum of base cations, ANC, and base cation surplus. Surface water chemistry
integrates the sum of upstream soil and water processes and reflects the results of watershed-scale
terrestrial effects of S and N deposition including, N saturation, forest decline, and soil acidification
(Stoddard et al., 2003). In many cases, surface water chemistry indicates the effects of acidification on
biotic species and communities found in fresh water ecosystems.
Surface water chemistry can be examined and reported as chronic chemistry or episodic chemistry.
Chronic chemistry refers to annual average conditions, which are often represented as summer and fall
chemistry for lakes and as spring baseflow chemistry for streams. Episodic chemistry refers to conditions
during rainstorms or snowmelt when proportionately more drainage water is routed through upper soil
horizons, which tend to provide less neutralizing of atmospheric acidity as compared with deeper soil
horizons. Surface water chemistry has lower pH and ANC during storm runoff or snowmelt than during
baseflow conditions. One of the most important effects of acidifying deposition on surface water
chemistry is the short-term change in chemistry that is termed "episodic acidification." Some streams may
have chronic or average chemistry that is suitable for aquatic biota, but be subject to occasional episodic
acidification with lethal consequences. Episodic declines in pH and ANC are nearly ubiquitous in
3-45
-------
drainage waters throughout the eastern U.S. caused partly by acidifying deposition and partly by natural
processes.
Acidification effects on aquatic biota are often evaluated using measures of either inorganic A1 or
pH. ANC is also used because it is an indicator of acid base status (although ANC does not relate directly
to the health of biota). The usefulness of ANC lies in the association between ANC and the surface water
constituents that directly contribute to or ameliorate acidity-related stress, in particular pH, Ca, and
inorganic Al.
SO42", NO3", and Base Cations
Changes in water chemistry resulting from acidifying deposition typically include changes in
S042 . N03 . and base cation concentrations. Each plays an important role in the acid-base chemistry of
water, but none are directly toxic at concentrations commonly encountered in natural waters.
¦ S042 is the primary inorganic anion found in most acid sensitive waters. Continued decreases in
S emissions should cause further decreases in S042 concentrations in surface waters. However,
the rate of decrease in surface water S042 concentrations may be delayed as accumulated S
leaches from watershed soils in some regions of the country, especially the Blue Ridge
Mountains.
¦ The importance of N03 as an agent of acidification varies by region, but it is particularly
important during periods of high hydrologic flow from soils to streams such as those that occur
during snowmelt and rain events. The relationship between N deposition and surface water N03
concentration is complex and involves the terrestrial and aquatic cycling of N and other elements.
N03 contributes to the acidity of many lakes and streams in the eastern U.S., but there is no
apparent relationship between recent trends in N deposition and trends in N03 concentrations in
these surface waters (in contrast to observed responses for S deposition and S042
concentrations). This suggests that the time scales ofN saturation may be longer than previously
considered (e.g., centuries rather than decades). Nevertheless, long-term retention ofN deposited
in forested regions and consequent dampening of deposition effects on surface waters is unlikely
to continue indefinitely (Aber et al., 2003).
¦ Decreases in base cation concentrations in eastern U.S. surface waters over the past two to three
decades are ubiquitous and are closely tied to trends in S042 concentrations. Rates of base cation
depletion have been similar to those for S042 plus N03 in most areas (Shenandoah National
Park is a notable exception). Decreasing trends in base cation concentrations do not necessarily
indicate further acidification or recovery of surface waters, but may indicate either lower base
cation leaching rates in soils or depletion of base cations from the soil system.
pH, Acid Neutralizing Capacity, and Aluminum
Acidification of surface water causes changes in pH, ANC, and inorganic aluminum concentration.
Low pH and high inorganic Al concentration can be directly toxic to aquatic biota.
¦ The pH of freshwater streams and lakes is a common measure used to link acidification to
adverse effects on aquatic biota. Decreases in pH below values of 6.0 typically result in species
loss of benthic invertebrates, plankton species, and fish. A number of synoptic surveys indicated
loss of species diversity and absence of several fish species in the pH range of 5.0 to 5.5. If pH
decreases to lower values, there is a greater likelihood that more aquatic species could be lost
without replacement, resulting in decreased richness and diversity. (See the following discussion
on biota).
3-46
-------
¦ ANC reflects the difference between base cations and anions of strong acids in solution and is the
most widely used measure of acid sensitivity, acidification, and chemical recovery of surface
waters in response to changes in acidifying deposition. Acidic waters are defined as those having
ANC equal to or below zero. Waters with ANC of <50 j^ieq/L are considered "extremely acid
sensitive" (Schindler, 1988) and are vulnerable to episodic acidification (DeWalle et al., 1987;
Eshleman, 1988). Lake and stream ANC values decreased throughout much of the 20th century in
a large number of acid-sensitive lakes and streams throughout the eastern U.S. Since about 1990,
the ANC of many affected lakes and streams has increased slightly. The number of acidic surface
waters has decreased in some areas of the Northeast, but not in the mid-Appalachian Mountains.
¦ Dissolved inorganic Al is an important chemical indicator of the effects of acidifying deposition
on surface water because it is toxic to aquatic life and generally does not leach from soils in the
absence of acidification. When pH falls below approximately 5.5, inorganic Al generally
becomes a greater health risk to biota. Limited data suggest that acid-sensitive regions of the
northeastern U.S. have elevated inorganic Al concentrations in surface waters induced by years of
acidifying deposition, posing a threat to aquatic life. Concentrations have decreased slightly in
some surface waters in the northeastern U.S. during the last two decades in response to decreased
levels of acidifying deposition.
3.2.3.3. Biological Effects
Aquatic effects of acidification have been well studied in the U.S. and elsewhere at various trophic
levels. These studies indicate that aquatic biota have been affected by acidification at virtually all levels of
the food web in acid sensitive aquatic ecosystems. Effects have been most clearly documented for fish,
aquatic insects, other invertebrates, and algae.
Biological effects are primarily attributable to a combination of low pH and high inorganic Al
concentration. Such conditions occur more frequently during rainfall and snowmelt that cause high flows
of water and less commonly during low-flow conditions, except where chronic acidity conditions are
severe. Biological effects of episodes include reduced fish condition factor, changes in species
composition, and declines in aquatic species richness across multiple taxa, ecosystems and regions. These
conditions may also result in direct mortality as was shown from results of in situ bioassays (Van Sickle
et al., 1996). High concentrations of Ca, and to a lesser extent other base cations, can lessen the toxicity of
low pH and high inorganic Al concentration where they occur (Baker et al., 1990a).
Biological effects in aquatic ecosystems can be divided into two major categories: effects on health,
vigor, and reproductive success; and effects on biodiversity. The first category includes changes in
biological indicators such as individual condition factor and recruitment success. The latter can be
described by changes in species composition and taxonomic richness.
The following sections define concepts used to measure and evaluate acidification-related effects
on aquatic biota. Measures are presented of changes in health, vigor, and reproductive success; and
biodiversity for fish. Finally, the general effects literature is summarized for phytoplankton, zooplankton,
benthic invertebrates, amphibians, and fish-eating birds. Specific reference is made to the biological
indicators outlined above where such information exists.
Measures of Health, Vigor, and Reproductive Success
There are few measures of the effects of acidification on the health, vigor, and reproductive success
of aquatic species. Condition factor is one measure of sublethal acidification stress that has been used to
quantify effects of acidification on an individual fish. Condition factor is an index that describes the
relationship between fish weight and length. Expressed as fish weight/length3, multiplied by a scaling
constant, this index reflects potential depletion of stored energy (Dennis and Bulger, 1995; Everhart and
Youngs, 1981; Goede and Barton, 1990). Condition factor is interpreted as depletion of energy resources
3-47
-------
such as stored liver glycogen and body fat in response to increased stress at sublethal levels (Goede and
Barton, 1990). Fish with higher condition factor are more robust than fish having low condition factor.
Field studies have shown lower condition factor in fish found in more acidic streams (Dennis and Bulger,
1995).
Measures of Biodiversity
Species composition refers to the mix of species that are present in a particular ecosystem.
Acidification alters species composition in aquatic ecosystems. There are a number of species common to
many oligotrophic waters that are sensitive to acidifying deposition and that cannot survive, compete, or
reproduce in acidic waters. In response to small to moderate changes in acidity, acid-sensitive species are
often replaced by other more acid-tolerant species, resulting in changes in community composition, but
little or no change in total community abundance or biomass. The extent of alteration of surface water
biological community composition increases as surface waters become more acidic. There is also a
common pattern of lower community diversity with increased acidification.
One important tool that aids in the determination of effects on species composition is the ASI
developed by (Baker et al., 1990b). This index uses fish bioassay survival data to predict the probability
of fish survival expressed as a percent mortality. Separate ASI models were developed for tolerant,
intermediate, and sensitive fish species.
Taxonomic richness is a metric that is commonly used to quantify the effects of an environmental
stress on biota. It can be applied at various taxonomic levels. For example, the number of fish species
present in a lake or stream can be used as an index of acidification (Bulger et al., 1999). Similarly,
acidification effects on aquatic insects can be evaluated on the basis of the number of families or genera
of mayflies (order Ephemeroptera) (Sullivan et al., 2003). In the latter cases, the mayfly order was
selected for study because it includes a number of genera and species having varying degrees of
sensitivity to acidification.
Decreases in ANC and pH and increases in inorganic Al concentration have been shown to
contribute to declines in species richness and abundance of zooplankton, macroinvertebrates, and fish
(Keller and Gunn, 1995; Schindler et al., 1985). Species richness is positively correlated with pH and
ANC (Baker et al., 1990b; Rago and Wiener, 1986) primarily because of the elimination of acid-sensitive
species at lower pH and ANC (Schindler et al., 1985). Interpretation of species richness can be difficult
because more species tend to occur in larger lakes and streams as compared with smaller ones,
irrespective of acidity (Sullivan et al., 2003). Nevertheless, decreases in species richness have been
observed for all major trophic levels of aquatic organisms (Baker et al., 1990b), even after adjusting for
lake size (Frenette et al., 1986; Harvey and Lee, 1982; Matuszek and Beggs, 1988; Rago and Wiener,
1986; Schofield and Driscoll, 1987).
Health, Vigor, and Reproductive Success of Fish
Fish populations in acidified streams and lakes of Europe and North America have declined, and
some have been eliminated as a result of atmospheric deposition of acids and the resulting changes in
water quality (Baker et al., 1990b). A variety of water chemistry variables, including inorganic Al, DOC,
and Ca, along with the timing and magnitude of episodic fluctuations in toxic acid and inorganic Al
concentrations, are related to the degree to which surface water acidification influences fish survival in
natural systems (Baker et al., 1990b; Baldigo and Murdoch, 1997; Gagen et al., 1993; Siminon et al.,
1993; Van Sickle etal., 1996).
The effects of acidification on the health, vigor, and reproductive success are manifested through a
range of physiological effects on individual life stages and fish species. The primary mechanism for the
toxic effects of low pH and elevated inorganic Al on fish involves disruption of normal ion regulation at
the gill surface, resulting in increased rates of ion loss and inhibition of ion uptake (Bergman et al., 1988;
Leivestad, 1982; McWilliams and Potts, 1978; Wood and McDonald, 1987). Additional effects might
include disruption of Ca metabolism (Gunn and Noakes, 1987; Peterson and Martin-Robichaud, 1986;
3-48
-------
Reader et al., 1988); and decreased hatching success (Gunn and Noakes, 1987; Haya and Waiwood, 1981;
Peterson et al., 1980; Reader et al., 1988; Runn et al., 1977; Waiwood and Haya, 1983).
There is marked variability among species, and among life stages within species, in the specific
levels of pH and inorganic Al that produce measurable responses. In general, early life stages are more
sensitive to acidic conditions than the young-of-the-year, yearlings, and adults (Baker and Schofield,
1985; Baker et al., 1990b; Johnson et al., 1987). Also, small fish, especially swim-up fry, are probably
less mobile and less able to avoid exposure to adverse chemical conditions than the relatively larger adults
(Baker et al., 1996). Here, effects are described by life stage. Several studies have shown that the earliest
reproductive stages are highly sensitive to low pH. The processes of oogenesis and fertilization in fish are
especially sensitive (Havas et al., 1995; Muniz, 1991), most likely due to adverse effects on the female
spawner. For instance, Beamish (1976) reported that reduced serum and plasma Ca in female fish in
acidified Canadian lakes caused a higher probability of failure in producing viable eggs. Depletion of Ca
from bone and increased numbers of females with unshed eggs have also been linked to acidification at
this life stage (cf. Muniz, 1991; Rosseland, 1986).
After fertilization, the embryo seems to be susceptible to acidic waters throughout the whole period
of development, although periods shortly after fertilization and before hatching seem to be most critical
(Rosseland, 1986). The susceptibility of the embryo can be the result of direct exposure to elevated H
concentrations and to the toxic effects of inorganic Al at intermediate pH-values. Low pH in the
surrounding water also results in pH-depression inside the egg, leading to either prolongation of hatching
or to reduced hatching success (Rosseland, 1986). Eggs lying in gravel on stream and lake beds are, to
some extent, protected from exposure to rapid changes in pH (Gunn and Keller, 1984b; Lacroix, 1985).
Nevertheless, they can experience high mortality during periods of acid runoff, such as snowmelt (Gunn
and Keller, 1984a). Yellowstone cutthroat trout (O. c. bouveri) were exposed to 7-day pH depressions by
Farag et al. (1993). Of the four life stages studied, eggs were most sensitive to low pH. Eggs exposed for
seven days to pH 5.0 test water showed a statistically significant reduction in survival compared with
eggs exposed for seven days to pH 6.5 test water. Survival of alevin and swim-up larvae was significantly
reduced from near 100% at pH 6.5 to near 0% at pH 4.5. Intermediate pH values (6.0, 5.5) in all cases
showed reduced survival compared with the control (6.5), but not by statistically significant amounts
(p >0.05).
Emergent alevins show susceptibility to the adverse effects of inorganic Al and H that increases
with age (Baker and Schofield, 1982; Wood and McDonald, 1982). Rosseland (1986) indicated that this
increasing sensitivity results from changes that take place in the respiratory system. Shortly after
hatching, alevins still respire through their skin but gradually gills become the primary organ of gas and
ion exchange. Gills are the locus for interference of hydrogen ion and inorganic Al with iono-regulatory
exchange.
Woodward et al. (1989) exposed cutthroat trout (Oncorhynchus clarki) from the Snake River in
Wyoming to pH depressions from pH 4.5 to 6.5 in the laboratory and found that reductions in pH from 6.5
to 6.0 in low-Ca water (70 j^icq/L) did not affect survival, but did reduce growth of swim-up larvae. The
eggs, alevin, and swim-up larval stages showed significantly higher mortality at pH 4.5 than at pH 6.5.
Mortality was also higher at pH 5.0 than at pH 6.5, but only statistically higher for eggs. The authors
concluded that the threshold for effects of acidity on greenback cutthroat trout in the absence of inorganic
Al was pH 5.0 (Woodward, 1991).
In juvenile, young-of-year and adult fish there is an energy cost in maintaining physiological
homeostasis; the calories used to respond to stress are a part of the fish's total energy budget and are
unavailable for other functions, such as growth and reproduction (Schreck, 1981, 1982; Wedemeyer et al.,
1990). Observed differences in condition factor may occur because maintenance of internal chemistry in
the more acidic streams would require energy that otherwise would be available for growth and weight
gain (Dennis and Bulger, 1999; Sullivan et al., 2003). The energy costs to fish for active iono-
osmoregulation can be substantial (Bulger, 1986; Farmer and Beamish, 1969).
Prominent physiological disturbances to fish exposed to acid waters are iono- and osmoregulatory
failure, acid-base regulatory failure, and respiratory and circulatory failure. Most of these effects can be
3-49
-------
directly attributed to effects on gill function or structure. The acute toxicity of low pH in acidic waters
results in the loss of Ca from important binding sites in the gill epithelium, which reduces the ability of
the gill to control membrane permeability (Exley and Phillips, 1988; Havas, 1986; McDonald, 1983). A1
has been shown to accumulate on the gill surface when fish are exposed to water having high inorganic A1
concentration.
Cumulative sublethal physiological effects can be expressed by changes in condition factor.
Condition factor has been developed and applied mainly for blacknose dace. This fish species is widely
distributed in Appalachian Mountain streams and is moderately tolerant of low pH and ANC, relative to
other fish species in the region. However, the condition factor concept is probably applicable to other
species as well. Condition factor may be a useful metric for many species in aquatic ecosystems that are
only marginally affected by acidification. Bulger et al. (1999) observed a positive relationship between
dace condition factor and pH in streams in Shenandoah National Park. Dennis and Bulger (1995) found a
reduction in the condition factor for blacknose dace in waters near pH 6.0. The four populations with the
lowest condition factor had mean habitat pH values within or below the range of critical pH values at
which Baker and Christensen (1991) estimated that negative population effects are likely for the species.
The mean condition factor of fish from the study stream with the lowest ANC was about 20% lower than
that of the fish in best condition. In addition to effects on blacknose dace, condition factor, reduced
growth rates have been also attributed to acid stress in a number of other fish species, including Atlantic
salmon (Salmo salar), Chinook salmon (Oncorhynchus tshawytscha), lake trout (Salvelinus namaycush),
rainbow trout (Oncorhynchus mykiss), brook trout, brown trout (Salmo trutta), and arctic char (Salvelinus
alpines) (Baker etal., 1990b).
In summary, some studies have been conducted on changes in the health, vigor, and reproductive
success of fish exposed to water having low pH and high inorganic Al concentration. Blacknose dace
have been most thoroughly studied regarding the sublethal effects of acidity on fish condition. Effects
tend to vary by life stage; early life stages tend to be particularly sensitive. Adverse effects often involve
disruption of gill function, partly due to Al toxicity.
Fish Biodiversity
Biodiversity loss is a predictable consequence of acidification and there are abundant examples of
this in North America and Europe, mostly focused on fish (Bulger et al., 2000). Population-level fish
response to acidification is primarily through recruitment failure, a result of increased mortality of early
life stages or indirect effects through the food chain (loss of prey species). Changes in inorganic Al, pH,
and Ca most likely have the greatest influence on fish community structure. These changes in water
chemistry can alter species composition and species richness, both of which are components of
biodiversity.
By 1990, it was well established that changes in pH in the range of 4.0 to 6.5 could cause
significant adverse biological effects on fish community composition. As described above, the toxicity of
low pH was, in most cases, the result of impaired body salt regulation. Decreased water pH inhibited the
active uptake of Na+ and CI and stimulated the passive loss of these ions from the bloodstream (Baker
et al., 1990b). Species vary in terms of their sensitivity to such disruptions of physiological condition.
The response of fish to pH, ANC, and inorganic Al is not uniform across species. A number of
synoptic surveys indicated loss of species diversity and absence of several fish species in the pH range of
5.0 to 5.5. If pH is lower, there is a greater likelihood that more fish species could be lost without
replacement, resulting in decreased richness and diversity. In general, populations of salmonids are not
found at pH levels less than 5.0, and smallmouth bass (Micropterus dolomieu) populations are usually not
found at pH values less than about 5.2 to 5.5. Bioassay experiments using brook trout eggs and fry have
demonstrated greater mortality in chronically acidic stream water as compared to water having higher
ANC.
The ASI is an index of acidification that uses fish bioassay survival data fitted to a regression
model of exposure to water chemistry (pH, Al, and Ca) to predict the probability of fish survival.
3-50
-------
Approximate ASI reference levels were reported by (Baker et al., 1990b) for various fish species, based
on logistic regression of fish presence as a function of the sensitive, intermediate, and tolerant ASI values
for brown bullhead (Ameiurus nebulosus), brook trout, lake trout, and common shiner (Luxilus cornutus).
Fish species richness is an important indicator of acidification response, in part because the public
tends to place relatively high value on fisheries. As discussed in the previous section, lakes and streams
having pH below about 5.0 or ANC below about 0 generally do not support fish. There is often a positive
relationship between pH and number of fish species, at least for pH values between about 5.0 and 6.5, or
ANC values between about 0 and 50 to 100 j^ieq/L (Bulger et al., 1999; Cosby et al., 2006; Sullivan et al.,
2006a). Such observed relationships are complicated, however, by the tendency for smaller lakes and
streams, having smaller watersheds, to also support fewer fish species, irrespective of acid-base
chemistry. This pattern may be due to a decrease in the number of available niches as stream or lake size
decreases. Nevertheless, fish species richness is relatively easily determined and is one of the most useful
indicators of biological effects of surface water acidification.
Some of the most in-depth studies of the effects of acid stress on fish species richness have been
conducted in the streams in Shenandoah National Park, Virginia and the lakes in the Adirondack
Mountains, New York. These regions are examined in detail below. However, note that effects on fish
species richness have also been documented in acid-sensitive streams of the Catskill Mountains of
southeastern New York (Stoddard and Murdoch, 1991) and the Appalachian Mountains from
Pennsylvania to Tennessee and South Carolina ((SAMAB), USDA 1996; Bulger et al., 1999; 2000).
The Shenandoah National Park Fish in Sensitive Habitats (FISH) Project evaluated the effects of
streamwater acidification on fish communities in streams in Shenandoah National Park (Bulger et al.,
1995; Dennis and Bulger, 1995; Dennis et al., 1995; MacAvoy and Bulger, 1995). A statistically robust
relationship between stream ANC and fish species richness was documented. Numbers of fish species
were compared among 13 Shenandoah National Park streams spanning a range of pH and ANC
conditions. There was a highly significant (p <0.0001) relationship between stream acid-base status
(during the 7-year period of record) and fish species richness among the 13 streams. The streams with the
lowest ANC hosted the fewest species (Figure 3-15). The 3-year FISH study of stream acidification
demonstrated negative effects on fish from both chronic and episodic acidification (Bulger et al., 1999).
Bulger et al. (1999) concluded that the most important cause of the observed decline in species richness
with decreasing ANC was acid stress from acidification. However, an additional causal factor may have
been a decrease in the number of available aquatic niches when moving from downstream locations
(which are seldom low in pH and ANC) to upstream locations (which are often low in pH and ANC in
this region)(Sullivan et al., 2003).
South of Shenandoah National Park, the effects of surface water acidification on fish species
richness have been studied in some detail in the St. Marys River in Virginia. Fish species richness was
closely associated with surface water acid-base chemistry. The number of fish species in the St. Marys
River within the wilderness declined from 12 in 1976 to 4 in 1998. Three of the four species present in
1998 (brook trout, blacknose dace, fantail darter [Etheostoma flabellare] are tolerant of low pH and are
typically the only fish species present in streams having similar levels of acidity in nearby Shenandoah
National Park (Bulger et al., 1999).
3-51
-------
Dynamic water chemistry model projections have been combined with biological dose-response
relationships to estimate declines in fish species richness with acidification. A relationship derived from
the Shenandoah National Park data was used by Sullivan et al. (2003), along with stream ANC values
predicted by the MAGIC model to provide estimates of the expected number of fish species in each of the
modeled streams for the past, present and future chemical conditions simulated for each stream. Results
suggest that historical loss of species had been greatest in the streams located on the most sensitive
geological class (siliciclastic bedrock; 1.6 species lost), w ith fewer lost species 011 granitic bedrock and
basaltic bedrock (average of 0.4 species lost).
TE
i
$
i
£
LA 1 1
1
1
a m
« ~
M' * *
it *
1 *
-
U I
It Mil ( f f
1 1 1 1 1 1 i i i i i
-25 0 25 50 75 100 125 150 175 200 225 250 275 300
Average ANC (Meq'L)
Source: Redrawn from Bulger et al. (1999)
Figure 3-15. Number of fish species as a function of mean stream ANC among 13 streams in Shenandoah
National Park, Virginia. Values of ANC are means based on quarterly measurements, 1987—
1994. The regression analysis showed a highly significant relationship (p <0.0001) between
mean stream ANC and number offish species. Streams having ANC consistently <50 peq/L
had three or fewer species.
In the Adirondack Mountains, lakewater acidification and the associated elevated concentrations of
inorganic Al have adversely affected fish populations and communities in sensitive areas (Baker and
Schofield, 1982; Baker et al., 1990a; Johnson et al., 1987; Schofield and Driscoll, 1987; Simmon et al.,
1993). Of the 53 fish species recorded in Adirondack lakes by the Adirondack Lakes Survey Corporation,
about half (26 species) were absent from lakes with pH below 6.0. Among the absent species were several
important recreational species (Baker et al., 1990a), plus ecologically important minnows that serve as
forage for sport fish. Fully 346 of 1,469 lakes surveyed supported no fish at all at the time of the survey.
These lakes were significantly lower in pH, dissolved Ca, and ANC, and had higher concentrations of
inorganic Al than lakes hosting one or more species of fish (Gallagher and Baker, 1990). Among lakes
with fish, there was an unambiguous relationship between the number of fish species and lake pH,
ranging from about one species per lake for lakes having pH less than 4.5 to about six species per lake for
lakes having pH higher than 6.5 (Baker et al., 1990a; Dnscoll et al., 2001b).
High-elevation lakes are more likely to be Ashless than larger lakes at low elevation (Gallagher and
Baker, 1990). This observation has been attributed to the fact that high elevation lakes tend to have poor
access for fish immigration, poor fish spawning substrate, or low pH, or they may be susceptible to
3-52
-------
periodic winter kills. Small, high-elevation Adirondack lakes with fish also had significantly higher pH
compared with Ashless high-elevation lakes; acidity is likely to play an important role in the absences of
fish from such lakes (Driscoll et al., 2001a).
Sullivan et al. (2006b) developed a relationship between fish species richness and ANC class for
Adirondack lakes. Under chronically acidic conditions (summer index or annual average ANC <0 (ieq/L),
Adirondack lakes are generally without fish. There was a marked increase in mean species richness with
increases in ANC up to values of approximately 50 to 100 j^ieq/L (Figure 3-16). The asymptote for the fish
species equation was 5.7 species. This analysis suggests that there could be loss of fish species with
decreases in ANC below approximately 50 to 100 j^icq/L. The response functions from Shenandoah
National Park (Figure 3-15) and the Adirondack Mountains (Figure 3-16) are generally similar at low
ANC values, below about 100 |icq/L. Fish species richness was somewhat higher in Shenandoah National
Park at higher ANC values. The reasons for this difference are not known.
-200
-100
100
200
300
400
500
ANC(petj/L)
Source: Sullivan et al. (2006b).
Figure 3-16. Number of fish species per lake versus acidity status, expressed as ANC, for Adirondack
lakes. The data are presented as mean (filled circles) and range (bars) of species richness
within 10 |jeq/L ANC categories, based on data collected by the Adirondack Lakes Survey
Corporation.
The absence of fish from a given lake or stream in an area that experiences surface water
acidification does not necessarily imply that acidification is responsible for the absence of fish. For
example, results of fisheries research in the Adirondacks have indicated that many Adirondack lakes
always had marginal spawning habitat for brook trout (Schofield, 1993). However, multivariate regression
of the presence or absence of brook trout in Adirondack waters produced a ranking of factors that
appeared to influence the presence of brook trout when biological factors (stocking, presence of
associated species, presence of competitors) were excluded from the analysis. Among contributing
factors, including silica (Si), ANC, DOC, substrate type, and distance to the nearest road, pH ranked first
as a predictor of brook trout presence. The results of this analysis supported the conclusion that 1990
levels of pH and related variables restricted the distribution of fish in some Adirondack lakes.
3-53
-------
In summary, acidic conditions characterized by low pH, low ANC, and high inorganic A1 exert
considerable influence on the fish species composition of sensitive surface waters, particularly in the
eastern U.S. Low pH and ANC, and high inorganic A1 concentrations, contribute to loss of the most acid-
sensitive fish species. Species richness is a common indicator used to reflect the effects of water
acidification on aquatic biota. This index is most often applied to fish. Few or no fish species are found in
lakes and streams that have very low ANC (near zero) and low pH (near 5.0). The number of fish species
generally increases at higher ANC and pH values. This relationship is complicated, to some extent, by the
tendency of smaller lakes and streams (which are more likely to have low ANC and pH) to host fewer fish
species, regardless of acid-base chemistry. Nevertheless, available data strongly suggest that acid stress is
a major factor governing the observed relationship between fish species richness and surface water
acidity.
3.2.3.4. Summary of Biological Effects
The evidence is sufficient to infer a causal relationship between acidifying deposition and changes in
aquatic biota. The strongest evidence for a causal relationship comes from studies of aquatic systems
exposed to elevated levels of acidifying deposition that support fewer species of fishes,
macroinvertebrates, and diatoms. Although there are few studies of the response of higher trophic levels
to pH changes resulting from acidifying deposition, piscivorous birds are known to be affected by
acidifying deposition (see Table 3-9). Consistent and coherent evidence from multiple species and studies
shows that acidification can result in the loss of acid-sensitive species, and more species are lost with
greater acidification. Biological effects are linked to changes in water chemistry including ANC, pH, and
inorganic Al. Decreases in ANC and pH and increases in inorganic A1 concentration contribute to declines
in taxonomic richness of zooplankton, macroinvertebrates, and fish. Chemical changes can occur over
both long- and short-term time scales, with additional effects on biological systems. Short-term (hours or
days) episodic changes in water chemistry can have biological effects, including reduced fish condition
factor, changes in species composition, and declines in aquatic species richness across multiple taxa,
ecosystems and regions.
Species
¦ High levels of acidification (to pH values below 5) virtually eliminate all mayflies, crustaceans,
and mollusks from some streams.
¦ In general, populations of salmonid fish are not found at pH levels less than 5.0, and smallmouth
bass (Micropterus dolomieu) populations are usually not found at pH values less than 5.5 to 5.2.
¦ Twenty percent mortality of young-of-the year brook trout was documented during a 30-day
period with a median inorganic Al concentration of 2 (imol/L (Baldigo et al., 2007). It was
estimated that 90% mortality would occur over 30 days with a median inorganic Al concentration
of 4.0 (imol/L.
Community
Community-level effects were observed in two well-studied ares, the Adirondacks and the
Shenandoah National Park, where taxonomic richness is lower in lakes and streams having low ANC and
pH.
¦ Decreases in pH and increases in inorganic Al concentrations have reduced the species richness
of plankton, invertebrates, and fish in acid-affected surface waters.
3-54
-------
Invertebrate taxathat are most sensitive to acidification include mayflies, amphipods, snails, and
clams.
¦ In the Adirondacks, a positive relationship exists between the pH and ANC in lakes and the
number of fish species present in those lakes. A number of synoptic surveys indicated suggested
loss of species diversity and absence of several sensitive fish species in the pH range of 5.0 to 6.0.
¦ In Shenandoah National Park streams, the fish species richness decreased with decreasing stream
ANC. On average, richness is lower by one fish species for every 21 j^ieq/L decrease in ANC.
¦ Short-term episodes of acidification are particularly harmful to aquatic biota. Early life stages are
more sensitive to acidic conditions than the young-of-the-year, yearlings, and adults. Episodes are
most likely to affect biota if the water had pre-episode pH above 5.5 and minimum pH during the
episode of less than 5.0. Episodic acidification can have long-term adverse effects on fish
populations.
3.2.4. Most Sensitive and Most Affected Ecosystems and Regions
3.2.4.1. Characteristics of Sensitive Ecosystems
The principal factor governing the sensitivity of terrestrial and aquatic ecosystems to acidification
from S and N deposition is geology (particularly surficial geology). Geologic formations having low base
cation supply generally underlie the watersheds of acid-sensitive lakes and streams. Bedrock geology has
been used in numerous acidification studies (Bricker and Rice, 1989; Stauffer, 1990; Stauffer and
Wittchen, 1991; Vertucci and Eilers, 1993; Sullivan et al., 2007b). Other factors contribute to the
sensitivity of soils and surface waters to acidifying deposition, including topography, soil chemistry, land
use, and hydrologic flowpath.
Several studies have confirmed the importance of geology in regulating terrestrial and aquatic
ecosystem sensitivity to acidification, and highlighted other key factors responsible for terrestrial and
aquatic sensitivity to acidifying deposition throughout the southeastern U.S. Sensitive terrestrial
ecosystems include high-elevation spruce-fir forests dominated by relatively nonreactive bedrock in
which base cation production via weathering is limited (Elwood et al., 1991). Soils in such areas tend to
have thick organic horizons, high organic matter content in the mineral horizons, and low pH (Joslin
et al., 1992). Because of the largely nonreactive bedrock, base-poor litter and organic acid anions
produced by the conifers, high precipitation, and high leaching rates, soil base saturation in these high-
elevation forests tends to be below about 10% and the soil cation exchange complex is generally
dominated by Al (Eagar et al., 1996; Johnson and Fernandez, 199).
Galloway (1996) further attributed forest soil sensitivity to acidification in the southeastern U.S. to
atmospheric deposition level, soil age, weathering rate, and S adsorption capacity. Moncoulon et al.
(2004) suggested that forest ecosystem sensitivity to acidification varies mainly with weathering rate. In a
review of 241 ecosystem types in France (classified by pedologic and geologic characteristics), the
ecosystems most susceptible to acidification were those with low weathering rates and thus limited
buffering capacity (Moncoulon et al., 2004).
In hardwood forests, species nutrient needs, soil conditions, and additional stressors work together
to determine sensitivity to acidifying deposition. Stand age and successional stage also can affect the
susceptibility of hardwood forests to acidification effects. In northeastern hardwood forests, older stands
exhibit greater potential for Ca depletion in response to acidifying deposition than younger stands. Thus,
with the successional change from pin cherry (Prunus pensylvanica), striped maple (Acer pensylvanicum),
white ash (Fraxinus americana), yellow birch and white birch (Betulapapyrifera) in younger stands to
3-55
-------
beech and red maple in older stands, there is an increase in sensitivity to acidification (Hamburg et al.,
2003).
Land use influences watershed sensitivity to acidification mainly through disturbance and
consequent exposure of S-bearing minerals to oxidation, loss of base cations through erosion and timber
harvesting, and change in N status of the forest through timber management. Each of these types of
activity can influence the relative availability of mobile mineral acid anions (S042 , N03 ) in soil solution
and base cations (Ca, Mg, K, Na) on the soil ion exchange sites and in drainage water.
The movement of water through the soils into a lake or stream, and the interchange between
drainage water and the soils and sediments, strongly regulate the type and degree of watershed response
to acidic inputs (Sullivan, 2000a). Surface waters in the same setting can have different sensitivities to
acidification depending on the relative contributions of near-surface drainage water and deeper
groundwater (Chen et al., 1984; Driscoll et al., 1991; Eilers et al., 1983).
Movement of a strong acid anion, such as S042 or N03 , through an acidic soil can mobilize H
and Al3+ because these cations are available on soil exchange sites. There is no time lag in this exchange
reaction and it is instantly reversible if input of strong acid anions is ceased (Turner et al., 1991). It is
necessary, however, for appreciable mobilization of H and Al3+ that the soil be acidic, either naturally or
because of soil acidification from acidifying deposition.
In summary, lakes and streams in the U.S. that are sensitive to episodic and chronic acidification in
response to SOx, and to a lesser extent NOx, deposition tend to occur at relatively high elevation in areas
that have base-poor bedrock, high relief, and shallow soils. For example, in the Southern Appalachian
region (Sullivan et al., 2002a, 2007b) determined that underlying bedrock geology dominated by
sandstone or related rock types and elevations greater than 1000 m (3250 ft) could be used to identify
landscapes in the region most likely to contain acidic streams.
3.2.4.2. Extent and Distribution of Sensitive Ecosystems
Surface Waters
Several regions of the U.S. contain appreciable numbers of lakes and streams with low ANC (less
than about 50 (ieq/L), including portions of the Northeast (especially New England, the Adirondacks, and
the Catskill Mountains), Southeast (the Appalachian Mountains and northern Florida), Upper Midwest,
and western U.S. (Charles, 1991). The Adirondack and Appalachian Mountains, and to a lesser extent the
Upper Midwest, include many acidified surface waters that have been affected by acidifying deposition.
Portions of northern Florida also contain many acidic and low-ANC lakes and streams, although the role
of acidifying deposition in these areas is less clear. The western U.S. contains many of the surface waters
most susceptible to potential acidification effects, but with the exception of the Los Angeles Basin and
surrounding areas, the levels of acidifying deposition in the West are low in most areas, acidic surface
waters are rare, and the extent of chronic surface water acidification that has occurred to date has likely
been very limited.
Several national assessments were conducted to estimate the distribution and extent of surface
water acidity in the U.S. During summer baseflow of 2004, the U.S. EPA conducted a National Wadeable
Stream Assessment (WSA) survey of 1,392 randomly selected sites across the conterminous 48 U.S. to
assess the ecological condition of wadeable streams (U.S. EPA, 2006d). Because this sampling was
conducted during baseflow in the summer (which exhibits the least acidic conditions of the year), only the
most chronically acidified streams were identified as acidic. Therefore, the extent of potential seasonal
acidification was underestimated by this approach (Lawrence et al., 2008). Overall, less than 1% of the
1,020,000 km of stream in the target population (based on the 1:100,000-scale U.S. Geological Survey
(USGS) map blue line network) was acidic due to acidifying deposition. No acidic streams were observed
in the Mountainous West, Xeric West, Upper Midwest, Northern Plains, Southern Plains, or Temperate
Plains ecoregions. Acidic streams attributable to acidifying deposition were found in the Northern
Appalachians (2.8% of 96,100 km of stream), and the Southern Appalachians (1.8% of287,000 km). Very
3-56
-------
low ANC (0-25 (ieq/L) streams likely exposed to episodic acidification were found in the Northern
Appalachians (2.7% of 96,100 km of stream), the Coastal Plain (6.3% of 119,000 km), and the
Mountainous West (0.6% of 204,000 km).
Even though the WSA had over 1,300 sample sites, it was still a very coarse sample of the nation's
streams with respect to acidifying deposition effects, which are only observed in spatially restricted
subpopulations. More precise survey estimates of the effects of surface water acidification were made in
the National Surface Water Survey (NSWS) in the mid 1980s. By statistically selecting representative
lakes and streams in each surveyed region, the NSWS estimated chemical conditions of 28,300 lakes and
56,000 stream reaches (Baker et al., 1990b). The NSWS concluded that 4.2% of lakes larger than 4 ha and
2.7% of stream segments in the acid-sensitive regions of the eastern U.S. were acidic. The NSWS
documented the status and extent of surface water acid-base chemistry during probability surveys of lakes
and streams conducted from 1984 through 1988 in the major acid sensitive regions of the U.S. (Kaufmann
et al., 1988; Landers et al., 1987; Linthurst et al., 1986a).
The stream component of the NSWS, the National Stream Survey (NSS), was focused in the
northern and southern Appalachians and Coastal Plain of the eastern U.S. (Kaufmann et al., 1991). The
NSS included 500 stream reaches selected from 1:250,000 scale USGS topographic maps using a
systematic, randomized sample. Study reaches were sampled at both the upstream and downstream end of
each selected reach. Population estimates were made for chemistry at both reach ends and for stream
length by interpolating chemical results between reach ends.
Overall, out of the estimated 57,000 stream reaches in the NSS, after excluding streams acidic due
to acid mine drainage, 6.2% of the upstream and 2.3% of the downstream reach ends were acidic during
spring baseflow (Kaufmann et al., 1991). After interpolation, this corresponded to 2.7% of the
201,000 km of stream in the study region. In acidic and low-ANC NSS reaches, ANC usually increased
with downstream distance. Acidic (ANC < 0) streams were located in the highlands of the Mid-Atlantic
Region (southern New York to southern Virginia, 2320 km), in coastal lowlands of the Mid-Atlantic
(2530 km), and in Florida (461 km). Acidic streams were rare (less than 1%) in the highlands of the
Southeast and Piedmont. Inorganic monomeric Al concentrations were highest in acidic streams of the
Mid-Atlantic Highlands, where over 70% of the acidic streams had inorganic Al greater than 3.7 (.iM
(100 (ig/L), a concentration above which deleterious biological effects have frequently been reported.
Anion composition of the NSS stream samples was examined to evaluate the most probable
sources of stream acidity in acidic and low-ANC sites (Baker et al., 1991b; Herlihy et al., 1991). Acidic
streams that had minimal organic influence (organic anions constituted less than 10% of total anions), and
S042 and N03 concentrations indicative of evaporative concentration of atmospheric deposition, were
classified as acidic due to acidifying deposition. These acidic streams were located in small (<30 km2)
forested watersheds in the Mid-Atlantic Highlands (an estimated 1980 km of stream length) and in the
Mid-Atlantic Coastal Plain (1250 km). Acidic streams affected primarily by acidifying deposition but also
influenced by naturally occurring organic anions accounted for another 1210 km of acidic stream length
and were mainly located in the New Jersey Pine Barrens, plateau tops in the Mid-Atlantic and Southeast
Highlands, and the Florida Panhandle. The total length of streams that were identified as acidic due to
acid mine drainage in the NSS (4590 km) was about the same as the total length of acidic streams likely
affected by acidifying deposition (4455 km). Acidic streams whose acid anion composition was
dominated by organics were mainly located in Florida and the Mid-Atlantic Coastal Plain. In Florida,
most of the acidic streams were organic-dominated, whereas about half of the acidic streams in the Mid-
Atlantic Coastal Plain were organic-dominated. Organic-dominated acidic streams were not observed in
the Mid-Atlantic or Southeast Highlands.
Stoddard et al. (2003) presented a map of acid-sensitive regions of the eastern U.S. where lakes and
streams occur that are likely to be affected by acidifying deposition (Figure 3-17). The map shows
considerable overlap with the areas of high interest identified by Baker et al. (1990b). Surface waters in
most other regions of the U.S. are not sensitive to the effects of acidification due largely to the nature of
the local geology (Stoddard et al., 2003). An exception is the region surrounding the Los Angeles Basin,
3-57
-------
which receives high N deposition ( >20 kg N/ha/yr in some areas) and includes streams with very high
NOb" concentrations ( >50 (.ieq/L; Bytnerowicz and Fenn, 1996; Fenn and Poth, 1999, 2001).
Acid Sensitive Regions of the Northern and Eastern United States
New England
( Northern
y /Appalachian PI
Source: Stoddard et al. (2003).
Figure 3-17, Regions of the northern and eastern U.S. that contain appreciable numbers of lakes and
streams that are sensitive to acidification from acidifying deposition.
In addition to the large water chemistry databases developed by the U.S. EPA, which help to
identify the spatial distribution of acid-sensitive and acid-affected surface waters in the U.S., there are
also some important supplemental regional databases in New England, the Adirondacks, the mid-
Appalachian region, the Florida Panhandle, the Upper Midwest, and the western U.S. Results from these
studies are summarized in the following paragraphs.
New England
For the New England region, results from the U.S. EPA TIME program indicate that 5.6% of the
regional lake population (386 lakes) in New England exhibited ANC <0 ucq/L during the period 1991 to
1994. This result is similar to the EMAP findings, which indicate that 5% of lakes in New England had
ANC values less than 0 (.ieq/L. The EMAP survey was a probability based survey representative of lakes
with surface area greater than 1 ha (1,812 lakes). The survey was conducted during low-flow summer
conditions, and the results therefore likely reflect the highest ANC values for the year. The EMAP
analysis also estimated that an additional 10% of the population had low ANC values, between 0 and
50 (.ieq/L, and were probably sensitive to episodic acidification (Driscoll et al., 2001b).
Adirondacks
A study by Driscoll, et al. (2001b) used EMAP data from 1991 to 1994 to evaluate the extent of
acidic lakes in the Adirondacks for that period. Results from the survey indicate that 10% of the
population of Adirondack lakes were chronically acidic (ANC values of less than 0) and 31% were
3-58
-------
sensitive to episodic acidification (ANC values between 0 and 50) during the study period (Driscoll et al.,
2001b).
The Adirondack Lake Survey Corporation conducted a comprehensive survey of Adirondack lakes
greater than 0.2 ha in surface area between 1984 and 1987 (Baker et al., 1990a). Of the 1,489 lakes
surveyed, 24% had summer pH values below 5.0, 27% were chronically acidic (ANC <0) and an
additional 21% were probably susceptible to episodic acidification (ANC between 0 and 50; Driscoll
et al., 2007a).
Mid-Appalachian Region
A compilation of survey data from the mid-Appalachians yields a consistent picture of the acid-
base status of streams. In the subpopulation of upland forested streams, which comprises about half of the
total stream population in the mid-Appalachian area, data from various local surveys showed that 5% to
20% of the streams were acidic and about 25 to 50% had ANC <50 j^ieq/L (Herlihy et al., 1993). NSS
estimates for the whole region showed that there were 2330 km of acidic streams and 7500 km of streams
with ANC <50 (ieq/L. In these forested reaches, 12% of the upstream reach ends were acidic and 17% had
pH <5.5. S042 from atmospheric deposition was the dominant source of acid anions in acidic mid-
Appalachian streams.
Cosby et al. (2006) provided a detailed characterization of streamwater acid-base chemistry in
Shenandoah National Park, Virginia, which has been the most thoroughly studied area within the mid-
Appalachian Mountain region with respect to acidification from acidifying deposition. Based on MAGIC
model simulations and extrapolation using landscape characteristics, Cosby et al. (2006) developed maps
showing the distribution of streamwater conditions in the Park for the preindustrial past, current
conditions, and anticipated future conditions.
Florida Panhandle
According to the U.S. EPA eastern lakes survey conducted in 1984, 75% of the Florida Panhandle
lakes were acidic at that time, as were 26% of the lakes in the northern peninsula. Most of the acidic lakes
were clear water (DOC <400 (.iM) seepage lakes in which the dominant acid anions were chloride and
S042 . Most of the acidic and low-ANC lakes were located in the Panhandle and north central lake
districts. Acidic streams were located in the Panhandle, were mildly acidic (mean pH 5.0), and extremely
dilute, with very low sea salt-corrected sum of base cations (mean 21 (ieq/L) and sea salt-corrected S042
concentrations (mean 16 (ieq/L). One-fourth of these acidic Panhandle streams were organic-dominated
but the remaining sites all had DOC <2 mg/L. Inorganic monomeric Al concentrations in these acidic
streams were very low (mean 11 (ig/L). In these low DOC, low ANC Panhandle streams, it was suggested
that the degree of S042 and N03 retention in soil was an important control on streamwater ANC (Baker
et al., 1990b).
Upper Midwest
Based on the eastern lakes survey, the Upper Midwest has a large population of lakes having ANC
< 200 (ieq/L, (Linthurst et al., 1986a; 1986b); only 6% of the lakes had ANC < 50 j^icq/L. Groundwater
recharge lakes (those having Si concentration less than 1 mg/L, indicating little groundwater input)
constituted 71% of the seepage lakes in the Upper Midwest, and were more frequently low in pH and
ANC. Five percent were acidic and 9% had pH <5.5. Nearly 90% of Upper Midwestern lakes that had
ANC < 50 (ieq/L were in the groundwater recharge category (Baker et al., 1991b). Such lakes tend to be
susceptible to acidification from acidifying deposition.
Acidic lakes in the Upper Midwest are primarily small, shallow, seepage lakes that have low
concentrations of base cations and Al and moderate S042 concentrations. Organic anions, estimated by
both the Oliver et al. (1983) method and the anion deficit, tend to be less than half the measured S042
3-59
-------
concentrations in the acidic lakes (Eilers et al., 1988), but much higher in many of the drainage lakes that
are less sensitive to acidification from acidifying deposition.
West
Landers et al. (1987) identified subregions in the West with acid-sensitive lakes, based on results of
the U.S. EPA Western Lakes Survey. The surface water chemistry data for the West indicate that the Sierra
Nevada and Cascade Mountains constitute the mountain ranges with the greatest number of sensitive lake
resources. Surface waters in this region are among the most poorly buffered surface waters in the U.S.
(Landers et al., 1987; Melack and Stoddard, 1991). The hydrologic cycle is dominated by the annual
accumulation and melting of a dilute, mildly acidic snowpack.
Many Cascade and Rocky Mountain lakes are highly sensitive to potential acidifying deposition
effects (Nelson, 1991; Turk and Spahr, 1991). It does not appear that chronic acidification has occurred to
any significant degree, although episodic acidification has been reported for lakes in the Colorado Front
Range (Williams and Tonnessen, 2000).
Along the eastern edge of the Continental Divide in Colorado and southeastern Wyoming,
Musselman et al. (1996) conducted a synoptic survey of surface water chemistry in the mountainous areas
that are exposed to relatively high (by western standards) deposition of N. A total of 267 high-elevation
lakes situated in watersheds having a high percentage of exposed bedrock or glaciated landscape were
selected for sampling. None of the lakes were chronically acidic (ANC <0), although several had
ANC <10 (ieq/L, and more than 10% of the lakes had ANC <50 (ieq/L.
Forest Ecosystems
No systematic national survey of terrestrial ecosystems in the U.S. has been conducted to
determine the extent and distribution of terrestrial ecosystem sensitivity to acidifying deposition. The
scarcity of information on sensitive terrestrial ecosystems is due in part to sparse soils data. In general,
forest ecosystems of the Adirondack Mountains of New York, Green Mountains of Vermont, White
Mountains of New Hampshire, the Allegheny Plateau of Pennsylvania, and high-elevation forests in the
southern Appalachians are considered to be the regions most sensitive to terrestrial acidification effects
from acidifying deposition.
One national and a few regional efforts have been undertaken to characterize forest sensitivity to
acidifying deposition using a critical loads approach. In this context, acid-sensitive soils are those which
contain low levels of exchangeable base cations and low base saturation. On a broad national scale,
McNulty et al. (2007) used a simple mass balance equation and available national databases to estimate
forest soil critical acidic loads (for wet and dry deposition of S and N) and exceedances for forest soils.
Exceedances are pollutant loads that are greater than the estimated critical load for that location. They
found that approximately 15% of forest soils in the U.S. receive acidifying deposition that exceeds the
estimated critical load of wet and dry deposition of S and N by more than 250 eq ha/yr (McNulty et al.,
2007). The areas where exceedances reach this level could be considered to represent those areas that are
likely most sensitive to continued high levels of acidifying deposition. Thus, there is not a national survey
of soil sensitivity to acidification, but there are approaches available with which to identify areas likely to
include sensitive soils.
Note that the McNulty et al. (2007) paper represents the beginning of an iterative process to
identify more precise critical loads for terrestrial acidity. The authors note that the actual area in
exceedance of the forest soil critical acid load may be higher than the mapped estimates for several
reasons (McNulty et al., 2007). First, their estimated total deposition did not include cloud deposition.
Second, base cation deposition to near-coastal areas was not corrected for marine aerosol contributions.
Third, the 1-km squared grid size of the mapping resulted in averaging of soil and deposition data, which
removed extreme values from the analysis (McNulty et al., 2007). The authors take care to describe their
results as "preliminary" and note that a more systematic analysis of model-predicted and measured forest
3-60
-------
soil critical acid load exceedance is needed before this approach can be used as a tool for identifying areas
of potential forest health concern (McNulty et al., 2007). For these reasons, and because of the significant
uncertainty associated with many of the large national databases used in the analysis, the appropriate use
of this information is not for the actual determination of critical loads at specific locations or for
predictions of forest health effects, but rather for increased understanding of relative differences in forest
soil sensitivity at a national scale. In general, the Northeast, the Southern Appalachians, parts of Florida
and the Upper Midwest have the highest proportion of soils that exceed the estimated critical acid loads
by at least 250 eq ha/yr and therefore could be termed vulnerable. Where the exceedances are highest,
forest soils are likely most sensitive to continued effects from acidifying deposition.
At a regional scale, Pardo et al. (2007) calculated critical loads of S and N deposition to forests in
Great Smoky Mountains National Park (GSMNP) based on available data. A simple mass balance model
and the Very Simple Dynamic model (VSD) were used to calculate a critical load for acidity (N+S) and N
nutrient. The authors concluded that current deposition exceeded the critical load at all four sites
evaluated (2 high elevation spruce-fir sites, a mid-high elevation beech site, and a lower elevation mixed
hardwood site). The exceedance for S + N deposition ranged from 150 eq/ha/yr for the low elevation
mixed hardwood site to 2300 eq/ha/yr at the upper spruce-fir site. The maximum acceptable deposition of
N ranged from 200 eq/ha/yr (3 kg N/ha/yr) for the low elevation mixed hardwood site to 500 eq/ha/yr
(7 kg N/ha/yr) at the upper spruce-fir site.
Another approach to identification of sensitive forest lands is to map the distribution of tree species
thought to be most sensitive to adverse effects. The effects of acidifying deposition are particularly well
documented for red spruce trees (Cronan and Grigal, 1995; Johnson and Lindberg, 1992; Johnson et al.,
1994b; Joslin et al., 1992) that occur in the northeastern U.S. and southern Appalachian Mountains
(Figure 3-6 shows the distribution). In the Northeast, red spruce grows at elevations from near sea level to
about 1,400 m. In the Appalachian Mountains, spruce-fir forests are generally found at relatively high
elevation, for example above about 1400 m in the southern portion of the range (SAMAB, 1996).
Northern hardwood forests have also been identified as forest resources experiencing air pollution effects.
Effects are best documented for sugar maples, which are broadly distributed across the northern
hardwood forests in the northeastern U.S. (Figure 3-6 shows the distribution). The areas where sugar
maples appear to be at greatest risk are along ridges and where this species occurs on nutrient-poor soils.
Model Simulations
In the eight-state Southern Appalachian Mountains region, Sullivan et al. (2005) modeled future
effects of atmospheric S and N deposition on aquatic resources. Modeling was conducted with the
MAGIC model for 40 to 50 sites within each of three physiographic provinces, stratified by stream water
ANC class. The model runs were based on three emissions control strategies (A2, Bl, and B3). A2 is the
base case that represents best estimates for air emission controls under regulations for which
implementation strategies were relatively certain at the time of the study (about the year 2000). This A2
strategy includes the acid rain controls under Title IV of the 1990 Amendments to the CAA, the 1-h ozone
(03) standard, NOx reductions required under the U.S. EPA call for revised State Implementation Plans
(SIPs), and several highway vehicle and fuel reductions. The Bl and B3 strategies assumed progressively
larger emissions reductions, targeted only to the eight states in the southern Appalachian Mountains
region, but covering all emissions sectors.
The results for the portion of the region south of Virginia and West Virginia suggest that the
percentages of streams having ANC below zero and below 20 j^ieq/L will actually increase through the
year 2040 under all except the most restrictive emissions control strategies (Sullivan et al., 2005). Most
simulated changes in stream water ANC from 1995 to 2040 were rather modest, given the very large
estimates of future decrease in S deposition. Few modeled streams showed projected change in ANC of
more than about 20 |_ieq/L (Sullivan et al., 2005). Some of the largest changes were simulated for some of
the streams that were most acidic in 1995. For such streams, however, even relatively large increases in
3-61
-------
ANC would still result in stream water having negative ANC, and therefore little biological improvements
would be expected from the simulated improvement in chemistry (Sullivan et al., 2005).
Sullivan et al. (2002b) used the NuCM model to evaluate potential changes in soil chemistry in
response to acidifying deposition in the southern Appalachian Mountains. The results suggest that spruce-
fir forests in the region are likely to experience decreased Ca:Al ratios in soil solution under virtually all
strategies of reduced future acidifying deposition considered. This result was partly because S042
adsorption in soils is likely to decline, even with dramatically reduced S deposition. In addition, many
spruce-fir forests in the region are N-saturated, and continued N deposition at moderate or high levels
would be expected to contribute to elevated NO3- concentrations in soil water, which could further
enhance base cation leaching and mobilization of Al from soils to soil solution.
In the Adirondacks, model results produced by several studies suggest that the trend of increasing
lakewater ANC for the most acid-sensitive lakes might not continue in coming decades. These results are
discussed above in the Adirondack case study.
In a regional application of PnET-BGC, Chen and Driscoll (2005) analyzed 60 DDRP (Direct
Delayed Response Project) lake watersheds within northern New England under three future emissions
reduction scenarios. Most of the lakes had surface water ANC values greater than 50 j^icq/L in 1984, and
were therefore not considered chronically acidic. The authors reported that ANC was projected to increase
under all three scenarios, with greater rates of recovery occurring with deeper emissions reductions. Soil
improvements were slow and modest under all scenarios. Simulations suggested that 80% of the northern
New England sites and 60% of the Maine sites will have soil base saturation below 20% in 2050 (Chen
and Driscoll, 2005). They concluded that the decreases in S042 and N03 concentrations in surface water
were coupled with nearly stoichiometric decreases in base cation concentrations. Simulated improvements
in ANC in response to reduced acidifying deposition were minor. Therefore, while further declines in
atmospheric deposition in S and N will bring some improvements, most ecosystems in the study were not
expected to recover to background conditions by 2050.
Bulger et al. (2000) developed model-based projections using the MAGIC model to evaluate the
potential effect of reductions in S deposition of 40% and 70% from 1991 levels using data from streams
in and near Shenandoah National Park. Projections were based on four brook trout stream categories:
Suitable, ANC >50 j^icq/L: Indeterminate, ANC 20 to 50 j^icq/L: Marginal, ANC 0 to 20 |_icq/L: and
Unsuitable, ANC <0 (ieq/L. Three scenarios of future acidifying deposition were modeled: constant
deposition at 1991 levels, 40% reduction from 1991 deposition levels, and 70% reduction from 1991
deposition levels. Based on observed 1991 ANC values, approximately 30% of all trout streams in
Virginia were marginal or unsuitable for brook trout because they were either episodically (24%) or
chronically (6%) acidic. In addition, another 20% of the streams were classified as indeterminate, and
brook trout in these streams may or may not have been affected. Based on the model simulations, 82% of
these streams would not have been acidic before the onset of acidifying deposition and would likely have
been suitable for brook trout.
The model projections suggested that neither the 40% nor the 70% reductions in acidifying
deposition would increase the number of streams that were suitable for brook trout above the ambient
50%. In fact, the results suggested that a 70% reduction in deposition would be needed in the long term
just to maintain the number of streams that were considered suitable for brook trout. Because of the length
of time required to restore buffering capacity in watershed soils, most of the marginal or unsuitable
streams were expected to remain marginal or unsuitable for the foreseeable future.
Results of modeling studies for lakes and streams in the Adirondack Mountains and in Shenandoah
National Park are presented in the case study sections of this report.
3.2.4.3. Levels of Deposition at Which Effects are Manifested
The effects of S and N deposition are manifested at a range of deposition levels, depending on the
inherent sensitivity of the natural resources, as described in the previous sections, and the historical
3-62
-------
deposition loading. The intersection among current deposition loading, historic loading, and sensitivity
defines the ecological vulnerability to the adverse effects of acidification. Few studies in the U.S. have
defined deposition levels that are associated with effects over large areas.
Some degree of surface water acidification, and perhaps also of soil acidification, can occur at very
low levels of S deposition (only a few kg S/ha/yr). These highly sensitive areas are characterized by very
low levels of exchangeable base cations and soil base saturation. Water pathways and soil depth can also
limit the capacity of these areas to neutralize acidifying deposition. These areas provide limited
neutralization of acidic drainage water.
Effects levels for N deposition can be established based on changes to stream and soil chemistry
that signal alteration of nutrient cycling, causing N03 leaching. Analyses have been conducted in the
northeastern U.S. and Europe to examine the relationships between N deposition and N03 leaching to
surface waters. The relationship between measured wet deposition of N and streamwater output of N03
was evaluated by Driscoll et al. (1989) for sites in North America (mostly eastern areas), and augmented
by Stoddard (1994). The resulting data showed a pattern of N leaching at wet inputs greater than
approximately 5.6 kg N/ha/yr. Aber et al. (2003) concluded that loss of N03 to surface waters during the
growing season in forested watersheds often occurs above a threshold of total (wet plus dry) atmospheric
N deposition of about 8 to 10 kg N/ha/yr.
The effects of N addition on forests have been shown to be wide-ranging. Additions of 25 kg
N/ha/yr to spruce plots in Vermont (ambient bulk deposition 5.4 kg N/ha/yr), in which net nitrification did
not occur before treatment, triggered net nitrification in the second year of treatment (McNulty et al.,
1996). Similar results were seen in Colorado, where additions of 25 kg N/ha/yr to old-growth spruce plots
in Loch Vale watershed (ambient bulk deposition 4 to 5 kg N/ha/yr) doubled N mineralization rates and
stimulated nitrification. In marked contrast to these results, concentrations of N03 plus NH/ were not
detected until the seventh year in hardwood plots in Harvard Forest, Massachusetts, which received
additions of 150 kg N/ha/yr (Magill et al., 2004). Concentrations of N03 plus NH/ in hardwood plots
receiving 50 kg N/ha/yr were not yet detectable in the 15th year of treatment.
Many of the changes in plant species composition, species diversity, and nitrification and
mineralization rates in response to atmospheric N deposition are associated with nutrient N fertilization,
rather than acidification. They are discussed in more detail in Section 3.3.
Chemical Response
As discussed in Section 3.2.1.6., surface water chemistry has responded to changes in emissions
and deposition of S over the past two to three decades and most recently also decreases in N. Monitoring
data collected within the U.S. EPA Long-Term Monitoring (LTM) and TIME projects, as well as other
monitoring programs, has been key to understanding chemical responses. See discussion of major
monitoring programs in Annex B. Surface water chemistry monitoring data generated through TIME and
LTM (Stoddard et al., 2003) suggest that the following important changes in lake and stream chemistry
have occurred over the past one to two decades in the eastern U.S.:
¦ S042 concentration has decreased as a percentage of total ion concentration in surface waters.
¦ ANC has increased modestly in three of the five regions studied.
¦ DOC and associated natural organic acidity increased, perhaps toward more natural pre-
disturbance concentrations, as surface water acidity contributed from acidifying deposition has
decreased.
¦ Inorganic Al concentrations appear to have decreased slightly in some sensitive aquatic systems.
The significant decreases in surface water S042 concentration, which have been observed in many
areas, have not necessarily brought large changes in the acidity of lakes and streams. For example, the
decline in Adirondack lakewater S042 during the 1980s was charge-balanced by a nearly equivalent
decrease in concentrations of base cations in many of the low-ANC lakes, and this limited the increase in
3-63
-------
ANC and pH that occurred in response to lower S042 concentrations. Overall, improvements in
lakewater acid-base chemistry since 1990 have been measurable but modest. Similar patterns have been
observed in most other regions. There are currently no data in the U.S. that indicate increases in soil pH
associated with recent declines in acidifying deposition levels.
Declines in S02 and NOx emissions have brought about measurable improvements in streamwater
chemistry in sensitive regions of the U.S. since 1990. However, model forecasts suggest that a reversal in
chemical recovery could occur in many sensitive ecosystems under current emissions and deposition
levels and that further reductions beyond those required by the 1990 Amendments to the CAA may be
needed to prevent continued adverse effects and to support biological recovery of terrestrial and aquatic
ecosystems (see discussion in Section 3.2.4.5.).
Biological Response
Biological recovery can occur only if chemical recovery is sufficient to allow survival and
reproduction of acid-sensitive plants and animals. The time required for biological recovery is uncertain.
For terrestrial ecosystems, it may be decades after soil chemistry is restored because of the long life of
many plant species and the complex interactions of soil, roots, microbes, and soil biota. For aquatic
systems, research suggests that stream macroinvertebrate populations may recover relatively rapidly
(within approximately 3 years), whereas lake populations of zooplankton recover more slowly (Gunn and
Mills, 1998).
Table 3-8 contains a general summary of pH levels at which biological changes are typically
manifested. Nevertheless, for aquatic ecosystems, there is currently no theoretical basis on which to
predict the pathway and timing of biological recovery. Biological recovery of previously acidified surface
waters can lag behind chemical recovery because of such factors as limits on dispersal and
recolonization; barriers imposed by water drainage patterns (Jackson and Harvey, 1995); the influence
of predation (McNicol et al., 1995); and other environmental stressors (Gunn et al., 1995; Havas et al.,
1995; Jackson and Harvey, 1995; McNicol et al., 1995; Yan et al., 1996a, 1996b). Full biological recovery
may take decades from the onset of chemical recovery. The results of biological recovery research from
the Sudbury region of Canada and several experimental lakes is summarized below.
Table 3-8. General summary of biological changes anticipated with surface water acidification,
expressed as a decrease in surface water pH.
^ General Biological Effects
Decrease °
6.5 to 6.0 Small decrease in species richness of plankton and benthic invertebrate communities resulting from the loss of a few highly
acid-sensitive species, but no measurable change in total community abundance or production.
Some adverse effects (decreased reproductive success) may occur for highly acid-sensitive fish species (e.g., fathead
minnow, striped bass).
6.0 to 5.5 Loss of sensitive species of minnows and dace, such as fathead minnow and blacknose dace; in some waters, decreased
reproductive success of lake trout and walleye, which are important sport fish species in some areas.
Visual accumulation of filamentous green algae in the near-shore zone of many lakes and in some streams.
Distinct decrease in species richness and change in species composition of plankton and benthic invertebrate communities,
although little if any change in total community abundance or production.
Loss of some common invertebrate species from zooplankton and benthic communities, including many species of snails,
clams, mayflies, and amphipods, and some crayfish.
3-64
-------
5.5 to 5.0
Loss of several important sport fish species, including lake trout, walleye, rainbow trout, and smallmouth bass, as well as
additional nongame species such as creek chub.
Further increase in the extent and abundance of filamentous green algae in lake near-shore areas and streams.
Continued shift in species composition and decline in species richness of plankton, periphyton, and benthic invertebrate
communities; decreases in total abundance and biomass of benthic invertebrates and zooplankton may occur in some waters.
Loss of several additional invertebrate species common in surface waters, including all snails, most species of clams, and
many species of mayflies, stoneflies, and other benthic invertebrates.
Inhibition of nitrification.
5.0 to 4.5
Loss of most fish species, including most important sport fish species such as brook trout and Atlantic salmon. A few fish
species are able to survive and reproduce in water below pH 4.5 (e.g., central mud minnow, yellow perch, and in some waters,
largemouth bass).
Measurable decline in the whole-system rates of decomposition of some forms of organic matter, potentially resulting in
decreased rates of nutrient cycling.
Substantial decrease in number of species of plankton and benthic invertebrates and further decline in species richness of
plankton and periphyton communities; measurable decrease in total community biomass of plankton and benthic invertebrates
of most waters.
Loss of additional species of plankton and benthic invertebrate species, including all clams and many insects and crustaceans.
Reproductive failure of some acid-sensitive species of amphibians, such as spotted salamanders, Jefferson salamanders, and
the leopard frog.
The Sudbury region of Ontario, Canada has been important for studying both the chemical and
biological effects of S deposition. Mining and smelting of copper-nickel ore began in the 1880s. By the
1950s and 1960s, S02 emissions from the mining and smelting operations peaked at over 5,000 tons/day
and extensive acidification of nearby surface waters was documented (Beamish and Harvey, 1972).
Emissions of S02 then decreased during the 1970s to less than one-third of the peak values. S emission
reductions resulted in improved water quality in many lakes (Keller and Pitblado, 1986; Keller et al.,
1986), and some fisheries recovery was also documented (Gunn and Keller, 1990; Keller and Yan, 1991).
Griffiths and Keller (1992) found changes in the occurrence and abundance of benthic invertebrates that
were consistent with a direct effect of reduced lakewater acidity. A more recent assessment of recovery of
ecosystems in Canada provided further evidence of biological recovery, but also showed that the spatial
extent of recovery was limited to lakes that had been severely acidified by the Sudbury smelter (Jeffries
et al., 2003). Research at Sudbury clearly documented that chemical recovery of lakes was possible upon
reduced emissions and deposition of S, and that biological recovery, involving multiple trophic levels,
could follow. Major findings of the research at Sudbury and elsewhere are summarized below.
Phytoplankton
Studies of phytoplankton recovery from experimental acidification indicate that there is an increase
in phytoplankton species richness and diversity as pH increases. In Lake 223 in the Experimental Lakes
area of Ontario, there was little increase in phytoplankton diversity as pH changed from 5.0 to 5.8 but a
strong recovery of diversity at pH above 6 (Findlay and Kasian, 1996). In Lake 302S, profound change
began at pH 5.5; phytoplankton assemblages at pH below 5.5 resembled acidified lakes.
3-65
-------
Zooplankton
Zooplankton recovery in response to experimental de-acidification has been reported for lakes in
Ontario, Canada and Minnesota. Zooplankton recovery in experimentally acidified Lake 223 as pH
returned back to 6.1 was reported by Malley and Chang (1995). Species diversity that had been reduced
during the acidification phase had partially returned to pre-acidification levels. Rotifers had recovered less
than crustaceans.
One decade after cessation of the experimental acidification of Little Rock Lake in Wisconsin,
recovery of the zooplankton community was complete (Frost et al., 2006). Recovery did not follow the
same trajectory as the initial acidification, however, indicating a substantial hysteresis in zooplankton
community recovery. About 40% of the zooplankton species in the lake exhibited a lag of 1 to 6 years to
recover to levels that occurred in the neutral reference basin.
Benthic Invertebrates
There has been some research conducted on the recovery of benthic invertebrate communities in
surface waters exhibiting chemical recovery from acidification. In Scotland, Soulsby et al. (1995)
reported an increase in acid-sensitive mayflies in some streams that showed recent ANC increases.
However, no increases in invertebrates were observed in the most acidic streams despite observed
increases in ANC. They suggested that further reductions in acidic deposition and sufficient time for
reversal of soil acidification may be required before aquatic biotic recovery can occur. The extent to
which benthic invertebrates in streams in the U.S. may have recovered in response to any recent increases
that may have occurred in stream ANC and pH is not known.
Fish
Fish populations have recovered in acidified lakes when the pH and ANC have been increased
through liming or reduction of acidifying deposition (Beggs and Gunn, 1986; Dillon et al., 1986; Gunn
et al., 1988; Hultberg and Andersson, 1982; Keller and Pitblado, 1986; Kelso and Jeffries, 1988; Raddum
et al., 1986). The timing of fish recovery is uncertain and probably depends heavily on dispersion.
Stocking could accelerate fish population recovery (Driscoll et al., 2001a). Limitations on dispersal and
recolonization can hamper biological recovery from acidification.
Continued periodic episodic acidification might hamper biological recovery of a lake or stream that
is experiencing improvement in chronic chemistry. If fish move into refugia during episodes of low pH
and then return, behavioral avoidance would reduce the overall effect of episodic acidification on fish
populations. If fish move out of the stream system in response to acidic episodes, as suggested by Baker
et al. (1996), and do not return or return in smaller numbers, then the population level effects of episodic
acidification would be greater than predicted based on mortality tests alone.
3-66
-------
Table 3-9.
Studies that either did or did not yield evidence that acidifying deposition affected certain
species of birds.
Species
Diet /
Foraging
YES NO
Breeding
Distribution
YES NO
Reproductive
Measures
YES NO
Reference
Common loon
X
X X
X X
Alvo et al. (1988); Blair (1990); Blancher and McNicol,
(1991); DesGranges and Houde (1989); Wayland and
McNicol (1990); Parker (1988)
Arctic loon
X
Eriksson (1987)
Common
merganser
X
X
McNicol et al. (1987)
Belted
kingfisher
X
Goriup (1989)
Osprey
X
X
X
Eriksson (1983,1986)
Black duck
X
X
xb
DesGranges and Darveau (1985); Harasmis and Chu (1987);
Hunter (1986); Rattner (1987)
Common
goldeneye
xb
DesGranges and Darveau (1985); McNicol et al. (1987)
Ring-necked
duck
X
X
McAuley and Longcore (1988a); McAuley and Longcore
(1988b)
Eurasian
dipper
X
X
X
Ormerod et al. (1985,1986); Ormerod and Tyler (1987)
Eastern
kingbird
X
X
Glooschenko etal. (1986)
Tree swallow
X
X
X
Blancher and McNicol (1988,1991); St. Louis etal. (1990)
b Effect was beneficial
Baker et al. (1990a) used field-based models to test the potential for biological recovery. For each
species considered, the current presence or absence of the species was analyzed as a function of the water
quality variables associated with acidification (e.g., pH, Al, Ca, ANC, and DOC) using maximum
likelihood logistic regression (Reckhow et al., 1987). The results from the various models were compared
to their prediction of the change in the number of Adirondack lakes with unsuitable acid-base chemistry,
given a 50% decrease or a 30% increase in S deposition relative to the existing conditions at the time of
the eastern lakes survey (1984). Most of the models provided similar results and suggest that a 30%
increase in S deposition would increase the unsuitable fish habitat in Adirondack DDRP lakes by 15% to
28% for brook trout, lake trout, and common shiner. A 50% decrease in S deposition was projected to
increase suitable habitat by 8% to 18%.
Waterfowl
Few studies have been conducted on the recovery of higher trophic level organisms such as birds
(Table 3-9). However, breeding distribution for the common goldeneye (Bucephala clangula), an
insectivorous duck, may be affected by changes in acidifying deposition (Longcore and Gill, 1993).
3-67
-------
Similarly, reduced prey diversity and quantity have been observed to create feeding problems for nesting
pairs of loons on low-pH lakes in the Adirondacks (Parker, 1988).
Logistic regression modeling with measured pH and species occurrence data for acid-sensitive
lakes in the Algoma region of Ontario showed that the occurrences of fish, common loons, and common
mergansers were positively related to lake water pH (McNicol, 2002). Predictions of common loon and
merganser recovery for this area were made using the Waterfowl Acidification Response Modeling
System (WARMS) under varying S emissions control scenarios targeted for 2010 (McNicol, 2002). The
number of lakes projected to be suitable for supporting breeding pairs and broods increased with Lake pH
and stricter emissions controls (McNicol, 2002). Marginal improvements in fish-eating bird habitat were
predicted to occur by 2010, with more significant improvements expected under hypothetical S emissions
reductions of 50% and 75% for lakes with pH below 6.5 (McNicol, 2002). Fundamental to the predicted
improvement of these fish-eating bird populations is the expected increase in food availability with lake
pH recovery.
3.2.4.4. Acidification Case Study #1: Adirondack Region of New York
In this and the following section, case studies are presented for two of the most thoroughly studied
regions of the U.S. that are known to have been affected by acidification from atmospheric S and N
deposition. Studies in these regions have focused on both chemical and biological effects, and have
included extensive model simulations of past acidification and projections of the likelihood of future
recovery as deposition levels decline. The Adirondack Mountain region is perhaps the most thoroughly
studied region in the world with respect to surface water acidification. Large numbers of Adirondack
lakes have been acidified over the past century, and many of those now show signs of chemical recovery.
Shenandoah National Park contains many acidified and acid-sensitive streams. Sensitivity in this region is
strongly controlled by geology and the extent to which deposited S is adsorbed to soils. These two case
studies are intended to be illustrative of the types of research that has been conducted and what that
research has revealed.
General Description of Region
The Adirondack Mountains are in northeastern New York State and are densely forested, have
abundant surface waters, and have 46 peaks that extend up to 1600 m in elevation. The Adirondack Park
has long been a nationally important recreation area for fishing, hiking, boating, and other outdoor
activities.
The Adirondacks, particularly the southwestern Adirondacks, are sensitive to acidifying deposition
because they receive high precipitation, have shallow base-poor soils, and are underlain by igneous
bedrock with low weathering rates . The Adirondacks are among the most severely acid-affected regions
in North America (Driscoll et al., 2003b; Landers et al., 1988) and have long been used as an indicator of
the response of forest and aquatic ecosystems to U.S. policy on atmospheric emissions of S02 and NOx
(GAO, 2000; NAPAP, 1998; U.S. EPA, 1995a).
Rates of Acidifying Deposition
Current rates of wet deposition of S and N in the western Adirondacks remain among the highest in
the nation. Spatial patterns in wet deposition of S and N from 1988 to 1999 were developed by Ito et al.
(2002), using data from 24 precipitation and 4 wet deposition monitoring stations. Results from this effort
suggest that wet S deposition ranged from 2.3 to 12.9 kg S/ha/yr and wet N03-N deposition rates ranged
from 1.7 to 5.1 kg N/ha/yr (Ito et al., 2002) (Figure 3-18). In general, deposition rates are highest in the
southwestern Adirondacks and decrease to the northeast. Rates of dry deposition are less well known, but
probably constitute 25% to 50% or more of total wet deposition (Sullivan et al., 2006b).
3-68
-------
Deposition trends have changed with the implementation of federal and state emissions control
regulations. For example, by 1990 average wet S deposition in the Adirondack region had declined by
approximately 30% from its peak in the 1970s (Sullivan et aL 1990). Deposition of S has continued to
decline (Figure 3-19) and (Figure 3-20) in response to implementation of the CAAAof 1990. Until
recently, wet N deposition had been fairly consistent over the previous two decades. N deposition now
appears to be decreasing (http: //nadp. sws .uiuc. edu/).
Annual S04 Deposition (kg S04/ha-yr)
6.9-10.4
10.4-13.9
13.9-17.4
_ 17.4-21.0
—] 21.0 - 24.5
¦ 24.5 - 28.0
28.0-31.6
131.6-35.1
35.1 -38.6
Not estimated
Annual NOs Deposition (kg N03/ha-yr)
7.6 - 9.3
9.3-10.9
10.9-12.6
12.6-14.2
14.2-15.9
15.9-17.6
17.6-19.2
19.2 - 20.9
20.9 - 22.5
Not estimated
Source: Ito et al. (2002). Reprinted with permission.
Figure 3-18. Spatial patterns in predicted wet SO42" and NO3" deposition in the Adirondack Park during the
period 1988 to 1999.
Soil Retention and Leaching of Sulfur and Nitrogen
As discussed in Section 3.2.4.4, acidifying deposition has resulted in the accumulation of S and N
in Adirondack soils. Although input-output budgets for S developed in the 1980s suggested that the
amount of S exported was approximately equal to the S inputs from atmospheric deposition, more recent
studies show that watershed loss of S04 now exceeds atmospheric S deposition inputs (Driscoll et al.,
1998). This pattern suggests that decades of atmospheric S deposition have resulted in the accumulation
of S in forest soils. With recent declines in atmospheric S deposition and a possible warming-induced
enhancement of S mineralization from soil organic matter, previously retained S is gradually being
3-69
-------
released to surface waters (Driscoll et al., 1998). This release of S042 from soils could contribute to a
delay in the recovery of surface waters in response to S02 emissions controls.
N dynamics are quite different from those of S. Because N is a growth-limiting nutrient for many
forest plants, retention in forest ecosystems under low levels of air pollution is generally high and NO;,
loss to streams is relatively low (Aber et al., 2003). However, recent research suggests that N has
accumulated in soils over time in the Adirondacks and that some forests have exhibited increasing
retention of N inputs (Driscoll et al., 2003a, c). The result has been increased leaching of N03 to surface
waters. The extent and degree of leaching appear to be linked to climatic variation, land-use history, and
vegetation type (see Section 3.2.1.3).
Huntington Forest Annual Sulfur Wet Deposition
y = -0.4233x + 860.39
R2= 0.6192
.>>
"ro
¦E
55
oi
*
1975 1979 1983
1987 1991
Year
i r
1995 1999 2003
Source: Sullivan et al. (2006b).
Figure 3-19. Measured wet deposition of sulfur at the Huntington Forest NADP/NTN monitoring station.
The leaching of both S042 and NO;, into drainage water has contributed to the displacement of
cations from soil, acidification of surface waters (Driscoll et al., 2001b), and the associated chemical and
biological effects discussed below.
Soil Acidification and Base Cation Depletion
Atmospherically deposited hydrogen ions can directly affect soil pH. Net uptake of nutrient cations
by vegetation can also generate acidity within the soil, and a considerable amount of natural organic
acidity is produced in the Oa horizon through the partial decomposition of organic matter and uptake of
nutrient cations. In the only repeated soil sampling in the U.S. in which the original sampling predated
acidifying deposition, (Johnson et al., 1994b) found significantly higher soil pH values in 1930 than in
1984 in the Oa horizon of Adirondack soils that had an initial pH of 4.0-5.5, but no decrease in pH in
soils with an initial pH of <4.0. (Johnson et al., 1994b) also documented a decrease in exchangeable Ca
concentrations in both the O (combined Oa and Oe horizons) and B horizons from 1930 to 1984. The
decrease in soil pH and Ca concentrations was attributed to a combination of acidifying deposition and
changing vegetation dynamics.
3-70
-------
1850
1900
1950
2000
2050
Year
2100
Source: Sullivan et al. (2006b).
Figure 3-20. Estimated time series of S deposition at one example watershed in the southwestern
Adirondack Mountains. Table used by Sullivan et al. (2006a) as input to the MAGIC model for
projecting past and future changes in lakewater chemistry attributable to acidifying
deposition. Future deposition estimates were based on three emissions control scenarios
(Base Case [solid line], Moderate Additional Controls [dotted line], Aggressive Additional
Controls [dashed line]).
In a statistically based regional assessment of changes in soil-exchange chemistry, Sullivan et al.
(2006a) found that base saturation and exchangeable Ca concentrations in the Adirondack region
appeared to have decreased in the B horizon between the mid 1980s and 2003 in watersheds of lakes with
acid-neutralizing capacity less than 200 (ieq/L. Although this study did not involve repeated sampling of
the same sites, the comparison could be made on a regional basis because the sampling locations were
selected randomly in both the mid 1980s and in 2003, and a large and similar number of sites were
included in both samplings.
Effects of Acidifying Deposition on Adirondack Surface Water Chemistry
The Adirondack Lake Survey Corporation conducted a comprehensive survey of Adirondack lakes
greater than 0.2 ha in surface area between 1984 and 1987 (Baker et al., 1990a). Of the 1,489 lakes
surveyed, 24% had summer pH values below 5.0, 27% were chronically acidic (ANC <0) and an
additional 21% were probably susceptible to episodic acidification (ANC between 0 and 50; Driscoll
et al., 2007a).
In addition to low pH and ANC, many acidic surface waters in the Adirondacks are characterized
by high concentrations of inorganic Al. For example, a study of 12 sub-basins in the watershed of the
North Branch of the Moose River by (Driscoll et al., 1987a) determined that the concentration of
inorganic Al in lakewater was higher in lakes having pH below 6.0. Recently, Lawrence et al. (2007)
determined that 66% of 188 streams sampled in the western Adirondack region during snowmelt in 2004
had measurable concentrations of inorganic Al, an indicator of acidification by acidifying deposition.
Historical changes in lakewater chemistry from the mid-1800s to recent times have been estimated
for the Adirondacks using paleolimnological techniques. Fossil remains of diatoms and chrysophytes in
3-71
-------
sediment cores have been used to reconstruct chemical histories. The PIRLAI and II projects
(Paleoecological Investigation of Recent Lake Acidification) used the remains of diatoms preserved in
lake sediments to estimate historical changes in lakewater chemistry across the Adirondack region. The
PIRLA-II project focused on lakes that are 4 ha or larger that represented a subpopulation of 675
Adirondack lakes. The results from these analyses suggest that nearly all lakes with estimated
preindustrial pH less than 6.0 had acidified between 0.3 and 1.0 pH units during the 20th century. Based
on an analysis of data from Cumming et al. (1992) and (Baker et al., 1990a), low-pH lakes were
uncommon or rare in the preindustrial Adirondacks; the number of lakes with pH less than 5.5 had at least
doubled by the mid 1980s and the number with pH less than 5.0 had increased by 5 to 10 times.
The PIRLA results are generally consistent with projections from model hindcasts. (Sullivan et al.,
2006b) modeled past changes in the acid-base chemistry of 70 Adirondack lake watersheds. These
included 44 that were statistically selected to be representative of the approximately 1,320 Adirondack
lake watersheds that have lakes larger than 1 ha and deeper than 1 m and that have ANC < 200 j^ieq/L.
Model hindcasts were constructed using both the MAGIC and PnET-BGC models. Based on MAGIC
model outputs, maximum past acidification occurred by about 1980 or 1990, with a median ANC for the
study population of about 61 j^ieq/L (reduced from a median of 92 j^ieq/L estimated for the preindustrial
period). By 1990, 10% of the population target lakes had decreased in ANC to below -16 j^ieq/L and 25%
had ANC <28 |_icq/L. The model simulations coupled with population-level extrapolations suggest that
none of the target lakes were chronically acidic (had ANC <0 j^ieq/L) under preindustrial conditions, but
that by 1980 there were about 204 chronically acidic Adirondack lakes.
PnET-BGC model simulations generated output that was generally similar to results provided by
MAGIC model simulations. Results from PnET-BGC suggest that none of the lakes in the Adirondack
population had preindustrial ANC below 20 |_icq/L. By 1990, there were 289 lakes having
ANC <20 |_ieq/L and 217 chronically acidic (ANC < 0 j^ieq/L) lakes according to PnET-BGC simulations.
There were 202 lakes in the population simulated to have had preindustrial ANC below 50 (ieq/L, and this
number increased 2.8 times by 1980 under the PnET-BGC simulations.
Zhai et al. (2008) reported PnET-BGC hindcasts for the 44 EMAP lakes. They report that simulated
median values of pH, ANC, and soil percent base saturation were 6.63, 67.7 (ieq/L, and 12.3%,
respectively, in 1850 compared to current measured values of 5.95, 27.8 (ieq/L, and 7.9%. They also
calculated F factors for the PnET-BGC model projections of historical acidification. The F-factor
(Henriksen, 1984; Husar and Sullivan, 1991) reflects the proportion of the increase in lakewater S042
plus N03 concentration that is charge balanced by an equivalent increase in base cation concentrations.
The remaining proportion (1-F) is attributed to increase in the potentially toxic cations, hydrogen ions
and inorganic Al. Based on PnET-BGC hindcast simulations, F-factors for the EMAP lakes ranged from
0.3 to slightly over 1.0, with a mean value of 0.7 (Figure 3-13). The F-factor increased with ambient
lakewater ANC. These results are in close agreement with paleolimnological analyses reported by
Sullivan et al. (1990), which showed historic F-factors for Adirondack lakes ranging from about 0.5 to
above 1.0.
While there were differences for some variables (i.e., MAGIC showed greater soil acidification)
and clear lake-to-lake differences, Sullivan et al. (2006b) observed that the population-level projections
with both models represent robust indicators of the extent and magnitude of changes in lake chemistry
associated with historical inputs of acidic deposition. The differences observed in the nature of
acidification between the two models may be linked to model structural differences. In MAGIC, soil
water is assumed to be in equilibrium with an aluminum trihydroxide mineral at all times (Cosby et al.,
1985a, b). Inputs of strong acids, therefore, drive the mobilization of Al. This Al strongly displaces
exchangeable base cations from the soil exchange complex. In PnET-BGC, the soilphase is in equilibrium
only when solutions are oversaturated with the aluminum trihydroxide solid phase. As a result, there is
less Al available in PnET-BGC to displace exchangeable soil base cations and greater acidification of the
solution is realized. Backward and forward projections of population-level estimates of lake water and
soil chemistry for the Adirondacks are presented as cumulative distribution functions in Figure 3-21.
3-72
-------
M/Vo|i. : I;;!'! i. .1:1"! :';i! Hf: Si jits
PntT-BGC Simulated ResuUs
40 60 80 100 120 140 160
U 02
B 10 0 10 20 30 40 50 60 70
£ 0.4
0 100 200 300 400 500 600 0 100 200 300 400 500 600 701 0.0
u 0.2
CALK
!,A .K
10 20 30 40 50 60 70 80 9C
10 20 30 40 5C
[/
/(
\(
1
1
/J
sec
7
2100
100 200 300 400 500 60
0 50 100 150 200 250 300 0 50 100 150 200 25c °'°
§ 0.4
rr
j
\
(i
) i
f i
BS%
-v
1850
rf
if
A
i j
(/
BS%
J J
1990
50 100 150 200 25
0 5 10 15 20 25 30 35 0 5 10 15 20 25 30 35
1.0
0.8
0.6
0.4
0.2
0,0
f
//
fj
/ J
(
I /
BS%
2100
10 15 20 25
Source: Sullivan et al. (2006b)
Figure 3-21. Cumulative distribution functions of selected major ions (peq/L), calculated ANC of lakewater
(fjeq/L), and B horizon soil % base saturation for the MAGIC and PnET-BGC models. Results
are shown for 1850,1990, and 2100.
3-73
-------
Biological Effects
The Adirondack region has a rich aquatic biota dataset from which to examine relationships among
lake water chemistry and species abundance, composition, and richness. In general, there tends to be a
negative relationship in Adirondack lakes between pH, ANC, and inorganic A1 chemical variables and the
diversity and abundance of fish (Baker and Laflen, 1983; Baker et al., 1990a; Havens et al., 1993) (Figure
3-16), phytoplankton, and zooplankton (Confer et al., 1983; Siegfried et al., 1989) (Table 3-10).
Table 3-10.
Observed relationships between zooplankton species richness and lakewater ANC
in the Adirondack Mountains.
Taxonomic Group
Equation
R2
P
Total zooplankton
Richness = 15.65 + 0.089 ANC
0.46
0.001
Crustaceans
Richness = 6.35 + 0.028 ANC
0.47
0.001
Rotifers
Richness = 9.04 + 0.053 ANC
0.30
0.001
Large Cladocerans
Richness = 1.95 + 0.017 ANC
0.41
0.001
Source: Sullivan et al. (2006b).
Through the Adirondack Lakes Survey, 1,469 lakes were sampled between 1984 and 1987,
representing 80% of the estimated population of Adirondack lakes larger than 1 ha in area (Whittier et al.,
2002). The goal of the survey was to characterize the biological, physical, and chemical characteristics of
the lakes and evaluate the relationships between fish communities and water chemistry. The major results
were reported by (Baker et al., 1990a). Key findings are:
¦ Seventy-six percent of the lakes had fish; 24% (346 lakes) were Ashless.
¦ The most common fish caught were native acid-tolerant species: brown bullhead, brook trout, and
white sucker.
¦ As pH decreases, fish diversity also decreases. The average number of fish species declines from
six fish species in lakes with pH higher than 6.5 to two or fewer fish species in lakes with pH of
5.0 or less.
¦ As pH decreases, the number of Ashless lakes increases. Few lakes with pH of 5.5 or higher are
Ashless. Below pH 5.0, approximately 75% of the lakes are Ashless.
Researchers in the Adirondacks were among the first in the U.S. to demonstrate that fish mortality
increases during acid episodes, which are common to lakes and streams in the Adirondacks during spring
runoff. (Driscoll et al., 1987b) documented surface water chemistry changes associated with periods of
high Aow. They found that pH and ANC decreased substantially during hydrological episodes and
inorganic Al concentrations commonly exceeded thresholds harmful to Ash. These relationships were
Anther documented by the Episodic Response Project as shown in the example for Bald Mountain Brook
in the Adirondacks (Wigington et al., 1996). Work by Van Sickle et al. (1996) and others linked these
chemical changes to Ash mortality in small streams. They determined that blacknose dace were highly
sensitive to low pH and could not tolerate inorganic Al concentrations above about 3.7 |_iM for extended
periods of time. After 6 days of exposure to high inorganic Al, dace mortality increased rapidly to nearly
3-74
-------
100% (Van Sickle et al., 1996). Brook trout were less sensitive, but still showed high mortality during
many acid episodes.
Several efforts have been made to link changes in fish populations with historical changes in water
chemistry associated with acidifying deposition. Among the most widely cited is the work of (Baker et al.,
1990a; 1996). They analyzed 988 Adirondack Lake Survey lakes for which data existed for the period
before 1970 and for the 1980s. Of the 2,824 fish populations confirmed by pre-1970 surveys, 30% had
apparently been lost by the 1980s (Baker et al., 1990a). An estimated 23% of the fish population losses
were related to acidifying deposition. This relationship was strengthened by evidence from the PIRLA
projects. In the 32 lakes that had both historic fish data and paleolimnological chemical reconstructions,
the lakes that had acidified the most or that were originally the most acidic were the same ones that were
judged to have lost fish populations (Baker et al., 1996).
Recent Trends in Surface Water Chemistry and Projections of Future Change
Several studies have been conducted to analyze trends in lake chemistry in the Adirondacks.
(Driscoll et al., 2003c) evaluated changes from 1982 to 2000 in the original 16 Adirondack LTM lakes
and from 1992 to 2000 in the complete set of 48 Adirondack LTM lakes. They found that nearly all study
lakes showed marked decreases in S042 concentration over the period of record and several lakes showed
declines in N03 concentration. Data for one example monitoring lake are given in Figure 3-22. They
found that 7 of the 16 original monitoring lakes showed a statistically significant increase in ANC (Figure
3-22), with a mean rate of increase of 0.78 (ieq/L/yr (Driscoll et al., 2003c). Twenty-nine of the group of
48 lakes showed increasing ANC trends from 1992 to 2000 with a mean rate of increase of 1.60 (ieq/L/yr
(Driscoll et al., 2003b). The authors attributed this recent increase in ANC to declines in both S042 and
N03 concentrations (Driscoll et al., 2003c).
Despite these recent improvements in lake water chemistry in the Adirondack Long-Term
Monitoring lakes, 34 of the 48 lakes still had mean ANC values less than 50 (ieq/L in 2000, including 10
lakes with ANC less than 0 (ieq/L. Thus, current chemistry data suggest that most of these lakes exhibit
chemical conditions that continue to pose a risk to aquatic biota. Model projections of future acid-base
chemistry of lakes in the Adirondack Mountains under three scenarios of future atmospheric emissions
controls were presented by (Sullivan et al., 2006a) to evaluate the extent to which lakes might be expected
to continue to increase in ANC in the future. Estimated levels of S deposition at one representative
watershed are shown in Figure 3-20 for the hindcast period and in the future under the three emissions
control scenarios. Model simulations for 44 statistically selected Adirondack lakes using the MAGIC and
PnET-BGC models were extrapolated to the regional lake population. Cumulative distribution frequencies
of ANC response projected by MAGIC are shown in Figure 3-24 for the past (1850), peak acidification
period (approximately 1990), and future (2100). Results for the future are given for each of the scenarios.
Forecasting results suggested that the ongoing trend of increasing lakewater ANC for the most
acid-sensitive lakes would not continue under future emissions and deposition levels anticipated as of
2003 (Base Case Scenario). The numbers of Adirondack lakes having ANC below 20 and below 50 j^ieq/L
were projected to increase between 2000 and 2100 under that scenario, and the number of chronically
acidic Adirondack lakes (i.e., ANC less than 0) was projected to stabilize at the level reached in 2000.
Thi3s projected partial reversal of chemical recovery of acid-sensitive lakes was due to a continuing
decline in the simulated pool of exchangeable base cations in watershed soils. Simulations suggested that
re-acidification might be prevented with further reductions in emissions and deposition.
Chen and Driscoll (2004) applied the PnET-BGC model to DDRP lake watersheds in the
Adirondacks. The model was applied to three future emissions scenarios: base case, moderate emissions
reductions, and aggressive emissions reductions. A case study for Indian Lake in the Adirondacks
illustrated that larger reductions in deposition caused greater decreases in S042 and base cation
concentrations in lake water and greater recovery in pH and ANC. Within the full population of lake-
watersheds, some showed decreasing ANC and pH values from 1990 to 2050 even under the moderate
and aggressive reduction scenarios. By 2050 to 2100, however, nearly all lakes were simulated to
3-75
-------
experience increasing ANC and pH. The modeled soil base saturation increased very slowly over the
modeled time period compared to changes in surface water chemistry. For 95% of the lake-watersheds
studied, simulated soil base saturation remained below 20% in 2100 under all emissions scenarios.
Darts lake (Adirondacks)
180
160
IS
*£ ©
z 3
£
~
s
+
E
3
'o
5
120
100
80
80
60
40
20
0 .
180
140
120
100
80-
20
0
3 -20
-40-
6.5
6.0
5.S
5.0
4.5
4.0
6
.A r * *
•» • '
~i 1 j i r
w
vvV *AS/ v\\aa
va
# »
i
*"• ...
% *"S
* /
/y ¦>* V "* \A/yVy"r VX*"* ^
% * 11 t ^ #
1 V* •" v^ y * *
o-
1/1/82 1/1/84 1/1/86 1/1/88 1/1/90 1/1/92 1/1/94 1/1/96 1/1/98 1/1/00 1/1/02
Source: Stoddard et al. (2003)
Figure 3-22. Time series data for S042~, N03", base cations [Ca plus Mg], Gran ANC, pH, and DOC in one
example of long-term monitoring in Darts Lake in the Adirondack Park. Shaded box indicates
time period of analyses reported by Stoddard et al. (2003).
3-76
-------
so4
no3
so4 & no3
CB
ANC
H*
AL,
-5 -4 -3-2-1012
Change in Lake Chemistry
(peq/L-yr)
Source: Driscoll et al. (2003b)
Figure 3-23. Mean rates of change in solute concentration in 16 lakes of the Adirondack Long-Term
Monitoring (ALTM) program from 1982 to 2000. Minimum, mean, and maximum changes in
concentrations and number of lakes showing significant trends are shown. All values are
in |jeq/L/yr, except for concentrations of inorganic monomeric aluminum (Al;), which are
expressed in |jM/yr.
Multipollutant Interaction: Biological Mercury Hotspots in the Adirondacks
The Adirondacks has been identified as a region at risk from the combined effects of acidifying
deposition and Hg deposition (Driscoll et al., 2007a). The relationship between atmospheric deposition of
S and enhanced Hg methylation is discussed in Section 3.4. In general, the solubility of Hg increases with
increasing sulfide concentrations in anoxic waters through complexation reactions, potentially increasing
the pool of Hg available for methylation (Benoit et al., 2003; Driscoll et al., 2007a). Evers et al. (2007)
identified a biological Hg hotspot in the western Adirondacks based on Hg concentrations in yellow perch
and common loons. Mean yellow perch Hg concentrations in the Adirondack hotspot were 1.5 to 2.5
times higher than the U.S. EPA and U.S. Food and Drug Administration's reference dose used for fish
consumption advisories (Evers et al., 2007). The authors hypothesized that the occurrence of the
biological hotspot was due in part to the combination of high Hg deposition and sensitive water
chemistry, such as low ANC and pH, which is associated with both natural acidity and the long-term
effects of acidifying deposition (Evers et al., 2007). Driscoll et al. (2007a) concluded that watersheds
sensitive to Hg deposition tend to be forested, have an abundance of wetlands, contain shallow hydrologic
flow paths and low nutrient concentrations, and are affected by acidifying deposition.
Min -Mean - Max
(16)
(16)
(16)
I "
(11)
3-77
-------
MAGIC Model Estimates of ANC Distribution
Adirondack Lakes with ANC < 200 jjeq/L
1850
1990
2100-Aggresive
2100-Base
2100-Moderate
-50 0 50 100 150 200 250
Predicted ANC (neq/L)
Source: Sullivan et al. (2006b).
Figure 3-24. Simulated cumulative frequency distributions of lakewater ANC at three dates for the
population of Adirondack lakes, based on MAGIC model simulations reported by Sullivan
et al., 2006b. Conditions for the year 2100 are presented for three emissions control
scenarios: Base Case, Moderate Additional Controls, and Aggressive Additional Controls.
(See Figure 3-15)
3.2.4.5. Acidification Case Study #2: Shenandoah National Park, Virginia
Shenandoah National Park is located along the crest of the Blue Ridge Mountains in Virginia. Air
pollution within Shenandoah National Park, including S and N deposition and O , concentration, is higher
than in most other national parks in the U.S. Measured wet S deposition in the park has ranged from 8 to
10 kg S/ha/yr in the early 1980s to near 6 kg S/ha/yr since 2000 (Figure 3-25). Dry S deposition may be
nearly as high as wet deposition (Sullivan et al., 2003). Most acidification effects in the park have been
linked with S deposition.
The sensitivity of streams in the park to acidification from acidifying deposition is determined
mainly by the types of rocks found beneath the stream and the characteristics of the watershed soils that
surround it. If the underlying geology is Si-based (siliclastic lithology), the soil and water in the
watershed generally have poor ability to neutralize acids deposited from the atmosphere. About one-third
of the streams in the park are located on this type of geology. Model estimates using the MAGIC model
suggest that such streams have typically lost most of their natural ANC, largely in response to a century of
industrial emissions and acidifying deposition. As a consequence, stream pH values in many streams are
low, especially during winter and spring. Before human-caused air pollution, most streams in Shenandoah
National Park probably had pH above about 6. Many park streams on siliclastic lithology currently have
pH as low as about 5 (Cosby et al., 2006; Sullivan et al., 2003). Other predominant lithologies in the park
include granite-based (granitic) lithologies typically characterized by intermediate ANC streams, and
3-78
-------
basalt-based (basaltic) lithologies typically characterized by relatively high stream ANC (Cosby et aL
2001; Sullivan et al., 2004, 2007, 2008). "
The effects of acidifying deposition on Shenandoah National Park streams have been studied for
over 25 years by the Shenandoah Watershed Study, the longest-running watershed study program in any
of the national parks (Cosby et al., 2006; see http: //swas. evsc .Virginia, edu). This program has determined
that the high rate of atmospheric deposition of S, combined with naturally low contributions from some
rock types of Ca and other base cations (that serve to neutralize acidity), are the most important causes of
low streamwater ANC in many park streams. Some park streams can also become temporarily acidic for
short periods (hours to days) during rainstorms or snowmelt.
12.0
tn 8.0
(/) 6.0
O
O 4.0
~i 1 1 1 1 1 1 1 1 1 1 1 r
1978 1980 1982 1984 1986 1988 1990 1992 1994 1996 1998 2000 2002 2004 2006
Year
Source: Sullivan et al. (2003)
Figure 3-25. Wet suifur deposition for the period of record at the Big Meadows NADP/NTN monitoring
station in Shenandoah National Park.
Hie acidification of streams in the Park is linked to effects that are occurring in the watershed soils.
Over time, the ability of soils to adsorb S, thereby effectively negating sulfur's potential to acidify water,
is decreasing due to the long term accumulation of S04~ on soil adsorption sites in response to a legacy
of acidifying deposition. In addition, the amount of stored Ca and Mg in the soil is gradually declining in
response to acidifying deposition. Therefore, streams are expected to acidify more in the future than they
have so far, relative to the amount of acidifying deposition received. This prognosis is consistent with
recent analysis of national lake and stream response to reductions in air pollution emissions (Stoddard
et al., 2003). Unlike a number of other regions of the country, streams in the region that includes
Shenandoah National Park are generally not recovering from acidification.
A great deal of research has been conducted in the park on the effects of S and N deposition on soil
and water acidification. This park was a major site of early research on acidification processes (Galloway
et al., 1983). This early work provided much of the foundation for development of the MAGIC model
(Cosby et al., 1985c), which has been the most widely used dynamic watershed acid-base chemistry
model worldwide for the past two decades.
Although research on many aspects of acidification effects science has been conducted in the park,
it has been particularly noteworthy for studies on episodic acidification; biological effects of stream
acidification; and dynamic modeling of acidification, recovery, and critical loads. Research within
Shenandoah National Park on each of these topics is discussed below.
3-79
-------
Episodic Acidification
A number of studies of episodic acidification have been conducted in streams within Shenandoah
National Park. Eshleman and Hyer (2000) estimated the contribution of each major ion to observed
episodic ANC depressions in Paine Run, Staunton River, and Piney River during a 3-year period. During
the study, 33 discrete storms were sampled and water chemistry values were compared between
antecedent baseflow and the point of minimum measured ANC (near peak discharge). The relative
contribution of each ion to the ANC depressions was estimated using the method of Molot et al. (1989),
which normalized the change in ion concentration by the overall change in ANC during the episode. At
the low-ANC (~0 j^ieq/L) Paine Run site on siliciclastic bedrock, increases in N03 and S042 . and to a
lesser extent organic acid anions, were the primary causes of episodic acidification. Increases in base
cations tended to compensate for most of the increases in acid anion concentration. ANC declined by 3 to
21 |_ieq/L (median 7 j^ieq/L) during the episodes studied.
At the intermediate-ANC (-60 to 120 j^ieq/L) Staunton River site on granitic bedrock, increases in
S042 and organic acid anions, and to a lesser extent N03 . were the primary causes of episodic
acidification. Base cation increases compensated these changes to a large degree, and ANC declined by 2
to 68 (ieq/L during the episodes (median decrease in ANC was 21 j^icq/L).
At the high-ANC (-150 to 200 j^ieq/L) Piney River site on basaltic (69%) and granitic (31%)
bedrock, base cation concentrations declined during episodes (in contrast with the other two sites where
base cation concentrations increased). S042 and N03 concentrations usually increased. The change in
ANC during the episodes studied ranged from 9 to 163 j^ieq/L (median 57 j^ieq/L) (Eshleman and Hyer,
2000). Changes in base cation concentrations during episodes contributed to changes in the ANC of Paine
Run, had little effect in Staunton River, and contributed to decreases in ANC in Piney River.
The most acidic conditions in Shenandoah National Park streams occur during high-flow periods,
in conjunction with storm or snowmelt runoff. There are several different mechanisms of episodic
acidification in operation in these streams, depending at least in part on the bedrock geology of the stream
watershed. The relative importance of the major processes that contribute to episodic acidification varies
among the streams, in part as a function of baseflow stream water ANC which is largely controlled by
bedrock geology. S-driven acidification was an important contributor to episodic loss of ANC at all three
study sites, probably because S adsorption by soils occurs to a lesser extent during high-flow periods.
This is due, at least in part, to diminished contact between drainage water and potentially adsorbing soil
surfaces along the shallow flow paths. Dilution of base cation concentrations during episodes, which is an
acidifying process, was most important at the high-ANC site.
Thus, episodic acidification of streams in Shenandoah National Park can be attributed to a number
of causes, including dilution of base cations and increased concentrations of sulfuric, nitric, and organic
acids (Eshleman et al., 1995; Hyer et al., 1995). For streams having low pre-episode ANC, episodic
decreases in pH and ANC and increases in toxic Al concentrations can have adverse effects on fish
populations. However, not all of the causes of episodic acidification are related to acidifying deposition.
Base-cation dilution and increase in organic acid anions during high-flow conditions are natural
processes. At least for streams in the Shenandoah National Park, the contribution of N, indicated by
increased N03 concentrations, evidently has been related to forest defoliation by the gypsy moth
(Eshleman et al., 1998; Webb et al., 1995). Significant contributions of H2S04, indicated by increased
S042 concentrations during episodes in some streams, is an effect of atmospheric deposition and the
dynamics of S adsorption on soils (Eshleman and Hyer, 2000).
A recent study by Deviney et al. (2006) used hourly ANC predictions over short time periods to
compute recurrence intervals of annual water-year minimum ANC values for periods of 6, 24, 72, and
168 h. They extrapolated the results to the rest of the catchments using catchment geology and
topography. On the basis of the models, they concluded that many streams in the park have 6- to 168-h
periods of low ANC values, which may stress resident fish populations (Deviney et al., 2006).
Specifically, on the basis of a 4-year recurrence interval, approximately 23% of the land area (44% of the
catchments) can be expected to have conditions for 72 continuous hours that are indeterminate with
3-80
-------
respect to brook trout suitability (ANC 20 to 50), episodically acidic (ANC 0 to 20), or chronically acidic
(ANC less than 0). Many catchments were predicted to have successive years of low-ANC values
potentially sufficient to extirpate some species (Deviney et al., 2006). The authors of the study reported
that smaller catchments are more vulnerable to adverse effects of episodic acidification than larger
catchments underlain by the same bedrock. Catchments with similar topography and size are more
vulnerable if underlain by less basaltic and carbonate bedrock.
Biological Effects of Acidification
A robust relationship between acid-base status of streams and fish species richness was
documented in Shenandoah National Park in the 3-year Fish in Sensitive Habitats (FISH) study (Bulger
et al., 1999). Numbers of fish species were compared among 13 streams spanning a range of pH and ANC
conditions. There was a highly significant (p <0.0001) relationship between stream acid-base status
(during the 7 year period of record) and fish species richness among the 13 streams. The streams with the
lowest ANC hosted the fewest species (Figure 3-15). This study demonstrated biological differences in
low- versus high-ANC streams, including species richness, population density, condition factor, age, size,
and field bioassay survival. Of particular note was that both episodic and chronic mortality occurred in
young brook trout exposed in a low-ANC stream, but not in a high-ANC stream (MacAvoy and Bulger,
1995), and that blacknose dace (Rhinichthys atratulus) in low-ANC streams were in poor condition
relative to blacknose dace in higher-ANC streams (Dennis et al., 1995; Dennis and Bulger, 1995).
Bulger et al. (1999) observed a positive relationship between condition factor and pH in streams in
Shenandoah National (Figure 3-26). Dennis and Bulger (1995) also found a reduction in condition factor
for blacknose dace in waters near pH 6.0. The four populations depicted in (Figure 3-26) with the lowest
condition factor had mean habitat pH values within or below the range of critical pH values at which
Baker and Christensen (1991) estimated that negative population effects for blacknose dace are likely for
the species. The mean condition factor of fish from the study stream with the lowest ANC was about 20%
lower than that of the fish in best condition. Comparisons with the work of Schofield and Driscoll (1987)
and Baker et al. (1990b) suggest that pH values in the low-pH streams are also near or below the limit of
occurrence for blacknose dace populations in the Adirondack region of New York (Sullivan et al., 2003).
MacAvoy and Bulger (1995) used multiple bioassays over 3 years in one of the low-ANC streams
as part of the FISH project to determine the effect of stream baseflow and acid episode stream chemistry
on the survival of brook trout eggs and fry. Simultaneous bioassays took place in mid- and higher-ANC
reference streams. Acidic episodes, with associated low pH and elevated inorganic Al concentrations and
high streamwater discharge, caused rapid fish mortality in the low-ANC stream, while the test fish in the
higher-ANC stream survived (Bulger et al., 1999).
3-81
-------
y = 1.194X +1.519, r2 = .777
10.5
o
o
n
LL
9.5
c
o
T3
C
o
u
8,5
7.5
5.2
5.4
5.6
5.8
6.2
6.4
6.6
6.8
7
7.2
6
mean pH
Source: Bulger et al. (1999).
Figure 3-26. Length-adjusted condition factor (K), a measure of body size in blacknose dace (Rhinichthys
atratulus) compared with mean stream pH among 11 populations (n = 442) in Shenandoah
National Park. Values of pH are means based on quarterly measurements, 1991-1994; K was
measured in 1994. The regression analysis showed a highly significant relationship
(p <0.0001) between mean stream pH and body size, such that fish from acidified streams
were less robust than fish from circumneutral streams.
Modeling of Acidification, Recovery, and Critical Loads
Dynamic models have been used in Shenandoah National Park to help determine whether the
changes in surface water chemistry that have occurred over the past one to two decades will continue and
whether they will reach levels needed to support biological recovery. The most commonly used models
are described in Annex B and details of these analyses are discussed below. In general, model forecasts
indicated that under base case conditions (those expected under existing or anticipated emissions controls)
surface water ANC in the southern Appalachians (and in parts of the Adirondacks) would be likely to
decline in the future. In terms of soil chemistry, projected future improvements in both regions appear to
be slow and in most cases do not reach a base saturation of 20% or more within the next 100 years.
MAGIC model simulations for streams in Shenandoah National Park by Sullivan et al. (2003)
suggested that acidifying deposition would have to be decreased substantially to improve and maintain
acid-sensitive streams at levels of ANC that would be expected to protect against ecological harm. In
addition, it took a long time for these streams to acidify in the past; because of complexities related to soil
conditions, it would take even longer for them to recover in the future. To protect against chronic acidity
in the year 2100, with associated probable lethal effects on brook trout, the authors predicted that
S deposition to the most geologically sensitive siliciclastic lithology watersheds in the park would have to
be kept below about 9 kg S/ha/yr for the next 100 years (Sullivan et al., 2007a). Before the Industrial
Revolution, most streamwater in the Park had ANC higher than about 50 |icq/L. To promote ANC
recovery to 50 j^icq/L in the future, to protect against general ecological harm, S deposition to Si-based
(siliciclastic) watersheds in the park would have to be kept below about 6 kg S/ha/yr. Finally, the authors
3-82
-------
predicted that some watersheds will likely not recover streamwater ANC to values above 50 j^ieq/L over
the next century even if S deposition is reduced to zero (Sullivan et al., 2007a).
Simulation and mapping of watershed responses to historical changes in acidifying deposition
(from preindustrial to current) by Cosby et al. (2006) suggest that large areas of Shenandoah National
Park have suffered deterioration of both soil and stream conditions. The changes in soil condition have
been relatively modest up to the present time, with areas in the southern district of the park moving from
classification of "moderate concern" (watershed average mineral soil percent base saturation 10% to 20%;
the historical baseline) to "elevated concern" (average mineral soil percent base saturation 5% to 10%) as
a result of leaching of base cations from the soils in response to S deposition. Simulation results indicated
that deterioration in stream conditions has been more severe than for soil conditions, with large areas in
the southern district and some smaller areas in the central and northern districts moving from "moderate
concern" (average stream ANC 50 to 100 j^ieq/L) to "elevated concern" (average stream ANC 0 to
50 (ieq/L). Neither soil nor stream conditions have shown any improvement from 1980 to the present in
response to the decline in acidifying deposition over the last 25 years.
Simulation and mapping of watershed responses to predicted future changes in acidifying
deposition by Cosby et al. (2006) were developed following U.S. EPA methods for preparation of
emissions inventory inputs into air quality modeling for policy analysis and rule making purposes. These
alternate emissions scenarios were based on existing emission control regulations and several proposed
alternatives. The model output suggested that the responses of soil conditions to changes in S deposition
are expected to be relatively slow. In the short term (by the year 2020), neither improvement nor further
deterioration is likely to be observed in soil condition regardless of the future deposition scenario
considered. However, model results suggested that constant deposition at 1990 levels would produce
worsening soil conditions in the park by the year 2100 with the development of areas of "acute concern"
(average percent soil base saturation below 5%) in the southern district. Although the scenarios of
possible reduced future deposition did not produce worsening soil conditions, neither did they indicate
any improvement in soil condition, even in the long term.
Simulated responses of stream conditions were more rapid than those of soils. In the short term (by
the year 2020), constant deposition at 1990 levels would likely produce further deterioration in stream
condition. The scenarios of future deposition reductions failed to reverse the deterioration of stream
condition that has occurred during the last century. In the long term (by year 2100), the effects of the
deposition reduction scenarios begin to diverge. The moderate S deposition reduction scenario (69%
reduction from 1990 values) did not produce improvement in stream chemistry relative to current
conditions. The larger deposition reduction scenario (75%), by contrast, produced modest improvements
in stream chemistry by 2100. However, even the relatively large S deposition reductions of this scenario
did not result in a simulated return of stream conditions to the preindustrial state.
To develop projections of probable past and future responses of aquatic biota to changing S
deposition in Shenandoah National Park, the MAGIC model was coupled by Sullivan et al. (2003) with
several empirical models that linked biological response to past and future model projections of water
quality. Unlike MAGIC, which is a geochemical, process-based model, the biological effects estimates
were based on observed empirical relationships rooted in correlation and expressed as linear relationships.
Correlation does not necessarily imply causality, but an observed pattern of covariation between variables
does provide a quantitative context for extrapolation. In this case, the projections did not require
extrapolation beyond the observed ranges of observations, and therefore the projections were statistically
robust. To the extent that the observed empirical relationships used in the coupled models do in fact
reflect the effects of acid stress on aquatic biota, the projections were also biologically robust.
Dynamic water chemistry model projections were combined with biological dose-response
relationships to estimate declines in fish species richness with acidification. A relationship derived from
the data in Figure 3-15 was used by Sullivan et al. (2003) with stream ANC values predicted by the
MAGIC model to provide estimates of the expected number of fish species in each of the modeled
streams for the past, present and future chemical conditions simulated for each stream. The coupled
geochemical and biological model predictions were evaluated by comparing the predicted species
3-83
-------
richness in each of the 13 streams with the observed number of species that occur in each stream. The
agreement between predicted and observed species numbers was good, with a root mean squared error in
predicted number of species across the 13 streams of 1.2 species. The average error was 0.3 species,
indicating that the coupled models were unbiased in their predictions. Model reconstructions of past
species richness in the streams suggested that historical loss of species had been greatest in the streams
located on the most sensitive geological class (siliciclastic). The average number of species lost from
streams on the three bedrock types examined was estimated as 1.6 species on siliciclastic bedrock, 0.4
species on granitic bedrock, and 0.4 species on basaltic bedrock. In the case of the siliciclastic streams,
the projected past changes were much larger than the average error and root mean squared error of the
coupled models, suggesting that the projections were reasonably robust.
3.2.5. Ecosystem Services
The evidence reviewed in this ISA supports that acidification of ecosystems is primarily driven by
NOx, NHX and SOx. Ecosystem services, as defined by Hassan et al. (2005), are broadly grouped into
four main categories (see Section 3.1.3). The specific effects of ecosystem acidification may be grouped
into these categories as follows.
¦ Supporting: altered nutrient cycling, decreased biodiversity, decline of productivity
¦ Provisioning: decline in the richness, abundance, and/or health of fish, other aquatic species and
some terrestrial trees
¦ Regulating: decline in water and soil quality
¦ Cultural: decline in forest aesthetics, fishing, ecotourism and cultural heritage values related to
ecosystem integrity and biodiversity
There are few publications that directly evaluate the ecological effects of acidification in terms of
valuation. Even though acidification is well documented in the scientific literature, the lack of valuation
studies related to them limits full assessments of terrestrial damages from NOx and SOx pollution. As
noted in the case study (see Section 3.2.4.4.), the Adirondack Park in New York is among the best
documented of all areas affected by acidic deposition in the U.S. This park has been the subject of
numerous valuation studies in recent decades on cultural ecosystem services (Morey and Shaw 1990;
Mullen and Menz 1985; Englin et al., 1991; Cameron and Englin 1997; Banzhaf et al., 2006). There are
currently no studies on the valuation of supporting or regulating ecosystem services.
3.3. Nutrient Enrichment Effects from Nitrogen Deposition
The ecological effects caused by atmospheric deposition of Nr are the main focus of this section. As
discussed previously, the scope of this ISA includes assessment of all forms of N compounds that
contribute to nutrient enrichment. The various chemical forms of N can be broadly divided into two
groups: nonreactive (N2 gas) and Nr. Nonreactive N2 gas composes 80% of the total mass of the Earth's
atmosphere, but it is not biologically available until transformed into reactive forms of N. Nr includes all
biologically and chemically active N compounds in the Earth's atmosphere and biosphere (Galloway
et al., 2003). The Nr group includes inorganic reduced forms (e.g., NH3 and NH/), inorganic oxidized
forms (e.g., NOx, HN03, N20, N03 ). and organic N compounds (e.g., urea, amine, proteins, nucleic
acids) (Galloway et al., 2003). Atmospheric N deposition may be composed of numerous chemical
species besides oxides, all of which contribute to ecosystem nutrient enrichment. The ISA evaluates the
nutrient effects of NOx in combination with all other forms of Nr deposition for which information is
3-84
-------
available. Agricultural lands are excluded from this discussion because crops are routinely fertilized with
amounts of N (100 to 300 kg N/ha) that exceed air pollutant inputs even in the most polluted areas
(U.S. EPA, 1993a). These high rates of fertilization can contribute to ground water N03 contamination
and eutrophication of some surface waters, especially estuaries. However, the environmental effects of
agricultural N fertilization is beyond the scope of this assessment.
Organisms in their natural environment are commonly adapted to a specific regime of nutrient
availability. Change in the availability of one important nutrient, such as N, may result in imbalance in
ecological stoichiometry, with effects on ecosystem processes, structure and function (Sterner and Elser,
2002). In general, ecosystems that are most vulnerable to nutrient enrichment from atmospheric N
deposition are those that receive high levels of deposition relative to non-anthropogenic N loading, those
that are N-limited, or those that contain species that have evolved in nutrient-poor environments. The
most common experimental designs to quantify the effects of N deposition on ecosystems are N addition,
N gradient, or observation correlated to changing pollution levels over time. N addition experiments often
use NH4NO3 or (NH4)2S04 additions to simulate the chemical species in atmospheric Nr deposition.
Deposition gradient experiments often only measure oxidized and reduced forms of N and therefore do
not adequately identify all components of Nr deposition. Therefore, publications addressing N additions or
deposition gradients often do not include data on all components of Nr.
The following discussion of N-nutrient deposition begins with the N cascade, which provides a
conceptual foundation for discussing the effects of N on the structure and function of ecosystems.
Subsequent sections include the effects of N deposition on: N cycling, C cycling, biogenic GHG
emissions, biodiversity, as well as the characterization of sensitive ecosystems and regions in the U.S.
Information is presented for ecosystems in which atmospheric deposition dominates total Nr input (i.e.
many non-managed terrestrial ecosystems) and ecosystems in which atmospheric deposition constitutes a
proportion of total Nr load (e.g., some wetlands and estuarine ecosystems).
3.3.1. Reactive Nitrogen and the Nitrogen Cascade
Nr is one of the most important nutrients in practically all ecosystems (Vitousek and Howarth,
1991). Nr is required by all organisms because it is a major constituent of both the nucleic acids that
determine the genetic character of all living things and the enzymes and proteins that drive the
metabolism of every living cell (U.S. EPA, 1993a; Galloway, 1998; Galloway and Cowling, 2002). It is of
critical importance in plant metabolism and it often governs the utilization of P, K and other nutrients.
3-85
-------
Stratospheric
effects
Atmosphere
Particulate
matter effects
Energy production
Ecosystem
Forests &
grasslands NHX
effects
Plant
• Agroecosystem
effects
Food production
Crop Animal
Soil
(terrestrial)
Coastal
effects
00 [J I'.:'
(food; fiber}
uatic)
Aquatic Ecosystem
Groundwater effects
The
Nitrogen
Cascade
Source: Galloway et al. (2003). Reprinted with permission.
Figure 3-27. Illustration of the N cascade showing the movement of the human-produced Nr as it cycles
through the various environmental reservoirs in the atmosphere, terrestrial ecosystems, and
aquatic ecosystems.
An increase in global Nr has occurred over the past century, largely due to three main causes:
widespread cultivation of legumes, rice, and other crops that promote conversion of N2 gas to organic N
through biological N fixation; combustion of fossil fuels, which converts both atmospheric N2 and fossil
N to NOx; and synthetic N fertilizer production via the Haber-Bosch process, which converts nonreactive
N2 to Nr to sustain food production and some industrial activities (Galloway and Cowling, 2002 Galloway
et al., 2003). Food production accounts for much of the conversion from N2 to Nr, and accounts for
geographic redistribution of N as food is shipped to meet population demands and often returned to the
environment via waste water.
Nr accumulates in the environment on local, regional, and global scales (Galloway, 1998; Galloway
and Cowling, 2002; Galloway et al., 2003). This accumulation occurs in the atmosphere, soil, and water
(Galloway and Cowling, 2002), with a multitude of effects on humans and ecosystems (Galloway, 1998;
Rabalais, 2002; Van Egmond et al., 2002; Townsend et al., 2003). The sequence of transfers,
transformations, and environmental effects is referred to as the "N cascade" (Galloway and Cowling,
2002; Galloway et al., 2003)(Figure 3-27).
In general, the results of the N cascade and the various transformations in the N cycle can be both
beneficial and detrimental to humans and to ecosystems (Galloway and Cowling, 2002; Galloway and
Aber, 2003). Among the most important effect of atmospheric N deposition are aquatic eutrophication,
changes in the structure of terrestrial plant communities, disruptions in nutrient cycling, increased soil
emissions of nitrous oxide (N20; a potent greenhouse gas), accumulation of N compounds in the soil,
soil-mediated effects of acidification (see Section 3.2), and increased susceptibility of plants to stress
factors (Aber et al., 1989, 1998; Bobbink, 1998; Dnscoll et al., 2003c; Fenn et al., 1998).
3-86
-------
3.3.2. Nitrogen Enrichment Effects on N Cycling
Given the complexity of the N cycle, a broadly applicable and well-tested predictive model of these
interactions has not yet been developed. There is scientific information with which to make
generalizations about how ecological and biogeochemical processes respond to N deposition. Significant
scientific advancements in recent years have included refinement of theoretical foundations of nutrient
limitation, development and improvement of analytical technologies, and improved understanding of the
role of N in regulating or influencing the cycling of other elements, especially C (see Section 3.3.3).
The key steps in the N cycle that are outlined in Figure 3-28. Processess within this cycle include N
fixation, assimilation, mineralization (conversion of organic N to simple inorganic forms), nitrification
(conversion of reduced inorganic N to oxidized inorganic N), and denitrification (the reduction of N 0; to
NO, N20, and N2 gas by microbes under anaerobic conditions). These steps generally require biologically
mediated transformations. Key organisms involved in transforming N from one form to another include
plants and microbes.
Deposition
Death
Soil
Runoff x
Bacterial
Nitrogen
Fixation
Animal
Proteins
Bacteria
Nitrites
Urea
(Ammonia)
Nitrates
In Soil
Plant
Utilization
Microbial
Utilization
Litter
Production
(Death)
Ammonia
Trace
Gas
Emissions
Gaseous
Nitrogen in
Atmosphere
Nitrogen
Oxides in
Atmosphere
Microbial
Decomposition
Photosynthesis
Nitrates in
Streams,
Rivers, Lakes,
and Oceans
Process altered by
nitrogen saturation
Source: Garner (1994)
Figure 3-28. N cycle (dotted lines indicated processes altered by N saturation).
3-87
-------
In addition to direct effects on the ecosystem where it is deposited, N can be exported from the
system and cause environmental effects in adjacent ecosystem. The two principal mechanisms for N
export or loss from ecosystems are leaching and denitrification. Leaching removes N from terrestrial or
transitional ecosystems, but adds it to aquatic ecosystems. Thus, an export from one ecosystem becomes
an import to another. Denitrification removes N from terrestrial, transitional, and aquatic ecosystems and
adds it to the atmosphere (Davidson et al., 2000; Seitzinger et al., 2006). Although denitrification
provides a pathway for removing excess N from ecosystems, incidental production of NO and N20 during
denitrification is of concern due to the roles of NO as a precursor in the p(03), and N20 as a potent
greenhouse gas (see Section 3.3.4 for N deposition effects on biogenic N20 flux). Here leaching and
denitrification are discussed in addition to other fundamentals of N cycling in terrestrial, transitional, and
aquatic ecosystems.
3.3.2.1. Terrestrial Ecosystems
N-deposition affects terrestrial ecosystems throughout large areas of the U.S. N-availability to
plants in soil is largely controlled by the process of N mineralization, or the microbial conversion from
organic N to simple amino acids and then to inorganic forms such as NH/ and N03 (Schimel and
Bennett, 2004). The two-step, aerobic, microbial process of autotrophic nitrification converts NH44" to
N03~. Nitrification is an acidifying process, releasing 2 mol hydrogen ion (H ) per mol NH/ converted to
N03 (Reuss and Johnson, 1986; see Section 3.2 and Annex B for the effects of acidifying deposition). As
the N cycle becomes enriched through cumulative N addition, N becomes more abundant, competition
among organisms for N decreases, net nitrification rates often increase, and N03 can leach from the
ecosystem (Aber et al., 1989 and 2003).
Numerous experimental 15N-addition studies have been conducted as a way of understanding how
N cycles through terrestrial ecosystems. These studies have shown that trees typically take up only a
small fraction of added 15N; the vast preponderance is retained in the soil (Nadelhoffer et al., 1999a;
Providoli et al., 2005; Templer et al., 2005; Tietema et al., 1998). This pattern persists even a decade after
15N application (Nadelhoffer et al., 2004), but these experiments have been criticized for applying 15N
directly to the soil surface, thereby precluding direct canopy uptake of N from wet or dry deposition
(Sievering, 1999; Sievering et al., 2000). Canopy 15N experiments are now underway, but have not yet
been published. Comparisons of rates of N deposition in throughfall and in total deposition suggest that
forest canopies can take up an average of 16% of total atmospheric N input (Lovett, 1992), but this
interception can be considerably higher (up to 90%) in some N-limited forests with large epiphyte loads
(e.g., Klopatek et al., 2006). It remains unclear how much of the N from deposition that is retained in
vegetation is then used in photosynthetic enzymes (e.g., Bauer et al., 2004).
N in forest ecosystems is stored primarily in the soil, and soil N often exceeds 85% of the total
ecosystem N (Bormann et al., 1977; Cole and Rapp, 1981). Most soil N is contained in organic matter,
typically bound in humic material or organo-mineral complexes that are resistant to microbial
degradation. This N is not directly available for biological uptake by plants or microbes or for leaching
loss into ground water or surface water.
Only what is termed the mineralizable, or labile, pool of N in the soil is considered to be
biologically active (Aber et al., 1989). Bioavailable N often controls photosynthesis and net primary
productivity (NPP) (e.g., Field and Mooney, 1986). Plants obtain N from the soil by absorbing NH/,
N03~, or simple organic N compounds through their roots, or N is taken up by symbiotic organisms (e.g.,
fungi, bacteria, cyanobacteria) in plant roots (cf. Lilleskov et al., 2001; Schimel and Bennett, 2004). Plant
roots, nitrifying bacteria, and microbial decomposers within the soil utilize, and compete for, this
available soil N pool. Plant uptake of N can be energetically costly, as N03 must be reduced to NH/, and
NH4+ fixed into amino acids before N can be used in plant processes. Some species reduce N03 in their
leaves, taking advantage of excess energy from photosynthesis, whereas other species are restricted to the
more energy expensive approach of reducing N03 in their roots.
3-88
-------
N Saturation
The term N-saturation refers to the condition whereby the input of N to the ecosystem exceeds the
requirements of terrestrial biota, and consequently an elevated fraction of the incoming N leaches from
soils to surface waters. The original description of N saturation by Aber et al. (1989) described four
stages. It was revised by Stoddard (1994) and Aber et al. (1998) (Figure 3-29). In Stage 0, N inputs are
low and there are strong N limitations on growth. Stage 1 is characterized by high N retention and a
fertilization effect of added N on tree growth. Stage 2 includes the induction of nitrification and some
N03 leaching, though growth may still be high. In Stage 3 tree growth declines, nitrification and N03
loss continue to increase, but N mineralization rates begin to decline. While not all terrestrial ecosystems
move through the stages of N saturation at the same rate or in response to the same N loading, several
experimental N addition studies and a survey of 161 spruce-fir stands along a N deposition gradient
support the concept of N saturation progressing from the onset of increase in net nitrification and N03
leaching loss to the eventual decline in tree growth and increase in tree mortality (Aber et al., 1998).
Foliar N
200
N Mineralization
150
£
5
> 100
NPP
Ca:Al & Mg;N Ratios
leaching
0
2
3
1
Stage
Source: Aberet al. (1998). Reprinted with permission.
Figure 3-29. Schematic illustration of the response of temperate forest ecosystems to long-term, chronic N
additions. Changes from initial hypotheses of Aber et al. (1989) include the reduction in N
mineralization in stage 3 and the addition of foliar Ca:AI and Mg:N ratios.
Decades of atmospheric deposition of N have increased the availability of N03 and NH/ in some
terrestrial ecosystems to levels where excess N availability results in net nitrification and associated N03
leaching in drainage water. Severe symptoms of N saturation have been observed in the northern
hardwood watersheds at Fernow Experimental Forest near Parsons, West Virginia (Peteijohn et al., 1996);
in high-elevation, nonaggrading spruce-fir ecosystems in the Appalachian Mountains (Cook et al., 1994);
throughout the northeastern U.S. (Aber et al., 1989; 1998); and lower-elevation eastern forests (Edwards
and Helvey, 1991; Peteijohn et al., 1996; Adams et al., 1997, 2000).
Mixed conifer forests and chaparral watersheds with high smog exposure in the Los Angeles Air
Basin also are N-saturated and exhibit the highest stream water N03 concentrations documented within
wildlands in North America (Bytnerowicz and Fenn, 1996; Fenn et al., 1998). In general, it is believed
that deciduous forest stands in the eastern U.S. have not progressed toward N-saturation as rapidly or as
far as coniferous stands. Deciduous forests may have a greater capacity for N retention than coniferous
forests. In addition, deciduous forests tend to be located at lower elevation and receive lower atmospheric
3-89
-------
inputs of N. Many deciduous forests have higher rates of N uptake and greater N requirement than
coniferous forests (Aber et al., 1998).
NO3- Leaching
Nitrate leaching is the process of nitrate moving from one ecosystem compartment to another. The
nitrate may first leach from terrestrial ecosystems to soil drainage waters where it then increases surface
water N03 concentrations and may be exported downstream. Because the pathway of NO, leaching
crosses over multiple ecosystem types is discussed in both the terrestrial, wetland and aquatic sections of
the ISA. Two of the primary indicators ofN enrichment in forested watersheds are the leaching of NO;,
in soil drainage waters and the export of NO; in stream water, especially during the growing season
(Stoddard, 1994). The concentration of N03 in surface water provides an indication of the extent to
which N deposited atmospherically or otherwise leaches from the terrestrial ecosystem.
In most upland forested areas in the U.S., most N received in atmospheric deposition is retained in
soil (Nadelhoffer et al., 1999a). Several different data compilations indicate that 80% to 100% of N
deposition is retained or denitrified within terrestrial ecosystems that receive less than about 10 N/ha/yr
(Aber et al., 2003; Dise and Wright, 1995; Kristensen et al., 2004; MacDonald et al., 2002; Sullivan,
2000a). In general, because much of the atmospherically deposited N is retained within the terrestrial
ecosystem or denitrified during export, a relatively small fraction of this N reaches downstream estuaries
(Alexander et al., 2002; Castro et al., 2001; Seitzinger et al., 2002; van Breemen et al., 2002).
Despite retention of most atmospheric N deposition within the terrestrial environment, N-related
adverse effects on aquatic life do occur (Driscoll et al., 2003d). For example, although 70% to 88% of
atmospheric N deposition was retained in the Catskill Mountains watersheds in upstate New York, fish
populations could not be sustained because high NO;, concentrations in stream water during high flows
caused the concentrations of inorganic Al to reach toxic levels (Lawrence et al., 1999; see Section 3.2).
Summer (n = 350) Spring (n - 212)
O Lakes
Streams
4 6 8 10 12 6 8 10 12
Estimated N deposition (kg per ha per yr)
Source: Aber et al. (2003). Reprinted with permission.
Figure 3-30. Surface water N03~ concentrations as a function of N deposition at the base of each
watershed in summer and spring. N deposition to the whole watershed may be 2 to 6 kg
N/ha/yr greater than at the base.
3-90
-------
In the northeastern U.S., an analysis by Aber et al. (2003) of data collected during the mid- to late
1990s from lakes and streams suggested that nearly all N deposition is retained or denitrified in
northeastern watersheds that receive less than about 8 to 10 kg N/ha/yr. An analysis of N deposition to
forestland in the northeastern U.S. based on Ollinger et al. (1993) suggested that approximately 36% of
the forests in the region received 8 kg N/ha/yr or more and may therefore be susceptible to elevated NO;,
leaching (Driscoll et al., 2003d). Aber et al. (2003) further found that surface water NO;, concentrations
exceeded 1 (ieq/L in watersheds receiving about 9 to 13 kg N/ha/yr of atmospheric N deposition (Figure
3-30). The lakes and streams found to have high NO; concentration were those receiving N deposition
above this range, but responses were variable among those receiving high N deposition. Above this range,
mean NO; export increased linearly with increasing deposition at a rate of 0.85 kg NO; kg N/ha/yr for
every 1 kg N/ha/yr increase in deposition, although there was considerable variability in N retention
among watersheds at higher rates of deposition (Figure 3-31) (Aber et al., 2003).
12
9
6
3
0
O
o
90%
O 50%
30%
5 7 9 11 13
N deposition (kg per ha per yr)
Source: Aber et al. (2003). Reprinted with permission.
Figure 3-31. a) N export in stream water as a function of N deposition at the base of sampled watersheds.
N export is represented by the equation NBexp = 0.85 Ndep - 5.8; r2 = 0.56; p <0.001. (b)
Watershed N retention decreases as N deposition at the base of the watersheds increases (N
retention = -0.07 N deposition + 1.44; r2 = 0.50).
3-91
-------
In the West, a study of mixed conifer forests of the Sierra Nevada and San Bernardino mountains
reported that the deposition load that causes the onset of N03 leaching is 17 kg N/ha/yr. Several studies
in the Rocky Mountains indicate that the onset of N03 leaching in alpine catchments occurs at
approximately 10 kg N/ha/yr (Baron et al., 1994; Williams and Tonnesen 2000).
In other studies, the isotopic signature of 180 in streamwater N03 indicates that only a small
percentage of the incoming N03 from atmospheric deposition leached directly to drainage waters (e.g.,
(Spoelstra et al., 2001; Burns and Kendall, 2002; Pardo et al., 2004). The rest of the N03 that leached
from the terrestrial ecosystem was cycled by biota in soils or streams before being exported. That cycled
N may have originated in atmospheric deposition, but its origin was not identified.
In general, field experiments have shown that N03 leaching can be induced by chronic addition of
N (Kahl et al., 1993, 1999; Norton et al., 1999; Edwards et al., 2002a; Peteijohn et al., 1996) (See Table
3-11). Several N-exclusion studies in Europe demonstrated that decreases in N deposition produced
immediate reductions in N03 leaching from forest stands (Gundersen et al., 1998; Quist et al., 1999). At
a regional scale, the leaching transport of N from terrestrial to freshwater systems has important
implications beyond its impact on upland lakes and streams, because N exports can ultimately also
contribute to the eutrophication of coastal ecosystems (Howarth et al., 1996; Driscoll et al., 2003d).
Denitrification
The role of denitrification in terrestrial ecosystems is important to understand the fraction of the
atmospheric deposition that is returned to the atmosphere and therefore does not have direct effects on
terrestrial and aquatic ecosystems. Denitrification has been difficult to measure directly in most
ecosystems, due to the difficulty of measuring small changes in N2 and to the great degree of spatial and
temporal heterogeneity inherent in the denitrification process (Davidson and Seitzinger, 2006). Additional
information on measurement techniques is available in Annex C.
Denitrification is accomplished by facultative anaerobic denitrifying bacteria, and occurs only
under anaerobic conditions in the presence of sufficient N03 and organic C. Hence, most terrestrial
denitrification occurs in "hotspots," that is, in sporadically wet places or times or in anaerobic soil
microsites (McClain et al., 2003; Seitzinger et al., 2006). The high organic matter content of terrestrial
soils provides an ample supply of C. Therefore, the factors that typically limit the rate of denitrification in
terrestrial ecosystems are the N03 supply and the occurrence of anaerobic conditions (See Annex C for
additional studies and Section 3.3.4 for an analysis of N deposition effects on N20 flux from terrestrial
and wetland ecosystems).
Using a simple model of the fate of global N inputs to terrestrial ecosystems, Seitzinger et al.
(2006) estimated that denitrification in terrestrial soils removed 46% (124 Tg/yr) of global N inputs from
all sources (N deposition, fertilizer, and N fixation). Half of this denitrification (66 Tg/yr) was estimated
to have occurred in agricultural systems. However, this model assumed that all N entering terrestrial
systems was leached as N03 if it was not taken up by plants. Hence, the model overestimated the
potential for denitrification by the extent to which N accumulated in soils or ground water (Seitzinger
et al., 2006).
Foliar N Concentration
The concentration of N in plant foliage, especially in forest trees, can provide an indicator of
nutrient enrichment (McNeil et al., 2007) (Table 3-11). This indicator may be especially relevant because
there is a potential to acquire regional-scale data on foliar N through remote sensing techniques. This
allows rapid assessment of N status across large land areas. The N content in tissue of some plant species
varies in proportion to N inputs (Baddeley et al., 1994; Hyvarinen and Crittenden, 1998; Pitcairn et al.,
2003). Similarly, species typical of nutrient-poor environments tend to accumulate the amino acid
arginine in plant tissue (Van Dijk and Roelofs, 1988). Therefore arginine concentration varies in
proportion to N inputs. Foliar N and foliar arginine concentrations both provide good indices of N
deposition effects.
3-92
-------
There are reported interactions between N content and insects. Increased N content has been shown
to increase palatability to defoliating insects and therefore increasing the extent of defoliation (Nordin
et al., 1998; Forkner and Hunter, 2000).
Soil Carbon-to-Nitrogen Ratio
The N and C cycles are tightly coupled in forest soils. For example, N03 leaching has been
correlated with forest floor C:N. Nitrification and N03 leaching rates are generally low on sites having
soil C:N ratios above about 22 to 25 (Lovett et al., 2002; Ross et al., 2004). The C:N ratio of the forest
floor can be changed by N deposition over time, although it is difficult to detect a change over time
against the background of spatial heterogeneity (Galloway et al., 2003). The forest floor C:N ratio has
been used as an indicator of ecosystem N status in mature coniferous forests (Table 3-11).
Table 3-11. Summary of biogeochemical indicators of N deposition to terrestrial ecosystems.
Region/Country Endpoint
Observations
Forest Type/ Species
Reference
White Mountains Soil C:N Ratio Observation: field relationships between soil C:N
New Hampshire Nitrification ra''0, canoPy lignin:N ratio, and high spectral reso-
lution remote sensing data was used to predict
spatial patterns in C:N. Remote-sensed data were
obtained from NASA's Airborne Visible and Infra-
red Imaging Spectrometer (AVIRIS) instrument.
Preliminary regional estimates of soil C:N ratio
suggested that 63% of the land area in the region
had C:N below 22, which was suggested as a
critical threshold for the onset of nitrification. Be-
low C:N = 22, increasing, but variable, rates of
nitrification were found.
Broadleaf and mixed conifer/ sugar
maple (Acer saccharum), red maple
(Acer rubrum), American beech
(Fagus grandifolia), yellow birch
(Betula alleghaniensis), paper birch
(Betula papyrifera), red spruce
(Picea rubens), balsam fir (Abies
balsamea), eastern hemlock (Tsuga
canadensis)
Ollinger et al.
(2002)
Europe
Soil C:N Ratio
NO3" Leaching
Observation and Modeling: data from 160 sites
across Europe was used to determine that NO3"
leaching when soil C:N ratio is above 30. NO3"
leaching occurs with foliar N <13 mg N/g. The
responses of soil solution nitrate concentration to
changes in N input are more pronounced in broad-
leaf than in coniferous forests, because in Euro-
pean forests broadleaf species grow on the more
fertile soils.
Broadleaf and mixed conifer/ three
broadleaf species (Quercus robur,
Quercus petraea, Fagus sylvatica)
and six conifer species (Picea abies,
Picea sitchensis, Abies alba, Pinus
syivestris, Pinus nigra, Pseudotsuga
menziesii)
Kristensen
etal. (2004)
Northeast U.S.
Soil C:N Ratio
Nitrification
surface water
[NO3-]
foliar [N ion]
Observation: In a compilation of soil C:N and
nitrification data from 250 plots across an N
deposition gradient of 3.3-12.7 kg N/ha/yr
showed a statistically significant but weak correla-
tions between both soil C:N and nitrification to
annual N deposition rate. Across plots, nitrification
increased sharply as C:N ratio (by mass) de-
creased below about 22. Significant foliar chemis-
try alterations only were attributable to effects of
climate and elevation. Strong relationship between
surface water [NO3"] and flux across N deposition
gradient.
Hardwood and conifer forests; Red Aber et al.
Spruce (Picea rubens), Sugar maple (2003)
(Acer saccharum)
3-93
-------
Region/Country Endpoint
Observations
Forest Type/ Species Reference
Continental Divide,
Colorado
Soil C:N Foliar
C:N
Mineralization
Soil %N
Foliar N:Mg
Foliar N:P
Deposition Gradient: Comparison of Englemann
spruce forest stands on the east (3 to 5 kg
N/ha/yr) and west (1 to 2 kg N/ha/yr) slopes of the
Continental Divide in Colorado. The east slope
receives higher levels of N deposition due to proxi-
mal agricultural and urban areas. East slope sites
showed (1) lower soil organic horizon C:N ratio,
(2) lower foliar C:N ratio, (3) higher potential net
mineralization, and (4) higher percent N, N:Mg
ratio, and N:P ratio in foliage. These results sug-
gested that even moderate levels of N deposition
input can cause measurable changes in spruce
forest biogeochemistry.
Englemann spruce (Picea
englemannii) dominated mesic
spruce-fir forests; Vaccinium spp.
Understory, subalpine fir (Abies
lasiocarpa) and lodgepole pine
(,Plnus contorta)
Rueth and
Baron (2002)
Adirondacks, New Foliar [
York
Deposition Gradient: Eight of nine major canopy
tree species had increased foliar N in response to
a gradient of roughly 4 to 8 kg N/ha/yr for wet
deposition (total N deposition gradient of 5 to 10 N
ha/yr). Species specific differences were strongly
related to two functional traits that arise from
within-leaf allocations of N resources: leaf mass
per area and shade
Broadleaf and mixed conifer/ paper
birch (Betula papyrifera), yellow
birch (Betula alleghaniensis), red
maple (Acer rubrum), sugar maple
(Acer saccharum), American beech
(Fagus grandifolia), white pine
(Pinus strobus), eastern hemlock
(Tsuga canadensis), balsam fir
(Abies balsamea), red spruce (Picea
rubens)
McNeil etal.
(2007)
Scotland Foliar [N] & Deposition Gradient: Total tissue N and arginine
[arginine] concentrations were closely correlated with both
atmospheric NH3 concentration and estimated N
deposition (r2 >1.97 and >0.78, respectively), and
this affected species composition in a gradient of
nitrophilous plant species closer to the N source
replacing the more N sensitive species
Mixed woodland; plantation of Pinus Pitcairn et al.
syivestris; Fagus sylvatica', various (2003)
mosses, ferns, forbs
Vermont
Nitrification
Basal area
growth
Mineralization
Foliar %N
Field Addition: Additions of 25 kg N/ha/yr to spruce
plots (ambient bulk deposition 5.4 kg N/ha/yr), in
which net nitrification did not occur before treat-
ment, triggered net nitrification in the second year
of treatment, whereas nitrification was not trig-
gered until the third year in plots receiving 19.8 kg
N/ha/yr. Positive correlation between forest floor %
N, nitrification potential and foliar %N increased
with N addition
High elevation spruce-fir stand; Red McNulty et al.
spruce (Picea rubens), maple (Acer (1996)
spp.) and birch (Betula spp.)
Colorado
Mineralization
Nitrification
Foliar [N] Or-
ganic Soil [N]
andC:N
Field Addition: Additions of 25 kg N/ha/yr to plots
in Loch Vale watershed (ambient bulk deposition
-3.2-5.5 kg N/ha/yr) doubled N mineralization
rates and stimulated nitrification, while the addition
of the same amount to plots receiving ambient
bulk deposition of ~1.7 kg N/ha/yr in Fraser
Experimental Forest elicited no microbial response
but significantly increased foliar and organic soil
horizon N, and decrease soil C:N.
Old-growth spruce-fir; Engelmann Rueth et al.
spruce (Picea engelmanni) Subal- (2003)
pine fir (Abies lasiocarpa) Lodge-
pole pine (Pinus contorta) Vaccinium
spp.
3-94
-------
Region/Country Endpoint
Observations
Forest Type/ Species Reference
Harvard Forest, MA Soil [NO3"
Soil [NH4+
Field Addition: Elevated concentrations of NO3"
plus NH4+ in soil water observed after 1 year of
150 kg N/ha/yr doses, and after 5 years of 50 kg
N/ha/yr doses. In plots that received additions of
150 kg N/ha/yr, elevated concentrations were de-
tected on the seventh year. In plots receiving
50 kg N/ha/yr, elevated soil concentrations were
not observed after 15 years of treatment.
Red pine plantation and mixed Magill et al.
hardwood/ Pinus resinosa (2004)
Bear Brook Water- NO3" Leaching Field Addition: Ammonium sulfate ([NhU^SCU)
shed, ME fertilization of a forested watershed resulted in
long-term increases in NO3" concentration in
stream water and high annual export of N. The
annual retention of N decreased from 96 % to
81 %, with a cumulative retention of 82% of N in-
puts, mostly in soil. The export of N from the refer-
ence watershed has declined from 178 to 23 kg
N/ha/yr during the treatment period.
Mixed hard and soft wood
Kahl et al.,
(1999); Norton
etal. (1999)
Tunk Mountain NO3" Leaching Observation: Changes in watershed processing of Mixed hard and soft wood
Watershed, ME nitrogen may influence acid-base status of surface
waters. The very low concentrations of NH4+ and
NO3" in lake waters in Maine (Kahl etal., 1991)
suggest that changes in N-status have not played
a major role in changes in ANC. However, N loss
to surface waters is well documented in Adiron-
dack surface waters (Driscoll et al., 1991), coinci-
dent with the decline in ANC. The Wild River data
also showed increased NO3" concentrations dur-
ing 1964-83. Different responses in soil reactions
among sites could account for the different re-
sponse of ANC to changing deposition patterns.
(Kahlet al.,
1993)
Fernow Experimen-
tal Forest, WV
NO3" Leaching
Ca Leaching
Observation: The percentage of conductivity attrib-
utable to NO increased similar to concentration. In
contrast, the percentage of conductivity attribut-
able to Ca2+ decreased slightly overtime. The Ca2+
is believed to be pairing with the NO as the NO is
leaching through the soil. While nitrification in ma-
ture stands can be strongly inhibited, limited nitrifi-
cation, especially in forest gaps, and high
anthropogenic inputs of NO3" probably were pri-
mary sources of leached NO3" Preferential
adsorption of SO42", rather than NO3", on soil
colloids is given as an explanation for the lack of
retention of NO3" in the soil system and subse-
quent leaching to the stream.
Mixed hardwoods/ principally north-
ern red oak (Quercus rubra L.),
American beech, red maple (A
rubrum L.), sugar maple, sweet
birch (6. lenta L.), and black cherry
(Prunus serotina)
Edwards and
Helvey (1991)
Fernow Experimen- NO3" leaching Observation: N saturation observed. Progressive
tal Forest, WV increases in streamwater NO3"and Ca concentra-
tions were measured at the Fernow Experimental
Forest in the 1970s and 1980s. This watershed
has received higher N deposition (average
throughfall input of 22 kg N/ha/yr of N in the
1980s) than is typical for low-elevation areas of
the eastern U.S., however (Eagar et al., 1996),
and this may help to explain the observed N
saturation.
Mixed hardwoods
Peterjohn
(1996)
3-95
-------
Region/Country
Endpoint
Observations Forest Type/ Species
Reference
Fernow Experimen-
NO3" leaching
Field Addition: Annual experimental additions of 40 Mixed hardwoods
Adams et al.
tal Forest, VW
kg S/ha/yr and 35 kg N/ha/yr as ammonium sulfate
fertilizer were applied to a 34 ha watershed with a
25-year-old stand of central Appalachian hard-
woods. An adjacent watershed served as the con-
trol. After 5 years of treatment (total additions of
275 kg S/ha and 220 kg N/ha), stream water NO3",
Ca2+, Mg2+ concentrations and export increased.
Soil solution concentrations provide evidence that
the treatment watershed is nitrogen-saturated,
which was unexpected for such a young stand.
(1997; 2000)
Fernow Experimen-
NO3" Leaching
Field Addition: Three times per year N was added Mixed hardwoods
Edwards et al.
tal Forest, VW
as (NH4)2S042- to watersheds. In the spring and
autumn 7.1 and 8.1 kg N/ha/yr were added, and in
the summer 21.3 and 24 kg N/ha/yr were added.
Anion adsorption on WS3 apparently delayed
increases in SO42" leaching, but resulted in en-
hanced early leaching losses of CI and NO3".
Leaching of Ca and Mg was strongly tied to NO3"
and SO42" leaching. F-factors for WS3 baseflow
and peakflow indicated that the catch ment was
insensitive to acid neutralizing capacity reductions
both before and during treatment, although NO3"
played a large role in reducing the treatment pe-
riod F-factor.
(2002b)
Greater Los Ange- N Saturation Observation: plant communities exposed to air
les area, CA pollution received sufficiently high levels of atmos-
pheric N deposition to be N saturated. Symptoms
of N saturation were evident in mixed conifer or
chaparral sites receiving atmospheric deposition of
20 to 25 kg N/ha/yr or higher. In the more highly
polluted study site (1) accumulation of NO3" in
foliage of plants (2) accumulation of NO3" in soil
(3) NO and N2O emissions higher (4) higher N
mineralization (5) low foliar and soil C:N, high
foliar N:P.
Sierra Nevada and NO3-Leaching Critical load for increased NO3" leaching Mixed conifer forests; Ponderosa Fennetal.
San Bernardino, CA calculated as 17 kg N/ha/yr and the empirical pine (Pinus ponderosa) Jeffrey pine (2008)
critical load for adverse effects on lichen (P. jeffreyi.); white fir (Abies con-
communities is 3.1 kg N/ha/yr. Much of this study color);- California black oak (Quer-
area is far above such levels of N deposition. cus kelloggii), incense cedar (Calo-
cedrus decurrens, the lichen Le-
tharia vulpina
Mixed conifer forest; Ponderosa Fenn et al.
pine (Pinus ponderosa)', Jeffrey pine (2000)
(P. jeffreyi); white fir (Abies con-
coiof)', California black oak (Quercus
kelloggii)', incense cedar (Calo-
cedrus decurrens)
and lower ends, respectively, of Devil Canyon
West Fork
Chaparral and mixed conifer forest; Fenn et al.
Ponderosa pine (Pinus ponderosa), (1996)
Jeffrey pine (P. jeffreyi), white fir
(Abies concolor), California black
oak (Quercus kelloggii), iincense
cedar (Calocedrus decurrens),
sugar pine (Pinus lambertiana),
bracken fern (Pteridium aquilinum
var. pubescens Underw.)
San Bernardino N Saturation Deposition Gradient: over the range of 12.1 to
Mountains, CA 31.7 kg N/ha/yr, the site nearest to urban area
(Los Angeles) received much more N deposition,
as well as other pollutants (i.e. S deposition), and
received much more fog, coinciding with much
more wet deposition of N in that site. Ecosystem
was N saturated, as evidenced by high streamwa-
ter NO3" concentration, 151 and 65 [jeq/L at upper
3-96
-------
Region/Country Endpoint
Observations
Forest Type/ Species
Reference
San Bernardino DIN Export Deposition Gradient: over the range of 11 to 40kg
Mountains, CA N/ha/yr, dissolved inorganic N (DIN) export was
scale dependent, with highest export occurring in
watersheds of ~150/ha. Differences attributed to
temporal asynchrony between N availability and
biological demand
Mixed forest- chaparral, hardwood,
coniferous; White alder (Alnus
rhombifolia)', California Bay Laurel
(iUmbellularia californica)] Scrub oak
('Quercus dumosa)', Coast live oak
('Quercus agrifblia)
Meixner and
Fenn (2004)
Rocky Mountain N Retention Results from several studies suggest that the ca-
Alpine Catchments pacity of Rocky Mountain alpine catchments to
sequester N is exceeded at input levels of about
10 kg N/ha/yr
Rocky Mountain alpine
Baron et al.
(1994); Wil-
liams and
Tonnessen
(2000)
Disturbance and stand age effects on N retention
The varying degree of N assimilation, leaching and microbial transformation often reflect
differences in N status among treatment sites. These variations have most often been attributed to
disturbance history, dating back a century or more (Goodale and Aber, 2001). Sites which have undergone
disturbances that cause loss of soil N, such as logging, fire, and agriculture, tend to be most effective at
retaining atmospheric and experimental inputs of N. Fire causes substantial N losses from ecosystems
(see Table 3-12). Timber harvest contributes to nutrient removal from the ecosystem via biomass export
and acceleration of leaching losses (Bormann et al., 1968; Mann et al., 1988). In particular, logging
contributes to loss of N and Ca2+ from the soil (Tritton et al., 1987; Latty et al., 2004). N retention
capability often decreases with stand age, which suggests that older forests are more susceptible than
younger forests to becoming N-saturated (Hedin et al., 1995). Aber et al. (1998) surmised that land use
history may be more important than cumulative atmospheric deposition of N in determining the N status
of a forest ecosystem. See Annex C for a more detailed discussion of how disturbance affects N cycling.
Table 3-12. Effects of fire on nutrient concentrations in forests in Nevada and California.
Region/Country Endpoint
Observations
Grassland Type/ Species Reference
Lake Tahoe Basin, Nutrient concen- Field Measurement: Compared runoff from fixed
Nevada tration in runoff plots within wildfire-burned and unburned areas in
both summer and winter seasons. Wildfire in-
creased the frequency and magnitude of elevated
nutrient in discharge runoff for all 3 parameters
studied: NO3" N, ammonium nitrogen, phosphate
P. The mobilization of nutrients was increased
due to wildfire, but the lack of 0 horizon material
(surface organic layer of mineral soils) after burn-
ing may ultimately reduce discharge concentra-
tions over time
Jeffery pine, white fir, sugar pine,
Sierra chinquapin, currant, and snow
brush, bitterbrush
Soils: Cagwin series
Miller (2006)
3-97
-------
Region/Country Endpoint
Observations
Grassland Type/ Species Reference
Lake Tahoe, Ne- Leaching, N Field Measurement: Fire and post-fire erosion
vada concentrations in caused large and statistically significant losses of
forest floor and C, N, P, S, Ca, and Mg from the forest floor; Be-
soil fore the burn, there were no significant differ-
ences in leaching, but during the first winter after
the fire, soil solution concentrations of NH4+,
NO3", ortho-P, and (especially) SO42" were ele-
vated in the burned area, and resin lysimeters
showed significant increases in the leaching of
NH4+and mineral N. The leaching losses of min-
eral N were much smaller than the losses from
the forest floor and A11 horizons. The major
short-term effects of wildfire were on leaching,
whereas the major long-term effect was the loss
of N from the forest floor and soil during the fire.
Sierra Nevada mixed conifer forest:
Jeffrey pine (Pinus jeffreyi), white fir
(Abies concoloi), sugar pine (Pinus
lambertiana) and incense-cedar
('Calocedrus decurrens). Understory
vegetation: green leaf manzanita
(.Arctostaphyios patuia), snowbrush
('Ceanothus velutinus).
Soils: Cagwin series: coarse, loamy
sand
Murphy
etal. (2006)
Sierra Nevada, Forest floor and Field Experiment: investigated the effect of forest
California nutrient content, thinning treatments and prescribed burning on C,
soil chemical N, ortho-P, and SO42" in the forest floor organic
properties, and layer and surface soil mineral horizons. The study
soil leaching included a prescribed fire and three timber
harvest treatments: whole-tree thinning (WT) cut-
to-length thinning (CTL), and no harvest (CONT).
There were no statistically significant effects of
burning on soil C, N, C:N ratio, Bray-extractable
P, exchangeable Ca2+, K+, or Mg2+ Burning had
no significant effect on soil solution pH, ortho-P,
SO42", NO3", or NH4+ as measured by ceramic
cup lysimeters and no effect on the cumulative
leaching of ortho-P, NO3", or NH4"* as measured
by resin lysimeters. Prescribed fire had little
impact on total and soluble nutrients in the upper
mineral soil layer. Loss of N capital from the
forest floor appears to be the major effect of
prescribed burning.
Jeffery pine (Pinus jeffreyi) forest
Murphy
etal. (2006)
Sierra Nevada, Nutrient budget of Field Measurement/Modeling: effects of fire, post-
California C, N, Ca, P, K, S, fire salvage logging, and revegetation on nutrient
Mg budgets were estimated for a site that burned in a
wildfire in 1981.2 decades after the fire, the
shrub ecosystem contained less C and more N
than the adjacent forest ecosystem. C was ex-
ported in biomass during salvage logging and will
not be recovered until forest vegetation occupies
the site again. Most N was lost via volatilization
during the fire rather than in post-fire salvage
logging (assuming that foliage and 0 horizons
were combusted). Comparison of the pre-fire and
present day N showed the lost N was rapidly
replenished in 0 horizons and mineral soils,
probably due to N-fixation by snowbush. No
differences in ecosystem P, K, or S contents or in
soil extractable P or S between the shrub and
forested plots. K+, Ca2+, and Mg2+were greater in
shrub than in adjacent forested soils. The large
increase in Ca resulted from either the release of
Ca from non-exchangeable forms in the soil or
the rapid uptake and recycling of Ca by post-fire
vegetation.
110-130 year old Jeffery pine (Pinus Johnson
jeffreyi] etal. (2005)
3-98
-------
Region/Country Endpoint
Observations
Grassland Type/ Species Reference
Little Valley, Ne- C and N loss Field Measurement/ Modeling: On an ecosystem
vada level, the fire consumed approximately equal
percentages of C and N (12 and 9%, respec-
tively), but a greater proportion of aboveground N
(71%) than C (21%). Salvage logging was the
major factor of C lost, and C lost will not be re-
plenished until forest vegetation is reestablished.
N2 fixation by Ceanothus velutinus in the post-fire
shrub vegetation appears to have more than
made up for N lost by gasification in the fire over
the first 16 year, and may result in long-term
increases in C stocks once forest vegetation
takes over the site. N loss from the fire equaled
>1,000 years of atmospheric N deposition and
>10,000 years of N leaching at current rates.
Calculations of C and N losses from theoretical
wildfires in the IFS sites show similar patterns to
those in Little Valley. Calculated losses of N in
most of the IFS sites would equal many centuries
of leaching. Conceptual models of biogeochemi-
cal cycling in forests need to include episodic
events such as fire.
Jeffery pine (Pinus jeffreyi) Mesic Johnson
forests in the Integrated Forest et al. (2004)
Study (IFS).
N. Lake Tahoe,
Nevada;
Truckee, California
(Tahoe National
Forest);
Glenbrook, Nevada
(Lake Tahoe Basin)
C and N loss Field measurement: The quantities of C and N
volatilized from the forest floor by prescription fire
in the Sierra Nevada were measured at three
sites: Marlene, Sawtooth and Spooner. C losses
calculated by the weight method were 6.12, 7.39,
and 17.8 mg C/ha at the Sawtooth, Marlene, and
Spooner sites. N losses calculated by the weight
method were 56.2,60.8, and 362 kg N/ha, at the
Sawtooth, Marlene, and Spooner sites, respec-
tively. N volatilization during prescribed fire is the
dominant mechanism of N loss from these sys-
tems.
Marlene: Jeffery pine, white fir, Caldwell
snowbrush, squawcarpet, greenleaf et al. (2002)
manzanita, pinemat manzanita,
Soils: Cagwin series
Sawtooth: Jeffery and Ponderosa
pine, Soils: Kyburz series
Spooner: mixed confer, red fir, white
fir, snowbrush and manzanita,
Soils: Tahoma series
3.3.2.2. Wetland Ecosystems
N dynamics in wetland ecosystems vary in time, with type of wetland and with environmental
factors, especially water availability (Howarth et al., 1996). A wetland can act as a source, sink, or
transformer of atmospherically deposited N (Devito et al., 1989) and these functions can vary with season
and with hydrological conditions. Vegetation type, physiography, local hydrology, and climate all play
significant roles in determining source and sink N dynamics in wetlands (Devito et al., 1989; Koerselman
etal., 1993; Arheimer and Wittgren, 1994; Mitchell et al., 1996).
N Fixation and Mineralization
N fixation and mineralization are two mechanisms by which N becomes available for plants. It is
documented that ecosystems may derive substantial amounts of new N inputs viaN2-fixation (Hurd et al.,
2001). N mineralization has been shown to increase with N addition, and this can cause an increase in
wetland N export to adjacent surface water (Groffman, 1994). Drought has been shown to inhibit
mineralization and nitrification in soils leading to a decrease in N03 concentration (Foster et al., 1992).
However, drying may stimulate mineralization upon re-wetting (Kieft et al., 1987). A laboratory study
3-99
-------
showed that within 24 h of re-wetting, extractable N03 concentration in dried peat increased
approximately 7-fold as compared to continuously moist peat (Watmough et al., 2004).
NO3- Leaching
Leaching losses from wetlands are rarely considered separately from leaching losses from upland
terrestrial environments and the wet environments that occur in upland catchments. That is, when
leaching losses from terrestrial ecosystems are quantified based on stream exports, estimated leaching
losses implicitly include the net loss of N03 from both terrestrial ecosystems and adjacent wetlands.
Leaching losses of N03 in water derived directly from wetlands are often small because of N03 removal
by denitrification. However, hydrologic flowpaths that deliver water to streams by bypassing wetland
soils can deliver substantial quantities of N03~-rich water from terrestrial uplands.
Denitrification
Transitional ecosystems can remove significant quantities of N03 from water because they
represent a convergence of conditions ofN03 . 02, and C that are requisite for denitrification.
Denitrification is frequently optimized when N03 from more oxic upland areas passes through wet, often
C-rich and anoxic wetlands. In some cases N03 concentration was found to be a better predictor of
denitrification rates than soil moisture (Groffman, 1994), and there is evidence that in some cases
denitrification is limited by C and N03 supply in wetlands (Ashby et al., 1998).
Denitrification has been studied in riparian zone ecosystems (Lowrance, 1992; Pinay et al., 1993;
Watts and Seizinger, 2001; Hefting et al., 2003) and seems to be related to C availability. Generally,
riparian soils that are both rich in organic matter and anaerobic have high denitrification potential. Where
riparian soils are aerobic, however, nitrification, rather than denitrification, can be the dominant process
(Stevens etal., 1997).
Table 3-13. Summary of N cycling studies for wetlands.
Region/Country
Endpoint
Observations
Wetland
Type/
Species
Reference
Canada
Soil [N]
Deposition Gradient: N deposition ranged from 2.7 to 8.1 kg
N/ha/yr across 23 ombrotrophic peatlands. Soil [N] (g/m2/yr)
increased linearly with deposition (y[wet deposition] =3.50(soil N)
+0.64; r2=0.29, p<0.001)
Peatland
Moore et al.
(2004)
Adirondack
Mountains, New York
N sources
used by
vegetation
Isotopic tracer: The study estimated N2 fixation by speckled alder
in five wetlands by the 15N natural abundance method and by
acetylene reduction using a flow-through system. The study of
alder-dominated wetlands showed that alder derived >85% of leaf
N from N-fixation at an estimated rate of 43 kg N/ha/yr.
Conclusion: speckled alder in wetlands of northern New York State
relies heavily on N2 fixation to meet N demands, and symbiotic N2
fixation in speckled alders adds substantial amounts of N to alder-
dominated wetlands in the Adirondack Mountains. These additions
may be important for watershed N budgets, where alder-
dominated wetlands occupy a large proportion of watershed area
Alnus
incana
Hurd et al.
(2001)
Rhode Island
Denitrification
Observation: the highest rates of denitrification (4 to 135 kg
N/ha/yr) were observed in very poorly drained soils on nutrient-rich
parent material, with lower rates (1.2 to 5.3 kg N/ha/yr) in soils that
Groffman
(1994)
were better-drained or less nutrient-rich.
3-100
-------
Region/Country Endpoint
Observations
Wetland
Type / Reference
Species
Catskill Mountain, Denitrification Field Addition: Higher rates of denitrification per unit area
New York associated with soils with higher organic matter content and water-
filled pore spaces. Instantaneous NO3" concentration did not
correlate with denitrification rate, suggesting that the rates of NO3"
supply through microbial production or hydrologic transport were
more important than in situ NO3" concentration. Denitrification was
most stimulated by amendments with glucose alone or glucose
plus NO3", suggesting limitation by labile C and NO3" supply.
Adirondack Denitrification Ecological Gradient: Changes in stream N were measured in one
Mountains, New York, riparian wetland and one beaver meadow: Strong effects of
Archer Creek peatlands on local N concentrations, but little effect of peatlands
Watershed on adjacent stream chemistry, since peatland ground water
contributed little to streamflow.
Smith Creek, Denitrification Field Additions/Measurement: Over a two-year period >1400 Riparian Hedin et al.
Michigan individual samples of subsurface waters were analyzed. Both wetlands (1998)
spatial patterns of water chemistry and additions of labile C to
demonstrate that the supply of degradable C from shallow
flowpaths limited rates of NO3" removal via denitrification in near-
stream zones. Thus, the immediate near-stream region may be
especially important for determining the landscape-level function
of many riparian wetlands.
Ashby et al.
(1998)
Peatlands McHale
etal. (2004)
N deposition may stimulate biogenic emissions if the N supply is limiting the rate of denitrification
in wetland soils (Hayden and Ross, 2005). Previous studies suggest that elevated N inputs to wetlands
will often increase the rate of denitrification (Dierberg and Brezonik, 1983; Broderick et al., 1988;
Cooper, 1990). This process increases the emission of nitrous oxide, a greenhouse gas, to the atmosphere
(see Section 3.3.4).
In a review of the effects of riparian zones on N03 removal from ground water, Hill (1996)
concluded that there are large losses of N03 to denitrification within riparian zones. However, there are
important limitations to the generalization that riparian wetlands prevent the leaching of N03 to streams.
Not all water entering streams passes directly through adjacent riparian zones, and denitrification in deep
subsurface flowpaths is often limited by the supply of labile C. In addition, not all stream water passes
through riparian zones, and large amounts of water may follow flowpaths beneath organic-rich riparian
zones, allowing significant transport of N03 to streams (McHale et al., 2004). The supply of degradable
C from shallow flowpaths has been shown to limit rates of N03 removal via denitrification in near-
stream zones (Hedin et al., 1998).
In summary, wetland soils can be hotspots of N03 removal by denitrification in anoxic sites rich in
N03 and labile C, but denitrification rates can be limited by suboptimal conditions of any single
biogeochemical factor, and deep water flowpaths can bypass wetland denitrification altogether (see
Table 3-13).
3.3.2.3. Freshwater Aquatic Ecosystems
As previously noted in Section 3.3.2.1., a large fraction of atmospheric N deposition is retained in
most terrestrial ecosystems. Nevertheless, the fraction that does leach to streams can make a substantial
contribution to total N inputs of downstream rivers and estuaries, especially in the eastern U.S. (Driscoll
et al., 2003d).
3-101
-------
N03- Leaching
After NO; leaches from terrestrial ecosystems it arrives in adjacent aquatic systems and may then
be transported downstream. The following discussion is intended to build on the information introduced
in Section 3.3.2.1 by addressing elevated NO3 concentration in surface water; and NO;, leaching
downstream. The concentration of NO;, in surface water can serve as a chemical indicator of N input in
excess of ecosystem requirements, and is relevance to acidification and eutrophication effects on surface
water.
The relationship between wet deposition of N and streamwater output of NO, was evaluated by
Driscoll et al. (1989) for sites in North America (mostly eastern areas), and augmented by Stoddard
(1994). The data showed a pattern of N leaching at wet inputs greater than approximately 5.6 kg N/ha/yr,
which probably corresponds with a total N deposition input of about 8 to 10 kg N/ha/yr. In the Northeast,
a survey of 230 lakes and streams documented NO, concentrations ranging from less than 2 up to
42 fieq/L, with the highest median values occurring in the Adirondacks (Aber et al., 2003) (Figure 3-32).
Figure 3-32. Mean annual NQ3~ concentrations in 230 lakes and streams across the northeastern U.S. Inset
indicates the median, quartile, and 90% range of mean annual NO3" in the Adirondacks (ADK),
the Catskills (CAT), Vermont (VT), New Hampshire (NH), and Maine (ME).
In the western U.S., NO ; concentrations of freshwater ecosystems have been shown to increase
with proximity to urban areas. Results from the Western Lake Survey (WLS) conducted by the U.S. EPA
Mean annual N03 (pmol/L)
0 0-10.5
o 10.5-21.0
© 21.0-31.5
• 31.5-42.0
ADK CAT VT HN ME
(n = 66) (11+48) (n+74) (n = 14)
Source: Aber et al. (2003). Reprinted with permission.
3-102
-------
(Eilers et al., 1987), document increased N concentrations in high elevation lakes adjacent to and
downwind of urban centers (Fenn et al., 2003a), such as those found in the Sierra Nevada and Colorado
Front Range (see Figure 3-33). For example, NO; concentrations in streamwater during the growing
season in the Sierra Nevada were reported to range from 4 to 19 ueq/L (Fenn et al., 2003b).
Concentrations above 10 j.ieq/L are generally considered high.
An interesting example from the Colorado Front Range indicates that lakes on the eastern and
western slopes can experience significantly different levels of NO.; . A survey of 44 lakes east and west of
the Continental Divide indicated that lakes on the western side of the Continental Divide averaged
6.6 }.ieq/L of NO; , whereas lakes on the eastern side of the divide averaged 10.5 ueq/L of NO;
concentration. NO; concentrations above 15 (.ieq/Lhave commonly been measured m lakes on the
eastern slope of the Front Range, suggesting some degree of N saturation (Baron, 1992), and extreme
values as high as 40 (.icq/L have also been reported (Campbell et al., 2000). Williams et al. (1996a)
concluded that N -saturation is occurring throughout high-elevation catchments of the Colorado Front
Range. Many lakes in the Colorado Front Range have chronic NO ; concentrations greater than 10 ueq/L
and concentrations during snowmelt are frequently much higher, due at least in part to leaching from
tundra, exposed bedrock, and talus areas.
In the Unita Mountains of Utah and the Bighorn Mountains of central Wyoming, 19% of the lakes
included within the Western Lakes Survey had NO; concentrations greater than 10 (.ieq/L. This pattern
suggests that N deposition in these areas may have exceeded the capability of these lakes to assimilate N.
It is unknown if these concentrations of NO, represent effects from anthropogenic sources or if this
constituted a natural condition associated with inhibited NO; assimilation in cold alpine environments.
Figure 3-33. NO3" concentrations in high-elevation lakes in western North America. Circles represent cities
with a population greater than 100,000.
Lake nitrate
concentration
(|jeq per L)
0-2
• 2-6
•6-12
• 12-20
• 20-33
Source: Fenn (2003a). Reprinted with permission.
3-103
-------
Dissolved Inorganic Nitrogen
Dissolved inorganic nitrogen (DIN) concentrations in surface waters have been directly linked to
wet deposition. Bergstrom and Jansson (2006) evaluated the relationship between wet-deposition and
DIN concentration in 4296 lakes across Canada, Europe and the U.S. They calculated a significant
correlation showing that DIN concentration increased with increasing N wet-deposition (log (y)=
1.341og(x) - 1.55, r2 = 0.70, p <0.001).
Denitrification
Denitrification in freshwater aquatic ecosystems has been studied in small streams most
intensively, though some work has been done at larger scales. N is cycled rapidly within streams,
especially small streams with large relative areas for contact with benthic surfaces and hyporheic zones.
For example, Peterson et al. (2001) found that 15N-NH4+ added to streams of various sizes was taken up
most rapidly in the smallest streams, and that these headwater streams exported less than 50% of their
added NH/. Nevertheless, the long-term fate of this removed or transformed and recycled N is more
difficult to assess. Mulholland et al. (2004) found that addition of '"N-N03 to a headwater stream at
Walker Branch, TN, indicated a mean uptake length of 35 m under ambient conditions. The uptake length
extended three-fold (i.e., reduced uptake) under a modest fertilization treatment, which employed N03
addition of approximately 500 |_ig N/L. Direct measurements of denitrification of added 15N indicated that
denitrification accounted for 16% of the N03 loss under the ambient treatment, and only 1% of N03
uptake under the fertilized treatment. Nearly all of the denitrification occurred as reduction to N2 gas
rather than to N20.
Hyporheic losses of N03 to denitrification may be largely controlled by supplies of labile
dissolved organic carbon (DOC). Bernhardt and Likens (2002) found that adding 6 mg/L of DOC as
acetate to a small stream at Hubbard Brook, NH, reduced stream N03 concentrations from ~5
to <1 (imol/L. In experimental mesocosms designed to mimic hyporheic flowpaths of a small river in the
Catskill Mountains, NY, Sobczak et al. (2003) found that adding just 0.5 to 1.0 mg/L DOC from leaf litter
resulted in the net consumption of nearly all of the 40 (imol/L N03 in solution. Acetylene block
measurements indicated that most N03 loss was due to microbial assimilation rather than denitrification,
consistent with the isotopic tracer results of Mulholland et al. (2004).
At large spatial scales, water residence time is the variable most frequently identified as a controller
of N loss from aquatic ecosystems examined. Examples include lakes of various sizes (Howarth et al.,
1996) and large river basins spanning the northeastern U.S. (Seitzinger et al., 2002). Compiling N loss
data sets from a wide range of aquatic ecosystems, Seitzinger et al. (2006) found that water residence time
alone explained 56% of the variance in rates of N loss across lakes, rivers, estuaries, and continental
shelves, from fast-flowing river reaches (residence time of hour) with 0% to 15% N loss to century-scale
turnover lakes that eventually incur 80% to 100% N loss.
N Transport Downstream: Urbanization and Determination of N Sources
The transport of N via rivers and streams represents an important source of N to downstream
ecosystems. The transport and loss of N is determined by the net balance of delivery of N by direct
atmospheric deposition and from upland terrestrial and associated transitional ecosystem sources, minus
the uptake and gaseous loss of that N during transport. Alexander et al. (2002) showed -70% of the N in
headwater streams is from N deposition and the net transport of N from headwater streams is between 40-
65% of the total N flux to lower order streams. Numerous studies have illustrated correlations between
water quality or ecological conditions and various measures of the extent of urbanization, such as human
population density or percent impervious surface (Hachmoller et al., 1991; Charbonneau and Kondolf,
1993; Johnson et al., 1997; Thorne et al., 2000; Alberti et al., 2007); see additional discussion of
urbanization in Annex C). In many higher order streams and estuaries, atmospheric N combines with
fertilizer N in agricultural areas and with N from wastewater treatment facilities in urban areas, and the
3-104
-------
role of atmospheric deposition in residential and urban ecosystems is rarely addressed (see Section 3.3.2.4
for additional discussion of inputs to estuaries).
Table 3-14.
Summary of studies on the effects of N deposition on freshwater aquatic ecosystems.
Region/
Country
Endpoint
.. Ecosystem Type /
Observations i . Jr
Species
Reference
Eastern U.S.
[NO3-]
leaching
Observation: NO3" leaching observed at 5.6 kg N/ha/yr wet-deposition, streams
which is equal to 8-10 kg N/ha/yr total N deposition.
Driscoll et al.
(1989); Stod-
dard (1994)
Western U.S.
surface
water [NO3"]
Observation: surface water [NO3"] is elevated downwind of urban areas in lakes
the West.
Fenn et al.
(2003b)
Colorado
Front Range,
U.S.
surface
water [NO3"]
Deposition gradient: higher levels of deposition on the eastern slope lake
caused elevated lake water [NO3"]
Baron (1992)
New England N sources in Modeling: Application of the statistical model SPARROW (SPAtially Refer- Rivers Alexander
rivers and enced Regression On Watershed attributes) showed that first-order head- et al. (2002;
streams waters contributed 65%, 55%, and 40% of the N flux to 2nd, 4th, and 2007)
higher-ordered catchments, respectively. Atmospheric deposition ac-
counted for almost 70% of the total simulated N load to these headwater
streams.
Neuse water [N] Observation: Trends in N and P concentrations from 1998 to 2002 could be Watershed/ estuary Burkholder
Estuary, NC |\| sources exP'ained mainly by a combination of climate, management policies, and etal. (2006)
urban/agricultural development. Nutrient loading reductions did occur in
response to imposed management practices in the watershed, but they
were affected by increases in human and livestock population in the water-
shed. Thus, goals for estuarine and coastal nutrient loading reduction must
consider the influence of within-watershed development
Chesapeake Watershed N Modeling: The Choptank tributary of the Chesapeake Bay had become Watershed/estuary Fisher etal.
Bay system sources eutrophic over the last 50-100 years. Systematic monitoring of nutrient (1998; 2006)
inputs began in 1970, and there have been 2- to 5-fold increases in
nitrogen (N) and P inputs during 1970-2004 due to sewage discharges,
fertilizer applications, atmospheric deposition, and changes in land use.
Hydrochemical modeling and land-use yield coefficients suggest that
current input rates are 4-20 times higher for N and P than under forested
conditions existing 350 year ago. The Choptank watershed (1756 km2) is
dominated by agricultural land use (62%), with only 5% urban
development. O2 concentration in bottom waters of the Patuxent estuary is
consistently below 3 mg/L in summer; O2 levels have been steadily
decreasing in the Choptank estuary over the past two decades and now
approach 3 mg/L in wet years
Chesapeake Watershed N Modeling: The Patuxent watershed (2260 km2) is dominated by forest Watershed/estuary Fisher etal.
Bay system sources (64%), with significant urban land use coverage (16%) and less intensive (1998; 2006)
agricultural development (20%). Sewage is a major cause of nutrient en-
richment. The low N: P of sewage inputs to the Patuxent results in an N-
limited, P-saturated system, whereas the Choptank is primarily limited by
N, but with P limitation of phytoplankton during spring river flows. Reduced
eutrophication in dry years suggests that both estuaries will respond to
significant decreases in nutrients
3-105
-------
In lowland areas, other terrestrial sources of N, such as fertilizer, livestock waste, septic effluent,
and wastewater treatment plant outflow, often become much more important than in upland areas. In
lowland areas, it is difficult to determine the percent of atmospheric N that leaches to drainage water
because there are other ill-defined sources of N to drainage waters. In Table 3-14 studies are summarized
that evaluate how atmospheric deposition of N to the estuary surfaces and to the terrestrial watershed
interact with the other anthropogenic sources of N to make up the total anthropogenic N load to the
system.
3.3.2.4. Estuarine and Coastal Marine Ecosystems
Estuaries and coastal marine environments tend to be N-limited, and many currently receive high
levels of N input from human activities (Howarth et al., 1996; Vitousek and Howarth, 1991). The nature
and extent of the impacts on estuarine and coastal environments is related, in part, to the export of N from
upland systems to coastal environments, as discussed in previous sections. Denitrification is the primary
mechanism of N output from the estuary and back to the atmosphere (See Annex C). Important
environmental effects of NCAM include increased algal blooms, depletion of dissolved 02 in bottom
waters, and reduction in fisheries and sea grass habitats (Boynton et al., 1995; Howarth et al„ 1996; Paerl,
1995, 1997; Valiela and Costa, 1988; Valiela et al., 1990). The general process of estuarine eutrophication
is depicted in Figure 3-34.
p
10-
Z
0-
|
¦M
|
VJ
jE
o
—
3
Overall eutrophic condition
¦ ¦
No Problem /low Moderate low Moderate Moderate high
High
Few symptoms occur Symptoms occur Symptoms occur Symptoms occur Symptoms occur
at more than episodically and/or less regularly less regularly and/or periodically or
minimal levels. over a small to and/or over a over a medium to persistency and/or
medium area. medium area extensive area. over an extensive area.
Key to symbols:
Submerged aquatic
vegetation
Chlorophyll a
• Nuisance/toxic
blooms
Macroalgae
Dissolved oxygen
Influencing factors
(luudi and mu-pUhUity)
Source: Bricker et al. (2007)
Figure 3-34. A conceptualization of the relationship between overall eutrophic conditions, associated
eutrophic symptoms, and influencing factors (N loads and susceptibility). Overall eutrophic
condition was assessed for estuaries throughout the U.S.
There is broad scientific consensus that N-driven eutrophication of shallow estuaries in the U.S.
has increased over the past several decades and that environmental degradation of coastal ecosystems is
now a widespread occurrence (Paerl et al., 2001b). For example, the frequency of phytoplankton blooms
and the extent and severity of hypoxia have increased in the Chesapeake Bay (Officer et al., 1984)
Pamlico estuary in North Carolina (Paerl et al., 1998), and along the continental shelf adjacent to the
Mississippi and Atchafalaya River discharges to the Gulf of Mexico (Eadie et al., 1994). A recent national
3-106
-------
assessment of eutrophic conditions in estuaries found that 65% of the assessed systems had moderate to
high overall eutrophic conditions (Bricker et al., 2007). Estuaries with high overall eutrophic conditions
were generally those that received the greatest N loads from all sources, including atmospheric and land-
based sources (Bricker et al., 2007). The relative importance of the various N sources varies from estuary
to estuary. Atmospheric sources are proportionately more important to estuaries that exhibit large surface
area relative to watershed drainage area, and in those estuaries that drain watersheds dominated by natural
ecosystems rather than agricultural or urban lands (Boyer et al., 2002).
Assessing the contribution of atmospheric N deposition to total N loading
The importance of atmospheric deposition as a cause of estuary eutrophication is determined by the
relative contribution of atmospheric versus non-atmospheric sources of N input. Anthropogenic sources of
N to estuarine and coastal ecosystems include atmospheric deposition, wastewater discharge, agricultural
runoff, and urban runoff. Valigura et al. (2000) estimated that direct atmospheric deposition to the estuary
surface generally constitutes at least 20% of the total N load for estuaries that occupy more than 20% of
their watershed. The U.S. EPA (2000b) estimated that between 10% and 40% of the total N input to
estuaries in the U.S. is typically derived from atmospheric deposition
(http://www.epa. gov/ttn/oarpg/t3/reports/head 2kf.pdf).
Estimates of the relative contribution of each major source have been developed by using the
Watershed Assessment Tool for Evaluating Reduction Scenarios for Nitrogen (WATERS-N) model
(Castro and Driscoll, 2002). Driscoll et al. (2003a) estimated annual net anthropogenic N inputs to eight
large watersheds in the Northeast for the year 1997 (see Figure 3-35). Input values of total atmospheric
plus non-atmospheric anthropogenic N ranged from 14 kg N/ha/yr in the watershed of Casco Bay in
Maine to 68 kg N/ha/yr in the watershed of Massachusetts Bay (Driscoll et al., 2003a). In all eight
watersheds, net import of N in food for humans (input into estuaries as wastewater) was the largest
anthropogenic input. Atmospheric deposition was estimated to be the second largest anthropogenic N
input, ranging from 5 to 10 kg N/ha/yr, or 11% to 36% of the total inputs, with four watersheds ranging
from 34% to 36% (Driscoll et al., 2003a) (Table 3-34). These results are broadly consistent with estimates
by Boyer et al. (2002), who used a similar N budgeting approach for 16 large northeastern U.S. river
basins and reported that N deposition contributes approximately 31% of the total N load to large river
basins, although this fraction varies regionally (Boyer et al., 2002). Boyer et al. (2002) considered only
the portions of each basin above USGS gauging stations, which often occurred above large population
centers. Hence, the Driscoll et al. (2003a) budgets included regions with greater human food consumption
than those considered by (Boyer et al., 2002).
Castro and Driscoll (2002) studied 10 estuaries along the U.S. east coast and found total
atmospheric N inputs (watershed runoff plus direct deposition to the surface of estuary) accounted for
15-42% of the total N inputs. Simulated reductions of atmospheric N deposition by 25% and 50% of
current deposition rates reduced the contribution made by atmospheric N deposition to the total N loads
by 1-6% and 2-11%, respectively. Overall, results from the simulated reductions suggested that
considerable reductions ( >25%) in atmospheric N deposition were needed to significantly reduce the
contribution made by atmospheric N deposition to the total N loads. In a later study, Driscoll et al.
(2003a) estimated that the implementation of aggressive controls on both mobile N emissions sources and
electric utilities would produce an estimated reduction in estuarine loading in Casco Bay, ME of 13%
(Driscoll et al., 2003a).
3-107
-------
(/)
.22 T
3 -
I a
® T5
o 0)
£
(/) C/)
3 0>
o
= 3
^ N in food
m Atmospheric N deposition
Fertilizer N
80 ~ N in feed
H N fixation
60
o
« n
c £
g. ¦-
3) 0
o a
o z
¦£ ro
=
<
z 40-
20 -
0 -1—1
Northeast
Til
i
J? ^ ^ J?
Mid-Atlantic
^ ^ .? c>
-------
3.3.2.5. Summary of N Effects on Biogeochemical Cycling of N and Associated
Chemical Indicators
Terrestrial Ecosystems
The evidence is sufficient to infer a causal relationship between N deposition and the alteration of
biogeochemical cycling of N in terrestrial ecosystems. This is supported by numerous observational,
deposition gradient and field addition experiments. The main source of new N to non-managed terrestrial
ecosystems is atmospheric deposition. N deposition disrupts the nutrient balance of ecosystem. The
chemical indicators that are typically measured are summarized in Table 3-11 and include N03 leaching,
C:N ratio, N mineralization, nitrification, denitrification, foliar N and soil water N03 . andNTL^. Values
for these indicators that represent a threshold for the onset of a related biogeochemical or biological effect
are also summarized. Note that N saturation does not need to occur to cause adverse effects on terrestrial
ecosystems. However, in some regions N saturation is a plausible mechanism of net nitrification and
associated N03 leaching in drainage water. Leaching of N03 from forest soils to stream water can
acidify downstream waters and deplete soils of nutrient base cations, especially Ca and Mg (see Section
3.2).
In northeastern watersheds, Aber et al. (2003) suggested that nearly all N deposition is retained or
denitrified in areas that receive less than about 8 to 10 kg N/ha/yr. Aber et al. (2003) further found that
surface water N03 concentrations exceeded 1 j^ieq/L in watersheds receiving about 9 to 13 kg N/ha/yr of
atmospheric N deposition (Figure 3.30). The lakes and streams found to have high N03 concentration
were those receiving N deposition above this range, but responses were variable among those receiving
high N deposition. Above this range, mean N03 export increased linearly with increasing deposition at a
rate of 0.85 kg N03 kg N/ha/yr for every 1 kg N/ha/yr increase in deposition, although there was
considerable variability in N retention among watersheds at higher rates of deposition (Aber et al., 2003).
In the West, a study of mixed conifer forests of the Sierra Nevada and San Bernardino mountains in
California reported that the deposition load that causes the onset of N03 leaching is 17 kg N/ha/yr.
Several studies in the Rocky Mountains indicate that the onset of N03 leaching in alpine catchments
occurs at approximately 10 kg N/ha/yr (Baron et al., 1994; Williams and Tonnesen 2000).
Wetlands
The evidence is sufficient to infer a casual relationship between N deposition and the alteration of
biogeochemical cycling of N in wetlands. N deposition contributes to total N load in wetlands that may also
receive N from terrestrial run-off. The chemical indicators that are typically measured include N03
leaching, N mineralization, and denitrification. N dynamics in wetland ecosystems are variable in time
and with type of wetland and environmental factors, especially water availability (Howarth et al., 1996).
A wetland can act as a source, sink, or transformer of atmospherically deposited N (Devito et al., 1989)
and these functions can vary with season and with hydrological conditions. Vegetation type,
physiography, local hydrology, and climate all play significant roles in determining source/sink N
dynamics in wetlands (Arheimer and Wittgren, 1994; Devito et al., 1989; Koerselman et al., 1993;
Mitchell et al., 1996).
N mineralization has been shown to increase with N addition, and this can cause an increase in
wetland N export to adjacent surface water (Groffman, 1994). In general, leaching losses of N03 in water
derived directly from wetlands are often small because of N03 removal by denitrification. Previous
studies suggest that elevated N inputs to wetlands will often increase the rate of denitrification (Dierberg
and Brezonik, 1983; Broderick et al., 1988; Cooper, 1990). This process limits the available N supply to
soils and drainage waters; but increases the biogenic emission of greenhouse gasses to the atmosphere
(see Section 3.3.4). Denitrification appears to be negligible in wetland environments that are typically
nutrient (including N) poor, such as some bogs and fens (Morris, 1991).
3-109
-------
Freshwater aquatic
The evidence is sufficient to infer a casual relationship between N deposition and the alteration of
biogeochemical cycling of N in freshwater ecosystems. N deposition is the main source of N to headwater
streams, lower order streams and high elevation lakes. The predominant chemical indicators are N03
concentration in surface waters and DIN. Elevated surface water N03 concentrations indicate N
saturation of the terrestrial ecosystem and occur in both the eastern and western U.S. In the East, 8-10 kg
N/ha/yr can cause increase N03 concentrations (Driscoll et al., 1989; Stoddard 1994). In the West,
elevated N03 concentrations have been observed down wind of urban centers (Fenn et al., 2003a) and on
slopes of the Colorado Front Range that are effected by elevated deposition from urban and agricultural
sources (Baron et al., 1992)
Estuaries and coastal marine
The evidence is sufficient to infer a casual relationship between N deposition and the biogeochemical
cycling of N in estuaries and coastal marine waters. The contribution of atmospheric N deposition to total
N load is calculated for some estuaries and can be greater than 40%. Atmospheric deposition N is not the
sole source of N loading to estuaries and it is unknown if atmospheric deposition alone is sufficient to
cause eutrophication. In general, estuaries tend to be N-limited, and many currently receive high levels of
N input from human activities to cause eutrophication (Vitousek and Howarth, 1991; Howarth et al.,
1996, Elser et al., 2007). The most widespread chemical indicators of eutrophication are submerged
aquatic vegetation, Chi a, algal blooms, macroalgae and dissolved 02.
3.3.3. N Deposition Effects on Productivity and C Budgets
The addition ofN from an exogenous source will alter the productivity of N-limited ecosystems.
In a meta-analysis including terrestrial, freshwater and marine ecosystems, Eisner et al. (2007) found
there were similar patterns of N and P limitation among ecosystem types. This finding is in contrast with
the existing paradigm that N-limitation dominates in terrestrial ecosystems and P-limitation dominates in
freshwater ecosystems. Marine ecosystems tend to be N-limited; however P-limitation can play an
important role. It is important to distinguish between effects on primary productivity and effects on C
sequestration. N addition to a given ecosystem may increase primary productivity; however this often
does not translate into greater C sequestration because C lost from the ecosystem by respiration may
offset the C gained by production. These topics are discussed for a range of ecosystem types in the
following section.
3.3.3.1.Terrestrial Ecosystems
The following section discusses the mechanisms by which atmospheric N deposition alters C
cycling in terrestrial ecosystems (see Figure 3-36). Although predicted values of atmospheric [C02] in the
future may alter the interaction between N and terrestrial C cycling (Hyvonen et al., 2007; Norby, 1998;
Schindler and Bayley, 1993) this topic is beyond the scope this review. Available studies include N
deposition gradients, N addition studies, modeling and time-trend analyses. Few studies have isolated the
effect of chronic N deposition on plant growth and ecosystem C balances. In those that have, it is difficult
to disentangle the effects of climate, disease and land use from N deposition effects. There are numerous
field studies ofN addition; however there are some difficulties in interpreting the results regarding
chronic N deposition effects on ecosystems. Some studies add N in a large pulse at one time, a technique
that does not simulate chronic N loads of smaller amounts that are characteristic of atmospheric
deposition. Ecosystem growth response to a pulse of 100 kg N/ha at one time may not be same as 10 kg
3-110
-------
N/ha/yr for 10 years, although the net load is the same. Additionally, there are few long-term addition
studies therefore limiting the evaluation of changes in ecosystem response over decadal time scales.
Carbon accumulation in terrestrial ecosystems occurs above and below ground. C cycling is a
complex process that can be quantified into ecosystem C budgets on the basis of net ecosystem
productivity (NEP), defined as gross primary productivity (GPP) after subtracting the ecosystem
respiration (vegetative + heterotrophic respiration). Factors that may increase terrestrial C02 sinks on a
regional scale are increased NPP, and decreased respiration of C02 from leaf or soil processes. These two
mechanisms may be altered by atmospheric deposition of N, tropospheric ozone exposure, increased C02
concentrations, land-use change and factors associated with climate warming (Beedlow et al., 2004;
Caspersen et al., 2000; Melillo et al., 2002; Myneni et al., 1997; Ollinger et al., 2002; Nowak et al., 2004;
Schimel et al., 2001). This adds to the uncertainty regarding the sources and sinks in the terrestrial
biosphere (Houghton, 2003). It should be noted that it is not known whether present terrestrial C
sequestration can be sustained, in view of limits of forest re-growth, nutrient availability and uncertainty
about changes in the frequency of disturbances such as fire (Schimel et al., 2001; Scholes and Noble,
2001).
CO,
CO,
photosynthesis
autotrophic =
respiration
N itrifi cation/de n itrifi cation
uptake
N deposition
mycorrhiza
heterotrophic
respiration Litter
1 coarse woody debris
I
heterotrophic
respiration
Immobilization^ Soi|mjnera|N - N fixation
^ <*
mineralization
' decomposition
*
Soil organic matter
N leaching
mineralization
C cycle
N cycle
Figure 3-36. Interactions between the C and N cycles.
Forests
C Allocation Interactions with Stressors
Addition of N is believed to decrease resistance to drought stress. This occurs because plants
balance their resources to maximize above-ground processes, such as light and C capture, with below-
ground processes, such as capture of water and other nutrients, including N (Sterner and Elser, 2002). N
3-111
-------
addition often causes trees to allocate less photosynthate to roots than shoots (Minnich et al., 1995).
Because shoot growth is more enhanced than root growth, the water supply from the roots can become
insufficient during periods of drought to support water loss via transpiration (Fangmeier et al., 1994a;
Krupa, 2003). Smaller root systems also cause greater susceptibility to wind-throw. For example, in
Switzerland, the amount of trees uprooting during a strong storm event was significantly correlated with
base saturation and N concentration in the leaves (positively) in beech trees (Braun et al., 2003). Across
Europe, soil acidification and soil N content since the 1980s have accentuated the storm sensitivity due to
changes in root architecture, including more superficial roots and loss of root ramification (Braun et al.,
2003; Godbold et al., 1988; Puhe, 2003; Nilsson et al., 2004).
Deposition of N is also believed to reduce frost hardiness of plants (Dueck et al., 1990), which may
play a role in red spruce die back (see Section 3.2). This is likely because the addition of N prolongs the
growth phase of the plants during autumn and delays winter hardiness. This can cause detrimental effects
if the first frost occurs early in the autumn period (Cape et al., 1991). Plant shoots also appear to be more
susceptible to pathogenic fungal infection under high N status or changed nutrient balance such as an
increase in the ratio of N to K+ (Ylimartimo, 1991; Krupa, 2003). As opposed to shoot diseases, addition
of N has been found to reduce mycorrhizal fungus colonization of roots (see more detailed discussion in
Section 3.3.5.1).
Above-Ground Processes
Nitrogen availability often limits rates of net primary production in temperate terrestrial
ecosystems (Vitousek and Howarth, 1991; LeBauer and Treseder, 2008); therefore there is an implicit link
between the C and N cycles (Figure 3-36). Over 50% of N taken up by plants is used for photosynthetic
enzymes. Rates of photosynthesis and net primary productivity (NPP) typically correlate with metrics of
N availability such as leafN content and netN mineralization rate (Field and Mooney, 1986; Reich et al.,
1997a, 1997b; Smith et al., 2002). A meta-analysis of 126 N addition experiments evaluated N limitation
of NPP in terrestrial ecosystems by evaluating above-ground plant growth in fertilized to control plots
(LeBauer and Treseder, 2008). The results showed that most ecosystems are N limited with an average
29% growth response to N. The response ratio was significant within temperate forests, tropical forests,
temperate grasslands, tropical grasslands, wetlands, and tundra, but not deserts (LeBauer and Treseder,
2008).
There is substantial evidence that N additions to trees cause increased leaf-level photosynthetic
rates. However, the potential for N deposition to increase above-ground C biomass is limited for reasons
related to the biogeochemical cycling ofN (see more detailed discussion in Section 3.3.2.1). Briefly, C:N
stoichiometry of the forest ecosystem compartments determines the C response to N deposition. Only a
small portion of added N is taken up by vegetation, thus only a small portion of N contributes to C
capture by trees (Nadelhoffer et al., 1999). A recent study reported that tree biomass (e.g., foliage, woody
tissue, and fine roots) accumulated 7 to 16% of N additions (Nadelhoffer et al., 2004). N may be
immobilized in the soil, leached out before biological assimilation, or, upon the addition of N, some other
factor may become limiting to growth (e.g., water or other nutrients). Even though only a portion of N
deposition is incorporated into vegetation, the general result of additional N is an increase in leaves,
wood, and root biomass (Nilsson and Wiklund, 1995). The biological endpoints typically measured to
evaluated plant growth in listed in Table 3-15.
3-112
-------
Table 3-15.
Summary of N effects on forest carbon cycling.
Region/Country
Endpoint
Observations
Forest type/
species
Reference
Europe
Growth
(biomass)
Modeling: Large regions of forests were used as empirical data
from which to build a model that forest biomass is accumulating,
largely in response to increases in forest area and improved
management, but other possible mechanisms of growth
enhancement (including N) cannot be ruled out
Forests from
Austria, Finland,
Sweden, France,
Germany, and
Switzerland
Kauppi
etal.
(1992);
Spiecker
etal. (1996)
Norway
Growth (tree
ring incre-
ments)
Deposition gradient: A comprehensive analyses of regional
forest growth trends analyzed tree increment cores from more
than 31,000 plots. In this study, growth increased during the
1960s and 1970s and then declined in the 1990s, especially in
southern regions exposed to the highest rates of N deposition
(Figure 3-37)
Boreal forest
(Picea abies and
Pinxus sylvestris)
Nellemann
and
Thomsen
(2001)
Sweden
Growth (stem
volume)
Field Addition: chronic fertilization at 30 kg N/ha/yr continued to
stimulate stemwood production even after 30 years, whereas a
higher application (90 kg N/ha/yr) decreased stem volume
growth, and an intermediate application (60 kg N/ha/yr) had little
positive or negative effect relative to the control plots
Boreal forest,
Scots pine forest
Hogberg
etal. (2006)
Bear Brook, ME,
U.S.
Growth (basal
area)
Field Addition: basal area increment of sugar maple was
enhanced 13 to 104% by addition of 25 kg N/ha/yr as ([NH4]
2SO4), whereas red spruce was not significantly affected.
Sugar maple and
red spruce
Elvir et al.
(2003)
Fernow
Experimental
Forest, VW, U.S.
Growth
Field Addition: The application of 35 kg N/ha/yr as (NH4) 2SO4
enhanced growth of Black cherry (Prunus serotina) and yellow
poplar (Liriodendron tulipifera) during the first 7 years, but led to
reduced growth of these species relative to control trees in years
9 through 12, with no change in red maple (Acerrubrum) or
sweet birch (Betula lenta)
Black cherry,
yellow poplar, red
maple, sweet
birch
DeWalle
etal. (2006)
Harvard Forest,
MA, U.S.
Mortality
Field Addition: chronic N addition levels of 50 and 150 kg N/ha/yr
for 15 years caused a 31 % and 54% decrease, respectively, in
red pine growth. As red pine has died, striped maple
(Acer pensylvanicum), black cherry, and black birch (Betula
lenta) have increased their contributions to annual litterfall
production.
Red pine, striped
maple, black
cherry, black
birch
Magill et al.
(2004)
Northeastern U.S.
Live basal area
Field Addition: In a high-elevation red spruce-balsam fir (Abies
balsamea) forest in the, N fertilization over 14 years led to a
decrease in live basal area (LBA) with increasing N additions. In
control plots, LBA increased by 9% over the course of the study,
while LBA decreased by 18% and 40% in plots treated,
respectively, with 15.7 kg N/ha/yr and 31.4 kg N/ha/yr.
Red spruce-
balsam fir
McNulty
etal. (2005)
Harvard Forest,
MA, U.S.
Growth
mortality
root production
Field Addition: Control, low and high N additions of NH4NO3 (0,
50 and 150 kg N/ha/yr) were made over 15 years (1988-2003) to
a pine plantation and mixed hardwood stand. Ambient deposition
was calculated to be 8 kg N/ha/yr. N addition stimulated
productivity, although the drought in 1995 induced significant
mortality in small red maple trees. Fine root biomass was
slightly, but not significantly, lower in highly fertilized stands
relative to controls in both red pine and oak/maple ecosystems.
Stem mortality of Pinius resinosa increased with N addition
Plantation: Pinus
resinosa
Hardwood stand:
Quercus velutina,
Q. rubra, Betula
lenta, Acer
rubrum, Fagus
grandiHora,
Prunus serotina
Magill et al.
(2004)
3-113
-------
Region/Country
Endpoint
Observations
Forest type/
species
Reference
Harvard Forest,
MA
Growth
Photosynthesis
Needle age
Field Addition: Ambient deposition to the site was 8 kg N/ha/yr. N
was applied as a concentrated solution of NH4NO3 divided into 6
equal monthly doses (May to Sept) totaling 50 (low N) and
150(high N) 8 kg N/ha from 1989-1999. Elevated nitrogen
additions caused decreased photosynthesis and decreased
needle age
Red pine (Pinus
resinosa)
Bauer et al.
(2004)
Ysselsteyn, The
Netherlands
Growth
Root
production
Field Addition and deposition exclusion: improvements in wood
accumulation rate, root production, and mycorrhizal associations
occurred when a "clean roof was installed at the site receiving
the highest rate of N deposition (>40 kg N/ha/yr). Decreased
production of fine roots may predispose N-fertilized plants to be
more sensitive to intermittent drought, as well as to nutrient
depletion exacerbated by acid deposition.
Coniferous- Picea
abies, Picea
sitchesis,
Pseudotsuga
menziesii, Pinus
sylvestris
Emmett
etal. (1998)
Boxman
etal.
(1998b)
Southern CA, U.S.
N- saturation-
reduced soil
base satura-
tion, and lack
of a growth
response
Observational: Areas of chaparral and mixed conifer forests that
receive very high levels of dry N deposition in southern
California have experienced significant environmental change
over the past several decades
Chaparral and
mixed conifer
Fenn et al.,
(1996,
2003a)
California
Growth
(productivity)
and mortality
Observational: high inputs of Nr appear to exhibit decreases in
productivity and increases in mortality (Fenn et al., 1998).
Conifer forests
Fenn et al.
(1998)
California
Litter
accumulation,
above-ground
woody
biomass, fire
susceptibility
Field Addition: N fertilization has been shown to cause increased
litter accumulation and C storage in above-ground woody
biomass, which in turn may lead to increased susceptibility to
more severe fires.
Ponderosa pine
Fenn et al.
(2003a)
California
Growth
Increased N deposition caused increased growth for Jeffrey
(,Pinus jeffreyi) and Ponderosa pine (Pinus ponderosa) stands,
Mixed conifers
Takemoto
etal. (2001)
North Carolina
and Virginia
Growth (basal
area), foliar
chemistry,
nitrification and
mineralization
Deposition Gradient: Results from a study of 46 forest plots on
six sites in North Carolina and Virginia dominated by American
beech, sugar maple, and yellow birch suggested that N
deposition is associated with changes in basal area, foliar
chemistry, and nitrification and mineralization rates. Growth
rates for the three tree species were similar at the lowest rates
of N deposition, and then diverged as N deposition increased,
with growth of yellow birch and American beech decreasing at
the high N deposition loads. These differential growth rates have
the potential to affect forest structure and biodiversity
American beech,
sugar maple, and
yellow birch
Boggs et al.
(2005)
Fernow
Experimental
Forest, VW, U.S.
Stem growth
Foliar N
dynamics
Field Addition: sites fertilized annually with 35 kg N/ha (7.1 in the
spring and 21.2 in the fall) of (NH4)2S04 over 11 years (1989-
2000). Responses among three species differed for foliar [N]
either showing no change or a decrease in the fertilized plots.
For all three species N/P ratio initially increased with a
subsequent decrease. Over a two year increment (1999-2001)
stem diameter growth was 37% lower for all three species in the
fertilized plots.
Acerrubrum
Liriodendron
tulipifera
Prunus serotina
May et al.
(2005)
California,
Sequoia National
Park
Growth
Field Addition: Aspen (Popuius tremuloides) have been reported
to show positive growth effects from fertilization at N deposition
rates as low as 10 kg N/ha/yr
Aspen
Bytnerowicz
etal. (2002)
3-114
-------
Region/Country Endpoint Observations Forest type/ Reference
° j r species
Michigan ANPP and Field Addition: Chronic N fertilization (30 kg N/ha/yr applied as Acer saccharum Pregitzer
surface soil NaNOs) for ten years (1994 to 2004) caused significant (sugar maple) et al. (2008)
organic matter increases in ANPP for 3 out of the 10 years. There was a
significant effect of year, however no consistent increase or
decrease through time. After ten years of addition woody
biomass and surface soil organic matter (0-10 cm) were both
significantly greater for the treatment than control.
10
9
8
7
G > 15 kg N/ha/yr (n= 1183}
• 7-15 kg N/ha/yr (n=2389)
A < 7 kg N/ha/yr (ri = 28034}
1940
1950
1980
1970
Year
1980
1990
2000
Source: Nellemann and Thomsen (2001). Reprinted with permission.
Figure 3-37. Mean 5-year radial increment from 31,606 core samples from Picea abies during the period
1945 to 1996 for three atmospheric N deposition zones (high, medium, and low wet N-
deposition in 1990), respectively. Note that the decline in radial increment after 1975
corresponds with the peak in exceedances for critical loads for the same areas. The increase
and subsequent decline from 1965-1996 is significant (p <0.01) using Kruskal-Wallis analysis
with Dunn's tests. S.E.s are all below 5% or 1-3.5 mm increment.
Experimental N additions to forest ecosystems have elicited positive growth responses in some, but
certainly not all, organisms (Emmett, 1999; Elvir et al., 2003; DeWalle et al., 2006; Hogberg et al., 2006).
Forest growth enhancement, can potentially exacerbate other nutrient deficiencies, such as Ca, Mg, or K.
Multiple long-term experiments have demonstrated transient growth increases followed by increased
mortality, especially at higher rates of fertilization (Elvir et al., 2003; Hogberg et al., 2006; Magill et al.,
2004; McNulty et al., 2005). An additional line of evidence comes from the experimental N removal
studies: removal of N and S from throughfall increased tree growth in Europe (Beier et al., 1995; Boxman
et al., 1998).
Decreased growth and increased mortality have more commonly been observed in high-elevation
coniferous stands than in lower elevation hardwood forests, and these differences have been partially
attributed to higher inputs of N at higher elevation and to response characteristics of coniferous, as
opposed to deciduous, trees (Aber et al., 1998). Conifer forests that receive high inputs of N appear to
exhibit decreases in productivity and increases in mortality (Fenn et al., 1998). For example, fertilization
experiments at Mount Ascutney, VT suggested that N saturation may lead to the replacement of slow-
3-115
-------
growing spruce-fir forest stands by fast-growing deciduous forests that cycle N more rapidly (McNulty
etal., 1996; 2005).
Below-Ground Processes
Soils contain the largest near-surface reservoir of terrestrial C. More than 50% of C captured
annually by plants may be allocated below ground (Kubiske and Godbold, 2001). Therefore,
understanding the factors that control soil C storage and turnover is essential for understanding the C
cycle and sequestration. Although there remains considerable uncertainty in the potential response of soil
C to increases in Nr additions (Neff et al., 2002), a meta-analysis by Johnson and Curtis (2001) suggested
thatN fertilization caused an 18% increase in soil carbon content.
There is also evidence of a relationship between N deposition and root production. Nadelhoffer
(2000) stated that N deposition functions to decrease forest fine-root biomass but to stimulate fine-root
turnover and production. However, very high levels of N ( >100 kg N/ha/yr) decreased root life span of
Pinusponderosa (Johnson et al., 2000).
Litter fall is usually the dominant source of soil organic C and a substantial source of organic N.
Decomposition of litter fall is often facilitated by heterotrophic bacteria and mycorrhizae. The quantity of
litter has been shown to increase with elevated N deposition (Schulze et al., 2000), with the result of
increased microbial metabolism in soil. It is also well demonstrated that increased N availability reduces
the ratio of C:N in leaf tissue. In turn, lower C:N in leaf litter has been shown to cause faster initial rates
of decomposition (Melillo et al., 1982), however higher N litter can actually decompose more slowly in
the long-term (Berg, 2000). N bound by leaf organic matter is released over a shorter period of time under
higher decomposition rates, this leads to lower N retention by the soil (De Vries et al., 2006). A 10 year
experiment that investigated decomposition in 21 sites from 7 biomes found net N release from leaf litter
is predominantly driven by the initial N concentration and mass remaining regardless of climate, edaphic
conditions, or biota (Parton et al., 2007). A recent meta-analysis by Knorr et al. (2005) indicated that, as
expected, litter decomposition was stimulated by additional N deposition, however only at sites with low
ambient N deposition rates (<5 kg N/ha/yr). Additional N deposition reduced decomposition at sites with
moderate levels of N deposition (5 to 10 kg N/ha/yr).
Soil respiration is the dominant means by which plant-assimilated C is returned to the atmosphere
as C02. Changes in the magnitude of soil C02 efflux due to changes in environmental conditions will
likely influence the global atmospheric C02 budget (Schlesinger and Andrews, 2000). Overall, the effects
of N addition on soil respiration are mixed; reductions at high levels of N (Lu et al., 1998; Bowden et al.,
2004), no effect (Vose et al., 1995), and increases (Griffin et al., 1997; Mikan et al., 2000) all have been
observed. At the Harvard Forest LTER Site Chronic Nitrogen Amendment Study, N additions increased
soil respiration for hardwood stand, but not for pine stand during the first year of fertilization. However,
continued N additions over a decade caused a 40% decrease in soil respiration for both stands and that
was attributed mostly to a decrease in microbial respiration (Bowden et al., 2004).
Long-Term N Addition
There are few long-term experiments (10+ years) on N addition effects on the C cycle. Pregitzer
et al. (2008) found that 30 kg N/ha/yr applied as NaN03 for ten years caused significant increases in
ANPP for 3 out of the 10 years. There was a statistically significant effect of year, however no consistent
increase or decrease through time. After ten years of addition, woody biomass and surface soil organic
matter (0-10 cm) were both significantly greater for the treatment than control. May et al. (2005) found
that 35 kg N/ha/yr applied as (NH4)2S04 over 11 years (1989-2000) caused different foliar [N] among
three species either showing no change or a decrease in the fertilized versus control plots. For all three
species N/P ratio initially increased with a subsequent decrease. Over a two year increment (1999-2001)
stem diameter growth was 37% lower for all three species in the fertilized plots. At Harvard forest
average atmospheric deposition of nitrogen was 8 kg N/ha/yr. N was applied as a concentrated solution of
3-116
-------
NH4NO3 for treatments that totaled 50 kg N/ha/yr (low N) and 150 kg N/ha/yr (high N) from 1989-1999.
Bauer et al. (2004) found that the nitrogen additions caused decreased photosynthesis and decreased
needle age. At the same site Magill et al. (2004) found that fine root biomass was slightly, but not
significantly, lower in highly fertilized stands relative to controls in both red pine and oak/maple
ecosystems. Stem mortality of Pinius resinosa increased with N addition.
Regional Trends in NEP and NEE
An analysis of >100 young and mature forest stands from around the world indicated that annual
values of C02 exchange varied from approximately -100 to 250 g C/m2/yr for boreal forests and 250 to
700 g C/m2/yr for temperate forests (Malhi et al., 1999). Net ecosystem exchange (NEE), defined as the
difference between NPP and heterotrophic respiration, was positive when the forest was a sink that took
up C02. Townsend et al. (1996) and Holland et al. (1997) modeled the impact of NOY and NHX
deposition on ecosystem C budgets by combining estimates of emissions with three dimensional transport
models. They used spatially explicit estimates of N inputs and climate data as drivers for a process-based
biogeochemical model to simulate ecosystem C dynamics globally. Their simulations predicted that
C02-C uptake due to NOY deposition on land surfaces ranged from 0.3 to 1.4 Pg C uptake/yr (Townsend
et al., 1996; Holland et al., 1997). The model allowed for variations in the degree of ecosystem N
retention. The highest C uptake was calculated when trees were assumed to uptake 80% of N inputs,
which is a likely overestimation because field studies suggested trees only took up a small portion (7-
16%) of N deposition (Nadelhoffer et al., 1999).
Satellite observations of canopy greenness over the last 20 years across North America suggest
enhancement of NEP in some regions, corresponding to observed changes in climate and forest
management. Few such changes were observed in the northeastern U.S., where rates of N deposition are
relatively high (Hicke et al., 2002). In another study, evaluation of tree growth rates in five states
(Minnesota, Michigan, Virginia, North Carolina, and Florida) found little evidence for growth
enhancement due to any factor examined, including N deposition, C02 fertilization, or climate change
(Caspersen et al., 2000). Potential effects ofN deposition on boreal forests of North America are of
concern in part due to the large size of this terrestrial biome. Climate warming and N deposition may
increase NPP and C sequestration in the boreal forest, but may also stimulate decomposition of soil
organic matter, potentially leading to a net loss of C from the ecosystem (Kirschbaum, 1994; Makipaa
et al., 1999).
A recent European study suggested that N deposition increased forest growth (Magnani et al.,
2007). Magnani et al. (2007) reported a strong correlation between estimated average long-term NEP and
estimated 1990 N wet-deposition (Holland et al., 2005) for 20 forest stands mostly in western Europe and
the conterminous U.S. The authors reported that when confounding effects of disturbance were factored
out, carbon sequestration was found to be increased by moderate N deposition (estimated up to 9.8 kg
N/ha/yr). However, this study did not evaluate forest stands that receive higher levels of N deposition that
may be showing negative symptoms of N saturation. Several responses to this study have been critical of
the methods and conclusions have been published (De Schrijver et al., 2008; De Vries et al., 2008; Sutton
et al., 2008b). For example, Sutton et al. (2008b) re-analyzed the data from Magnani et al. (2007) and
concluded the NEP response to N response reported by Magnani et al. (2008) was implausibly high
(725 kg C/kg wet-deposited N). Magnani (2008) responded by calculating NEP sensitivity to total N
deposition of approximately 175-225 kg C/kg N deposition and noted that one of the main effects of
increased N availability is an increased allocation to woody tissues (with a high C:N ratio of up to 500:1).
After considering the uncertainties in wet and dry N deposition and climate variability, Sutton et al.
(2008b) reported the estimated NEP response to N deposition was 68 kg C/ deposited N. Sutton et al.
(2008b) concluded that N deposition remains an important driver of NEP, but did not find support that the
NEP was overwhelmingly driven by N deposition.
The U.S. EPA conducted a meta-analysis of 17 observations from 9 studies in U.S. forests to
examine the effect of N fertilization on forest ecosystem C content (EC). Here EC was defined as the sum
3-117
-------
of C content of vegetation, forest floor and soil (Johnson et al., 2006). To avoid possible confounded
variability caused by site conditions, this meta-analysis only included studies of which control and
treatment sites experienced the same climatic, soil and vegetation conditions. Studies on N nutrient effects
along a deposition gradient, such as Magnani et al. (2007), were not included. The U.S. EPA meta-
analysis revealed that while there was a great deal of variation in response, overall N addition increased
EC by 6% for U.S. forest ecosystems (see Figure 3-38). Different from Magnani et al. (2007), this study
did not find any correlation between the amount ofN addition and the response magnitudes of EC.
It is uncertain if short term C accumulation may lead to long term C sequestration. N fertilization
could reduce the capacity of ecosystem to sequester decay resistant soil C. Giardina et al. (2004) found
that although N fertilization significantly increased plant production, the C flux moving to mineral soil
was reduced by 22% in a humid tropical forest in Hawaii. Mycorrhizal biomass comprises a substantial
carbon pool - represent up to 15% of soil organic matter in some ecosystems (Vogt et al., 1982). A meta-
analysis by Treseder (2004) suggested that mycorrhizal abundance decreased 15% under N fertilization.
17 -
V)
0)
_Q
E
3
c
c
o
• mean response ratio
I coniferous
W77X deciduous
CO
£
-------
community composition from graminoid tundra dominated by the tussock-forming sedge, Eriophorum
vaginatum, to shrub tundra dominated by Betula nana (Shaver et al., 2001). Consequently, this greatly
increased above-ground NPP, but had a larger effect on decomposition than on plant production, resulting
in a net loss of almost 2,000 g C/m2 from this ecosystem over 20 year (p <0.04). Carbon storage increased
above ground because of the accumulation of woody shrub biomass and litter, but this was offset by a
larger decrease of C in below-ground pools due to a pronounced decrease in the C contained in deep
organic (>5 cm depth) and upper mineral soil layers. This study clearly showed that increased nutrient
availability enhanced decomposition of below-ground C pools in deep soil layers more than it increased
primary production, leading to a substantial net loss of C from this ecosystem.
The key process responsible for the C loss was identified as increased deep soil C decomposition in
response to increased nutrient availability. The authors noted that increasing temperatures may amplify
these effects and further stimulate C losses from high-latitude systems. The amount of N released due to a
3 to 7 °C increase in mean annual temperature (MAT) is likely to range in magnitude from 7 to 9.4 g
N/m2/yr, respectively (Mack et al., 2004). This will cause species shifts in the vegetation community from
tussock to increased shrub abundance and lead to decreased ecosystem C storage. Finally, the decreased
soil moisture and increased depth of thaw that accompany temperature rise are predicted to have a
positive effect on decomposition (Shaver et al., 2001), releasing more C02.
Grasslands
Below-Ground Factors
An investigation by Neff et al. (2002) of long-term effects (10+ years) of N deposition (10 kg
N/ha/yr) in a dry meadow ecosystem indicated that N additions significantly accelerated decomposition of
soil C fractions with decadal turnover times while further stabilizing soil C compounds in mineral-
associated fractions with multi-decadal to century lifetimes. Despite these changes in the dynamics of
different soil pools, no significant changes in bulk soil C were observed, highlighting a limitation of the
single-pool approach for investigating soil C responses to changing environmental conditions (Neff et al.,
2002). The authors noted that it remains to be seen if the effects that were caused by relatively high,
decadal-term fertilizer additions are similar to those which would arise from lower, longer-term additions
of N to natural ecosystems from atmospheric deposition.
Interactions with Fire
Several lines of evidence suggest that N deposition may be contributing to greater fuel loads and
thus altering the fire cycle in a variety of ecosystem types (Fenn et al., 2003a). Invasive grasses, which
can be favored by high N deposition, promote a rapid fire cycle in many locations (D'Antonio and
Vitousek, 1992). The increased productivity of flammable understory grasses increases the spread of fire
and has been hypothesized as one mechanism for the recent conversion of coastal sage scrub (CSS) to
grassland in California (Minnich and Dezzani, 1998).
High grass biomass has also been associated with increased fire frequency in the Mohave Desert
(Brooks, 1999; Brooks and Esque, 2002; Brooks et al., 2004). This effect is most pronounced at higher
elevation, probably because the increased precipitation at higher elevation contributes to greater grass
productivity. Increased N supply at lower elevation in arid lands can only increase productivity to the
point at which moisture limitation prevents additional growth. Fire was relatively rare in the Mojave
Desert until the past two decades, but now fire occurs frequently in areas that have experienced invasion
of exotic grasses (Brooks, 1999).
3-119
-------
Wetlands
Above-Ground Processes
The 1993 NOxAQCD showed that N applications, ranging from 7 to 3120 kgN/ha/yr, stimulated
standing biomass production by 6-413% (U.S. EPA 1993a). However, the magnitude of the changes in
primary production depended on soil N availability and limitation of other nutrients. The degree of N
limitation to growth is varied among wetlands across the U.S. (Bedford et al., 1999).
The genus Sphagnum dominates ombrotrophic bogs and some nutrient poor fens in the Northern
U.S. and Canada. These mosses efficiently capture atmospheric deposition with retention rates between
50-90%, much of the variation due to the depth of the water table (Bedford et al., 1999). Studies
conducted on 4 species of Sphagnum in Maine (2 to 4 kg N/ha/yr ambient deposition) and New York (10
to 13 kg N/ha/yr ambient deposition) document that higher N deposition resulted in higher tissue N
concentrations and greater NPP (Aldous, 2002), but lower bulk density (Aldous, 2002). A study of
Sphagnum fuscum in six Canadian peatlands showed a weak, although significant, negative correlation
between NPP and N deposition when deposition levels were greater than 3 kg N/ha/yr (y = 150 - 3.4(x);
1^=0.01, p = 0.04) (Vitt et al., 2003). A study of 23 ombrotrophic peatlands in Canada with deposition
levels ranging from 2.7 to 8.1 kg N/ha/yr showed peat accumulation increases linearly with N deposition
(y = 2.84(x) +0.67, r2 = 0.32, p <0.001), however in recent years this rate has begun to slow indicating
limited capacity for N to stimulate accumulation (Moore et al., 2004).
Primary production of plant species from intertidal wetlands typically increases with N addition,
however most studies apply fertilizer treatments that are several orders of magnitude larger than
atmospheric deposition (Mendelssohn, 1979; Wigand et al., 2003; Tyler et al., 2007; Darby and Turner,
2008). N loads brought by tidal water and ground water (565-668 kg N/ha/yr) are much larger than N
depositing directly to the surface of coastal marshes, which suggested that direct N deposition may have
limited impacts on this ecosystem (Morris, 1991). On the other hand, indirect atmospheric deposition that
is N deposited to the watershed and transported via surface or ground water, could be the major sources of
the total N load to coastal marshes. For example, model calculation suggested that the contribution from
the atmosphere (36 million kg N/yr) was about 21-30% of the total N loading (170 million kg N/yr) in
Chesapeake Bay waters (U.S. EPA, 2000b). Therefore 30% of the N delivered to wetlands via estuarine
tides would originate from atmospheric deposition. Future studies are needed to determine the role of
indirect atmospheric N deposition on the nutrient budget of intertidal wetlands.
Below-Ground Processes
Bragazza et al. (2006) investigated the decomposition rates of recently formed litter peat samples
collected in nine European countries under a natural gradient of atmospheric N deposition from 2 to
20 kg N/ha/yr. They found enhanced decomposition rates for material accumulated under higher
atmospheric N supplies resulted in higher C02 emissions and dissolved organic carbon release. The
increased N availability favored microbial decomposition by removing N constraints on microbial
metabolism and through a chemical amelioration of litter peat quality with a positive feedback on
microbial enzymatic activity. In a follow-up study, Bragazza and Freeman (2007) evaluated whether there
was a relationship between N deposition and a decay-inhibiting polyphenol in the Sphagnum tissue. They
found that as N deposition level increased the polyphenol concentration decreased. This observation
implies the lower concentration of decay-inhibiting polyphenols would lead to accelerated peat
decomposition. Although some uncertainty remains about whether Sphagnum will continue to dominate
litter peat, these findings indicated that even without such changes, increased N deposition poses a serious
risk to the valuable peatland C sinks.
3-120
-------
Reduced vs. Oxidized N
The form of added N may regulate wetland response to N deposition. Experimental applications of
N03 appear to have been less effective at stimulating wetland plant productivity than applications of
NH4+ (U.S. EPA, 1993a). This may reflect higher rates of denitrification in response to the added N03 .
suggesting the importance of competition between plants and microbes for bioavailable N. Plants appear
to compete more successfully for NH44" and microbes to compete more successfully for N03 . An
important caveat expressed by U.S. EPA (1993a), however, was that the results of relatively short-term N
fertilization experiments are not necessarily good predictors of long-term wetland community responses
to increased N inputs.
NEE of Grassland, Tundra and Wetlands
In the meta-analysis of 16 observations from 9 publications on the relationship between N addition
and C sequestration of non-forest ecosystems, N addition had no significant effect on NEE of non-forest
ecosystems (Figure 3-39). N limitation to NPP is globally distributed and therefore plant productivity is
normally enhanced by N addition. A meta-analysis by Lebauer and Treseder (2008) indicated that N
fertilization increased aboveground NPP (ANPP) in all non-forest ecosystems except for desert. However,
N addition also has been observed to stimulate ecosystem C loss. For example, N fertilization stimulated
soil organic carbon decomposition in arctic tundra. Increasing N deposition led to higher C loss in
temperate peatlands (See Section 3.3.3.1). In an agricultural experiment site, Khan et al. (2007) observed
that 40 to 50 years N fertilization resulted in a net decline in soil C despite massive residue C
incorporation. This meta-analysis indicated that N addition had no significant impact on C sequestration
in non forest ecosystems, which may be due to C gain via NPP was exceeded by C loss via heterotrophic
respiration.
-------
3.3.3.2. Freshwater Aquatic
The biogeochemical cycles of N, P and C are linked in freshwater ecosystems (Figure 3-40),
therefore N additions alter the balance of all three cycles. InN-limited aquatic systems, atmospheric
inputs of N increase productivity and alter biological communities, especially phytoplankton. N
deposition effects on productivity are discussed in this Section. The results of numerous publications
addressing the experimental additions of N are tabulated in Annex C. Evidence that altered productivity
leads to altered community structure is discussed in Section 3.3.5.
Generally, the dose-response data for aquatic organisms such as those cited below are expressed in
concentration units, as mg/L or (iinol/L of N, for example. Such concentration data cannot be directly
related to ecosystem exposure (i.e. deposition), which is generally expressed in such units as kg N/ha.
This is because N deposition can result in widely varying concentrations of N compounds (especially
NO, ) in water. For convenience, a concentration of 1 mg/L of N (as, for example, in the case of NO;, -N
or NFL -N) is equal to 71.4 f^mol/L or 71.4 ueq/L of NO;, or NHL
CO,
deposition
Terrestrial input
C02
tl
CO,
CO,
uptake
excretion
L respiration! photosynthesis
deposition
I N fixation
i
\ * i
Primary^ N
f producers ^cretlon
*1
Grazers,
predators
and viruses
C cycle
N cycle
P cycle
decomposition
Detritus
Sediments
water surface
Benthic
decomposition
Figure 3-40. N cycle in freshwater ecosystem.
N-Limitation
A freshwater lake or stream must be N-limited to be sensitive to N-mediated eutrophication. There
are many examples of fresh waters that are N-limited or N and P co-limited (e.g., Elser et al., 1990; Fenn
et al., 2003a; Tank and Dodds, 2003; Bergstrom et al., 2005; Baron, 2006; Bergstrom and Jansson, 2006).
In a meta-analysis that included 653 datasets, Elser et al. (2008) found that N-limitation occurred as
frequently as P-limitation in freshwater ecosystems. Recently, a comprehensive study of available data
from the northern hemisphere surveys of lake along gradients of N deposition show increased inorganic N
concentration and productivity to be correlated with atmospheric N deposition (Bergstrom and Jansson,
3-122
-------
2006). The results are unequivocal evidence ofN limitation in lakes with low ambient inputs of N, and
increased N concentrations in lakes receiving N solely from atmospheric N deposition (Bergstrom and
Jansson, 2006). These authors suggested that most lakes in the northern hemisphere may have originally
been N-limited, and that atmospheric N deposition has changed the balance of N and P in lakes so that P-
limitation is generally observed today. If this is correct, the role of atmospheric N deposition as an
influence on aquatic primary production may have been underestimated throughout the entire history of
limnology.
Recent research (e.g., Wolfe et al., 2001, 2003, 2006; Lafrancois et al., 2003; Das et al., 2005;
Saros et al., 2005) has provided additional evidence indicating that N deposition has played an important
role in influencing the productivity of oligotrophic, high-elevation lakes in the western U.S. and Canada,
and the Canadian arctic. There is evidence suggesting historical N-limitation of some lakes based on
paleolimnological studies conducted in mountainous regions of the western U.S. that have been the
recipient of elevated levels of N, but not S, deposition over background values (see Section 3.3.4).
Interactions between N and P loading are discussed in Annex C.
Productivity investigations have included gradient studies in which the relationship between lake N
concentration and primary productivity (reported as Chi a, NPP, or an index such as the lake chemistry
ratio of dissolved inorganic N [DIN] to total P, DIN: TP) was surveyed and correlated with atmospheric N
deposition. Productivity studies have also included lake and stream bioassays in which N was added to
waters in field or laboratory to measure the response. The most common, and easiest to document,
indicators of change in algal productivity are measures of the concentration of Chi a and water clarity.
However, clarity is also strongly influenced by erosional inputs of fine sediment to the lake or stream
system. Chi a concentration is generally more directly tied to algal productivity than is water clarity.
Phytoplankton Biomass
Studies have shown an increase in lake phytoplankton biomass with increasing N deposition in
several regions, including the Snowy Range in Wyoming (Lafrancois et al., 2003), the Sierra Nevada
Mountains in California (Sickman et al., 2003), and across Europe (Bergstrom and Jansson, 2006).
Gradient studies of undisturbed northern temperate, mountain, or boreal lakes that receive low levels of
atmospheric N deposition found strong relationships between N-limitation and productivity where N
deposition was low, and P and N+P limitations where N deposition was higher (Fenn et al., 2003a;
Bergstrom et al., 2005; Bergstrom and Jansson, 2006).
Bergstrom and Jansson (2006) concluded the eutrophication caused by inorganic N deposition
indicates that phytoplankton biomass in a majority of lakes in the northern hemisphere is limited by N in
their natural state. Chemical data from 3,907 lakes and phytoplankton biomass data from 225 lakes from
Swedish monitoring programs showed a clear north-south gradient of increasing lake concentrations and
algal productivity related to the pattern of increasing N deposition input (Bergstrom et al., 2005). The
lowest productivity was found at sites where wet N deposition was about 1.3 kg N/ha/yr; increasing
productivity occurred at greater than 2.2 kg N/ha/yr (Bergstrom et al., 2005). Although these lakes are all
in Sweden, the study size and the strong correlation between productivity and atmospheric N deposition
makes the results likely relevant to North American audiences.
Experiments conducted with mesocosms in lakes where N03 was below the detection level found
a strong response in phytoplankton biomass with additions of N (bringing concentrations to -1.0 mg N/L)
and even stronger responses to additions of N plus P, but not P alone (Lafrancois et al., 2004). The reverse
was found in Colorado Front Range lakes with ambient N03 concentrations of-1.0 mg/L: productivity
increased with additions of N plus P or P only, but not N03 alone (Lafrancois et al., 2004).
A meta-analysis of enrichment bioassays in 62 freshwater lakes of North America, including many
of the studies described above, found algal growth enhancement from N amendments to be common in
slightly less than half the studies (Elser et al., 1990). There was a mean increase in phytoplankton biomass
of 79% in response to N enrichment (average of 46.3 j^ieq/L N) (Elser et al., 1990). This meta-analysis
was recently repeated with a much large data set and similar results (Elser et al., 2007). Freshwater
3-123
-------
enrichment bioassay studies from 990 separate studies worldwide were gleaned from the literature. The
natural log (ln)-transformed response ratio (RRx), a frequently used effect metric in ecological meta-
analysis, was equal at about 0.3 for N and P experiments with stream benthos (periphyton) bioassays, and
approximately equal at about 0.2 for lake phytoplankton. There was a stronger response to P than N in
lake benthos studies, but the RRX for N was still about 0.3, showing that many sites increased
productivity when fertilized with N alone (Elser et al., 2007). Additional studies are discussed in Table
3-16.
Table 3-16. Summary of additional evidence of N effects on productivity of freshwater ecosystems.
Region
Endpoint
Observation
Ecosystem
Type
Reference
Lake Tahoe, CA
Primary
productivity
Water clarity
Long-term (28 years) measurements showed that primary
productivity has doubled, while water clarity has declined,
mostly as a result of atmospheric N deposition
Lake
Goldman
(1988); Jassby
(1994)
Alaska
Primary production
Fungal biomass
Decomposition
rates
Benthic
macroinvertebrate
N amendment experiments with 6.4 |jM N to elicited
responses throughout the ecosystem, including enhanced
primary production, enhanced fungal biomass and elevated
leaf litter decomposition rates, and a fourfold to sevenfold
greater benthic macroinvertebrate abundance
Small arctic
streams
Benstead et al.
(2005)
Rocky Mountain
Lakes, Colorado
Growth
In situ mesocosm: incubations the growth of the diatom A.
formosa has been stimulated with N amendments during from
6.4 to 1616 [JM N
Diatoms
McKnight
et al., (1990)
Yellowstone
National Park,
Wyoming
Growth
In situ incubations in large lakes: stimulated F. crotonensi.
This publication did not reveal how much N was added to the
incubations
Lakes
Interlandi and
Kilham, (1998)
Central Rocky
Mountains of
North America
Growth
The N requirements for A. formosa and F. crotonensis were
determined to be 0.041 |jM and 0.006 |jM, respectively, and
higher concentrations stimulated growth
Alpine lakes
Michel et al.
(2006)
Chlorophyll a
The most widely used index of biological change in response to nutrient addition is measurement
of Chi a concentration in water. Surveys and fertilization experiments show increased inorganic N
concentration and aquatic ecosystem productivity (as indicated by Chi a concentration) to be strongly
related. For example, a series of in situ meso- and microcosm N amendment experiments more than 30
years ago showed increases in lake algal productivity. Lake 226S in Ontario's Experimental Lake Area
(ELA) showed doubling of average epilimnetic Chi a over five years of fertilization. However, because
the response to P fertilization was much greater, the effects of N received less attention (Schindler, 1980).
Other ELD lakes that had relatively low N to P concentration ratios experienced 3 to 10 times greater
increases in Chi a than Lake 226S (Schindler, 1980).
Similar experiments at Castle Lake, California, the Snowy Range of southern Wyoming, and
Alaskan arctic foothill lakes yielded measurable increases in Chi a, primary productivity and algal
3-124
-------
detritus with N amendments (Axler and Reuter, 1996; Levine and Whalen, 2001; Nydick et al., 2003;
Lafrancois et al., 2004; Nydick et al., 2004b).
Periphyton Biomass
N effects have been observed in periphyton, which grows on rocks or sediment in lakes and
streams where there is sufficient light for photosynthesis. No studies have reported resource requirements
for periphyton, although several papers described stimulated growth with N amendments from
ecosystems throughout the U.S. (Annex C), including streams in Alaska, Arizona, Iowa, Texas,
Minnesota, Missouri, and lakes in California, Colorado, and Massachusetts. Growth stimulation occurred
with N additions ranging from 8 to 50 (iM/L, or with exposure to 0.5 M N concentrations on agar
substrate (Bushong and Bachmann, 1989; Allen and Hershey, 1996; Wold and Hershey, 1999; Smith and
Lee, 2006). Additional lake bioassay experiments that enriched the water column down into the sediments
found enhancement of periphyton growth on bioassay container walls in experiments in California,
Wyoming, and Massachusetts (Axler and Reuter, 1996; Nydick et al., 2004a; Smith and Lee, 2006).
Strong N limitation of benthic algae has also been inferred in streams of Arizona (Grimm and Fisher,
1986), California (Hill and Knight, 1988), Missouri (Lohman et al., 1991), and Montana (Lohman and
Priscu, 1992; Smith and Nicholas, 1999).
Trophic Status Indices
Nutritional responses of aquatic ecosystems to atmospheric N deposition are heavily dependent on
surface water P concentrations. Thus, chemical ratios of N to P can be very useful in evaluating
eutrophication potential. A series of papers, described below, has been published exploring nutrient
limitations and offering indices that describe the trophic state of freshwaters. Valuable insights have been
gained from several indices, including total N to total P (TN:TP), dissolved inorganic N to total P
(DIN:TP), dissolved inorganic N to total dissolved P (DIN:TDP), dissolved inorganic N to soluble
reactive P (DIN:SRP), and dissolved inorganic N to the ratio of Chi a to total P (DIN:[Chl a:TP]). While
there are publications that compare the effectiveness of some of these indices, it appears that different
indices are useful for different purposes; we make no attempt to favor one over another.
Algal growth was reported to be limited at DIN:TP values between 5 and 20 (Schindler, 1980;
Grimm and Fisher, 1986; Morris and Lewis, 1988; Downing and McCauley, 1992; Bergstrom and
Jansson, 2006). When DIN:TP ratios are greater than reference values, growth stimulation, N and P
colimitation, or P limitation commonly occurs (Sickman et al., 2003). In a Swedish lake survey, N-
limitation occurred in lakes where the DIN:TP ratio was less than 7 (concentrations <33 (.iM N/L). Co-
limitation of both N and P were found in lakes with DIN:TP ratio between 8 and 10, and P-limitation at
DIN:TP values greater than 10. This corresponds roughly to N concentrations of 45 to 80 |_iM N/L for co-
limited lakes, and concentrations >80 (.iM N/L for P-limited lakes (Bergstrom et al., 2005). Other
thresholds for N-limitation were reported in the literature to occur at DIN:SRP ratios <4 (Lohman and
Priscu, 1992) and <10 (Wold and Hershey, 1999).
Bergstrom et al. (2005) reported a new index, (DIN: [Chi a:TP]) to indicate the eutrophication of
lakes from N deposition. The choice of DIN/[Chl a:TP] was based on whole lake experiments in Sweden
(Jansson et al., 2001) and permits the assessment of a possible eutrophication effect of N deposition
independent of differences in P input between lakes in different regions. These researchers found that the
mean Chi a:TP ratios increased more than three times from low N to high N deposition areas, indicating
that N deposition contributed to eutrophication.
3.3.3.3. Estuarine and Marine
In coastal marine ecosystems, the nutrients most commonly associated with phytoplankton growth
are N, P, and Si (see Annex C for interactions between hydrology and nutrient cycling). Interactions
3-125
-------
among the supplies of these nutrients can affect phytoplankton species composition in ways that might
effect ecosystem function (Paerl et al., 2001a; Riegman, 1992). The relative proportions of these nutrients
are important determinants of primary production, food web structure, and energy flow through the
ecosystem (Dortch and Whitledge, 1992; Justic et al., 1995a; Justic et al., 1995b; Turner et al., 1998).
There is a strong scientific consensus that N is the principal cause of coastal eutrophication in the U.S.
(NRC, 2000) On average, human activity has likely contributed to a sixfold increase in the N flux to the
coastal waters of the U.S., and N now represents the most significant coastal pollution problem (Howarth
et al., 2002b; Howarth and Marino, 2006). Atmospheric deposition is responsible for a portion of the N
input (Howarth et al., 2002a).
Ecological effects of accelerated estuarine eutrophication and climatic perturbations such as
droughts, floods, and hurricanes are often observed most closely at the level of the primary producers.
Phytoplankton can be divided into functional groups that reflect ecological change, for example:
chlorophytes, cryptophytes, cyanobacteria, diatoms, and dinoflagellates (Pinckney et al., 2001).
The relative abundances of these groups can be determined using photopigment indicators which
can be easily measured in the laboratory (Paerl et al., 2003). Changes in phytoplankton community
composition, which may affect food web interactions, can have important effects on nutrient cycling. For
example, if the growth of phytoplankton species that are more readily grazed by zooplankton (i.e.,
diatoms) is favored, trophic transfer will occur in the water column from diatoms to fish and nutrient
export will take place as fish move to the ocean. However, if the phytoplankton that are favored by
nutrient addition and disturbance are not readily grazed (i.e., cyanobacteria and dinoflagellates), trophic
transfer will be poor. In that case, more unconsumed algal biomass will settle to the bottom where it can
contribute to 02 consumption and associated hypoxia (Paerl et al., 2003).
To evaluate the impacts of eutrophication, five biological indicators were used in the recent
national assessment of estuary trophic condition: Chi a, macroalgae, dissolved 02, nuisance/toxic algal
blooms, and submerged aquatic vegetation (SAV) (Bricker et al., 2007) (Figure 3-41). Each of these
indicators is discussed below and/or within the biodiversity section of this document (see Section 3.3.5.4).
N-Limitation
Estuaries and coastal waters tend to be N-limited and are therefore inherently sensitive to increased
N loading (D'Elia et al., 1986; Howarth and Marino, 2006). There is a scientific consensus that N-driven
eutrophication of shallow estuaries has increased over the past several decades and that environmental
degradation of coastal ecosystems is now a widespread occurrence (Paerl et al., 2001a).
For example, the frequency of phytoplankton blooms and the extent and severity of hypoxia have
increased in the Chesapeake Bay (Officer et al., 1984) and Pamlico estuaries in North Carolina (Paerl
et al., 1998) and along the continental shelf adjacent to the Mississippi and Atchafalaya rivers discharges
to the Gulf of Mexico (Eadie et al., 1994). It is partly because many estuaries and near-coastal marine
waters are degraded by N enrichment that they are highly sensitive to potential adverse impacts from N
addition from atmospheric deposition.
N enrichment of marine and estuarine waters can alter the ratios among nutrients and affect overall
nutrient limitation. The sensitivity of estuarine and coastal marine waters to eutrophication from
atmospheric N deposition depends on the supply of, and relative availability of, N and P. At upstream
freshwater locations in Chesapeake Bay, P is often the limiting nutrient (Larson and Moore, 1985). At the
transition between fresh water and salt water, N and P may be co-limiting, whereas the saltwater
environments of the outer bay are usually N-limited (Fisher et al., 1998; Rudek et al., 1991). Nutrient
limitation varies in space and over time, in response to changes in discharge and temperature that interact
with estuarine morphology and hydrology (Paerl et al., 2006).
3-126
-------
Primary symptoms Description
A measure used to indicate the amount of microscopic algae
(phytoplankton) growing in a water body. High concentrations can lead
to low dissolved oxygen levels as a result of decomposition.
Large algae commonly referred to as "seaweed." Blooms can cause
losses of submerged aquatic vegetation by blocking sunlight.
Additionally, blooms may smother immobile shellfish, corals, or other
habitat. The unsightly nature of some blooms may impact tourism due
to the declining value of swimming, fishing, and boating.
Secondary symptoms Description
Low dissolved oxygen is a eutroph ic symptom because it occurs as a
result of decomposing organic matter (from dense algal biooms), which
sinks to the bottom and uses oxygen during decay. Low dissolved
oxygen can cause fish kills, habitat loss, and degraded aesthetic values,
resulting in the loss of tourism and recreational water use.
Loss of submerged aquatic vegetation (SAV) occurs when dense algal
blooms caused by excess nutrient additions (and absence of grazers)
decrease water clarity and light penetration. Turbidity caused by other
factors (e.g, wave energy, color) similarly affects SAV. The loss of SAV can
have negative effects on an estuary's functionality and may impact
some fisheries due to loss of a critical nursery habitat.
Thought to be caused by a change in the natural mixture of nutrients
that occurs when nutrient inputs increase over a long period of time.
These blooms may release toxins that kill fish and shellfish. Human
health problems may also occur due to the consumption of
contaminated shellfish or from inhalation of airborne toxins. Many
nuisance/toxic blooms occur naturally, some are advected into
estuaries from the ocean; the role of nutrient enrichment is unclear.
Source: Brickeret al. (2007))
Figure 3-41. Description of the eutrophic symptoms included in the national estuary condition
assessment.
The data for 92 worldwide coastal marine sites analyzed by Smith (2006), for which measurements
of both total N and total P were available, illustrated that about half of the sites had total nitrogen
(TN):total phosphorus (TP) above the Redfield ratio, which is commonly used to evaluate nutrient
limitation in freshwater (TN:TP = 16). As was emphasized in earlier work on nutrient limitation in fresh
waters by Redfield (1958) and Reiners (1986), elemental stoichiometry is a fundamental property of life
that probably stems from the shared phylogenetic histories of marine and freshwater autotrophs (Smith,
2006; Sterner and Elser, 2002).
In general, the scientific community is at an early stage in development of an understanding of the
effects of anthropogenic activities on the stoichiometry of nutrient loading to estuaries and marine waters
(Dodds, 2006; Turner, 2002). Changes in nutrient stoichiometry in estuarine and marine ecosystems could
alter algal assemblages and cascade to higher trophic levels (Frost et al., 2002).
Chlorophyll a
(Phytoplankton)
Macroalgal blooms
°2
Dissolved
oxygen
Submerged
aquatic vegetation
Nuisance/toxic
blooms
3-127
-------
~
¦ah
a
*i 40
t 30
a? 20
rt
~
Low Moderate High
Chlorophyll o
0 200 400
¦EZZ) Kilometers
Miles
0 100 200
N
,^-tlD
.y
op
/
A
A
¦¦I High: symptoms occur periodically or persistently and/or over an extensive area.
I I Moderate high: symptoms occur less regularly and/or over a medium to extensive area.
i i Moderate: symptoms occur less regularly and/or over a medium area.
^¦1 Moderate low: symptoms occur episodically and/or over a small to medium area.
' i Low: few symptoms occur at more than minimal levels.
I I Unknown: insufficient data for analysis.
Change in eutrophic condition since 1999 assessment
A Symptoms improved since 1999 assessment.
O No change in symptoms since 1999 assessment.
V Symptoms worsened since 1999 assessment.
~ Insufficient data to show trend
Source: Bricker et al. (2007)
Figure 3-42. A high Chi a rating was observed in a large number of the nation's estuaries. White squares
indicate that data were not available for a particular estuary.
Chlorophyll a
Chi a concentration in estuarine or marine water is an indicator of total phytoplankton biomass. It
can signal an early stage of water quality degradation related to nutrient loading. High concentration of
Chi ci suggests that algal biomass is sufficiently high that it might contribute to low dissolved 02
concentration due to increased decomposition of dead algae. In the national estuary condition assessment,
high Chi a concentration was the most widespread documented symptom of eutrophication (Bricker et al.,
2007) (see Figure 3-42). Half of the estuaries for which there were available data exhibited high Chi a
concentration (Bricker et al., 2007).
San Francisco Bay, California is an example of an estuary that has experienced considerable
increases in Chi a concentrations in recent years. Phytoplankton biomass in much of the bay has increased
by more than 5% per year from 1993 to 2004. During this time, modeled primary production has doubled
and nutrient loading is identified as one of eight possible causes (Cloern et al., 2006).
Macroalgal Abundance
Macroalgae are generally referred to collectively as seaweed. Macroalgal blooms can contribute to
loss of important SAV by blocking the penetration of sunlight into the water column. Although
3-128
-------
macroalgal data for estuaries in the U.S. were generally sparse, the national estuary condition assessment
reported that conditions were moderate or high for 33 of the estuaries evaluated (Bricker et al., 2007).
Dissolved O2
Hie decomposition of organic matter associated with increased algal abundance consumes
dissolved 02 and can reduce dissolved 02 concentrations in eutrophic waters to levels that cannot support
aquatic life. Decreased dissolved 02 can lead to development of hypoxic or anoxic zones that are
inhospitable to fish and other life forms. Perhaps the most important environmental effect of N input to
coastal waters is the development of hypoxia. The largest zone of hypoxic coastal water in the U.S. has
been documented in the northern Gulf of Mexico on the Louisiana-Texas continental shelf. During
midsummer, this hypoxic zone has regularly been larger than 16,000 km" (Rabalais, 1998). The timing,
duration, and spatial extent of hypoxia in this case are related mostly to the nutrient flux from the
Mississippi River (Justic et al., 1993, 1997; Lohrenz et al^ 1997; Paerl et al., 2001a; Rabalais et al.,
1996).
Connecticut
T I.
New York \ .
Long Island Sound
^ sil
Atlantic Ocean S5S-S
am eo-90
M90-100 |
Source: Bricker et al. (2007))
Figure 3-43. Frequency of hypoxia in Long Island Sound, 1994 to 2002.
Although impacts on dissolved 02 can be quite severe in the areas where they are manifested, the
national assessment reports that the severity of dissolved 02 impacts related to eutrophication are
relatively limited in many of the systems assessed (Bricker et al., 2007). In the shallow estuary of Long
Island Sound, the existence of extended periods of low dissolved 02 is a notable problem, and
atmospheric deposition is considered to comprise a significant fraction of the total N loading. Dissolved
02 levels below 3 mg/L are common, and levels below 2 mg/L also occur. During some years, portions of
the Long Island Sound bottom waters become anoxic ( <1 mg/L; see Figure 3-43).
Nuisance/Toxic Algal Blooms
Nuisance or toxic algal blooms reflect the proliferation of a toxic or nuisance algal species that
negatively affects natural resources or humans. Such blooms can release toxins that kill fish and shellfish
and pose a risk to human health. Unlike the other indicators of estuarine eutrophication, the role of
nutrients in stimulating toxic algal blooms is less clear. Of the 81 estuary systems for which data were
available, 26 exhibited a moderate or high symptom expression for nuisance or toxic algae (Bricker et al.,
2007).
3-129
-------
3.3.3.4. Summary of Nitrogen Effects on Carbon Cycling
The evidence is sufficient to infer a casual relationship between N deposition and alteration to the
biogeochemical cycling of C in terrestrial, wetland, freshwater and marine ecosystems. N deposition causes
alteration to the C cycle in forest ecosystems. Experimental N addition studies show a range of responses
in terms of mortality and productivity. In general, moderate to high additions of N lead to either no
significant change in growth rates or transient growth increases followed by increased mortality,
especially at higher rates of fertilization (See Table 3-15). Although there remains considerable
uncertainty in the potential response of soil C to increases in Nr additions (Neff et al., 2002), a meta-
analysis by Johnson and Curtis (2001) suggested that N fertilization caused an 18% increase in soil
carbon content. The effects of N addition on soil respiration are mixed; reductions at high levels of N (Lu
et al., 1998), no effect (Vose et al., 1995) and increases (Griffin et al., 1997; Mikan et al., 2000) all have
been observed.
Region trends in NEP have been documented through models based on observational gradient
studies. Magnani et al. (2007) recently reported a strong correlation between estimated average long-term
NEP (NEPav) and estimated 1990 wet N deposition (up to 9.8 kg N/ha/yr) for 20 forest stands mostly in
Western Europe and the conterminous U.S. Sutton et al. (2008b) critiqued the methods of Magnani et al.
(2007). After considering the uncertainties in N deposition and climate variability, Sutton et al. (2008b)
reported the estimated NEP response to N deposition was 68 kg C/wet-deposited N. Sutton et al. (2008b)
agreed with Magnani et al. (2007) that N deposition remains an important driver of NEPav, but did not
support the NEPav was overwhelmingly driven by N deposition. The ISA staff conducted a meta-analysis
to examine the effect of N fertilization on forest ecosystem C content, defined as the sum of C content of
vegetation, forest floor and soil (Johnson et al., 2006) and found that N addition increased ecosystem C
by 6%. In summary, it remains unclear to what extent N deposition at current levels could potentially
increase C sequestration in forest ecosystems in the U.S.
N addition causes alterations to the C cycle in Tundra. Mack et al. (2004) examined C and N pools
in a long-term fertilization experiment at the arctic Long-Term Ecological Research site near Toolik Lake,
AK. This study showed that increased nutrient availability enhanced decomposition of below-ground C
pools in deep soil layers more than it increased primary production, leading to a substantial net loss of C
from this ecosystem.
N deposition causes alteration to the C cycle in freshwater wetlands. In Sphagnum-dominated
ombrotrophic bogs, higher N deposition resulted in higher tissue N concentrations and greater NPP
(Aldous, 2002), but lower bulk density (Aldous, 2002). A study of 23 ombrotrophic peatlands in Canada
with deposition levels ranging from 2.7 to 8.1 kg N/ha/yr showed peat accumulation increases linearly
with N deposition (y = 2.84(x) + 0.67, r2 = 0.32, p <0.001), however in recent years this rate has begun to
slow indicating limited capacity for N to stimulate accumulation (Moore et al., 2004). Soil respiration has
been studied in European countries under a natural gradient of atmospheric N deposition from 2 to 20 kg
N/ha/yr. They found enhanced decomposition rates for material accumulated under higher atmospheric N
supplies resulted in higher C02 emissions from soil. Primary production of plant species from intertidal
wetlands typically increases with N addition, however most studies apply fertilizer treatments that are
several orders of magnitude larger than atmospheric deposition (Darby and Turner, 2008; Mendelssohn,
1979; Tyler et al., 2007; Wigand et al., 2003).
N deposition causes alteration to the C cycle in freshwater aquatic ecosystems. The productivity of
many freshwater ecosystems is currently limited by the availability of N (Elser et al., 2007). European
and North American lakes may have been N-limited before human-caused disturbance, and remote lakes
may have remained N-limited until slight increases in atmospheric N deposition brought about an increase
in phytoplankton and periphyton biomass and induced P limitation. Numerous studies investigated the
relationship between lake N concentration and primary productivity (reported as Chi a, NPP, or an index
such as the lake chemistry ratio of dissolved inorganic N [DIN] to total P, DIN: TP) and atmospheric N
deposition. N addition experiments of lake and stream bioassays in which N was added to waters in field
or laboratory to measure the response. Gradient studies of undisturbed northern temperate, mountain, or
3-130
-------
boreal lakes that receive low levels of atmospheric N deposition found strong relationships between N-
limitation and productivity where N deposition was low, and P and N+P limitations where N deposition
was higher (Bergstrom et al., 2005; Bergstrom and Jansson, 2006; Fenn et al., 2003a).
N deposition causes alteration to the C cycle in near coastal marine ecosystems. As previously
mentioned, N deposition is not the sole source of N loading to estuaries and its contribution varies across
the U.S. Estuaries and coastal waters tend to be N-limited and are therefore sensitive to increased
atmospheric N loading (Elser et al., 2007; D'Elia et al., 1986; Howarth and Marino, 2006). This is at least
partly because denitrification by microbes found in estuarine and marine sediments releases much of the
added N inputs back into the atmosphere (Vitousek et al., 1997). However, other limiting factors occur in
some locations and during some seasons. Levels of N limitations are affected by seasonal patterns. N-
limited conditions are likely to be found during the peak of annual productivity in the summer.
Numerous studies evaluate the relationship between N loading, eutrophication and ecological
endpoints including Chl-a, macroalgal abundance, dissolved 02, nuisance/toxic and algal blooms In the
national estuary condition assessment, high Chi a concentration was the most widespread documented
symptom of eutrophication (Bricker et al., 2007) (see Figure 3-42).
3.3.4. Biogenic Trace Gases: Nitrous Oxide, Methane, Nitric Oxide and
VOCs
Methane (CH4) and nitrous oxide (N20) are greenhouse gases (GHGs) contributing to global
warming. Although atmospheric concentrations of CH4 (1774 ppb) and N20 (319 ppb) are much lower
than C02 (379 ppm), their global warming potential is 23 and 296 times that of C02, respectively. Human
activities have dramatically increased atmospheric concentration of CH4 by 48% and N20 by 18% since
pre-industrial times (IPCC 2007). The continuing increase of those GHGs concentrations have been
shown to threaten human and ecosystem health. Anthropogenic N deposition to natural ecosystem is a
primary component of global change. Additional N input not only changes the global N cycle, but also
has profound effects on biogeochemical processes associated with GHGs emission (Vitousek et al., 1997;
Dalai et al., 2003; Bodelier and Laanbroek 2004). In the following section, the effects of N addition on
CH4 and N20 emissions were reviewed and quantitatively synthesized by meta-analysis. Further details
on this meta-analysis including study site, ecosystem type, N addition level, chemical form of N,
experimental conditions, relationship between N addition and CH4 flux, are in included Annex C.
3.3.4.1. Methane
Atmospheric CH4 originates mainly (70-80%) from biogenic sources (Le Mer and Roger, 2001).
Methane is produced in anaerobic environment by methanogenic bacteria during decomposition of
organic matter. Once produced in soil, CH4 can then be released to the atmosphere or oxidized by
methanotrophic bacteria in the aerobic zone (Le Mer and Roger, 2001). Methane production and
oxidation processes occur simultaneously in most ecosystems. Wetland soils are generally CH4 sources,
accounting for about 20% of global CH4 emission (see Annex C for a more detailed discussion of
methane in wetlands). Non-flooded upland soils are the most important biological sink for CH4,
consuming about 6% of the atmospheric CH4 (Le Mer and Roger, 2001). Numerous studies have
demonstrated that N is an important regulatory factor for both CH4 production and oxidation (Bodelier
and Laanbroek, 2004).
The U.S. EPA conducted a meta-analysis, including 61 observations from 27 publications, to
evaluate the relationship between N addition and CH4 flux. Details on those publications, including study
site, ecosystem type, N addition level, chemical form ofN, experimental condition, are given in Annex C.
The impact of N addition on CH4 source and sink strength were estimated by CH4 emission and CH4
3-131
-------
uptake respectively. The result suggested that N addition significantly increased CH4 emission by 115%
for grasslands and wetlands (Figure 3-44). This response ratio did not differ among vegetation type, N
addition level, chemical form of N and experiment condition.
14 -
CO
5
response ratio
Figure 3-44. Effects of N addition on biogenic CH4 emission. The bars show the distribution of the number
of studies categorized by ecosystem type. The dot with error bars shows the overall mean
response ratio with 95% CI.
Overall, N addition significantly reduced CH4 uptake by 38% (Figure 3-45). Ecosystem type, N
form and experiment condition influenced the degree of CH4 uptake response to N addition (Figure 3-46).
Methane uptake was reduced for all ecosystems, but this inhibition was significant only for coniferous
and deciduous forest, with a reduction of 28% and 45%, respectively (Figure 3-46).
All forms of N fertilizer except urea were shown to reduce CH4 uptake (Figure 3-46). Several
possible mechanisms have been proposed to explain the inhibition in CH4 oxidation by N addition.
Besides the oxidation of CH4, methane monooxygenase (MMO) can convert NH4+ to N03 . and NH44"
therefore usually inhibits CH4 oxidation by competing for MMO (Bodelier and Laanbroek, 2004).
Methanotrophic bacteria are sensitive to osmotic stress induced by salts. Inhibition of CH4 uptake by
nitrogenous salts (e.g., KN03, NH4CI, and NH4NO3) and non-nitrogenous salts (e.g., K2S04, KC1 and
NaCl) has been observed in field and laboratory studies (Bodelier and Laanbroek, 2004; King and
Schnell, 1998). Other mechanisms, such as toxicity of nitrite (N02 ) produced by nitrification or
denitrification processes, may also involve in the inhibition of CH4 oxidation (Schnell and King, 1994).
3-132
-------
47 -
• mean response ratio
i i coniferous
Y////AI deciduous
kkski tropical forest
Bsaggg grassland
1 1 heathland
mTTTII tundra
LLLLMJI wetland
0.0
0.2
0.4
0.6
0.8
1.0
1.2
>1.4
response ratio
Figure 3-45. Effects of N addition on biological CH4 uptake. The bars show the distribution of the number
of studies categorized by ecosystem type. The dot with error bars shows the overall mean
response ratio with 95% CI.
The mean response ratio of CH4 uptake from laboratory incubation studies was significantly lower
than that from field studies (Figure 3-46). This difference could be due to that the spatially and chemically
heterogeneous field conditions resulted in large experimental errors (Crill et al., 1994; Gulledge and
Schimel, 2000; Weitz et al., 1999). Also laboratory microcosms were characterized by a closed and
incomplete N cycle, where N loss by leaching was very small and no plant competed for N with soil
microbes. Therefore, N addition may result in stronger impacts on microbial processes under laboratory
condition than under field condition
Several laboratory incubation studies found that CH4 uptake rates decreased with increasing N
input (Schnell and King 1994; King and Schnell 1998). This meta-analysis did not find significant
correlation between the amount of N addition and the response ratio of CH4 production/consumption
(Figure 3-46). The lack of dose response relationship is probably because CH4 production is influenced by
multiple interactions of soil N content, soil moisture, pH and temperature (Le Mer and Roger, 2001), and
varies greatly over small spatial and temporal scales (IPCC, 2001).
3-133
-------
coniferous (18)
deciduous (10)
grassland (7)
wetland (6)
tropical forest (2)
heathland (3)
NH4N03 (27)
NH4+ (8)
urea (5)
N03- (5)
field (36)
incubation (11)
0.2 0.4 0.6 0.8 1.0 1.2 1.4
response ratio
Figure 3-46. Effects of N addition on biological CH4 uptake. The data are expressed as the mean response
ratio with 95% confident intervals. The numbers of studies included are indicated in
parentheses.
3.3.4.2. Nitrous Oxide
Biogenic sources are the dominating contributors (>90%) to atmospheric N20. Terrestrial soil is the
largest source of atmospheric N20, accounting for 60% of global emissions (IPCC, 2001). Nitrous oxide
production in soil is mainly governed by microbial nitrification and denitrification (Dalai et al., 2003).
The contribution of each process to the total N20 production varies with environmental conditions.
Denitrifying bacteria reduce N03 orN02 into N20 or N2 under anaerobic condition. In submerged soils
such as wetland soil, denitrification should be the dominant contributing process to N20 emission
(Conrad, 1996). Increasing N03 input generally increases denitrification rate under suitable condition of
temperature and organic C supply. High soil N03 concentrations also inhibit N20 reduction to N2 and
result in high N20/N2 ratio (Dalai et al., 2003). Under aerobic environment, autotrophic nitrifying bacteria
obtain energy by reducing NH4+. Nitrous oxide is an intermediate product of the oxidation of NH44" to
N02 or decomposition ofN02 . The increase in N20 emission following NH44" addition has been
observed in many lab and field experiments (Aerts and Toet, 1997; Aerts and de Caluwe, 1999; Keller
et al., 2005).
vegetation
I • 1
I • 1
I ft
I
1 •
1
1 •
1
1 *
h* 1
N form
1—•—1
1 • 1
1 •
1
I—* 1
experimental condition
1—• 1
1—• 1
t 1 1 1 r
3-134
-------
The U.S. EPA conducted a meta-analysis on the effects of N addition on N20 emissions from non-
agricultural ecosystems, including 99 observations from 30 publications. N addition normally increased
N20 emissions, but some studies also observed N20 emission was decreased by N addition (Ambus and
Robertson, 2006; Ambus et al., 2006; Borken et al., 2002; Curtis et al., 2006; Skiba et al., 1999).
Although some natural ecosystems can be an N20 sink (Chapuis-Lardy et al., 2007), very limited
publications assessed the impact of N addition on N20 uptake. Thus, only changes in N20 production
were estimated in this meta-analysis. Overall, the results of the meta-analysis indicated that N addition
increased N20 emission by 215% (Figure 3-47). The response of N20 emissions was influenced by
ecosystem type, the form and the amount of the N addition (Figure 3-48).
47 -
1.4
Figure 3-47. Effects of N addition on biogenic N2O emission. The bars show the distribution of the number
of studies categorized by vegetation type. The dot with error bars shows the overall mean
response ratio with 95% CI.
Compared to other ecosystems, tropical forest emitted more N20 under N enrichment condition
(+735%) (Figure 3-48). This greater response may be because tropical forests are often P limited rather
than N limited (IPCC, 2001). Hall and Matson (1999) measured N20 emissions after adding N fertilizer in
two tropical rainforests in Hawaii. They found that N20 emissions from P-limited sites was 54 times
greater in the short term N addition experiment and 8 times greater in the chronic N addition experiment
compared to that from N-limited sites. The P-limited soil had higher inorganic N concentration than the
N-limited soil (Hall and Matson, 1999) which increased N availability to the nitrifying and denitrifying
bacteria. However, climatic conditions, especially temperature and precipitation, can also be important
factors driving N20 emissions from tropical forest ecosystem.
N03 caused a higher stimulation (+494%) o N20 emissions than NH/ did (+95%) (Figure 3-48),
which was consistent with the previous field studies (Keller et al., 1988; Russow et al., 2008; Wolf and
Russow, 2000). By adding 15N labeled N03 and NH/to soil, Russow et al. (2008) found that N20 was
mainly emitted by denitrification and the contribution of denitrification to the total N20 production
increased from 54% in soil with normal SOM content to 76% in soil with high SOM.
The Intergovernmental Panel on Climate Change (IPCC) issued a guideline for national GHGs
inventories of biogenic N20 (IPCC, 2000). In this guideline, the default N20 emission factor is 1.25% for
3-135
-------
N fertilizer applied to agricultural fields (i.e., 1.25% of the amount of N applied to a field will be emitted
to the atmosphere as N20). Several studies have questioned the validity of this emission factor. Some
studies suggested a much lower emission factor, such as 0.25% for rice paddy field (Yan et al., 2003) and
0.02% for semi-arid regions (Barton et al., 2008), while others found the amount of N20 emission was not
clearly related to the amount of N addition (Akiyama et al., 2005; Barnard et al., 2005; FAO/IFA, 2001).
The U.S. EPA compiled ambient N20 emission data from 36 studies and did not find any correlation
between N20 emission and the level of N deposition (Figure 3-49). In this meta-analysis, although the
mean response ratio increased with the amount of N addition, the differences among the three levels (<75,
75-150 and >150 kg N/ha/yr) were not significant (Figure 3-48). The weak correlation is probably due to
that the effect of N addition on N20 emission is affected by many other biotic and abiotic factors such as
fertilizer type, vegetation, temperature, soil drainage etc. (Dalai et al., 2003).
coniferous (34)
deciduous (17)
grassland (22)
wetland (19)
tropical forest (11
heathland (3)
NH4N03 (47)
NH4+ (19)
N03- (21)
urea (9)
<75 (30)
75-150 (19)
>150 (24)
I—•-
H
I—* 1
I—• 1
I-
I—• 1
I—• 1
^
vegetation
N form
N addition level
—/ /-H
1 1 1—i 1 1 1 1 1—i 1 1 1 1—r
012345678 9 10 11 12 13 14 15 16 60
response ratio
Figure 3-48. Effects of N addition on biogenic N20 emission. The data are expressed as the mean
response ratio with 95% confident intervals. The numbers of studies included are indicated in
parentheses.
3-136
-------
0
deciduous forest o
•
coniferous forest
A
tropical forest
T
heathland
0
0 CO 1
0' '>o * •• •
•
0 10 20 30 40 50 60
N deposition (kg N ha"1 yr"1)
Figure 3-49. The relationship between N2O emission and N deposition. The data were compiled from 36
independent studies.
3.3.4.3. Nitric Oxide and VOCs
N saturation caused by chronic N input often leads to increased soil NO emission (Hall and Matson
1999; Venterea et al., 2003; Venterea et al., 2004; Kitzler et al., 2006). Although nitrification and
denitrification both contribute to NO production, many studies have found that NO flux increased in a
manner that was consistent with nitrification rate increases, suggesting that NO emission is more
triggered by nitrification activity rather than by denitrification activity (Wolf and Russow 2000; Venterea
et al., 2004).
Monthly NO monitoring at Harvard Forest showed that NO emission increased at coniferous plots
treated with 50 and 150 kg N/ha/yr and hardwood plots treated with 150 kg N/ha/yr (Venterea et al.,
2003). NO emission accounted for 3.0-3.7% of N inputs to the high N plots and 8.3% of inputs to the
coniferous low N plots. At two paired watersheds subjected to elevated N input, 35.5 kg N/ha/yr at
Fernow Experimental Forest in Western Virginia and 25.2 kg N/ha/yr at the Bear Brook Watershed in
Maine, NO emissions at the N amended watersheds (0.61-6.8 |ig NO-N /m2/h ) were higher than the
reference watersheds (0.21-1.4 |ig NO-N /m2/h). Mean NO fluxes at both watersheds were positively
correlated with mean soil N03 concentrations (Venterea et al., 2004). Similarly, (Pilegaard et al., 2006)
found that NO emission from 7 coniferous forests in Europe were highly correlated with the rates of N-
deposition (NO flux (fig N/m2/h) = 25.52 xN deposition (g N/m2/yr) -13.93; r2=0.82).
The response of NO production rates to N deposition may vary significantly between vegetation
types and climatic regimes. The measurements of 15 forest sites across Europe showed that coniferous
forest emitted more NO (31.0±15.6 |ig NO-N/m2/h) than deciduous forest (4.1±1.9|ig NO-N/m2/h),
although the average N deposition to coniferous forests (16 ± 5 kg N/ha/yr) was similar to deciduous
forests (15±2 kg N/ha/yr) (Pilegaard et al., 2006). Several studies, such as ButterbachBahl et al. (1997)
and Venterea et al. (2003), also found that NO emissions from coniferous forest were generally higher
than from deciduous forest. This difference is probably due to that the characteristics of coniferous forest
floor, such as low soil moisture and low pH, favour nitrification and thus NO emission (Pilegaard et al.,
2006).
3-137
-------
In addition to vegetation type, climate and soil nutrient conditions also influence NO flux under N
enrichment. Natural forests in tropic region are generally considered P limited, contrasting with N-limited
temperate forests (Vitousek et al., 1997). After adding N (50 kg N/ha/yr) to three tropical rainforests (N
limited, NP-limited and P limited) in Hawaii, Hall and Matson (2003) found that all N addition plots
emitting more NO than control plots and responses of NO production to N addition increased in the order
of: P limited sites >NP-limited sites >N limited sites. Hall and Matson (2003) work suggested that
P-limited and N-limited ecosystems responded differently to N deposition. Rather than retaining N in
ecosystems, anthropogenic N addition may immediately increase N trace gas loss if N availability
exceeds biological demand.
3.3.4.4. Volatile Organic Compounds (VOCs)
Volatile organic compounds (VOCs) comprise a wide range of chemical compounds including
hydrocarbons, halocarbons and oxygenates. Methane is one of the most important VOCs and the effect of
N addition on CH4 emission was discussed in Section 3.3.4.1. The available information for non-methane
VOCs emissions is much sparser than that for CH4, with only a few studies measuring isoprene emissions
under N enrichment. Isoprene is the primary non-methane hydrocarbon emitted from temperate deciduous
forest and tropical forest (Funk et al., 2006). Both pot and field studies found that there was a strong
positive correlation between isoprene emission rates and leaf N concentrations (Litvak et al., 1996; Funk
et al., 2006). N addition could significantly increase plant isoprene emissions by increasing leaf N
concentrations (Harley et al., 1994; Litvak et al., 1996).
3.3.4.5. Summary of N Effects on Biogenic Trace Gases
Integrating 160 observations across 57 independent studies, this meta-analysis suggested that N
addition tended to increase CH4 emission, reduce CH4 uptake and increase N20 emission. Overall, N
deposition may result in higher CH4 and N20 concentrations in atmosphere and exacerbate global
warming, but these responses can also be influenced by many environmental factors, such as vegetation
type, N form, and climate.
The evidence is sufficient to infer a casual relationship between N deposition and the alteration of
biogeochemical flux of CH4 in terrestrial and wetland ecosystems. N addition ranging from 10 to 560 kg
N/ha/yr reduced CH4 uptake by 38% across all ecosystems (Figure 3-45), but this inhibition was
significant only for coniferous and deciduous forest, with a reduction of 28% and 45%, respectively
(Figure 3-46). Wetlands are generally net sources of CH4, but some wetlands can be net sinks depending
on environmental conditions such as drainage and vegetation (Crill et al., 1994; Saarnio et al., 2003). The
meta-analysis indicated that N addition, ranging from 30 to 240 kg N/ha/yr, increased CH4 production by
115% from the source wetlands (Figure 3-44), but had no significant effect on CH4 uptake of the sink
wetlands (Figure 3-46).
The evidence is sufficient to infer a casual relationship between N deposition and the alteration of
biogeochemical flux of N20 in terrestrial and wetland ecosystems. Overall, the results of the meta-analysis
discussed in Section 3.3.4 indicated that N addition ranging from 10 to 560 kg N/ha/yr. increased N20
emission by 215% in terrestrial ecosystems (Figure 3-47). The response of N20 emission to N for
coniferous forest, deciduous forest and grasslands was significant (Figure 3-48). In the meta-analysis of
19 observations from studies that evaluated the effects of N additions ranging from 15.4 to 300 N kg
N/ha/yrwas shown to increase the production of N20 by 207% in wetlands (Figure 3-48).
3-138
-------
3.3.5. Species Composition, Species Richness and Biodiversity
A common response to environmental stress is the tendency for the more sensitive species to
decrease in abundance, or to be eliminated, while the more tolerant species increase in abundance
(Woodwell, 1970). Species composition and species richness, as well as impacts on rare or threatened
species, indicate changes to biodiversity. The ecological consequences of changing species composition,
richness and/or biodiversity can be profound. Selective removal of certain species can result in an
impairment of ecosystem function, change in community structure and food web dynamics, and decrease
in species richness and diversity. Such changes in species composition can occur in response to N
addition to terrestrial, aquatic, and transitional ecosystems.
Weiss et al. (2006) presented an overview of potential biodiversity loss from N enrichment. A
survey by Stohlgren et al. (1999) of variables that contribute to species richness and invisibility of sites
found negative relationships between soil N, and species richness and numbers of nonnative plant species.
The implication of this work is that N fertilization alters competitive interactions that may cause native
species to be lost, with subsequent decrease in species richness.
Alteration of plant productivity and growth by N deposition (see Section 3.3.3) causes a cascading
effect on the competitive interactions among species. Atmospheric deposition of N is expected to benefit
those species that are best able to take advantage of the increased nutrient availability. Other species may
experience decreased growth, reproduction and population size, because they are out-competed by species
that are more successful under conditions of enhanced N availability. Numerous studies evaluate
ecosystem response to levels of N addition that far exceed the range of N deposition levels in the U.S.
This assessment focuses on the information most relevant to the review of the NAAQS, therefore research
conducted at N loading levels that greatly exceed current conditions ( >150 kg N/ha/yr) are excluded from
the discussion.
3.3.5.1. Terrestrial Ecosystem Biodiversity
Atmospheric inputs ofN can alleviate deficiencies and increase growth of some plants at the
expense of others. Thus, N deposition can alter competitive relationships among terrestrial plant species
and therefore alter species composition and diversity (Ellenberg, 1987; Kenk and Fischer, 1988;
U.S. EPA, 1993a). Wholesale shifts in species composition are easier to detect in short-lived ecosystems
such as annual grasslands, in the forest understory, or mycorrhizal associations than for long-lived forest
trees where changes are evident on a decadal, or longer, time scale. Note species shifts and ecosystem
changes can occur even if the ecosystem does not exhibit signs of N saturation.
Forests
Trees
There is very little information on the effect of N deposition on the biodiversity of overstory trees
in the U.S. However, the altered growth rates caused by N enrichment have the potential to affect forest
structure and biodiversity. The life span of many trees is 100 years or more, therefore observation of how
growth rates effect biodiversity within established forests are difficult to observe on a decadal time scale.
N deposition has been observed to cause tree invasion into grasslands, also called forest
encroachment. A study of the northern edge of the Great Plains (southern Canada), showed that increasing
N deposition over a range of 8 to 22 kg N/ha/yr to aspen-dominated forest and boreal forest caused an
increase forest expansion into the grasslands (Kochy and Wilson, 2001). The following mechanisms have
been document to facilitate forest encroachment. Due to their height, trees and shrubs can intercept more
airborne particulate N than grasses and they should therefore benefit most from N deposition (Kellman
and Carty, 1986; Binkley, 1995). Fertilization also increases the water-use efficiency of woody invaders
3-139
-------
(Bert et al., 1997) and this may enable them colonize temperate grasslands on dry, coarsely textured soils.
Accelerated N cycling following deposition (Berendse, 1994; Carreiro et al., 2000; Hogbom and
Hogberg, 1991) decreases competition for N and increases competition for light (Wilson and Tillman,
1991), and may give further advantage to tall or fast growing trees (Aerts et al., 1999). Overall, therefore,
increased deposition rates may result in a self-maintaining positive feedback that allows trees to establish
in grasslands (Wilson et al., 1998).
Understory Herbaceous Plants and Shrubs
Studies in Europe have generally been based on natural gradients, whereas findings in the U.S.
have mostly been based on experimental N addition. The effects of increasing N deposition on herbaceous
plants were reviewed by (Gilliam et al., 2006b). Reported effects include species shifts towards
nitrophilous and more acid-tolerant plant species along a deposition gradient from 6 to 20 kg N/ha/yr in
Swedish oak forests; a decline in abundance and cover of ericaceous shrubs along a deposition gradient
from <3 to >12 kg N/ha/yr in the boreal forest in Sweden; and a decline in herbaceous cover under
hardwoods following 3 years of N additions applied as (NFL^SC^ at rates ranging from 14 to 28 kg
N/ha/yr. The decline in herbaceous cover in the latter study was attributed to increased shading by ferns,
and the effect was more pronounced at sites that experienced lower ambient atmospheric N inputs.
Van Breemen and Van Dijk (1988) noted that over the previous several decades of N deposition the
composition of plants in forest herb layers in The Netherlands had shifted toward species commonly
found in N-rich areas. Brunet et al. (1998) and Falkengren-Grerup (1998) reported the effects of excessive
N deposition on mixed-oak forest vegetation along a depositional gradient. Results of this study suggest
that N deposition had affected non-woody vegetation directly by increased N availability and, indirectly,
by accelerating soil acidity. Time series studies indicated that 20 of the 30 non-woody plant species that
were associated most closely with high N deposition had increased in abundance in those areas in Europe
that received high N deposition.
Mixed results have been reported in other studies. Research at Fernow Experimental Forest, West
Virginia, indicated that application of 35 kg N/ha/yr applied as (NFL^SC^ for 6 years had no significant
impact on the herbaceous layer in an Appalachian hardwood forest (Gilliam et al., 2006a). Fernow has
been the recipient of high levels of N deposition for decades, raising the possibility that the herbaceous
layer responded long ago to changes in N availability.
Mycorrhizal and Microbial Diversity
Mycorrhizal and microbial biodiversity can also be affected by N enrichment. Relationships among
plant roots, mycorrhizal fungi, and microbes are critical for N cycling and for the growth and health of
plants. Mycorrhizal fungal diversity has been shown to be associated with above-ground plant
biodiversity and ecosystem productivity (Wall and Moore, 1999) and to be adversely affected by
increased N availability (Egerton-Warburton and Allen, 2000). The loss of mycorrhizal function has been
hypothesized as a key process contributing to reduced N uptake by vegetation and increased N03
mobility from soil into drainage water under conditions of high N supply (U.S. EPA, 2004).
Progressive decline in ectomycorrhizal fungal species richness in Alaskan coniferous forest (white
spruce [Picea glauca\ dominant) occurred along a local N deposition gradient, from 1 to 20 kg N/ha/yr,
downwind from an industrial complex (Lilleskov et al., 2002). Ectomycorrhizal fungal communities are
important in tree nutrition, and ectomycorrhizal fungal trees tend to be dominant in N-limited forest
ecosystems.
N fertilization at rates of 54 and 170 kg N/ha/yr (as NH4NO3) led to a decline in ectomycorrhizal
fungal diversity and species composition in an oak savanna at Cedar Creek Natural History Area in
Minnesota (Avis et al., 2003). In the reference plots, five species collectively accounted for more than
40% cover versus four plant species in the lower N addition plots. In the higher N addition plots, a single
plant species accounted for more than 40% cover.
3-140
-------
Compton et al. (2004) investigated the effects of 11 years of experimental N addition on forest soil
microbial ecology. Experimental N addition decreased the C content of microbial biomass in the O
horizon of both experimental forest stands, based on chloroform fumigation-extraction. In addition, the
use of N-containing substrates by microbes appeared to be reduced by N addition in the pine stand, but
not in the hardwood stand. In addition, the use of N-containing substrates by microbes appeared to be
reduced by N addition in the pine stand, but not in the hardwood stand. The mechanisms responsible for
such changes are not clear (Arnebrandt et al., 1990; Compton et al., 2004). It is possible that added N has
both direct (nutrient) and indirect (soil chemistry, organic matter quality, and quantity) effects on
microbial ecology. Effects can be manifested on mycorrhizal fruiting body abundance, hyphal networks,
and community composition (Frey et al., 2004; Lilleskov et al., 2002).
Grasslands
Reduced biodiversity in response to N deposition is reported for grasslands in the U.S. and Europe.
Clark and Tilman (2008) recently evaluated the effects of chronic N deposition over 23 years in
Minnesota prairie-like successional grasslands and in a native savanna grassland, each originally
dominated by a species-rich mixture of native C4 grasses and forbs (Cedar Creek Long Term Ecological
Research Site, Minnesota). Experimental N addition ranged from 10 to 95 kg N/ha/yr above ambient
atmospheric N deposition (6 kg N/ha/yr). The N addition rate reduced plant species numbers by 17%
relative to controls receiving ambient N deposition. Moreover, species numbers were reduced more per
unit of added N at lower addition rates and relative species number was reduced at all addition levels.
This suggests that chronic but low-level N deposition may decrease diversity below the lowest addition
levels tested (the critical load was calculated as 5.3 kg N/ha/yr with an inverse prediction interval of 1.3-
9.8 kg N/ha/yr). A second experiment showed that a decade after cessation of N addition, relative plant
species number, although not species abundances, had recovered, demonstrating that some effects of N
addition are reversible.
Change in species composition in response to N deposition has been observed regardless of soil
type in European grasslands. Such effects have been found in calcareous, neutral, and acidic
environments, species-rich heaths, and montane-subalpine grasslands (Bobbink et al., 1992b; 1998;
Bobbink, 1998; Stevens et al., 2004). A transect of 68 acid grasslands across Great Britain, covering the
lower range of ambient annual N deposition (5 to 35 kg N/ha/yr), indicates that long-term, chronic N
deposition significantly reduced plant species richness. Species richness declined as a linear function of
the rate of inorganic N deposition, with a reduction of one species per 4-m2 quadrant for every
2.5 kg N/ha/yr of chronic N deposition.
Grasslands are well known to respond to increased N availability through changes in growth rates
of both native and exotic species. Under high N supply, exotic grasses often out-compete other species,
and cause changes in plant community composition (Lowe et al., 2002). A summary of studies, which
have shown altered plant community composition, or growth rates of plant species that have implications
for community composition is in Table 3-17. Increased availability of N to grasses can also affect
herbivores that feed on grasses by altering food quality, quantity, and phenology, and perhaps by changing
the relationships between herbivores and their predators (Throop and Lerdau, 2004).
3-141
-------
Table 3-17.
Summary of N effects on grassland biodiversity.
Region/Country Endpoint
San Francisco Bay Area, Bay
CA Checkerspot
Populations
Colorado Growth
Plant Research Growth and
Laboratory, University of Tissue Quality
Illinois at Chicago
Jasper Ridge Biological Species
Preserve in California Richness and
Diversity
Lund University, Lund, Growth
Southern Sweden
Observations
Observation: Serpentinitic soils sustain native
grasses that support populations of the
endangered Bay Checkerspot butterfly. Several
lines of evidence indicate that dry N deposition is
responsible for grass invasion and subsequent
decline of the butterfly population. However, this
relationship is uncertain.
Greenhouse experiment: they tested the
response of grassland species to increased N
availability (0,10,40, 70, or 100 kg N/ha/yr) over
75 days. All of the grass species exhibited
increased growth with increased N availability.
Native species did not consistently grow better at
low N levels than the exotic species. Two of the
exotic grasses exhibited the greatest increase in
growth, while another of the exotics exhibited the
smallest increase in growth.
Field Addition: N additions in (0.1,1,3 mmol N
addition for 80 days), caused species-specific
growth and plant tissue quality changes. C3
grasses (Elymus virginicus L, E. Canadensis L)
showed a greater positive growth response to N
additions than C4 grasses (Andropogon geradii
Vitmanm, Schizachyrium scoparium Michx.) and
forbs (Solidago nemoralis Ait., S. rigida L).
Species with smaller initial biomass exhibited the
greatest increase in biomass, with a sevenfold to
eightfold increase in S. nemoralis and £.
canadensis and only a threefold increase in S.
rigida.
Grassland Type/ Species Reference
Serpentine Grasslands Weiss (1999)
Bay Checkerspot Butterfly
(.Euphydryas Editha Bayensis)
Two North American Native Lowe et al.
Species (Blue Grama And (2002)
Western Wheatgrass) and Four
Exotic Species (Cheatgrass,
Leafy Spurge, Canada Thistle,
and Russian Knapweed)
Tallgrass Prairie Lane and
BassiriRad
(2002)
Nine Annual Species: Avena Zavaleta
Barbata, Bromus Hordeaceus, et al. (2003)
Loiium Muitifiorum, Avena
Fatua, Bromus Diandrus,
Anagaiiis Arvensis, Geranium
Dissectum Erodium Botrys,
Vicia Sativa, and one Biennial
Species, Crepis Vesicaria
15 Herb and 13 Grass Species Falkengren-
Grerup
(1998)
Field Addition: 70 kg N/ha/yr over three years led
to decreased to a decline in total species richness,
species diversity decreased by 5% and all three
N-fixing forbs disappeared.
Greenhouse experiment: Sand-solution
experiments studying how growth was effected by
N concentrations of 50,250 and 1250 |jM in a
simulated acid forest soil solution, similar to
naturally occurring soil solutions is Southern
Sweden. 46% of grasses displayed a significantly
greater biomass at 250 than at 50 |jM N as
compared with only 7% for the herbs. Some
species attained their highest biomass at 1250 |jM
N and others at 50 |jM N. Grasses grew better
than herbs in response to experimental addition of
N. At the highest experimental N deposition rates,
growth was limited for most species by the supply
of nutrients other than N.
3-142
-------
Region/Country Endpoint
Observations
Grassland Type/ Species Reference
Minnesota, Cedar Creek Species Field Addition: N enrichment over a 12 year
Natural History Area Composition period on 162 plots using a N addition gradient
from 0 to 30 g N/m2/yr. Plots initially dominated by
native warm-season grasses shifted to low-
diversity mixes of species dominated by cool-
season grasses at all but the lowest rates of N
addition. Grasslands with high N retention and C
storage rates were the most vulnerable to loss of
species and major shifts in N cycling in response
to experimental N enrichment.
San Francisco Bay Area,
CA
Species
Composition
Observational: N deposition levels of 10 to 15 kg
N/ha/yr, exotic nitrophilous grasses have been
reported to have displaced native grass species,
likely due to greater N availability from deposition
and from the cessation of grazing, which
previously exported N out of the system
Grasslands
Fenn et al.
(2003a)
Jasper Ridge Biological
Preserve in California
Npp
Field Addition: 70 kg N/ha/yr applied as
(CaN03)2 increased NPP by 30%
Grasslands
Shawet al.
(2003)
Jasper Ridge Biological
Preserve in California
Herbivory, Leaf
Tissue N And
Growth Rates
Field Addition: 70 kg N/ha/yr applied as
(CaN03)2 caused altered herbivory by gastropod
which differed by species (J,, f and no change), 5
out of 6 species had increased leaf tissue, 5 out of
6 species had increased growth rates
Grasslands
Cleland et al.
(2006)
Michigan old field
Biomass
Field Addition: 120 kg N/ha/yr applied as
NH4NO3 pellets had a significant positive growth
effect on annual dicot biomass but no significant
growth effect on annual grass biomass.
Grasslands
Huberty
etal. (1998)
Field Additions and Greenhouse experiment: N AmbrosiaMemisiifolia Throop
addition to common ragweed (Ambrosia (Common Ragweed) and two of (2005)
artemisifolia) led to increased vegetative and seed its Insect Herbivores; a Leaf
biomass and decreased root: shoot ratios. N Beetle, Ophraella Communa
deposition may indirectly affect biomass Lesage {Coieoptera:
production and allocation through affecting insect Chrysomelidae), and an Aphid,
herbivory. The particularly strong influence of both Uroleucon Tuataiae Olive
herbivory and N deposition on A. artemisiifolia (Hemiptera: Aphididae)
reproduction suggests potential population and
community-level consequences.
Not all studies have shown an effect of N addition on species richness or diversity. In old
agricultural fields in Michigan, increased N deposition changed neither the successional timing nor the
gain or loss of species numbers (Huberty et al., 1998). A lack of response in species richness may have
been due to application of mid-growing season fertilization in the experimental design. Huberty and
colleagues (1998) suggested that N additions may change the dominance structure instead of the species
composition, of these successional old-field communities. Other studies in Michigan on successional
grasslands showed no response to N application of 10 kg N/m2/yr, equivalent to about 2.5 times current
deposition rates (Ambus and Robertson, 2006).
Arid and Semi-Arid Land Ecosystems
Some arid and semi-arid ecosystems in the southwestern U.S. are considered sensitive to N
enrichment effects and receive high levels of atmospheric N deposition. However, water is generally more
Three N-Limited Minnesota Wedin and
Grasslands with Varying Tilman
Successional Age, Species (1996)
Composition, and Total Soil C
Brookhaven Nat. Lab.,
Long Island, New York
and the Sciences
Greenhouse Facility,
SUNY, Stony Brook, NY
Vegetative and
Seed Biomass
and Decreased
Root: Shoot
Ratios
3-143
-------
limiting than N in these systems. Nevertheless, enhanced N may play a role in the observed invasion of
some exotic plant species and associated changes in ecosystem function, especially where water supply is
adequate.
In semi-arid ecosystems, results from several N fertilization experiments showed increased
biomass of nonnative plant species over native species; decreased soil moisture under some conditions;
and increased fire risk where dense mats of grasses replaced shrub cover (See Table 3-18).
Much of the arid land data are from the coastal sage scrub (CSS) communities of southern
California, down-wind of the Los Angeles Basin, where dry N deposition is very high. The CSS
community in California has been declining in land area and in shrub density over about the past 60 years
and is being replaced in many areas by Mediterranean annual grasses (Padgett and Allen, 1999; Padgett
et al., 1999). N deposition is considered a possible cause or contributor to this ecosystem alteration. More
than 30 kg N/ha/yr of atmospheric N is deposited to portions of the Los Angeles Air Basin (Bytnerowicz
and Fenn, 1996). The CSS community is of particular interest because about 200 sensitive plant species
and several federally listed threatened or endangered animal species are found in the area.
Native shrub and forb seedlings in the CSS community are unable to compete with dense stands of
exotic grasses, and thus are gradually replaced by the grasses, especially following disturbances such as
fire (Cione et al., 2002; Eliason and Allen, 1997; Yoshida and Allen, 2001). Biodiversity impacts have
also been documented for microbial communities in coastal sage scrub ecosystems. It has been
hypothesized that the decline in coastal sage shrub species could be linked to the decline of the arbuscular
mycorrhizal community (Egerton-Warburton and Allen, 2000).
Table 3-18. Summary of N effects on arid and semi-arid ecosystems.
Region/Country Endpoint
Observations
Ecosystem Type/
species
Reference
Southeast Idaho Species com- Field Addition: 6 or 12 kg N/ha/yr applied as NH4NO3 for 6
position and years (in addition to ambient inputs of 1.3 to 1.4 kg N/ha/yr)
cover resulted in a decrease in soil moisture caused by shifts in
plants species composition and cover. However, there were
no effects on perennial grass cover in response to experi-
mental N additions
Southern California Plant com-
munity
Observation: Dry N deposition is above 30 kg N/ha/yr in
some places. Native shrub and forb seedlings in this plant
community are unable to compete with dense stands of
exotic grasses, and thus are gradually replaced by the
grasses, especially following disturbances such as fire
Sagebrush (Artemisia
tridentata) steppe eco-
system
Coastal sage scrub
Inouye (2006)
Eliason and
Allen, (1997);
Yoshida and
Allen (2001);
Cione et al.
(2002)
Southern California Plant com- Greenhouse and deposition gradient experiments: N
munity deposition is considered a possible cause or contributor to
declining shrub density over about the past 60 years and is
being replaced in many areas by Mediterranean annual
grasses
Coastal sage scrub
Allen et al.
(1998); Padgett
and Allen (1999);
Padgett et al.
(1999)
3-144
-------
Region/Country Endpoint
Observations
Ecosystem Type/
species
Reference
Coastal sage scrub
Padgett and Allen
(1999)
Southern California Response of Greenhouse Experiment: N (as 5.4 g/L NH4CI or 10g/L
native and KNO3) was added to obtain soil N concentrations of 2, 20,
nonnative 40, and 80 |jg/g. The grasses demonstrated a 1.5- to 2.5-
plants. fold growth increase when soil N levels increased from 20 to
40 |jg/g. To achieve a comparable growth increase, shrubs
required higher soil N levels (between 20 and 80 |jg/g).
These lab experiments agree with observations in the field,
where exotic grasses, especially once established, have
replaced native shrubs under elevated N deposition.
Southern California Plant root-to- Greenhouse Experiment: Changes in plant root-to-shoot Coastal sage scrub Padgett and Allen
shoot growth growth ratios were observed in the plant community, which is (1999); Padgett
ratios composed largely of the drought-resistant deciduous shrubs et al. (1999)
Artemisia californica, Encelia farinosa, and Eriogonum fas-
ciculatum.
Southern California Response of Observation: More than 30 kg N/ha/yr of atmospheric N is Coastal sage scrub Bytnerowicz and
native and deposited to this ecosystem in portions of the Los Angeles Fenn (1996)
nonnative Air Basin). Decreases in the diversity of native plants paral-
plants. leled increases in exotic grass biomass.
Colorado Plateau
Response of
Field Addition: For 2 years, plots were treated with 0,10, Arid grassland
Schwinning etal.
native and
20, or 40 kg N/ha/yr as a KNO3 solution. Galleta (Hilaria
(2005)
nonnative
jamesii] and Indian ricegrass (Oryzopis hymenoides) showed
plants.
no increase in leaf photosynthesis or tiller size, but ricegrass
showed a 50% increase in tiller density in the second year at
the 20 and 40 kg N/ha/yr application levels. For both spe-
cies, the increased N application hastened the onset of water
stress. Unexpectedly, a non-native species, Russian thistle
(Salsola iberica) showed a rapid growth response to the
highest fertilization rate in the first summer, when rainfall was
above average. The authors suggested that the timing and
amount of N deposition could facilitate noxious weed inva-
sion and thus change community composition.
Joshua Tree Na-
Response of
Deposition Gradient: 18 locations, chosen to cover the Arid grassland
Allen et al. (2007)
tional Park, Califor-
native and
dominant vegetation types (Creosote Bush Scrub, Joshua
nia
nonnative
plants.
Tree Woodland, Pinyon Juniper Woodland), were sampled
for atmospheric concentrations of NO, NO2, NH3, and soil
[N], The relationship between reactive atmospheric N con-
centrations and soil N were consistent in most sites.
Observations along the N gradient did not reveal a clear rela-
tionship between non-native grass cover and soil N
concentration up to 20 |jg/g.
Joshua Tree Na-
Response of
Field Addition: Was applied at levels of 5 and 30 kg N/ha/yr Arid grassland
Allen et al. (2007)
tional Park, Califor-
native and
at four sites over a 2 year period. Low-elevation sites were
nia
nonnative
plants.
dominated by creosote bush scrub and higher-elevation sites
by pinyon-juniper woodland. Non-native grass biomass in-
creased significantly at three of four treatment sites that
received 30 kg N/ha/yr, but not at the sites that received
5 kg N/ha/yr. A soil N concentration of 23 |jg/g was conserva-
tively considered the low threshold for significant plant N
response based on this fertilization study.
3-145
-------
Region/Country Endpoint
Observations
Ecosystem Type/
species
Reference
Western U.S. Fire cycle Observation: Vegetative changes stimulated by nutrient
enrichment from N deposition may affect the frequency and
severity of subsequent disturbance. Several lines of evi-
dence suggest that N deposition may be contributing to
greater fuel loads, thus altering the fire cycle in a variety of
ecosystem types. Invasive grasses, which can be favored by
high N deposition, promote a rapid fire cycle in many loca-
tions.
San Francisco Bay Response of Observation: N deposition levels of 10 to 15 kg N/ha/yr,
area native and exotic nitrophilous grasses have displaced native grass
nonnative species, likely due to greater N availability from deposition
plants and from the cessation of grazing, which previousy exported
N out of the system. Since this change in species composi-
tion, populations of the rare and threatened bay checkerspot
butterfly (Euphydryas editha bayensis) have declined greatly.
It has been hypothesized that the response of the butterfly
has been due to the vegetative changes.
Arid grassland
Fenn et al.
(2003a)
Grasslands dominated
by exotic annuals such
as wild oat (Avena
fatua), brome (Bromus
mollis), and ryegrass
(.Lolium multiflorum).
Fenn et al.
(2003a)
Mycorrhizal and Microbial Diversity
It has been hypothesized that the decline in coastal sage shrub species in California could be linked
to the decline of the arbuscular mycorrhizal community (Egerton-Warburton and Allen, 2000). The au-
thors discerned a shift in arbuscular mycorrhizal community composition with decreased species richness
and diversity along a deposition gradient (2 to 57 (.ig N/g as soil N03 ). These shifts in mycorrhizal fungal
communities may facilitate replacement of native plant communities by Mediterranean annual grasslands.
Larger-spored fungal species (Scutellospora and Gigaspora) have decreased in number due to a failure to
sporulate, with a concomitant proliferation of small-spored species. This pattern suggests selective pres-
sure favoring the smaller spored species of fungi (Egerton-Warburton and Allen, 2000), and that N enrich-
ment of the soil might alter the arbuscular mycorrhizal species composition and diversity.
Desert Ecosystems
Some desert ecosystems in the southwestern U.S. are considered sensitive to N enrichment effects
and receive high levels of atmospheric N deposition. However, water is generally more limiting than N in
these systems. Nevertheless, N deposition can stimulate plant growth and cause the observed invasion of
some exotic plant species and associated changes in ecosystem function, especially where water supply is
adequate. Most evidence is from field additions of N, levels ranging from 10-100 kg N/ha/yr (Table 3-
19).
Fertilization experiments in the Mojave Desert showed that increased levels of N deposition could
favor the establishment of nonnative species where the non-natives are already prevalent (Brooks, 2003).
There is also evidence that N deposition decreases the growth of desert legumes (Baez et al., 2007). A link
between N deposition and decrease in legumes has been found across other North American sites (Suding
et al., 2005). The effect on legumes may be attributable because legumes, which are N fixers, often com-
pete better under low N supply.
There is evidence from the desert ecosystems that N accumulates during periods of drought, and
that more N is immobilized during periods of high precipitation (Stursova et al., 2006). Thus, where water
and N appear to be co-limiting factors, the observed pattern of higher rates of N deposition during months
with higher precipitation may result in a stronger fertilization effect than if N deposition were independ-
3-146
-------
ent of precipitation (Hooper and Johnson, 1999; Asner et al., 2001; Knapp and Smith, 2001; McLain and
Martens, 2006).
Table 3-19. Summary of N effects on desert ecosystems.
Region/Country
Endpoint
Observations
Grassland
type/ spe-
cies
Reference
Chihuahuan Desert
Growth
response of
native
species
Field Addition: 20 kg N/ha/yr addition in one season showed
blue gramma (Bouteloua gracilis) was favored over black
gramma (Bouteloua eriopoda), the current dominant species
blue
gramma and
black
gramma
Baezet al.
(2007)
Jornada Basin, New
Mexico
Growth
response of
native
species
Field Addition: Black grama and another dominate species,
creosote bush (Larrea tridentate) did not significantly increase
biomass after experimental additions of 25 kg N/ha/yr, but did
after additions of 100 kg N/ha/yr
black grama
and creo-
sote bush
Ettershank
etal. (1978);
Fisher et al.,
(1998)
Chihuahuan Desert,
Growth
response of
native
species
Field Addition: Additions of 100 kg N/ha/yr over about a
decade, resulted in percent soil N that was 15-61% higher,
extractable NO3- that was 25-175% higher, and extractable NH4
that was 247-1721 % higher compared to control plots (Stursova
et al., 2006). The resultant biologic effects were a 30% increase
in cover of warm season grasses and a 52% reduction in cover
of legumes (Baez et al., 2007).
Grasses
and
legumes
Stursova
etal., (2006);
Baezet al.
(2007)
Mojave Desert
Native vs.
non-native
Field Addition: At application rates of 32 kg N/ha/yr over 2
years, both density and biomass of non-native plants increased
(54% increased biomass), while native species biomass
declined by about 39%. Plant responses were influenced by
rainfall events rather than by average annual rainfall, with the
annual plants thriving in a year when high rainfall events
triggered germination.
Grasses
Brooks
(2003)
Great Basin Desert,
sites near Mono
Lake, CA
Growth and
seed
viability
Field Addition: Sarcobatus vermiculatus, a desert shrub found
demonstrated a twofold to threefold increase in stem growth, a
2.5 to 4 fold increase in viable seed production, and a 17% to
35% increase in leaf N with N additions. N was applied in March
and November as NH4NO3, at a cumulative addition rate of
233.6 g N per plant.
Sarcobatus
vermiculatu
Drenovsky
and Richards
(2005)
Mojave desert
Growth
response
Field Addition: The shrub Larrea fridentata showed no
increased growth response to N additions (at 10 and 40 kg
N/ha/yr CaNOs) but did respond to increased water
Larrea
tridentata
Barker et al.
(2006)
Mojave and
Sonoran deserts.
Native vs.
non-native
Observation: Invasive annuals showed a greater response to
elevated N than native species, and have recently invaded.
Though their invasion is correlated with greater N deposition, no
causation has been established.
Fenn et al.
(2003a)
Lichens
Lichens are frequently used as indicators of air pollution and atmospheric deposition levels (see
Annexes A and C for an additional discussion). In addition to being good subjects for biomonitoring, they
3-147
-------
constitute important components of the forest ecosystem by contributing to biodiversity, regulating
nutrient and hydrological cycles, and providing habitat elements for wildlife (McCune and Geiser, 1997).
Little is known about the mechanisms that control growth and resource partitioning in lichens,
which are complex symbiotic systems comprised of a fungus (mycobiont) and a green alga and/or
cyanobacterium (photobionts) (Palmqvist, 2000; Sundberg et al., 2001). Organic compounds required for
growth are produced via photobiont photosynthesis. Production is strongly coupled with N utilization.
The non-photosynthetic fungus comprises much of the lichen biomass and requires N for protein
synthesis, nucleic acids and fungal cell wall (chitin) synthesis (Palmqvist et al., 1998). Carbon and N
uptake must be balanced for coordinated development of lichen thalli (Sundberg et al., 2001).
Lichens can be classified on the basis of their response to atmospheric pollution. Nitrophytic
lichens occur in areas that receive high atmospheric N deposition; acidophytic lichens are prevalent in
areas that receive low N input (Rouss 1999)(Gaio-01iveira et al., 2005; van Herk, 2001). Lichens differ
with respect to N requirements. Many lichens that have a cyanobacterial photobiont are N-fixing, whereas
those with a green algal photobiont are dependent on atmospheric deposition for their N supply.
N-fixing lichen species are particularly affected by N deposition (Dahlman et al., 2002).
Cyanobacteria have been shown to grow on either N03 or NH44" sources when administered at non-toxic
concentrations. More rapid growth was observed with NH/ fertilization as compared with N03
fertilization (Ruckert and Giani, 2004). Ammonium is more easily assimilated; both N03 and nitrite must
first be reduced to NH/ before assimilation (Ruckert and Giani, 2004).
Lichens with a green algal photobiont are solely dependent on atmospheric deposition as a source
of N. However, a buildup of N within the thallus can lead to toxicity. Lichens exhibit varying degrees of
sensitivity to increasing N deposition, owing to diverse mechanisms of responding to high N supply by
reducing N uptake or assimilating N into non-toxic forms such as arginine (Dahlman et al., 2003; Gaio-
Oliveira et al., 2005).
Lichens that contain a cyanobacterial photobiont appear to be more sensitive to adverse effects
from atmospheric N deposition than most other lichens (Hallingback and Kellner, 1992; Hallingback,
1991). In Sweden, the proportion of cyanobacterial lichens that has disappeared or is threatened is three
times as large as the corresponding proportion of lichens having green algal photobionts (Hallingback,
1991). Low pH may be the most important effect of air pollution on Peltigera aphthosa in Sweden.
Nevertheless, there is some indication that NH/ in combination with S042 is more detrimental than low
pH per se (Hallingback and Kellner, 1992). The decline of lichens containing cyanobacteria in parts of
northern Europe has been associated with N deposition in the range of 5 to 10 kg N/ha/yr (Bobbink et al.,
1998). In fact, epiphytic cyanobacteria-containing lichens may be among the most sensitive species in
humid forested ecosystems to atmospheric N deposition (Bobbink et al., 1998; Hallingbaeck, 1991).
Epiphytic macro lichens (those that grow attached to trees or other plants) exhibit different
sensitivities to atmospheric pollutants, with some species being adversely impacted at air pollution levels
that may not be considered high relative to other sensitive receptors. Particularly sensitive genera include
Alectoria, Bryoria, Ramalina, Lobaria, Pseudocyphellaria, Nephroma, and Usnea (Blettetal., 2003;
McCune and Geiser, 1997).
Community composition of epiphytic lichens in the U.S. can be altered by relatively small
increases in N deposition (Fenn, 2003a). Most epiphytic lichens meet their nutritional requirements from
atmospheric deposition and can store N in excess of their nutritional needs (van Herk, 1999). Early work
in the San Bernardino Mountains, CA indicated that lichen cover was inversely related to estimated
oxidant doses (Sigal and Nash, 1983). In recent analysis it has been determined that up to 50% of lichen
species that occurred in the region in the early 1900s have disappeared, with a disproportionate number of
locally extinct species being epiphytic cyanolichens (Fenn et al., 2003a; Nash and Sigal, 1999). The
calculated critical load for lichen communities in mixed conifer forests in California is 3.1 kg N/ha/yr
(Fenn et al., 2008).
The Pacific Northwest retains widespread populations of pollution-sensitive lichens (Fenn, 2003a).
However, in urban areas, intensive agricultural zones, and downwind of major urban and industrial
centers in the Pacific Northwest, there are few air pollution-sensitive lichen species, such as epiphytic
3-148
-------
cyanolichens, and high N concentrations have been measured in lichen tissue (Fenn, 2003a). With N
enrichment, especially around urban and agricultural areas, there is a shift towards weedy, nitrophilous
lichen species (Fenn, 2003a). Replacement of sensitive lichens by nitrophilous species has undesirable
ecological consequences. In late-successional, naturally N-limited forests of the Coast Range and western
Cascade Mountains, for example, epiphytic cyanolichens make important contributions to mineral cycling
and soil fertility (Antoine, 2001, 2004; Pike, 1978; Sollins et al., 1980), and together with other large,
pollution-sensitive macrolichens, are an integral part of the food web for mammals, insects, and birds
(McCune and Geiser, 1997). Sensitive lichen species appear to be negatively affected by N inputs as low
as 3 to 8 kg N/ha/yr (Fenn, 2003a). (A summary of additional experiments on lichens is given in
Table 3-20.)
Table 3-20.
Summary of N effects on lichens.
Region/
Country
Endpoint
Observations
Ecosystem
Type/
Species
Reference
Netherlands
Species
richness
Deposition Gradient: Van Dobben, et al. (2001) recorded epiphytic lichen
presence, tree bark chemical composition, and atmospheric
concentrations of SO2, NO2 and NH3 at 123 sites along depositional
gradients. Relationships between atmospheric and bark chemistry and the
composition of the lichen vegetation were evaluated (ter Braak and Wiertz,
1994). Results showed nearly all lichen species investigated were
negatively affected by exposures to SO2 and NO2, collectively decreasing
lichen species richness. Of somewhat less importance were the ecological
factors such as bark pH, host tree species and tree diameter.
Epiphytic
lichens
Van
Dobben,
etal. (2001)
ter Braak
and Wiertz
(1994)
Scotland and
northern
England
Community
composition
Deposition Gradient: The authors suggested that the empirical critical
load of N deposition for protection of community composition of lichens
and bryophytes was in the range of 11 to 18 kg N/ha/yr
Atlantic oak
woods
Mitchell
etal. (2005)
Umea,
Vasterbotten,
Sweden
NH4+ vs.
NO3 uptake
rate
Isotopic Tracer: In a study of 15N uptake in 14 lichen associations (for
simplicity, designated as "species"), found that NH4"*uptake was
significantly greater, and to a higher extent passive, relative to amino acid
or NO3" sources of N. Differences were also observed in NO3" uptake,
depending on photobiont group; cyanobacterial lichens had a lower NO3"
uptake rate than green algal lichens. Morphology and microhabitat were
not found to be associated with N uptake
Cyanobacterial
lichens and
green algal
lichens
Dahlman
etal. (2004)
Sweden
NH4+ vs.
NO3 uptake
The assimilation and allocation of externally added N was investigated for
two N-fixing tripartite (possessing both green algal and cyanobacterial
phytobionts) lichen species, Peltigera aphthosa and Nephroma articum. N
uptake ranged from 2 to 27 percent of the 5 kg N/ha/yr that was applied
during the experiment over a 3 month period. Atmospheric deposition in
this part of Sweden (~5 kg N/ha/yr was about one-fourth the total experi-
mental N application rate. NH4+was absorbed to a greater extent than was
NO3". In general, 15N levels of NH4"* treated thalli were about four times
higher than for NO3" treated thalli. To some extent, this may reflect the
increased energy requirements of NO3" reduction as compared with NH4"*
assimilation (Raven et al., 1992) and/or the adsorption of positively
charged NH4"* on the negatively charged functional groups present on
hyphal cell walls.
Peltigera
aphthosa and
Nephroma
articum
Dahlman
etal.
(2002),
Sweden
NH4+vs. N03
uptake
NhUio be the preferred N source for the green algal foliose lichen
Plathismatia glauca, followed by glutamine and then NO3". This species
responded to increased N availability by increasing growth rate and C
assimilation capacity through increased investment in the photobiont cells
Plathismatia
glauca, a
foliose lichen
Palmqvist
and
Dahlman
(2006)
3-149
-------
Region/
Country
Endpoint
Observations
Ecosystem
Type/
Species
Santo
Antonio,
near Serra
de Aire
Candeeiros
Natural Park,
central
Portugal
NH4+ uptake
Physiological
responses
Comparison of the physiological responses of two lichens to increased N
supply. Uptake was quantified using 15N labeled NH4+. Cholrophyll a and
ergonsterol were used as indirect markers of algal and fugal activity,
respectively. The acidophytic lichen Evernia prunastri showed greater N
uptake from NH4+ than the nitrophytic lichen Xanthoria parietina. In the
acidophytic lichen, but not the nitrophytic lichen, ergosterol concentrations
decreased with increasing N uptake, and an increase in the NH4+ pool was
also observed at the highest N doses (216 kg N/ha/yr, applied as nine
applications over a 2-month period). These differences can partially
explain the higher tolerance of X parietina to high N deposition.
Evernia
prunastri and
Xanthoria
parietina
Reference
Pampulha NH4+vs. Cyanobacteria have been shown to grow on either N03~ or NH4+sources Cyanobacteria vonRuckert
reservoir NO3 growth when administered at non-toxic concentrations. More rapid growth was and Giani
(Belo observed with NH4+ fertilization as compared with NO3" fertilization. (2004)
Horizonte, Ammonium is more easily assimilated; both NO3 and nitrite must first be
Brazil) reduced to NH4+ before assimilation
Sweden NH4+ in Low pH may be the most important effect of air pollution on Peitigera Peitigera Hallingback
combination aphthosa in Sweden. Nevertheless, there is some indication that NH4+ in aphthosa andKellner
with SO42" combination with SO42" is more detrimental than low pH per se (1992)
(Hallingback and Kellner, 1992).
Gaio-
Oliveira
etal. (2005)
Alpine
Herbaceous plants in alpine communities are considered very sensitive to changes in N deposition.
A combination of short growing season, strong seasonal variation in moisture and temperature, shallow
and poorly developed soils, steep terrain, sparse vegetation, and low rates of primary productivity
generally limit the N uptake and retention capacity of herbaceous plant species in alpine ecosystems
(Burns, 2004; Fisk et al., 1998). Alpine herbaceous plants are generally considered N-limited and changes
in alpine plant productivity and species composition have been noted in response to increased N inputs
(Bowman et al., 2006; Vitousek et al., 1997) (See Table 3-21).
Research on N enrichment effects on alpine and subalpine ecosystems in the Western U.S. has
mainly been limited to studies at the Loch Vale Watershed in Rocky Mountain National Park and the
Niwot Ridge LTER site, both located east of the Continental Divide in Colorado (see review by Burns,
2004). Changes in alpine plant species composition on Niwot Ridge have included increased cover of the
plant species that tend to be most responsive to N fertilization in some of the long-term monitoring plots
(Fenn, 2003a; Korb and Ranker, 2001). These changes are likely due to response to changes in N
deposition. However, the influences of climatic change, particularly changes in precipitation (Williams
et al., 1996b) and pocket gopher disturbance (Sherrod and Seastedt, 2001) could not be ruled out as
contributors to vegetation change (Fenn, 2003a). Other environmental factors also affect the species
make-up of alpine ecosystems, but long-term experimental fertilization plots demonstrate a clear response
of alpine flora to N, including shifts toward graminoid plants that shade smaller flowering species, and
accompanying changes in soil N cycling (Bowman et al., 2006).
3-150
-------
Table 3-21. Summary of N effects on alpine ecosystems.
Region/Country
Endpoint
Observations
Ecosystem Type /
Species
Reference
Niwot Ridge, CO
and Southern
Wyoming
Community
shift
Field Addition: 25 kg N/ha/yr added during
summer caused a community shift towards
greater dominance of hairgrass (Deschampsia
sp.) in wet alpine meadows, but the increase in
plant biomass (+67%) and plant N content
(+107%) following N fertilization was higher in
graminoid-dominated dry meadows than in forb-
dominated wet meadows (+53% plant biomass,
+64% standing N crop, respectively)
Wet and dry alpine meadow,
alpine tundra, talus, alpine
and subalpine forest—
Englemann spruce,
Bristlecone pine, surface
waters, algae, amphibians
Bowman et
al. (1995);
Burns (2004)
Niwot Ridge, CO
Plant foliage
productivity
species
richness.
Field Addition: Showed that 4 years of N
addition to alpine vegetation at rates ranging
between 100 and 200 kg N/ha/yr (depending on
the year) caused marginal increases in alpine
plant foliage productivity but reduced species
richness.
Wet and dry alpine tundra-
sedge Kobresia myosuroides.
Acomastylis rossii,
Polygonum viviparurn
Trifolium. A.rossii and
Deschumnpsia caespitosa. D.
caespitosa, Caltha
leptosepala, Sibbaldia
procurnbens and Trifolium
parryi.
Seastedt and
Vaccaro
(2001)
Niwot Ridge, CO
Species
composition
species
diversity plant
biomass tissue
[N]
Field Addition: Additions of 20,40, and 60 kg
N/ha/yr (on top of ambient N deposition near 5
kg N/hafyr) over an 8 year period to a dry alpine
meadow led to a change in plant species
composition, an increase in species diversity and
plant biomass, and an increase in tissue N
concentration at all treatment levels within 3
Dry alpine meadow
Bowman
etal. (2006)
years of application. Much of the response was
due to increased cover and total biomass of
sedges (Carex spp.). There was a significant
decrease in Kobresia biomass with increasing N
input. Vegetation composition appeared to
respond at lower N input levels than those that
caused measurable changes in soil inorganic N
content. Changes in an individual species (Carex
rupestris) were estimated to occur at deposition
levels near 4 kg N/ha/yr. Changes in the plant
community, based on the first axis of a detrended
correspondence analysis, were estimated to
occur at deposition levels near 10 kg N/ha/yr. In
contrast, increases in NO3" leaching, soil solution
NO3" concentration, and net nitrification occurred
at levels above 20 kg N/ha/yr. The authors
concluded that changes in vegetation
composition preceded detectable changes in soil
indicators of ecosystem response to N
deposition.
3-151
-------
Mycorrhizae across ecosystems
Microbial biodiversity can also be affected by N enrichment. Interactions between atmospherically
deposited N and terrestrial vegetation frequently occur in the rhizosphere. The rhizosphere includes the
soil that surrounds and is influenced by plant roots (Wall and Moore, 1999). Relationships among plant
roots, mycorrhizal fungi, and microbes are critical forN cycling and for the growth and health of plants.
The plant provides shelter and C; the fungi and bacteria provide access to potentially limiting nutrients,
particularly N and P.
A meta-analysis of the effect of N and P fertilization on mycorrhize observed a 15% decrease in
mycorrhizal abundance due to N fertilization across 16 studies at 31 sites, covering a range of grassland,
shrubland, temperate and boreal forest ecosystems (Treseder, 2004). Declines in mycorrhizal abundance
were slightly higher at higher rates of N fertilization, but there was significant variation across all studies.
The loss of mycorrhizal function has been hypothesized as a key process contributing to reduced N uptake
by vegetation and increased N03 mobility from soil into drainage water (U.S. EPA, 2004).
NO3" versus NhV deposition
Plants also exhibit different degrees of response to N03 versus NH44" deposition. In general, fast-
growing annual species, including many agricultural crops, and fast growing pioneer trees such as birch
(Betula spp.) prefer N03 (Pearson and Stewart, 1993). Slow-growing perennial plant species generally
prefer NH/. There are also many plant species which readily utilize both N03 and NH44" (Krupa, 2003).
These include members of the family Ericaceae (e.g., Calluna, Erica, Vaccinium), conifer trees, and
climax species such as Quercus and Fagus (Krupa, 2003).
3.3.5.2. Transitional Ecosystems
Wetlands in the U.S. support over 4200 native plant species, of which 121 are federally threatened
or endangered (http://plants.usda.gov/). Wetlands can be divided into three general categories based on
hydrology. Hydrologic pathways are often the same pathways of N input; therefore they are useful for
discussing the N sources and sensitivity to atmospheric N deposition. Nearly all new N comes from
atmospheric deposition in ombrotrophic bogs because they only receive water inputs via precipitation and
they develop where precipitation exceeds evapotranspiration and where there is some impediment to
drainage of the surplus water (Mitsch and Gosselink, 1986). Fens, marshes and swamps are characterized
by ground and surface water inputs that are often on the same order of magnitude as precipitation
(Koerselman et al., 1989). Lastly, intertidal wetlands receive water from precipitation, ground/surface
water and marine/estuarine sources.
The balance of competition among plant species in some sensitive wetland ecosystems can be
altered by N addition, with resulting displacement of some species by others that can utilize the excess N
more efficiently (U.S. EPA, 1993a). The sensitivity of wetlands is particularly important given that they
contain a disproportionately high number of rare plant species that have evolved under N-limited
condition (Moore et al., 1989) (See Annex C). In general these include the genus Isoetes sp., of which
three species are federally endangered; insectivorous plants like the endangered green pitcher Sarracenia
oreophila; and the genus Sphagnum, of which there are 15 species are listed as endangered by eastern
U.S. states. Roundleaf sundew (Drosera rotundifolia) is also susceptible to elevated atmospheric N
deposition (Redbo-Torstensson, 1994). This plant is native to, and broadly distributed across, the U.S. and
is federally listed as endangered in Illinois and Iowa, threatened in Tennessee, and vulnerable in New
York (http://plants.usda.gov/').
Freshwater wetlands
Peatlands and bogs are among the most vulnerable transitional ecosystems to adverse nutrient-
enrichment effects of N deposition (Krupa, 2003). The sensitivity of peatland Sphagnum species to
3-152
-------
elevated atmospheric N deposition is well documented in Europe (Berendse et al., 2001; Tomassen et al.,
2004). Sphagnum squarrosum and S. fallax have been observed to be negatively affected by
experimentally elevated atmospheric N and S inputs in Europe (Kooijman and Bakker, 1994). The genus
Sphagnum dominates ombrotrophic bogs and some nutrient poor fens in the northern U.S. and Canada.
These mosses efficiently capture atmospheric deposition with retention rates between 50-90%, much of
the variation due to the depth of the water table (Aldous, 2002). Studies conducted on 4 species of
Sphagnum in Maine (2 to 4 kg N/ha/yr ambient deposition) and New York (10 to 13 kg N/ha/yr ambient
deposition) document that higher N deposition resulted in higher tissue N concentrations and greater NPP,
but lower bulk density (Aldous, 2002). A study of Sphagnum fuscum in six Canadian peatlands showed a
weak, although significant, negative correlation between NPP and N deposition when deposition levels
were greater than 3 kg N/ha/yr (y = 150 - 3.4x, p=0.04, r2=0.01) (Vitt et al., 2003). A study of 23
ombrotrophic peatlands in Canada with deposition levels ranging from 2.7 to 8.1 kg N/ha/yr showed peat
accumulation increases linearly with N deposition (y = 2.84(x) + 0.67, r2 = 0.32, p <0.001), however in
recent years this rate has begun to slow indicating limited capacity for N to stimulate accumulation
(Moore et al., 2004).
The sensitivity of peatland Sphagnum species to elevated atmospheric N deposition is well
documented in Europe (Berendse et al., 2001; Tomassen et al., 2004). Sphagnum squarrosum and S. fallax
have been observed to be negatively affected by experimentally elevated atmospheric N and S inputs in
Europe (Kooijman and Bakker, 1994). Roundleaf sundew (Drosera rotundifolia) is also susceptible to
elevated atmospheric N deposition (Redbo-Torstensson, 1994).
Sarracenia purpurea is a long lived (30-50 years) northern pitcher plant and widely distributed in
bogs, fens and swamps across Canada and the eastern U.S. (Ellison and Gotelli, 2002). S. purpurea has
adapted to nutrient poor environments and very sensitive to increasing N input. In a study of S. purpurea
in Vermont and Massachusetts, Ellison and Gotelli (2002) conducted a series of N enrichment
experiments by augmenting N availability to leaves with 0, 0.1 and 1 mg N/L NH44" solution for one
growing season. Population growth rates, estimated by demographic survey, were positive for 0 and 0.1
mg N/L additions (equal to atmospheric deposition of 0-1.4 kg N/ha/yr)1 and negative for 1 mg N/ L
additions (equivalent to 10-14 kg N/ha/yr)1 (Gotelli and Ellison, 2006). Based on the annual demographic
rates, a non stationary matrix model forecasted that the extinction risk within the next 100 years increased
substantially if N deposition rate increase (1-4.7%) from the rate of 4.5-6.8 kg N/ha/yr (Gotelli and
Ellison, 2002).
Increasing N availability not only reduced population growth of S. purpurea, also dramatically
altered plant morphology. S. purpurea produces carnivorous leaves (pitcher) and photosynthesis efficient
leaves (phyllodia). N enrichment was shown to stimulate the photosynthesis rate and increase the
production of phyllodia relative to pitcher (Ellison and Gotelli, 2002). The field N deposition simulation
experiment (ranged from 0-35 kg N/ha/yr)1 revealed a positive linear relationship between N deposition
level and relative keel size (keel width/total width). This correlation was supported by the field surveys of
26 sites across Massachusetts and Vermont (Ellison and Gotelli, 2002), and 39 sites across Canada and
eastern U.S. (Ellison and Gotelli, 2002). The relative kneel size of northern pitcher plant increased with
increasing NEL^ concentration in soil water, and may be used as bioindicator (log [NEL^] = -1.57 + 1.78x
relative keel size).
In wet heathlands in Europe, changes in plant species composition have been attributed to elevated
atmospheric N deposition (Roem and Berendse, 2000). Diverse plant communities have been replaced by
monospecific stands Dutch wet heathlands (Aerts and Berendse, 1988; Houdijk et al., 1993). In other
studies, wetland species such as Calluna vulgaris can successfully compete with grasses even at relatively
1 N treatments were selected to represent annual N deposition measured at the nearest monitoring sites of National Atmospheric Deposition
Program (NADP). The unit of N treatments reported in the publication was precipitation-weighted mean concentrations (mg N/ L), from which
we calculated the level of deposition (kg N/ha/yr) using the equation: Deposition= Precipitation-Weighted Mean Concentrations / Annual
Precipitation. More detailed information on nitrogen deposition is available on the NADP website: http://nadp.sws.uiuc.edu/sites/ntnmap.asp?
3-153
-------
high rates of N deposition, as long as the vegetative canopies are closed (Aerts et al., 1990). However, N
deposition causes nutrient imbalances, and increased plant shoot-to-root ratio, and therefore increases in
the sensitivity of shrubs to drought stress, frost stress, and attack by insect pests (Heil and Diemont,
1983). These can result in gaps in the canopy of the shrub layer, which can then be readily invaded by
grasses that are more efficient in using the additional N and therefore gain a competitive advantage
(Krupa, 2003).
Riparian wetlands
Marler et al. (2001) evaluated the potential impacts of experimentally elevated stream water
nutrient concentrations on three riparian wetland tree species: Fremont Cottonwood (Populus fremontii),
Goodding willow (Salix gooddingii), and exotic saltcedar (Tamarix ramosissima) in the riparian zone of
the Salt River near Phoenix, AZ. The results from this 43-day experiment showed that growth of all three
riparian plant species responded positively to increased nutrient supply (treatment 3 and 4) (Marler et al.,
2001). The exotic and invasive salt cedar showed the greatest increases in biomass at high nutrient supply.
Other studies have also found that exotic plant species often respond more rapidly than native vegetation
to increased nutrient supply (Milberg et al., 1999; Paschke et al., 2000). This experiment was conducted
to simulate impacts of wastewater effluent on riparian zones, and N additions were therefore very large
nutrient supply to riparian systems via atmospheric N deposition in the U.S. is more typically in the range
of treatments in this experiment that showed minimal response to N addition.
Table 3-22. Summarized responses of coastal marshes ecosystem to N fertilization. The table includes
studies in which the lowest fertilization treatment is below 400 kg N/ha/yr, a value at the higher
end of the range that includes direct and indirect N deposition.
Site
Species
Responses
N enrichment
Reference
Walden Creek
(NC)
Spartina
alterniflora
Field Addition: (1) increased the growth of short Spartina, but
had no effect on tall Spartina; (3) biomass production of short
Spartina increased linearly with N addition; and (3) ammonium
showed higher growth stimulation on short Spartina than NO3
does
0,280,560,1120
kg N/ha/yr
Mendelssohn
etal. (1979)
Narragansett
Bay (Rl)
Spartina
patens;
Spartina
alterniflora
Field Addition: (1) decreased the density and extent of S
patens; (2) decreased the extent of tall S. alterniflora increased
with
N gradient from 9
to 3282 kg
N/ha/yr
Wigand et al.
(2003)
Great
Sippewissett
Marsh (MA)
Spartina
alterniflora
Field Addition: increased live above ground biomass, leaf
area coverage, evapotranspiration
396 kg N/ha/yr
Howes et al.
(1986)
Intertidal wetlands
Wetland eutrophication could significantly damage the structure and function of coastal marshes.
N enrichments were shown to facilitate the invasion of nonnative species (Tyler et al., 2007); shift the
competition between native species (Mendelssohn, 1979; Wigand et al., 2003; Crain 2007); increase
herbivore damage on plants (Bertness et al., 2008); stimulate evapotranspiration (Howes et al., 1986);
change microbial community and pore water chemistry (Caffrey et al., 2007); and alter carbon allocation
3-154
-------
between root and shoot (Darby and Turner, 2008). MostN fertilization experiments add levels ofN orders
of magnitude above that expected by atmospheric deposition. A summary of experiments that include
addition levels below 400 kg N/ha/yr is given in Table 3-22.
3.3.5.3. Freshwater Aquatic Ecosystems
Paleolimnological studies
The paleolimnological method of taxonomic identification of fossil diatoms in lake sediments has
been augmented in recent years with cell counts and pigment concentrations of Chi a and chlorophyll
derivatives, rendering inferences about trophic state from proxies preserved in sediments more robust than
before (Das et al., 2005). Paleolimnological studies of mountain lakes that have only been disturbed by
atmospheric deposition and climate change have reported changes in diatom species assemblages,
increases in cell numbers, and pigment-inferred increases in whole lake primary production. These
inferred changes have been coincident with regional surrogates for increased N deposition. Such changes
have included increases in human population, industrial animal production, and fossil fuel combustion
emissions (Das et al., 2005; Saros et al., 2003; Wolfe et al., 2001a, 2003). In most, but not all, of these
studies, the observed changes in ecology were inconsistent with changes in climate and more concordant
with effects from increased atmospheric N deposition.
Available data suggest that the increases in total N deposition do not have to be large to elicit an
ecological effect. For example, a hindcasting exercise determined that the change in Rocky Mountain
National Park lake algae that occurred between 1850 and 1964 was associated with an increase in wet N
deposition that was only about 1.5 kg N/ha (Baron, 2006). Similar changes inferred from lake sediment
cores of the Beartooth Mountains of Wyoming also occurred at about 1.5 kg N/ha deposition (Saros et al.,
2003). Pre-industrial inorganic N deposition is estimated to have been only 0.1 to 0.7 kg N/ha based on
measurements from remote parts of the world (Galloway et al., 1995; Holland et al., 1999). In the western
U.S., pre-industrial, or background, inorganic N deposition was estimated by (Holland et al., 1999) to
range from 0.4 to 0.7 kg N/ha/yr.
Bioassay, mesocosm, and laboratory experiment
Bioassay, mesocosm, and laboratory experiments have been conducted on algae (both
phytoplankton and periphyton), invertebrates, amphibians, and fish, to determine effects of N on sensitive
aquatic organisms (see Annex C). Some freshwater algae are particularly sensitive to the effects of added
nutrient N and experience shifts in community composition and biodiversity with increased N deposition.
For example, two species of diatom, Asterionella formosa and Fragilaria crotonensis, now dominate the
flora of at least several alpine and montane Rocky Mountain lakes and sharp increases have occurred in
Lake Tahoe (Baron et al., 2000; Interlandi and Kilham, 1998; Saros et al., 2003, 2005; Wolfe et al., 2001
and 2003). The timing of this shift has varied, with changes beginning in the 1950s in the southern Rocky
Mountains and in the 1970s or later in the central Rocky Mountains (Figure 3-50). These species are
opportunistic algae that have been observed to respond rapidly to disturbance and slight nutrient
enrichment in many parts of the world (See Annex C for additional discussion).
Further evidence for the relationship between N enrichment and algal changes has been provided
by N addition studies that include in situ mesocosm studies (Lafrancois et al., 2004; McKnight et al.,
1990; Saros et al., 2005) and in situ incubations in large lakes (Interlandi and Kilham, 1998). Differences
in resource requirements allow some species to gain competitive advantage over others upon nutrient
addition, causing changes in species composition (Lafrancois et al., 2004; Saros et al., 2005; Wolfe et al.,
2003). A summary of these experiments is given in Table 3-23. This is in keeping with findings of
Interlandi and Kilham (2001), who demonstrated that maximum species diversity was maintained when
N levels were low (<3 |_iM) in lakes in the Yellowstone National Park region.
3-155
-------
The implication of this research is that species diversity declines with increasing availability of N.
In studies of lake sediment diatom remains, typical oligotrophic species such as Aulacoseria perglabra,
Cyclotella steligera, and Achncmthes spp. declined coincident with the rise in dominance of A. formosa
and F. crotonensis (Wolfe et al., 2001b, 2003).
£
o
a.
J?
&
&
i i i i i
20 40 60 80 100
20
~ I I
20 40 60
-2000
-1990
-1970
-1930
-1880
20 20
10 12
1 I
2 4
Source : Saros (2003). Reprinted with permission.
Figure 3-50. Diatom assemblage sediment patterns in Emerald Lake, WY.
Community shifts in phytoplankton other than diatoms have also been observed under conditions
of elevated N availability (Lafrancois et al., 2004). For example, a positive correlation between the
proportion of the phytoplankton comprised of chrysophytes and the concentration of NO;, in lake water
was found in a survey of 15 Snowy Range lakes (Lafrancois et al., 2003). Chlorophytes, like the two
diatom species identified above, generally prefer high concentrations of N and are able to rapidly
dominate the flora when N concentrations increase (Findlay et al., 1999). This occurs in both
circumneutral and acidified waters (Findlay et al., 1999; Wilcox and Decosta, 1982).
In summary, survey data and fertilization experiments demonstrate that increase in algal
productivity, as well as species changes and reductions in biodiversity, have occurred at sensitive high
elevation lakes in the western U.S. in response to increased availability of N.
3-156
-------
Table 3-23.
N effects on algal species composition and biodiversity
Region
Endpoint
Observation
Ecosystem
Type
Reference
Snowy Range,
Wyoming
Community shifts
in phytoplankton
Mesocosm Experiment: correlation between the proportion of the
phytoplankton comprised of chrysophytes and the concentration of
NO3" in lake water was found in a survey of 15 lakes. Chrysophytes
were favored in lakes having lower N and cyanophytes and
chlorophytes favored in lakes having higher N
Lakes
Lafrancois
etal. (2003)
Beartooth
Mountains of
Montana-
Wyoming
Diatom
community
Paleo and Observation: evaluation of resource requirements for
dominant diatom species with paleolimnological reconstructions
and contemporary surveys of the flora of seven lakes. Results
reinforced the likelihood that recent increases in dominant diatom
numbers have been the result of N enrichment rather than climatic
change
Lakes
Saros et al.
(2005)
Colorado Front
Range
Community shifts
in phytoplankton
Paleo: sediment cores showed increasing representation of
mesotrophic diatoms in recent times, as compared with pre-
development conditions
Lakes
Wolfe etal.,
(2001b)
Lake Tahoe,
CA
Community shifts
in phytoplankton
Paleo: there has been a sharp increase in the ratio of araphidinate
pennate to centric diatoms since about 1950 (largely due increases
in Fragilaria crotenensis), associated with increased N loading to
the lake. Jassby et al. (1994) showed that atmospheric deposition
supplies most of the N to Lake Tahoe.
Lakes
Goldman
(1988)
3.3.5.4. Estuarine and Marine Ecosystems
In coastal ecosystems, eutrophication can cause changes in marine biodiversity and species
composition. Phytoplankton production and community composition in estuarine and marine
environments also respond to differences in the form of atmospheric N input. Major algal functional
groups, including diatoms, dinoflagellates, cyanobacteria, and chlorophytes, may show different
responses to changing mixtures of added N (Paerl et al., 2002).
Phytoplankton
In addition to causing increased phytoplankton biomass, as indicated by Chi a measurements (see
Section 3.3.3.2), excess N can contribute to changes in phytoplankton species composition. High loadings
of N and P can also increase the potential for Si limitation, with associated changes in diatoms. Such
changes to the phytoplankton community can also affect higher trophic levels. For example, Officer and
Ryther (1980) and Turner et al. (1998) suggested that a shift in the Si-to-N atomic ratio to less than 1
would alter the marine food web. Specifically, the diatom-to-zooplankton-to-higher tropic level ratios
would decrease, whereas flagellated algae (including those that often contribute to hypoxia) would
increase (Paerl et al., 2001a).
Changes in phytoplankton species abundances and diversity have been further documented through
in situ bioassay experiments such as the results reported by Paerl et al. (2003) for the Neuse River Estuary
in North Carolina. Effects were species-specific and varied dramatically depending on whether, and in
what form, N was added. The findings illustrate the potential impacts of N additions on phytoplankton
community structure (see Figure 3-51).
3-157
-------
Changing phytoplankton community composition has numerous potential ecological ramifications,
including modifications to the ecosystem food web and nutrient dynamics. For example, if the nutrient
mix favors species that are not readily grazed (e.g., cyanobacteria, dinoflagellates), trophic transfer will
be poor and relatively large amounts of unconsumed algal biomass will settle to the bottom, which could
stimulate decomposition, 02 consumption, and the potential for hypoxia (Paerl et al., 2003).
Reduced vs. Oxidized N
The form of N input to coastal aquatic ecosystems has an important influence on its effects.
Atmospheric deposition of reduced N has increased relative to oxidized N in the eastern U.S., and this
trend is expected to continue in the future under existing emissions controls. Such patterns can influence
marine eutrophication responses. Some studies suggest that large diatoms tend to dominate coastal waters
when N03 is supplied (Paerl et al., 2001a; Stolte et al., 1994), whereas smaller diatom species have a
greater preference for NH44" uptake. Thus, ongoing trends of decreasing N03 deposition and increasing
NH4+ deposition might lead to changes in species distributions and size distributions of phytoplankton,
with cascading effects on trophic structure and biogeochemical cycling (Paerl et al., 2001a).
Not all studies have found variation in algal response with the form of N applied. For example,
Richardson et al. (2001) examined the effects of different forms of N application (N03 , NH/, urea) on
the structure and function of estuarine phytoplankton communities in mesocosm experiments in the Neuse
River Estuary, NC. Even though NH44" is more readily taken up by phytoplankton in this estuary than is
N03 (Twomey et al., 2005), the results of the Richardson et al. (2001) study suggested that
phytoplankton community structure was determined more by the hydrodynamics of the system than by
the form of N available for growth.
Twomey et al. (2005) measured Neuse River Estuary phytoplankton uptake rates of NH/, N03 .
and urea. Ammonium was the dominant form of N taken up, contributing about half of the total N uptake
throughout the estuary. Uptake varied spatially; in particular N03 uptake declined from 33% of the total
uptake in the upper estuary to 11% and 16%, respectively in the middle and lower estuary. Urea uptake
contributed least to the total in the upper estuary (16%) but comprised 45 and 37% of the total N uptake in
the middle and lower estuary. Therefore, N budgets based only on inorganic forms may seriously
underestimate the total phytoplankton uptake (Twomey et al., 2005).
Submerged Aquatic Vegetation (SAV)
SAV provides important nursery grounds to many estuarine fish. There are few data documenting
the long-term response of SAV in coastal ecosystems to N loading. The national assessment (Bricker
et al., 2007) suggested that only a small fraction of the estuary systems evaluated reported high severity of
SAV loss. Most of those that did report moderate or high loss were located in the Mid-Atlantic region.
However, where SAV loss is a problem, the results can be severe, and there is evidence suggesting a
correlation with increases in N loading. For example, at Waquoit Bay, Massachusetts, Valiela et al. (1990)
reported a strong negative relationship between modeled N loading and measured eelgrass area based on
measurements of eelgrass coverage from 1951 to 1992.
3-158
-------
400
350
! ¦ Centritractus
21 Ankistrodesmus
H Unidentified dinofiageliate
H Closterium I
~ Comphosphaeria I
m 200
S
150
100
a
50
0
Source: Paerl et al. (2003)
Figure 3-51. Microscopic counts of phytopiankton species composition in the Neuse River Estuary, NC
following 36-h in situ bioassays to manipulate available forms of N. Treatments included a
control (unamended estuarine water sample), all nutrients (N, P, vitamins, trace metals, and
Si), all with urea as the N form, all with ammonium (NIV) as the N form, all with N03~ as the N
form, and all with no N. Bars represent the mean density of cells present (three replicate
counts for each treatment).
3.3.5.5. Summary of N Effects on Species Composition, Species Richness and
Biodiversity
The evidence is sufficient to infer a causal relationship between N deposition and the alteration of
species richness, species composition and biodiversity in terrestrial ecosystems. The ecological effects of
N deposition were described for a variety of taxa and ecosystem types including: forests, grasslands, arid
and semi-arid, deserts, lichens, alpine, and mycorrhizae. Among the most sensitive terrestrial taxa are
lichens. Empirical evidence indicates that lichens in the U.S. are adversely affected by deposition levels
as low as 3/ha/yr. Among the most sensitive ecosystems are Alpine ecosystems; alteration of plant cover
of an individual species (Carex rupestris) in Alpine communities were estimated to occur at deposition
levels near 4 kg N/ha/yr and modeling indicates that deposition levels near 10 kg N/ha/yr alter plant
community assemblages.
The evidence is sufficient to infer a causal relationship between Nr deposition and the alteration of
species richness, species composition and biodiversity in wetland ecosystems. The effect of N deposition
on wetland ecosystems depends on the fraction of rainfall in its total water budget. The sensitivity to N
deposition was suggested as bogs >fens >intertidal wetlands (Morris, 1991).
Terrestrial
Wetland
3-159
-------
Excess N deposition can cause shifts in wetland community composition by altering competitive
relationships among species, which potentially leads to effects such as decreasing biodiversity, increasing
non-native species establishment and increasing the risk of extinction for sensitive and rare species.
Wetlands contain a high number of rare plant species (Bedford and Godwin, 2003; Moore et al., 1989;
U.S. EPA, 1993a). High levels of atmospheric N deposition increase the risk of decline and extinction of
these species that are adapted to low N conditions. In general these include the genus Isoetes sp., of which
three species are federally endangered; insectivorous plants like the endangered green pitcher Sarracenia
oreophila; and the genus Sphagnum, of which there are 15 species are listed as endangered by eastern
U.S. states. Roundleaf sundew (Drosera rotundifolia) is also susceptible to elevated atmospheric
N deposition (Redbo-Torstensson, 1994). This plant is native to, and broadly distributed across, the U.S.
and is federally listed as endangered in Illinois and Iowa, threatened in Tennessee, and vulnerable in New
York (http://plants.usda.gov/). In the U.S., Sarracenia purpurea can be used as a biological indicator of
local N deposition in some locations (Ellison and Gotelli, 2002).
Freshwater Aquatic
The evidence is sufficient to infer a causal relationship between Nr deposition and the alteration of
species richness, species composition and biodiversity in freshwater aquatic ecosystems. Evidence from
multiple lines of research and experimental approaches support this observation, including
paleolimnological reconstructions, bioassays, mesocosm and laboratory experiments. Increased N
deposition can cause a shift in community composition and reduce algal biodiversity. Elevated N
deposition results in changes in algal species composition, especially in sensitive oligotrophic lakes.
In the West, a hindcasting exercise determined that the change in Rocky Mountain National Park
lake algae that occurred between 1850 and 1964 was associated with an increase in wet N deposition that
was only about 1.5 kg N/ha (Baron, 2006). Similar changes inferred from lake sediment cores of the
Beartooth Mountains of Wyoming also occurred at about 1.5 kg N/ha deposition (Saros et al., 2003).
Some freshwater algae are particularly sensitive to added nutrient N and experience shifts in
community composition and biodiversity with increased N deposition. For example, two species of
diatom (a group of algae), Asterionella formosa and Fragilaria crotonensis, now dominate the flora of at
least several alpine and montane Rocky Mountain lakes. Sharp increases have occurred in Lake Tahoe
(Baron et al., 2000; Interlandi and Kilham, 1998; Saros et al., 2003; Saros et al., 2005; Wolfe et al.,
2001b; Wolfe et al., 2003). The timing of this shift has varied, with changes beginning in the 1950s in the
southern Rocky Mountains and in the 1970s or later in the central Rocky Mountains. These species are
opportunistic algae that have been observed to respond rapidly to disturbance and slight nutrient
enrichment in many parts of the world.
Estuarine Aquatic
The evidence is sufficient to infer a causal relationship between N deposition and the alteration of
species richness, species composition and biodiversity in estuarine ecosystems. Increased N deposition
can cause shifts in community composition, reduced hypolimnetic DO, reduced biodiversity, and
mortality of submerged aquatic vegetation. The form of deposited N can significantly affect
phytoplankton community composition in estuarine and marine environments. Small diatoms are more
efficient in using N03 than NH/. Increasing NH/ deposition relative to N03 in the eastern U.S. favors
small diatoms at the expense of large diatoms. This alters the foundation of the food web. Submerged
aquatic vegetation is important to the quality of estuarine ecosystem habitats because it provides habitat
for a variety of aquatic organisms, absorbs excess nutrients, and traps sediments. Nutrient enrichment is
the major driving factor contributing to declines in submerged aquatic vegetation coverage. The Mid-
Atlantic region is the most heavily impacted area in terms of moderate or high loss of submerged aquatic
vegetation due to eutrophication. Indicators to assess the eutrophic condition of estuarine and coastal
waters are given in Figure 3-34.
3-160
-------
3.3.6. N Deposition Effects on NO3- Toxicity
N03 in freshwater at extremely high concentrations can have direct adverse effects on many life
stages of fish, as well as on invertebrates and amphibians. These effects occur at levels that are typically
more than 30 times higher than those that would commonly be attributable to atmospheric deposition, and
therefore N03 concentration has not been defined as a primary biological indicator. These effects are
described in Annex C.
3.3.7. Critical Loads and Other Quantified Relationships between
Deposition Levels and Ecological Effects
This section highlights a variety of sensitive chemical and biological receptors that have been used
in developing critical loads for nutrient effects of N deposition on natural ecosystems. Sensitive receptors
for effects of excess nutrient N deposition on surface water could include water chemistry, productivity,
and the response of important taxa. Key sensitive receptors for assessing effects on soil include soil
chemistry and soil solution chemistry. Sensitive receptors for flora include macro-lichens and vascular
plant species that are adapted to nutrient-poor environments. In some cases, the chemical receptors may
be easier to characterize, although they likely also reflect important biological changes that may be more
difficult to document. Background information on critical loads is presented in Annex C. Empirical
models of critical loads for nutrient-N have been in use in Europe for some time (UNECE, 2004). Efforts
have begun to develop empirical relationships in the U.S.
3.3.7.1. Empirical Critical Loads for Europe
Within the United Nations Economic Commission for Europe (UNECE) Long Range
Transboundary Air Pollution (LRTAP) convention, empirical procedures have been developed to set
critical loads for atmospheric N deposition to protect against effects caused by nutrient enrichment.
Empirical critical loads ofN deposition for natural and semi-natural terrestrial and wetland ecosystems
were first presented in a background document for the 1992 LRTAP workshop on critical loads held at
Lokeberg, Sweden (Bobbink et al., 1992b). A number of European expert workshops have taken place to
reach agreement among specialists regarding the impacts of N deposition on various ecosystems and
related critical loads (Achermann and Bobbink, 2003; Bobbink et al., 1992b; Bobbink et al., 1996;
Hornung et al., 1995; Nilsson and Grennfelt, 1988).
Information from the period 1996-2002 on the effects of increased N deposition on the structure
and function of natural and semi-natural ecosystems in Europe was evaluated in Bobbink et al. (2003).
The updated N critical loads were discussed and approved by full consensus at the November 2002 expert
meeting held under the LRTAP Convention in Berne, Switzerland, (Achermann and Bobbink, 2003).
Values for areas with low N deposition were updated by a CLRTAP workshop on critical loads of N in
low-deposition areas (Stockholm, Sweden, March 2007) and adopted by ICP M&M and WGE in 2007.
The resulting values are given in Table 3-24.
3-161
-------
Table 3-24 Biological indicators for the effects of elevated N deposition and related empirical critical
loads for major ecosystem types (according to the eunis classification) occurring in Europe.
Ecosystem Type
Biological Effect Indicators
Empirical Critical Load
(kg N/ha/yr)
GRASSLANDS AND TALL FORB HABITATS (E)
Sub-Atlantic semi-dry calcareous
grassland
Increased mineralization, nitrification and N leaching; increased tall
grasses; decreased diversity
15-25
Non-Mediterranean dry acid and
neutral closed grassland
Increase in nitrophilous graminoids, decline of typical species
10-20
Inland dune grasslands
Decrease in lichens, increase in biomass, accelerated succession
10-20
Low and medium elevation hay
meadows
Increased tall grasses, decreased diversity
20-30
Mountain hay meadows
Increase in nitrophilous graminoids, changes in diversity
10-20
Moist and wet oligotrophic
grasslands
Increase in tall graminoids, decreased diversity, decrease in
bryophytes
10-25
Alpine and subalpine meadows
Increase in nitrophilous graminoids, changes in diversity
10-15
Moss and lichen dominated
mountain summits
Effects on bryophytes and lichens
5-10
HEATHLAND HABITATS (F)
Northern wet heaths
Decreased heather dominance, transition heather to grass, decline in
lichens and mosses
10-20
Dry heaths
Transition from heather to grass, decline in lichens
10-20
Arctic, alpine, and subalpine scrub
habitats
Decline in lichens, mosses, and evergreen shrubs
5-15
COASTAL HABITAT (B)
Shifting coastal dunes
Increased biomass, increased N leaching
10-20
Coastal stable dune grasslands
Increase in tall grasses, decreased prostrate plants, increased N
leaching
10-20
Coastal dune heaths
Increase in plant production, increased N leaching, accelerated
succession
10-20
Moist to wet dune slacks
Increase in biomass and tall graminoids
10-25
MIRE, BOG, AND FEN HABITATS (D)
Raised and blanket bogs
Changed species composition, N saturation of Spagnum
5-10
Poor fens
Increased sedges and vascular plant, negative effects on mosses
10-20
Rich fens
Increase in tall graminoids, decreased diversity, decrease of
characteristic mosses
15-35
Mountain rich fens
Increase in vascular plants, decrease in bryophytes
15-25
3-162
-------
Ecosystem Type
Biological Effect Indicators
Empirical Critical Load
(kg N/ha/yr)
FOREST HABITATS (G)
Mycorrhizae
Reduced sporocarp production, reduced below ground species
composition
10-20
Ground vegetation
Changed species composition, increased nitrophilous species;
increased susceptibility to parasites (insects, fungi, virus)
10-15
Lichens and algae
Increase in algae; decrease in lichens
10-15
Source: Achermann and Bobbink (2003). Reprinted with permission.
3.3.7.2. Empirical Critical Loads for U.S.
Efforts have begun to develop empirical relationships in the U.S., particularly for western
ecosystems, however there is currently no published assessment for the U.S or the continent of North
America. Table 4-4 summarizes publications of deposition levels and related ecological effects, presenting
critical loads when reported in the original publication. Table 4-4 includes N levels at which effects are
manifested in terrestrial and freshwater ecosystems that have been documented through N addition and
deposition gradient studies. Several important studies from Europe published after the assessment by
Bobbink et al. (2003) are included in addition to some publications from Asia. Dose-response
relationships between N and ecological indicators are given in Table 3-25.
In terrestrial ecosystems, the reported effect levels range from 4 to 5 kg N/ha/yr for changes in the
abundance of individual sensitive alpine plant species, to 20 kg N/ha/yr for community level changes in
alpine plant communities. Clark and Tilman (2008) calculate the CL for the onset of reduced relative
species number in grasslands to be 5.3 kg N/ha/yr with a 95% inverse prediction interval of 1.3-
9.8 kg N/ha/yr. A critical load of 3.1 kg N/ha/yr is considered protective of lichen communities in the
West (Fenn et al., 2008).
Differences in the levels at which increased nitrification and N03 leaching have been observed in
eastern and western watersheds. For example, Rueth (2002) observed increased rates of nitrification in
old-growth forests in Colorado at approximately 5 kg N/ha/yr, whereas Aber et al. (2003) associated the
onset of N03 leaching in eastern forests with deposition levels of 7 to 10 kg N/ha/yr. The critical load for
N03 leaching in western chapparal ecosystems is 17 kg N/ha/yr (Fenn et al., 2008).
There is evidence that freshwater wetlands in the U.S. and Canada that are dominated by
Sphagnum sp. are affected by N deposition. Most evidence documents N retention, peat accumulation and
changes in NPP, and is not sufficient to quantify a critical load. The suggested critical load for protecting
the population health of northern pitcher plant is 10-14 kg N/ha/yr (Gotelli and Ellison 2006). There are
no publications suggesting critical loads for coastal wetland ecosystems.
3-163
-------
Table 3-25.
Summary of dose-response curves for N deposition and ecological indicators.
Ecosystems
Deposition
Range
(kg N/ha/yr)
Effect
Indicator
Study Region
Response Curve
Reference
N DEPOSITION
Forest
1-75
Nutrient
enrichment
N leaching
65 forest sites
across Europe
Y = 0.48(X) - 2.17
(r2 = 0.69; p <0.001)
Y: N leaching (kg N/ha/yr)
X: total inorganic N deposition
(kg N/ha/yr)
Dise and
Wright (1995);
Lawrence
(1995,2007)
Forest 2-8
*wet deposition only Nutrifnt ,
enrichment
Acidification/ %N in the Oa horizon
12 red spruce
stands across
the northeastern
U.S.
Y = 0.097(X) + 1.03 (r2 = 0.63) Driscoll et al.
Y: N content in the (2001)
Oa horizon (%)
X: wet inorganic N deposition
(kg N/ha/yr)
Forest
0-50
"throughfall N
Acidification/ Soil solution NO3" 104 European
Nutrient monitoring site
enrichment
Conifers: Y = 0.06e° 132
(r2 = 0.59, p = 0.0001)
Deciduous: Y = 0.018e0 266(x>
(r2 = 0.65, p = 0.0001)
Y: NO3" concentration in soil
water (averages Jan 1996 to
Jan. 1998) (mg N/L)
X: throughfall N deposition
(averages 1993-1997)
(kg N/ha/yr)
Gundersen et
al. (2006)
Forest
0^0
"throughfall N
Acidification/
Nutrient
enrichment
Soil solution NO3" 104 European Conifers: log(Y) = 0.06(X)-
monitoring site 1.2 (r2 = 0.59, p = 0.0001)
Broadleaves:
log (Y) = 0.12(X) -1.8
(r2 = 0.65, p=0.0001)
Y: NO3" concentration in soil
water (averages Dec 1995 to
Feb. 1998) (mg N/L)
X: Throughfall N deposition
(averages 1993- 1997)
(kg N/ha/yr)
Kristensen et
al. (2004)
Forest
0-25
*bulk precipitation N
Acidification/ Soil solution NO3" 104 European
Nutrient monitoring site
enrichment
Conifers: log(Y) = 0.09(X) -
1.3 (r2 = 0.32, p<0.0001)
Broadleaves:
log(Y) = 0.09(X) -1.18
(r2 = 0.16, p=0.02)
Y: NO3" concentration in soil
water (averages Dec 1995 to
Feb. 1998) (mg N/L)
X: Bulk
precipitation N deposition
(averages 1993- 1997)
(kg N/ha/yr)
Kristensen et
al. (2006)
3-164
-------
Ecosystems
Deposition
Range
(kg N/ha/yr)
Effect
Indicator
Study Region
Response Curve
Reference
Forest
Relation between
throughfall and bulk
precipitation N input
104 European
monitoring site
Conifers: Y= 1 8(X) - 2.3
(r2=0.65,p <0.0001)
Broadleaves: Y=0.86(X) - 4.9
(r2 = 0.32, p = 0.0008)
Y: Throughfall N deposition
(averages 1993-1997)
(kg N/ha/yr)
X: Bulk precipitation N
deposition (averages 1993—
1997) (kg N/ha/yr)
Kristensen et
al. (2006)
Forest
0^0
"throughfall N
Acidification/
Nutrient
enrichment
Foliage N
104 European
monitoring site
Conifers: Y=0.14(X) +12.7
(r2 = 0.40 , p <0.0001)
Y: Foliage N concentration
(mg/g)
X: Throughfall N deposition
(averages 1993-1997)
(kg N/ha/yr)
Kristensen et
al. (2006)
Forest
0^0
"throughfall N
Acidification/
Nutrient
enrichment
C:N organic layer
104 European
monitoring site
Conifers: Y =-0.21 (X) +31.5
(r2 = 0.19 ; p = 0.0002)
Broadleaves: Y = -0.48(X) +
32.3 (r2 = 0.14; p = 0.036)
Kristensen et
al. (2006)
Y: C:N organic layer
X: Throughfall N deposition
(averages 1993- 1997)
(kg N/ha/yr)
Spruce forest 5 to 30 kg N/ha/yr
Nutrient
enrichment
N2O flux
showed a significant
and positive
correlations between
increasing N
deposition and
increasing N2O flux
2 sites in
Germany and
Ireland
Y = 4.7 +1,4(X)
(r2 = 0.38, p <0.001)
Y = N2O flux rates
(|jg N2O N/m2/h)
X= NH4+ input by wet
deposition (mmol/m2)
Butterbach-
Bahl et al.
(1998)
Spruce forest 5 to 30 kg N/ha/yr
Nutrient
enrichment
NO flux
showed a significant
and positive
correlations between
increasing N
deposition and
increasing NO flux
2 sites in
Germany and
Ireland
Y = 14.1 +16.7(X)
(r2 = 0.67, p <0.001)
Y= NO flux rates (|jg NO-
N/m2/h)
X= NH4+ input by wet
deposition (mmol/m2)
Butterbach-
Bahl et al.
(1998)
Spruce forest 5 to 30 kg N/ha/yr Nutrient
showed a significant
2 sites in
Y = -34.7 + 3.4(X)
Butterbach-
enrichment
and negative
Germany and
(r2 = 0.33, p <0.001)
Bahl et al.
CH4 oxidation
correlations between
Ireland
Y= CH4 oxidation rate
(1998)
rate
increasing N
deposition and
decreasing ChU
oxidation rate
(|jg CH4/m2/h)
X= NH4+ input by wet
deposition (mmol/m2)
3-165
-------
Ecosystems
Deposition
Range
(kg N/ha/yr)
Effect
Indicator
Study Region
Response Curve
Reference
Grassland
5-35
Acidification/
Nutrient
enrichment
Species richness
68 acid
grasslands
across UK
Y = 23.3 — 0.408(X)
(r2=0.70; p <0.0001)
Y: plant species richness
X: total inorganic N deposition
(kg N/ha/yr)
Stevens et al.
(2004)
Lakes
1-16*
*wet deposition only
Nutrient
enrichment
water [DIN]
4296 lakes
across USA,
Canada and
Europe
log(Y) = 1 34log(X) -1.55
(r2 = 0.70; p <0.001)
Y: water DIN (|jg/L)
X: wet deposition (kg N/km2/yr)
Bergstrom and
Jansson
(2006)
Lakes
1-16*
*wet deposition only
Nutrient
enrichment
Chi a: Tot-P
515 lakes
across USA,
Canada and
Europe
Log(Y) = 1,03ln(X)-1.43
(r2 = 0.52; p <0.001)
Y: Chi a: Tot-P
X: wet deposition (kg N/
km2/yr)
Bergstrom and
Jansson
(2006)
Surface water 3.1 to 17.6 *
* net anthropogenic
nitrogen inputs
(NAN I)
Nutrient Riverine exports of 16 watersheds Y = (0.00087(Q)—0.096)
enrichment anthropogenic N from Maine to NANI -101
V|rg|n|a Y: Riverine N flux (kg N
/km2/yr)
Q: riverine discharge
NANI: Net anthropogenic N
inputs (N/km2/yr)
Howarth and
Marino (2006)
Surface water 4-12
Acidification/
Nutrient
enrichment
Surface water [N03~^
220 lakes and
streams across
the northeastern
U.S.
Summer: Y=2.5(X)-14.4 Aberetal.
(r2 = 0.30, p <0.001) (2003)
Spring: Y = 6.7(X)-40.7
(r2 = 0.38, p <0.001)
Y: NO3" concentration in water
([jmol/L)
X: N deposition (kg N/ha/yr)
Stream water 5-13
Acidification/ N export
83 lakes and
Y=0.85(X) -5.8
Aber et al.
Nutrient
streams across
(r2 = 0.56, p = 0.01)
(2003)
enrichment
the northeastern
U.S.
Y: NO3" export (kg N/ha/yr)
X: N deposition (kg N/ha/yr)
Stream water 5-13
Acidification/ Inorganic N retention
83 lakes and
Y=-0.07(X) +1.44
Aber et al.
Nutrient
streams across
(r2 = 0.50, p = 0.01)
(2003)
enrichment
the northeastern
U.S.
Y: inorganic N retentions
(kg N/ha/yr)
X: N deposition (kg N/ha/yr)
Estuary
8-24*
*N loads from
atmospheric depo-
sition, human
waster water and
fertilization
application
Nutrient
enrichment
Eelgrass area (ha) Waquoit Bay
Y = -1,9(X) + 50.7 (r2 = 0.89) Driscoll et al.
Y: Eelgrass area (ha) (2003)
X: modeled N load (kg N/ha/yr)
3-166
-------
Ecosystems
Deposition
Range
(kg N/ha/yr)
Effect
Indicator Study Region Response Curve
Reference
Peatland
0-160
Acidification/
Nutrient
enrichment
NPP of Sphagnum
Aiscum
8 bogs in North
America and
Europe
Y = 831e-0185(x>-48
(r2 = 0.73, p <0.01)
Y: difference between NPP of
S. Aiscum under augmented N
and control treatment
(g/m2/yr);
X: N deposition (kg N/ha/yr)
Vitt et al.
(2003)
Peatland
2-20
Nutrient
Decomposition of litter 12 bogs from 9 Y = 0.98 +0.21 ln(X)
Bragazza et al.
enrichment collected from the field European
and incubated and countries
constant temperature
(r2 = 0.75, p <0.01) four days (2006)
incubation
Y= 0.49 +0.11 ln(X)
(r2 = 0.73, p <0.01) 10 days
incubation
Y: CO2 emission (mg/g/h)
X: N deposition (g/m2/yr)
Peatland
2-20
Nutrient Decomposition of litter 12 bogs from 9
enrichment collected from the field European
and incubated and countries
constant temperature
Y = 4.3 + 2.4ln(X)
(r2 = 0.61, p = 0.01)
Y: DOC concentration (mg/g)
X: N deposition (g/m2/yr)
Bragazza et al.
(2006)
Peatland 2.7 to 8.1 kg N/ha/yr Nutrient
enrichment
Showed N
accumulation peatlands in
increases linearly with Canada
N deposition
23 ombrotrophic Y = 3.50(X) +0.64
(r2 = 0.29, p <0.001)
Y: N accumulation in soil
(g/m2/yr)
X: N wet deposition (g/m2/yr)
Moore et al.
(2004)
3.3.8. Characterization of Sensitivity and Vulnerability
3.3.8.1. Extent and Distribution of Sensitive and Vulnerable Ecosystems
In general, ecosystems that are most responsive to nutrient enrichment from atmospheric N
deposition are those that receive high levels of N loading, are N-limited, or contain species that have
evolved in nutrient-poor environments. Species that are adapted to low N supply will often be more
readily outcompeted by species that have higher N demand when the availability of N is increased (Aerts,
1990; Krupa, 2003; Tilman and Wedin, 1991). As a consequence, some native species can be eliminated
byN deposition (Ellenberg, 1985; Falkengren-Grerup, 1986, 1989; Roelofs, 1986; Stevens etal., 2004).
Note the terms "low" and "high" are relative to the amount of bioavailable N in the ecosystem and the
level of deposition.
The following discussion of sensitive ecosystems is organized into three ecosystem categories:
terrestrial, transitional, and aquatic. Case studies are intended to highlight ecosystems and/or regions
where there are many publications documenting the effects of N deposition, thus they can provide
sufficient data for quantitative risk assessment.
3-167
-------
o
Terrestrial:
3. N enrichment or N saturation (e.g.soil, vegetation, & water are N enriched:
increased fluxes of nitrogenous trace gases)
4. Altered plant communities in response to N enrichment
5. Physiological perturbation of plants: combined effects of ozone and N deposition
6. Impacts on lichen communities.
7. Evidence that threatened and endangered species impacted
8. Decreased diversity of mycorrhizal communities
9. Forest expansion into grasslands (preliminary evidence for)
A Plot with lichen community affected by air pollution with a major N deposition component
High-elevation lake with elevated nitrate, reportedly from N deposition
Available data indicate elevated N deposition, but ecological effects have not been studied.
Source: (Fenn et al., 2003a). Reprinted with permission.
Figure 3-52. Map of the western U.S. showing the primary geographic areas where N deposition effects
have been reported. Eutrophication effects are more widespread and of greater importance
than acidification effects in western North America. Areas where effects of air pollution on
lichen communities have been reported in CAare represented by pink triangles. The plots in
north central CO where lichen community changes were observed are exposed to emissions
of both N and sulfur (S) from two large power plants in Craig, CO and Haydens, CO (Peterson
et al., 2001). The areas shown in red in OR and WA (lichen communities affected by N
deposition) are kriged data (Geiserand Neitlich, 2007). Only lakes at an elevation greater than
1000 m and with a N03 concentration of more than 5 peq/L (measured in fall surveys or on an
annual volume-weighted basis) are shown in thisfgure. Other high-elevation lakes in the West
also had elevated N03 concentrations, but were F excluded because N sources other than N
deposition may have contributed to the elevated concentrations of NO3.
Terrestrial
Most terrestrial ecosystems are N-limited, therefore they are sensitive to perturbation caused by N
additions (LeBauer and Treseder, 2008). Little is known about the full extent and distribution of the
terrestrial ecosystems in the U.S. that are most sensitive to adverse impacts caused by nutrient enrichment
Canada
fblumbia
tiver Gor<
Lake Tahoe
Alaska
1.2,3
(incipient stages), 4
2,
(circumstantial evidence,
text), 8
Ecological effects
Aquatic:
1. N enrichment or eutrophication of lakes
2. Elevated nitrate levels in runoff
River
City (.50,000)
Lake
National Forest/National
Park
3-168
-------
from atmospheric N deposition. Effects are most likely to occur where areas of relatively high
atmospheric N deposition intersect with N-limited plant communities. The factors that govern the
vulnerability of terrestrial ecosystems to nutrient enrichment from N deposition include the degree of N-
limitation, rates and form of N deposition, elevation, species composition, length of growing season, and
soil N retention capacity.
Regions and ecosystems in the western U.S. where N enrichment effects have been documented in
terrestrial ecosystems are shown on Figure 3-52 (Fenn, 2003a). The alpine ecosystems of the Colorado
Front Range (see case study), chaparral watersheds of the Sierra Nevada, lichen and vascular plant
communities in the San Bernardino Mountains (see case study) and the Pacific Northwest, and the
southern California coastal sage scrub community are among the most sensitive terrestrial ecosystems.
In the eastern U.S., the degree of N saturation of the terrestrial ecosystem is often assessed in terms
of the degree of NO -, leaching from watershed soils into ground water or surface water. Stoddard (1994)
estimated the number of surface waters at different stages of saturation across several regions in the
eastern U.S. Of the 85 northeastern watersheds examined, 40% were in N-saturation Stage 0, 52% in
Stage 1, and 8% in Stage 2 (stages are defined in Section 3.3.2.1). Of the northeastern sites for which
adequate data were available for assessment, those in Stage 1 or 2 were most prevalent in the Adirondack
and Catskill Mountains. Effects on individual plant species have not been well studied in the U.S. More is
known about the sensitivity of particular plant communities. Based largely on results obtained in more
extensive studies conducted in Europe, it is expected that the more sensitive terrestrial ecosystems include
hardwood forests, alpine meadows, arid and semi-arid lands, and grassland ecosystems.
0 200 KILOMETCBS
¦ Predominantly wetland
Area typified by a high
density of small wetlands
Source: Data were obtained from the National Land Cover Data (NLCD) (2001) (http://www.mrlc.gov/)
Figure 3-53. Location of wetlands in CONUS.
Transitional
About 107.7 million acres of wetlands are widely distributed in the conterminous U.S., 95 percent
of which are freshwater wetlands and 5 percent are estuarine or marine wetlands (U.S. FWS 2005)
(Figure 3-53). At one end of the spectrum, ombrotrophic bogs are very sensitive to N deposition because
they receive exogenous nutrients exclusively from precipitation, and the species in them are adapted to
3-169
-------
low levels of N (Bridgham et al., 1995, 1996; Shaver and Melillo 1984). Intertidal wetlands are at the
other end of the spectrum; in these ecosystems marine/estuarine water sources generally exceed
atmospheric inputs by one or two orders of magnitude (Morris, 1991). Data are not available with which
to evaluate the extent to which wetlands in the U.S. have been affected by nutrient enrichment from N
deposition. Wetlands are widely distributed, including some areas that receive moderate to high levels of
N deposition.
Peat-forming bog ecosystems are among the most sensitive transitional ecosystems to the effects of
N deposition. In the conterminous U.S., peat-forming bogs are most common in areas that were glaciated,
especially in portions of the Northeast and Upper Midwest (U.S. EPA, 1993a). In Alaska, these
ecosystems are common in poorly drained locations throughout the state.
Nutrient concentrations in wetland waters associated with the Great Lakes suggest that coastal
Great Lakes wetlands are N-limited. Hill et al. (2006) found that more wetlands were N- than P-limited at
each of the five Laurentian Great Lakes. This result is consistent with the apparent N-limitation of most
North American marsh lands (Bedford et al., 1999). Nutrient loading to lakeshore wetlands is a concern
throughout the lower lakes (Lakes Erie, Ontario, and the southern part of Lake Michigan) and in some
localized areas of the upper lakes (Hill et al., 2006). Both agricultural and atmospheric sources of
nutrients contribute to this stress.
Coastal marsh ecosystems, unlike bog ecosystems, often receive large N inputs in tidal water,
groundwater, and surface runoff. Atmospheric inputs to these systems are important because any N
addition has the potential to contribute to eutrophication of coastal marshes and nearby marine and
estuarine ecosystems (Galloway et al., 2003; Paerl, 2002). At many locations, especially along the
Atlantic and Gulf coasts, atmospheric N inputs probably contribute to eutrophication problems in coastal
marshes either by direct deposition to the wetland or from marine water inputs of N that originated from
atmospheric deposition.
Freshwater Aquatic
Aquatic systems in which N has been observed to influence ecological processes either receive
extremely high inputs (e.g., Dumont et al., 2005), or have very low initial N concentrations, and respond
rapidly to additional inputs (Baron et al., 2000; Bergstrom and Jansson, 2006). Eutrophication effects on
freshwater ecosystems from atmospheric deposition of N are of great concern in lakes and streams that
have very low productivity and nutrient levels and that are located in remote areas. In more productive
freshwaters, nutrient enrichment from N deposition usually does not stimulate productivity or community
changes because P is more commonly the limiting nutrient. Also, in many places with even minor levels
of human disturbance, nutrient enrichment with both N and P from non-atmospheric sources is common.
Thus, eutrophication effects from N deposition are most likely to be manifested in undisturbed, low-
nutrient surface waters such as those found in the higher elevation areas of the western U.S. The most
severe eutrophication from N deposition effects is expected downwind of major urban and agricultural
centers.
High concentrations of lake or streamwater NO? . indicative of ecosystem saturation, have been
found at a variety of locations throughout the U.S., including the San Bernardino and San Gabriel
Mountains within the Los Angeles Air Basin (Fenn et al., 1996), the Front Range of Colorado (Baron
et al., 1994; Williams et al., 1996b), the Allegheny Mountains of West Virginia (Gilliam et al., 1996), the
Catskill Mountains of New York (Murdoch and Stoddard, 1992; Stoddard, 1994), the Adirondack
Mountains of NY (Wigington et al., 1996), and the Great Smoky Mountains in TN (Cook et al., 1994). All
of these regions, except CO, received more than about 10 kg N/ha/yr atmospheric deposition of N
throughout the 1980s and 1990s. In contrast, the Front Range of Colorado receives up to about 5 kg
N/ha/yr of total (wet plus dry) deposition (Sullivan et al., 2005), less than half of the total N deposition
received at many of these other locations.
3-170
-------
Table 3-26. Changes in aquatic ecosystems associated with elevated N loadings in the Western U.S.
Ecological or
Environmental
Impact
Location
Level of Uncertainty
Possibility of
Broader
Occurrence (at
other sites)
Reference
EFFECTS IN AQUATIC SYSTEMS
Elevated NO3" in runoff;
most severe in southern
California and in
chaparral catchments in
the southwestern Sierra
Nevada
Transverse ranges of
southern California; low-
elevation catchments in
the Sierra Nevada; high-
elevation catchments in
the Colorado Front Range
Well-documented
response
It is unclear how
widespread this
phenomenon is
outside the
ecosystems listed,
because there is littler
information from low-
elevation systems in
the Sierra Nevada
and elsewhere.
Williams et al. (1996a);
Fenn and Poth (1999);
Fenn et al. (2003a)
N enrichment and shifts
in diatom communities in
alpine lakes
Colorado Front Range;
Lake Tahoe
(California/Nevada
border)
Documented for two
lakes east of the
Continental Divide and
Lake Tahoe
These effects seem
likely in other
N-enriched lakes but
have not been
investigated.
Baron et al. (2000); Wolfe
et al. (2001b); Goldman
(1988).
Reduced lake water
clarity and increased
algal growth
Lake Tahoe
(California/Nevada
border); high-elevation
lakes throughout central
and southern Sierra
Nevada
Well-documented
response;N and P
deposition believed to
be important factors
Lake Tahoe is an
unusual case
because of its
renowned lake clarity;
extent of occurrence
elsewhere in northern
Sierra Nevada is
unknown.
Jassby et al. (1994);
Sickman et al. (2003).
Increased NO3"
concentrations in high-
elevation lakes
Several regions, mainly
downwind of urban
centers
Fairly well established
from lake surveys, but
more data needed for
improved definition of
frequency and severity
Evidence suggests
that urban plumes
and agricultural
emissions affect lake
NO3" levels. There is
also evidence of
impacts on low-
elevation lakes.
Figure 2, Sickman et al.
(2002)
High concentrations of N03 in surface waters in the western U.S. are not widespread. N03
concentrations during the fall sampling season were low in most western lakes sampled in the Western
Lakes Survey. Only 24 sampled lakes were found to have N03 concentrations greater than 10 j^icq/L. Of
those, 19 lakes were situated at high elevation, most above 3,000 m (Eilers et al., 1987). Other effects on
aquatic ecosystems in the west are summarized in Table 3-26.
There is some evidence suggesting that reductions in atmospheric N deposition could decrease the
extent of eutrophication in at least some of the Great Lakes. It has generally been believed that the
Laurentian Great Lakes are P-limited (Downing and McCauley, 1992; Rose and Axler, 1998; Schelske,
1991). Water quality in the open waters of these lakes has been improving in recent years in response to
controls on point sources of P (Nicholls et al., 2001). Work by Levine et al. (1997), however, suggested a
more complicated pattern of response to nutrient addition for Lake Champlain. They added nutrients to in
situ enclosures and measured indicators of P status, including alkaline phosphatase activity and
3-171
-------
orthophosphate turnover time. Although P appeared to be the principal limiting nutrient during summer, N
addition also resulted in algal growth stimulation. P sufficiency appeared to be as common as
P deficiency. During spring, phytoplankton growth was not limited by P, N, or Si, but perhaps by light or
temperature (Levine et al., 1997).
Estuarine and Coastal Aquatic
N is an essential nutrient for estuarine and marine fertility. However, excessive N contributions can
cause habitat degradation, algal blooms, toxicity, hypoxia (reduced dissolved 02), anoxia (absence of
dissolved 02), reduction of sea grass habitats, fish kills, and decrease in biodiversity (Boynton et al..
1995; Howarth et al., 1996; Paerl, 1995; 1997; Valiela and Costa, 1988; Valiela et al., 1990). Each of these
potential impacts carries ecological and economic consequences. Ecosystem services provided by
estuaries include fish and shellfish harvest, waste assimilation, and recreational activities (Costanza et al.,
1997).
M
8
§
o
0
°P
&
S 40
TO
^ 30
4)
0 20
1 10
£
3 o
M
35
14 15
Low ModerateModerateModefate High
low high
200 400
J Kilometers
] Miles
100 200
N
o-
i
High: symptoms occur periodically or persistently and/or over an extensive area
Moderate high: symptoms occur less regularly and/or over a medium to extensive area,
Moderate: symptoms occur less regularly and/or over a medium area.
Moderate low: symptoms occur episodically and/or over a small to medium area.
Low: few symptoms occur at more than minimal levels.
Unknown: insufficient data for analysis.
Change in eutrophic condition since 1999 assessment
A Symptoms improved since 1999 assessment.
O No change in symptoms since 1999 assessment.
V Symptoms worsened since 1999 assessment
~ Insufficient data to show trend
Source: Bricker et al. (2007)
Figure 3-54. Overall eutrophication condition on a national scale.
3-172
-------
Estuaries and coastal waters tend to be N-limited and are therefore inherently sensitive to increased
atmospheric N loading (D'Elia et al., 1986; Howarth and Marino, 2006). The national estuary condition
assessment conducted by Bricker et al. (2007) found that the most impacted estuaries occurred in the mid-
Atlantic region and the estuaries with the lowest symptoms of eutrophication were in the North Atlantic.
N over-enrichment is a major environmental problem for coastal regions of the U.S., especially in the
eastern and Gulf Coast regions. Of 138 estuaries examined by Bricker et al. (1999), 44 were identified as
showing symptoms of nutrient over-enrichment. Estuaries are among the most biologically productive
ecosystems on Earth and provide critical habitat for an enormous diversity of life forms, especially fish.
Of the 23 estuaries examined in the Northeast, 61% were classified as moderately to severely degraded
(Bricker et al., 1999). Other regions had mixtures of low, moderate, and high degree of eutrophication
(See Figure 3-54).
The estuaries with the greatest extent of eutrophication corresponded with conditions related to
both the degree of N loading and the inherent sensitivity of the estuary, as influenced by morphology and
water flushing dynamics (see Annex C for a discussion of estuarine sensitivity). The most eutrophic
estuaries were generally those that had large watershed-to-estuarine surface area, high human population
density, high rainfall and runoff, low dilution, and low flushing rates (Bricker et al., 2007).
Bricker et al. (2007) evaluated the future outlook of the nations estuaries based on population
growth and future management plans. They predicted that trophic conditions would worsen in 48
estuaries, stay the same in 11, and improve in only 14 by the year 2020. Between 1999 and 2007, an equal
number of estuary systems have improved their trophic status as have worsened. The assessed estuarine
surface area with high to moderate/high eutrophic conditions have stayed roughly the same, from 72% in
1999 (Bricker et al., 1999), to 78% in the recent assessment (Bricker et al., 2007).
Studies linking changes in estuary nutrient status to atmospheric N deposition have been limited,
though it is noted that many states are addressing atmospheric inputs as part of their development of Total
Maximum Daily Load plans to address estuarine water quality impairments, including those associated
with low dissolved 02. In an effort to evaluate the contribution of atmospheric N deposition to the future
reduction in N loading to estuaries, Castro and Driscoll (2002) reported model calculations that suggested
that considerable reductions (more than 25%) in atmospheric N deposition will be needed to reduce the
contribution made by atmospheric N deposition to the total N loads to their study estuaries in the
northeastern U.S. A simulated reduction in atmospheric deposition of 25% of ambient deposition rates
reduced the contribution made by atmospheric deposition to the total estuarine N loads by only 1% to 6%
(Castro and Driscoll 2002). In a later study, Driscoll et al. (2003b) estimated that reduction of both mobile
N emissions sources and electric utilities would produce an estimated reduction in estuarine N loading in
Casco Bay, Maine of 13% (Driscoll et al., 2003c). Casco Bay receives the lowest atmospheric and non-
atmospheric N loading per unit area of watershed (4 kg N/ha/yr) of the eight estuaries in the northeastern
U.S. evaluated by (Driscoll et al., 2003a) (Figure 3-35).
3.3.8.2. Case Study: Alpine and Subalpine Communities of the Eastern Slope of the
Rocky Mountains
Some alpine plant communities occur in areas that receive moderately elevated atmospheric N
deposition; especially those proximal to urban areas (see Annex C for a map). Because alpine plant
species are typically adapted to low nutrient availability, they often are sensitive to effects from N
enrichment.
Research on N enrichment effects on alpine and subalpine ecosystems in the western U.S. has been
limited mainly to studies at the Loch Vale Watershed in Rocky Mountain National Park and the Niwot
Ridge Long-Term Ecosystem Research site; both located east of the Continental Divide in Colorado (see
review by Burns 2004). Research has been conducted in this region on both the terrestrial and aquatic
effects of nutrient enrichment.
3-173
-------
Terrestrial Effects
Changes in biomass production and N03 leaching are indicative of effects on the health and vigor
of plants in alpine and subalpine ecosystems. Biomass production responses of alpine communities to
increased N deposition are dependent on moisture regimes (Fisk et al., 1998) and are driven by shifts in
species composition. In a fertilization experiment, the addition of 25 kg N/ha/yr during summer caused a
community shift towards greater dominance of hairgrass (Deschampsia sp.) in wet alpine meadows.
However, the increase in plant biomass (+67%) and plant N content (+107%) following N fertilization
was higher in graminoid-dominated dry meadows than in forb-dominated wet meadows; +53% plant
biomass, +64% standing N crop, respectively (Bowman et al., 1995; Burns, 2004).
Alteration of plant productivity and species richness has been observed in fertilization experiments.
Seastedt and Vaccaro (2001) showed that four years of N addition to alpine vegetation at rates ranging
between 100 and 200 kg N/ha (depending on the year) caused marginal increases in plant foliage
productivity but reduced species richness. In a follow-up study at Niwot Ridge additions of 20, 40, and
60 kg N/ha/yr (on top of ambient N deposition near 5 kg N/ha/yr) over an 8 year period to a dry alpine
meadow led to an increase in plant biomass, and an increase in tissue N concentration at all treatment
levels within three years of application. Much of the response was due to increased cover and total
biomass of sedges (Carex spp ). There was a significant decrease in Kobresia myosuroides with increasing
N input.
High elevation alpine zones exhibit a relatively low capacity to sequester atmospheric deposition of
N because of steep slopes, shallow soils, sparse vegetation, short growing season and other factors (Baron
et al., 1994; Williams et al., 1996b). Results from several studies suggest that the capacity of Rocky
Mountain alpine catchments to sequester N is exceeded at deposition levels less than 10 kg N/ha/yr
(Baron et al., 1994; Williams and Tonnessen, 2000). The changes in plant species that occur in response to
N deposition in the alpine zone can result in further increased leaching of N03 from the soils, because
the plant species favored by higher N supply are often associated with greater rates of N mineralization
and nitrification than the preexisting species (Bowman et al., 1993; 2006; Steltzer and Bowman, 1998;
Suding et al., 2005).
Effects of N deposition to alpine terrestrial ecosystems in this region include community-level
changes in plants, lichens, and mycorrhizae. Alpine plant communities are sensitive to changes in species
composition in response to added N (Bowman et al., 1995; Seastedt and Vaccaro, 2001). Plant species
composition likely responds at lower N input levels than those that cause measurable changes in soil
inorganic N content. For example, Bowman et al. (2006) conducted a N-addition experiment in the
Colorado Front Range with 20, 40, or 60 kg N/ha/yr (see Figure 3-55). Experimental sites were monitored
for 8 years along with a reference site that received about 5 kg N/ha/yr total ambient deposition. Changes
in plant species composition associated with the treatments occurred within 3 years of the initiation of the
experiment, and were significant at all levels of N addition.
Using changes of individual species abundance and ordination scores to evaluate critical load, the
critical load for total N deposition was estimated for change in individual species to be 4 kg N/ha/yr and
for overall community change to be 10 kg N/ha/yr (Bowman et al., 2006). In contrast, increases in N03
leaching, soil solution inorganic N03 . and net nitrification were detectable at levels above 20 kg N/ha/yr
(Bowman et al., 2006). These results indicate that changes in plant species composition may be detectable
at lower N deposition rates than the level at which the traditional soil indicators signal ecosystem
responses to N deposition. This response suggests that changes in species composition are probably
ongoing in alpine dry meadows of the Front Range of the Colorado Rocky Mountains at current
atmospheric N deposition levels. This research also demonstrated that long-term experimental fertilization
plots illustrate a clear response of alpine flora to N addition, including shifts toward graminoid plants that
shade smaller flowering species, and accompanying changes in soil N cycling (Bowman et al., 2006).
Changes in alpine plant species composition have also been documented on Niwot Ridge, where
increased cover of plant species that are most responsive to N fertilization has occurred in some of the
long-term monitoring plots (Fenn et al., 2003a; Korb and Ranker, 2001). These changes have probably
3-174
-------
developed in response to changes in N deposition. However, the influences of climatic change,
particularly changes in precipitation (Williams et al., 1996b) and pocket gopher disturbance (Sherrod and
Seastedt, 2001) could not be ruled out (Fenn et al., 2003a). The altered N cycling provided the potential
for replacement of some native plant species by more competitive, faster growing native species (Baron
et al., 2000; Bowman, 2000; Bowman and Steltzer, 1998).
Aquatic Effects
Rocky Mountain National Park has been the site of research addressing the effects of N deposition
on algal species abundance in freshwater lakes. Wolfe et al. (2001) analyzed sediments from Sky Pond
and Lake Louise, two small alpine lakes located at more than 3300 m elevation on the east slope of the
Colorado Front Range in Rocky Mountain National Park. Before 1900, the diatom flora was typical of
oligotrophic Rocky Mountain lakes, dominated by such species as Aulacoseira distans, A. perglabra,
Fragilaria pinnata, F. construens, and various Achnanthes spp. The mesotrophic planktonic species
Astrionella formosa and Fragilaria crotonensis were present in trace frequencies, but became common
elements of the diatom flora during the 20th century. Between 1950 and 1970, A. formosa became the
dominant taxa in both lakes. It is known from studies in other locations as an opportunistic alga that
responds rapidly to disturbance and nutrient enrichment (Renberg et al., 1993; Anderson et al., 1995;
Reavie and Smol, 2001). This shift in diatom species is apparently the result of environmental
stimulation, rather than recent colonization, as evidenced by the presence of these mesotrophic taxa in the
older sediment record.
50
treatment x year P < 0.01
dP Or*
# &
Hadded
• 0
* 2
Source: Bowman et al. (2006))
Figure 3-55. Changes in plant species composition associated with N addition treatments in an alpine dry
meadow of the Colorado Front Range. Within 3 years of the initiation of the experiment,
statistically significant changes in the cover of Carex rupestris occurred at all treatment
levels.
Additional corroborative evidence for the linkage between atmospheric N deposition and the
observed diatom shifts in these alpine lakes is provided by the results of laboratory (Interlandi and
Kilham, 1998) and in-lake (McKnight et al., 1990) N addition experiments. In both sets of experiments,
growth of A. formosa and F. crotonensis was accelerated by experimental N addition. The post-1950
period of rapid shifts in diatom species composition in Sky Pond and Lake Louise corresponded with
intensification of agricultural practices, animal husbandry, and population growth in adjacent regions to
3-175
-------
the east of Rocky Mountain National Park (Wolfe et al., 2001). Nevertheless, N deposition at that time
was estimated to be low, probably less than 2 kg N/ha/yr (Baron, 2006).
3.3.8.3. Case Study: Chesapeake Bay
Chesapeake Bay is the largest estuary in the U.S. and one of the most sensitive to N inputs (Bricker
et al., 1999; Howarth, 2007). Eutrophication effects have been pronounced in Chesapeake Bay (Howarth,
2007) and it is perhaps the best known example in the U.S. of human activities leading to accelerated
estuarine eutrophication and its associated negative effects. In the recent national assessment of eutrophic
conditions in estuaries, the Chesapeake Bay stands out as a system with both physical features and N
loading levels that make it particularly vulnerable to eutrophication (Bricker et al., 2007).
The role of atmospheric N deposition was not considered as a factor in estuary eutrophication in
the U.S., and was ignored until Fisher and Oppenheimer (1991) suggested that it could constitute up to
40% of the total N inputs to the Chesapeake Bay. Recent studies have reported more conservative
estimates (Boyer et al., 2002); however, Fisher and Oppenheimer (1991) drew attention to the potential
for atmospheric deposition as an important contributor to the overall N budget of estuaries in the eastern
U.S. N inputs to the Chesapeake Bay have increased substantially over the last 50 to 100 years. The
increase is attributed to rapid acceleration of the use of chemical fertilizers in agriculture, the increasing
human population density and associated wastewater discharge, and rising atmospheric N emissions
within the airshed and consequent deposition within the Chesapeake Bay watershed. Atmospheric
deposition of N is currently estimated to contribute about one-fourth of the total N loading to Chesapeake
Bay (Boyer et al., 2002; Howarth, 2007)
Human activities have increased the susceptibility of the Chesapeake Bay to the effects of
atmospheric N deposition. For example, the filling in of wetlands and deforestation for agricultural and
urban development, have reduced the ability of natural ecosystem processes to remove or trap nutrients,
thereby further accelerating nutrient delivery to the bay. In addition, diseases and over-harvesting led to a
dramatic decline of the once highly abundant eastern oyster, seriously reducing the natural filtering of
algae and other organic matter from the water column.
As a result of these changing conditions, eutrophic symptoms intensified in the Chesapeake Bay
from the mid-1950s to the mid-1980s. The most apparent symptoms were high production of algae;
increasingly turbid water; major declines in SAV abundance and species; and increasingly worsening
anoxia and hypoxia (Boesch et al., 2001). The recent national estuary condition assessment (Bricker et al.,
2007) reported that Chi a, dissolved 02, nuisance/toxic algal blooms, and SAV rated "high" in
Chesapeake Bay in terms of severity of effects associated with eutrophication. In addition, macroalgae
and toxic algal bloom conditions have worsened since the previous national assessment in 1999 (Bricker
et al., 1999, 2007).
Concentrations of Chi a in the surface mixed layer have increased tenfold in the seaward regions of
the bay and one-and-one-half- to twofold elsewhere, paralleling estimates of increased loading of N and P
to the bay since 1945 (Harding and Perry, 1997).
SAV began to decline as a result of nutrient enrichment during the mid-1960s, disappearing entirely
from the Patuxent and lower Potomac Rivers. By 1980, many areas of the bay that once contained
abundant SAV beds had none or only very small remnants left (Orth and Moore, 1984). Research
indicated that the major driving factor in the decline of SAV was nutrient enrichment, which was causing
excessive growth of algae in the water column and on SAV leaf blades (epiphytic algae). This algal
growth decreased light availability to the submerged plants to the point that they could not survive (Kemp
et al., 1983; Twilley et al., 1985).
There is an annual cycle of 02 depletion in the Chesapeake Bay that begins as the water starts to
warm in spring, and 02 depletion accelerates during and following the spring freshet. The spring
accumulation of algal biomass is more than sufficient to create conditions for 02 depletion and summer
anoxia (Malone, 1991, 1992). Hypoxia (very low dissolved 02 concentration ~ <2mg/L) and anoxia
3-176
-------
(absence of dissolved 02) generally occur from May through September, with the most severe conditions
observed in mid-summer. Seasonal hypoxia has been a feature of the Chesapeake Bay since deforestation
during the colonial period (Cooper and Brush, 1991; Malone, 1991), but evidence suggests an increase in
the extent of the problem in recent decades (Malone, 1991; Officer et al., 1984). Estuarine eutrophication
is sometimes accompanied by increases in the populations of species of algae, often cyanobacteria that
produce toxins. Such chemicals can affect people, fish, shellfish, and other organisms. Blooms of algae
that produce toxins in Chesapeake Bay have become more extensive over approximately the past decade
(Bricker et al., 2007).
In 1983, the U.S. EPA, District of Columbia, and states of Virginia, Maryland, and Pennsylvania
signed the first Chesapeake Bay Agreement, which established the Chesapeake Bay Program—a
voluntary government partnership that directs and manages bay cleanup efforts. Scientific findings from
the program led to the signing of the second Chesapeake Bay Agreement in 1987, in which it was agreed
to reduce by 40% the N and P entering the Chesapeake Bay by the year 2000. Point source reductions
have been most successful, especially for P. Between 1985 and 1996, emissions from P point sources were
reduced by 58% and N by 15%. Nonpoint source reductions have been slower, largely because nonpoint
sources of nutrients are more difficult to control. Nonpoint source emissions of N and P have been
reduced by only 7% and 9%, respectively (Boesch et al., 2001). Strategies to reduce nonpoint source
nutrients include changes such as adoption of better agricultural practices, reduction of atmospheric N
deposition, enhancement of wetlands and other nutrient sinks, and control of urban sprawl.
3.3.8.4. Case Study: San Bernardino
The San Bernardino Mountains lie east of the Los Angeles Air Basin in California. Pollutants
generated in the greater LA metropolitan area are transported 60-100 km downwind and affect mid-
elevation forests in the San Bernardino Mountains and the San Gorgonio Class I Wilderness area. The
primary source of air pollution is fossil fuel combustion. Approximately half of the air pollution in the LA
air basin is generated from mobile sources including trucks, trains, cars, ships, and buses (South Coast Air
Quality Management District). On the western end of the San Bernardino Mountains, nearly half of the N
deposition is in reduced forms (Fenn and Poth, 2004), most of which is believed to originate from dairy
farms in the Chino/Norco area.
In the San Bernardino Mountains, a wide variety ofN species (NO, N02, HN03, HN02, N03 ) are
deposited to vegetation and soil surfaces, in gaseous, wet, and dry forms. Ammonium (NH4 +) can also be
transported moderate distances from feedlots for cattle and poultry (Bytnerowicz, 2002). Along a west to
east gradient in the San Bernardino Mountains, throughfall N deposition goes from averaging 71 kg
N/ha/yr at Camp Paivika (nearer to the pollutant sources), to 9 kg N/ha/yr at Barton Flats (farther from the
sources), 45 km east of Camp Paivika (Breiner et al., 2007; Fenn et al., 2008). These throughfall
measurements were made using ion exchange resin columns that measure total N03 and NH44" deposition
from precipitation, plus fog or dry deposition that has been scavenged by the overstory pine canopy and
then washed through the canopy (Fenn and Poth, 2004).
Several key ecological endpoints in the mixed conifer forests of the San Bernardino Mountains
have been linked to anthropogenic N deposition. Because air pollution has been high since 1945 in the LA
air basin (Lee et al., 2003), N deposition is in excess of plant and microbial demand (Fenn et al., 1996).
The cardinal symptom of excess N is the export of high N03 levels in streamwater (see 3.3.2.1), which is
well demonstrated for areas with N deposition above 17 kg N/ha/yr in the San Bernardino and San
Gabriel Mountains (Breiner et al., 2007; Fenn and Poth, 1999; Fenn et al., 2008; Michalski et al., 2004;
Riggan et al., 1985, Fenn et al., 2008). Other indicators of excess N in the ecosystem include lowered
litter C:N and elevated emissions of NO and N20 from the soil (see 3.3.4.2), which have been observed at
the more polluted sites in the San Bernardino Mountains. Lichen communities in the San Bernardino
Mountains have also been dramatically changed by the disappearance of up to 50% of the species that
occurred in the region in the early 1900s, due to N pollution (Fenn et al., 1996; Nash and Sigal 1999). A
3-177
-------
disproportionate number of the locally extinct lichen species are cyanolichens. N deposition has also been
tentatively linked to reduction in fine root biomass in Ponderosa pine (Pinus ponderosa C. Lawson) at
three sites (Grulke et al., 1998; Fenn et al., 2008). However, ozone is also believed to contribute to
decreased C allocation to fine roots (Grulke et al., 1998), and could be a confounding factor (see below).
Recently, (Grulke et al., 1998) have used the linkages between N deposition and effects described above
to calculate empirical and simulated (i.e. DayCent, Simple Mass Balance for N as a nutrient) critical loads
for N deposition for California mixed conifer forests.
The effects of high N deposition in the forests of San Bernardino Mountains are compounded by
the high ozone exposures that have occurred throughout the past 65 years. For example, from west to east,
03 concentrations were high at Camp Paivika (80 ppb/h, averaged over 24 h, from April 15 through
October 15, from 1993 through 1995(Grulke et al., 1998); moderately high (72 to 74 ppb/h) 5 km further
east near Rim Forest; and moderate (62-64 ppb/h) 45 km east of Camp Paivika, at Barton Flats. Both 03
exposure and N deposition reduce foliar retention (Grulke and Balduman, 1999) and alter tissue chemistry
of both needles and litter (Poth and Fenn, 1998). In addition, confounding factors such as drought and fire
suppression add to the complexity of ecosystem response (Arbaugh et al., 2003, Minnich et al., 1995;
Takemoto et al., 2001) . Extensive crown injury measurements have also been made, linking ambient 03
exposure data to chlorotic mottle and needle retention (Arbaugh et al., 1998). Ozone exposure and N
deposition reduce carbon allocation to stems and roots (Grulke et al., 1998), further predisposing trees to
drought stress, windthrow, root diseases, and insect infestation (Takemoto et al., 2001). Recently, Grulke
et al. (2008) reported that various lines of phenomenological and experimental evidence indicate that N
deposition and ozone pollution contribute to the susceptibility of forests to wildfire in the San Bernardino
Mountains by increasing stress due to drought, weakening trees, and predisposing them to bark beetle
infestation. Figure 3-56 shows the multiple factors contributing to susceptibility to wildfires in the San
Bernardino Mountains.
R»pM population incrtwt -
Change in taut use
locrssswl swcc-fss
of bilk beetle,
kwre»ed twctpliMliljf
to fire
Continued life suppreaa>on
-WILDFIRE
Source: Grulke et al. (2008 )
Figure 3-56. Diagram of multiple factors contributing to forest susceptibility to wildfire.
3-178
-------
3.3.9. Ecosystem Services
Ecosystem services are the benefits people obtain from ecosystems (Millenium Ecosystem
Assessment, 2005). This concept has gained recent interest and support because it recognizes that
ecosystems are valuable to humans and important in ways that are not generally appreciated (Daily 1997).
It also provides a context for assessing the collective effects of human actions on a broad range of the
goods and services upon which humans rely.
There are no publications at this time that focus on the ecosystem services specifically affected by
N deposition. Some valuation studies address the effects of N enrichment from multiple sources (see
Annex F). The evidence reviewed in this ISA illustrates that N deposition affects ecosystem services in
the following categories (defined by Hassan et al., 2005):
¦ Supporting: nutrient cycling, biodiversity
¦ Provisioning: forest yields, fishing yields in estuaries
¦ Regulating: water quality, air quality, climate regulation (interactions with greenhouse gases C02,
N20, CH4), fire frequency and intensity, disease resistance
* Cultural: swimming, boating, recreation, biodiversity
In general, both ecosystem structure and function play essential roles in providing goods and
services (Table 3-27) (Daily, 1997). Ecosystem processes provide diverse benefits including absorption
and breakdown of pollutants, cycling of nutrients, binding of soil, degradation of organic waste,
maintenance of a balance of gases in the air, regulation of radiation balance and climate, and fixation of
solar energy (Westman, 1977; Daily, 1997; World Resources Institute, 2000). These ecological benefits, in
turn, provide economic benefits and values to society (Costanza et al., 1997; Pimentel et al., 1997). Goods
such as food crops, timber, livestock, fish, and drinking water have market value. The values of
ecosystem services such as flood-control, wildlife habitat, cycling of nutrients, and removal of air
pollutants are more difficult to measure (Goulder and Kennedy, 1997).
EST \
j/
A Drivers &
Disturbance
A Ecosystem
Properties
A Ecosystem
Services
Individual
Decisions
Markets
Management
Regulations
Technology
Envif0nmental Drivers
j Presses and Pulses of Disturbance!
I Nutrient Acid Ozone
I loading Ideposition] exposure
Water
Use
Land
STRUCTURE
Species distribution and abundance
Food Webs Spatial Organization
FUNCTION
Nutrient Cycling, Sort Formation,
Competition, Reproduction, MortaMy
Farm & Forest
Production
Air Quality
Recreation
Water Quality
Climate (GHG)
Regulation
Fisheries
Production Biodiversity
Biogeochemical
cycling
Aquatic
Habitat
Human health
Farm, Fish &
Forest Harvest
Drinking Water
Provision
Swimming and
Recreation
Figure 3-57. Diagram of relationships of human actions, N loading, and ecosystem services.
3-179
-------
Particular concern has developed within the past decade regarding the consequences of decreasing
biological diversity (Ayensu et al., 1999; Chapin et al., 1998; Hooper and Vitousek, 1997; Tilman, 2000;
Wall, 1999). Human activities that decrease biodiversity also alter the complexity and stability of
ecosystems, and change ecological processes. In response, ecosystem structure, composition and function
can be affected (Pimm, 1984; Tilman and Downing, 1994; Tilman, 1996; Chapin et al., 1998; Levlin,
1998; Peterson et al., 1998; Daily and Ehrlich, 1999; Wall, 1999). Biodiversity is an important
consideration at all levels of biological organization, including species, individuals, populations, and
ecosystems. Human-induced changes in biotic diversity and alterations in the structure and functioning of
ecosystems are the two most dramatic ecological trends of the past century (Vitousek et al., 1997b; U.S.
EPA, 2004). The deposition of nutrient N from the atmosphere alters ecosystem structure and function by
altering nutrient cycling and changing biodiversity.
How does N deposition impact ecosystem services? Figure 3-57 illustrates some of the
relationships between human actions, N loading, and ecosystem services. Human actions that affect N
loading influence the disturbances and physical drivers. These changes influence ecosystem structure and
function, which in turn alter the production of ecosystem goods and services. In the bottom row, the
impacts on human benefits or endpoints are shown. These endpoints are highlighted because they are
most easily linked to valuation (Boyd and Banzhaf 2007). Changes in human benefits from ecosystems
can then lead to policy changes, resulting in a feedback between these endpoints and human actions.
Information supporting several of the benefits outlined in Figure 3-57 is developed. These relationships
are futher discussed for some outcomes below, drawing upon the evidence presented in the preceeding
sections.
Human Health. Increasing N deposition has a number of ecological effects that can be linked to
human health. Harmful algal blooms in some areas are linked to increased N loading, and these can cause
bans in swimming as they contain neurotoxins. In addition, links between acid deposition and mercury
mobilization are important for human health in areas where fish, birds and shellfish are consumed.
Swimming and Recreation. Appropriate uses are identified by taking into consideration the use and
value of the water body for public water supply, for protection of fish, shellfish, and wildlife, and for
recreational, agricultural, industrial, and navigational purposes. Increases in N in estuaries can result in
hypoxic zones and fish kills, with numerous negative impacts on commercial and recreation fishing and
shellfish harvest. Harmful algal blooms (e.g., red tide) can be linked to increased N loading, and lead to
swimming bans as they contain neurotoxins. Unpleasant odors and dead fish and wildlife can also reduce
the value of areas for swimming, recreation and even impact property values.
Drinking Water. Forests, wetlands and streams all store and remove nitrogen at various timescales
via plant and microbial uptake, biogeochemical stabilization, and denitrification. This N removal is a
valuable ecosystem service, protecting water quality. In particular, the provision of clear, cool and clean
water as a drinking water supply is an important service provided by forested watersheds, recognized by
many states (e.g., Ashendorff et al., 1997). Air pollution can threaten the provision of this service, by
directly increasing nitrogen loads, some of which will enter the drinking water supply, and also by
affecting forest health, which could in turn decrease water quality and affect other forest ecosystem
services. N inputs can impair the ability of terrestrial and aquatic ecosystems to retain and remove N
(Aber et al., 1989, 1998; Mulholland et al., 2008; see Section 3.3.2), which can lead to degradation of
water quality for many uses.
Forest Supporting and Provisioning Services. Nitrogen limits the primary production of most
terrestrial ecosystems on earth (Vitousek and Howarth 1991; LeBauer and Treseder 2008; see Section
3.3.3). Food and timber harvests have increased tremendously via inputs of properly timed inputs of
nitrogen fertilizers. However, nitrogen additions to agricultural and industrial forest lands typically far
exceed (100 to >300 kg N/ha/yr) the atmospheric deposition of N, which therefore only makes a small
contribution to total N loading. Managed ecosystems are not within the scope of this review. Increases in
availability of this limiting nutrient via atmospheric deposition could increase forest production over large
non-managed areas. Some studies suggest that chronic N additions via deposition can increase production
in forests (Holland et al., 1997; Magnani et al., 2007), but other studies show little effect on wood
3-180
-------
production (e.g., Nadelhoffer et al., 1999) (see Section 3.3.3.1). Leaching of nitrate can promote cation
losses, which in some cases create nutrient imbalances, slower growth and lessened disease and freezing
tolerances for forest trees, particularly in sensitive areas such as high elevation forests. The net effect of
increased N on forests in the U.S. is a key unknown.
Table 3-27. Primary Goods and Services Provided by Ecosystems
Ecosystem
Goods
Services
Coastal
Fish and shellfish
Moderate storm impacts (mangroves, barrier islands)
Ecosystems
Fish meat (animal feed)
Provide habitat and breeding areas/hatcheries/nurseries for wildlife
Seaweeds (for food and industrial use)
(marine and terrestrial)
Salt
Maintain biodiversity
Genetic resources
Dilute and treat wastes
Provide harbors and transportations routes
Provide human and wildlife habitat
Provide employment
Contribute aesthetic beauty and provide recreations
Forest Ecosystems
Timber
Remove air pollutants, emit O2
Fuel wood
Cycle nutrients
Drinking and irrigation water
Maintain array of watershed functions (infiltration, purification, flow
Fodder
control, soil stabilization)
Non timber products (vines, bamboos,
Maintain biodiversity
leaves, etc.)
Sequester atmospheric carbon
Food (honey, mushrooms, fruit, and other
Moderate weather extremes and impacts
edible plants; game)
Generate soil
Genetic resources
Provide employment
Provide human and wildlife habitat
Contribute aesthetic beauty and provide recreation
Freshwater
Drinking and irrigation water
Buffer water flow (control timing and volume)
Fish
Dilute and carry away wastes
Hydroelectricity
Cycle nutrients
Genetic resources
Maintain biodiversity
Provide aquatic habitat
Provide transportation corridor
Provide employment
Contribute aesthetic beauty and provide recreation
3-181
-------
Ecosystem
Goods
Services
Grassland
Ecosystems
Livestock (food, game, hides, and fiber)
Drinking and irrigation water
Genetic resources
Maintain array of watershed functions (infiltration, purification, flow
control, and soil stabilization)
Cycle nutrients
Remove air pollutants and emit O2
Maintain biodiversity
Generate soil
Sequester atmospheric carbon
Provide human and wildlife habitat
Provide employment
Contribute aesthetic beauty and provide recreations
Source: World Resources Institute (2000).
3.4. Other Welfare Effects
This section includes the non-acidification effects of sulfur and direct phytotoxic effects of gas-
phase NOx and SOx on vegetation. Materials and structures damage caused by NOx and SOx are
addressed in Annex E.
As discussed in Section 3.2, a number of environmental effects are associated with S deposition, in
particular soil and water acidification. However, S deposition also contributes to nutrient enrichment,
toxicity, and has secondary effects on the cycling and bioavailability of Hg, a highly neurotoxic
contaminant. High concentrations of S02 can harm vegetation by causing foliar injury, decreasing plant
growth, and eliminating sensitive plant species, although atmospheric concentrations of S02 are seldom
high enough to cause these effects on vegetation at ambient air pollution levels in the U.S. The
biogeochemical cycling of S is closely linked with the cycling of other important elements, including C,
N, P, Al, and Hg. Therefore, S deposition can influence the cycling of these elements in ways that
influence nutrient availability or contaminant toxicity. In particular, current research suggests that S
deposition influences the cycling of Hg in transitional and aquatic ecosystems by stimulating
S042 -reducing bacteria, which are responsible for the bulk of Hg methylation, a key process that
increases the bioavailability of Hg.
3.4.1.1. Biological Role of Sulfur
Effects on Plants
S is an essential plant nutrient. Low dosages of S serve as a fertilizer, particularly for plants
growing in S-deficient soil (Hogan et al., 1998). A certain level of foliar S042ls necessary for adequate
plant S nutrition (Johnson and Mitchell, 1998; Marschner, 1995), and S deficiency has been shown to
occur at foliar S042 levels below 80 jj.g/g in Pinus radiata (Turner and Lambert, 1980). Nevertheless, the
annual increment of S in vegetation is usually small compared to atmospheric deposition and leaching
fluxes. Plants require similar levels of S and P, but S is generally available in much higher concentrations
3.4.1. Non-Acidification Effects of Sulfur
3-182
-------
in soil. Storage of S in vegetation is of minor significance in the retention or loss of S in most forests
(Johnson and Mitchell, 1998; Mitchell et al., 1992a, 1992b).
Atmospheric deposition is an important component of the S cycle. This is true not only in polluted
areas where atmospheric deposition is very high, but also in areas of low S deposition. Biochemical
relationships between S and N are involved in plant protein synthesis and metabolism. S deficiency
reduces N03 reductase and glutamine synthetase activity. N uptake in forests, therefore, could be loosely
regulated by S availability, but S042 additions in excess of needs do not necessarily lead to injury (Hogan
et al., 1998; Turner and Lambert, 1980). Current levels of S deposition throughout much of the U.S.
exceed the capacity of most plant communities to immobilize the deposited S (Johnson, 1984; Lindberg,
1992). S excesses associated with acidic deposition have been found (Johnson et al., 1982a; Meiwes and
Khanna, 1981; Shriner and Henderson, 1978).
S deficiency in forest soil is rare, but has been reported in remote areas that receive very low levels
of atmospheric S deposition and that have inherently low S levels in soil (Kelly and Lambert, 1972;
Schnug, 1997; Turner et al., 1977, 1991). In such cases, atmospheric S deposition might be taken up by
vegetation, with little S042 leaching. Within areas of the U.S. influenced by acidic deposition, this is not
expected to be a common phenomenon. To some extent, plant uptake of S is determined by the
availability of N. This is because most S in plant tissue is in protein form, with a specific S:N ratio
(Johnson et al., 1982a; Turner et al., 1977, 1991).
S plays a critical role in agriculture, and is an essential component of fertilizers (Ceccotti and
Messick, 1997). It is particularly important for plants growing in S-deficient soil (Hogan et al., 1998). The
most important source of S to vegetation is SO/~, which is taken up from the soil by plant roots
(Marschner, 1995). There are few field demonstrations of foliar S042 uptake (Krupa and Legge, 1986,
1998, 1999; U.S. EPA, 2004). Rather, S042 in throughfall is often enriched above levels in precipitation.
The relative importance of the contribution of foliar leachate versus prior dry-deposited S042 particles to
this enrichment is difficult to quantify (Cape et al., 1992). The major factor controlling the movement of S
from the soil into vegetation is the rate of release through microbial decomposition of S from organic to
inorganic forms (Marschner, 1995; May et al., 1972; U.S. EPA, 1982c, 1993b).
S deposition can also have direct effects on plants via nutrient enrichment pathways. Sulfur is an
essential nutrient for protein synthesis in plants. Adequate S supply for sustaining plant health is 0.01% to
0.05% in soils (Nriagu, 1978). S042 is the dominant form of bioavailable S in soils. Plants can also
utilize volatile S compounds such as S02 in the atmosphere to fulfill nutrient requirements
(Rennenberg, 1984). This S is directly available for diffusive uptake through the leaf surface to support
plant growth (Jager and Klein, 1980), and can also become bioavailable in the soil for plant root uptake
(Moss, 1978). However, excess S inputs via atmospheric deposition can be toxic to plants and result in
delayed flowering, reduced growth, and mortality (Rennenberg, 1984; Roelofs, 1991; Smith, 1981;
Smolders and Roelofs, 1996). Plants that have exhibited reduced growth due to S toxicity have also been
observed to have reduced molybdenum (Mo) uptake and increased copper (Cu), manganese (Mn), and
zinc (Zn) uptake (Gupta and Mehla, 1980; Munro and Gupta, 1969). The threshold level of S toxicity is
variable among species (Mudd and Kozlowski, 1975).
Koch et al. (1990) found that hypoxia and high levels of sulfide (>1 mM) limited wetland plant
growth by inhibiting nutrient uptake. Sulfide toxicity to plants (e.g., Carex spp. Juncus acutiflorus,
Galium palustre, Gramineae) has also been observed in wetland mesocosm experimentally enriched with
S042 (Lamers et al., 1998). Biomass regrowth was significantly reduced for these species for both 2 and
4 mmol/L S042 treatments (Lamers et al., 1998). Van der Welle (2007) also showed that increased S042
loading had negative effects on aquatic macrophytes (Stratiotes aloides and Elodea nuttallii), via sulfide
toxicity. Though S. aloides was native to the study region (The Netherlands) of Van der Welle (2007), it is
considered a noxious invasive plant in the U.S. However, E. nuttallii is native to the U.S., widely
distributed across 33 states, and is considered threatened in Kentucky and a species of concern in
Tennessee (http ://plants .usda. gov/'). Negative impacts from elevated rates of atmospherically deposited
S042 on this species could be of concern. It is important to note, however, that the S042 concentrations
3-183
-------
reported in these studies were much higher than would generally be expected to occur in regions of the
U.S. exposed to elevated atmospheric S deposition.
Fe concentrations can influence the level of sulfide toxicity in wetland sediments (Lamers et al.,
2002; Smolders et al., 2001). Free sulfide produced through S042 reduction is able to bind with Fe,
forming insoluble Fe sulfide (FeS). If sufficient Fe is present, this complexation can reduce or eliminate
sulfide toxicity to plants by removing the free sulfide from solution. Van der Welle (2007) confirmed the
role of Fe in buffering against sulfide toxicity to plants by observing no toxic effects when sufficient Fe
was available to precipitate free sulfide.
However, the formation of FeS can disrupt, or compete with, Fe phosphate (FeP04) complexation,
resulting in P release and potential undesirable eutrophication effects on downstream receiving waters
(Caraco et al., 1989; Smolders et al., 2003). Iron(III) hydroxides and iron(III) phosphates are reduced in
anaerobic soils and highly insoluble FeS is formed, increasing phosphate (P043 ) mobility and
bioavailability in surface waters (Smolders et al., 2006). This process has been termed "internal
eutrophication" since P is mobilized from within the system and is not contributed from an external
source (Roelofs, 1991). Increased nutrient availability via S042 -induced P release from wetland
sediments can result in changes in aquatic vegetation community composition. Rooted aquatic
macrophytes can be out-competed by non-rooting floating species and filamentous algae (Smolders et al.,
2003). If Fe is available in high enough concentrations, it can prevent P release from saturated soils with
high S loading by providing adequate Fe to bind with sulfide without releasing P (Van der Welle, 2007).
The observation that N03 addition decreases P release in wetland enclosures provides further
indication that S-induced P release is related to redox conditions and microbial dynamics in the soil
profile (Lucassen et al., 2004). Sufficiently high N03 concentrations can prevent S042 reduction, and
subsequent interruption of Fe-P binding, by maintaining redox status above that suitable for S042
reduction (Lucassen et al., 2004). In the absence of a sufficient supply of N03 to act as a redox buffer,
S042 will undergo reduction and potentially trigger the internal eutrophication mechanism described
above (Lucassen et al., 2004). It is important to note that most research on the topic of internal
eutrophication of has occurred in Dutch peatlands that have historically experienced much larger N
loading than those in the U.S., making it difficult to extrapolate these findings to U.S. systems.
Effects on Methane-producing Microbes
Increased atmospheric S deposition and its impacts on microbial community structure can also
affect CH4 emissions from saturated soils. Early investigation into the effects of elevated pore water S042
concentrations on CH4 emissions from wetland soils involved the application of a single large dose of
S042 (Fowler et al., 1995). CH4 production was observed to be suppressed (40% less than the control)
three weeks after the addition of S042 . This was followed by a 4-week recovery period, after which CH4
production had returned to pre-treatment levels. These results led to the hypothesis that large single
addition of S042 , as applied by Fowler (1995), only stimulate S042 -reducing bacterial (SRB) activity for
a short time. Fowler concluded that studies that more closely approximated long-term S042 loading, as
with atmospheric S deposition, were necessary. Dise and Verry (2001) and Gauci et al. (2002) showed that
smaller and more numerous S042 additions sustained CH4 emission suppression in wetland soils. These
studies more closely approximated S042 enrichment associated with acidic deposition. These results
provided support to the hypothesis that continuous elevated S042 deposition, as encountered in areas
affected by acidic deposition, contributes to sustained suppression of CH4 emissions from wetland soils.
Gauci et al. (2004) considered both methods of S042 addition (a single large dose versus numerous
small doses) in the same experiment. Rates of S042 addition ranged between 15 and 100 kg S/ha/yr to
wetland soils previously exposed to 4 kg S/ha/yr of atmospheric S deposition. They observed that CH4
emissions from these wetland soils were almost equally suppressed under each treatment, and that each
treatment experienced the same CH4 emission "recovery" as found in the single-dose Fowler et al. (1995)
study. The two main conclusions from Gauci et al. (2004) were that 15 kg S/ha/yr is either at or above the
rate of S deposition required to achieve maximum CH4 emission suppression, and that a single large dose
3-184
-------
of S has similar effects on CH4 emission suppression as do numerous smaller doses. The authors observed
that CH4 emissions from treated soils recovered to levels that were observed from untreated soils during
the period of plant senescence. This led to the hypothesis that SRB will out-compete CH4 producing
bacteria under conditions of elevated S deposition and during vigorous plant growth when available C
substrate is limited, but that root exudates and root degradation during the period of plant senescence
provides adequate substrate to sustain both methanogenic and SRB bacteria populations. Although the
suppression of CH4 emissions can fluctuate based upon plant growth cycles, elevated S deposition is
considered to shift microbial community structure in favor of SRB over methanogenic bacteria, reducing
annual CH4 emissions from saturated soils (Granberg et al., 2001). However, climate change simulations
suggest that increased soil temperature may override the suppressive effect that elevated S deposition has
on CH4 emissions (Gauci et al., 2004; Granberg et al., 2001).
3.4.1.2. Cycling and Storage of Sulfur
Terrestrial Ecosystems
Considerable effort was devoted in the 1980s to the computation of S budgets for watersheds and
forest plots, with the objective of evaluating S retention and release. These budgets were subject to
complications from fluxes that could not be measured directly, such as dry deposition and weathering, but
they generally indicated net S retention at sites south of the line of glaciation — a result attributed to net
adsorption of S042 (Cappellato et al., 1998; Rochelle and Church, 1987). During the 1990s, little or no
decrease in S042 concentration occurred in streams in the Ridge and Blue Ridge physiographic
provinces, despite regional decreases in atmospheric deposition of S (Webb et al., 2004), and no evidence
of S addition from mine drainage. This lack of response in stream chemistry has been generally attributed
to a shift in S equilibrium between the adsorbed and solution phases under conditions of decreased
atmospheric inputs of S042 . This interpretation is supported by a decrease in concentrations of adsorbed
S042 from 1982 to 1990 in a Piedmont soil in South Carolina that received decreasing levels of S
deposition during this period (Markewitz et al., 1998). This same soil also experienced an increase in
adsorbed S042 from 1962 to 1972 (Markewitz et al., 1998). The only published S budget more recent
than 1992 for an unglaciated site in the U.S. (Castro and Morgan, 2000) also suggested a net release of
S042 . This upland Maryland watershed released 1.6 times more S042 than measured in throughfall in
1996-97.
Numerous S budgets were compiled in the 1980s for glaciated sites, and results generally indicated
that inputs approximately equaled outputs on an annual basis (Rochelle and Church, 1987). The
observation of little or no S retention at glaciated sites was attributed to relatively low S042 adsorption
capacity in soils. Balanced S budgets in glaciated regions implied that decreases in atmospheric
deposition of S would lead directly to decreases in S042 leaching. The strong correlation between recent
decreases in both atmospheric S deposition and S042 concentrations in surface waters is widely
recognized to be a result of this direct linkage (Stoddard et al., 2003). Nevertheless, considerable evidence
also indicates that S inputs in glaciated ecosystems do not behave conservatively, but instead are cycled in
part through microbial and plant biomass (Alewell and Gehre, 1999; David et al., 1987; Likens et al.,
2002). As a result, large quantities of S are stored in organic forms within the soil. David et al. (1987)
found that annual S deposition (wet plus dry) at a site in the central Adirondack region of New York was
about 1% of the organic S pool in the soil. Houle et al. (2001) estimated that annual S deposition at 11
sites in North America ranged from 1% to 13% of the organic S pool in soil.
The S cycle in forest ecosystems can be represented as a series of input, uptake, and output terms
(Figure 3-58). Some of the fluxes illustrated in this schematic drawing can be measured in the field,
including wet deposition, litterfall, and throughfall. Other fluxes must be calculated or estimated, which
involves considerable uncertainty (Johnson and Mitchell, 1998). Perhaps the most important uncertainty
concerns the amount of dry deposition, which can be substantial (Lindberg et al., 1990).
3-185
-------
WET DEPOSITION
TRANSLOCATION
VEGETATION
DRY DEPOSITION
LITTERFALL
FOLIAR LEACHING
ROOT DEATH
AND DECAY -
FOREST
FLOOR
REDUCTION AND
VOLATILIZATION
DECOMPOSITION
UPTAKE
MINERALIZATION
SOIL
ORGANIC
IMMOBILIZATION
LEACHING
~ so42- -S
I I ORGANIC S
I I INORGANIC S MINUS SO„2' -S
SOIL
MINERAL
ADSORBED
. SQ42- .
SOIL N
SOLUBLE
S042" J
—THROUGHFALL
SOd
Source: Johnson (1984)
Figure 3-58. Representation of the S cycle in forest ecosystems.
Atmospheric deposition is an important part of the S cycle, including in areas that are not exposed
to appreciable air pollution levels. In fact, although agricultural S and geologic S (especially associated
with mining activities) can be locally important or dominant, atmospheric S inputs may constitute the
major source of S input to many terrestrial ecosystems (Johnson and Mitchell, 1998; Probert and Asmosir,
1983).
Much of the organic S stored in soil is in C-bonded forms that are relatively unreactive, but can be
oxidized by bacteria or mineralized to S042 under oxic conditions, which are typically found in
moderately well drained to well drained soils (Johnson and Mitchell, 1998). Carbon-bonded S in forest
soils can be found in a variety of organic S compounds, including amino acids, sulfolipids, and sulfonic
acids. Carbon-bonded S can also be found in humic material in the form of aliphatic and aromatic
structures (Likens et al.. 2002). Furthermore, strong correlations have been shown between levels of
atmospheric deposition of S and concentrations of S in soil (Driscoll et al., 2001b; Novak et al., 2005).
Long-term increases in concentrations of total S in soils that are at least partially attributable to increases
in organic S have also been documented (Knights et al., 2000; Lapenis et al., 2004), although the study of
Houle et al. (2001) did not find a relation between these factors. A Swedish "clean roof' study also
provides some insight into the role of organic S in possibly delaying chemical recovery from acidification
due to S deposition (Morth et al., 2005). After 9 years of application of pre-industrial levels of
3-186
-------
S deposition, the amount of S042 in runoff still exceeded inputs by 30%. Most of the S in runoff was
attributed to mineralization of organic S in the O horizon.
Thus, research on the effects of atmospheric S deposition on soils has indicated pronounced
changes in soils from sustained S042 leaching, and accumulation of S through physical/chemical
adsorption and biological assimilation. The recent evidence of net loss of S from soils at a number of sites
is likely a response to decreased atmospheric inputs. The gradual loss of previously accumulated S
contributes to continued S042 leaching. Uncertainties in estimates of ecosystem fluxes such as
weathering and dry deposition, and complications in discerning the effects of desorption from
mineralization make it difficult to predict when S outputs will no longer exceed inputs as levels of S
deposition continue to decline. Research based on experimental reduction of S inputs suggests that this
process will occur on a decadal time scale (Martinson et al., 2005; Morth et al., 2005). The long-term role
of C-bonded S adds further uncertainty because enhancement of S mineralization by a warming climate
could also affect S retention and release (Driscoll et al., 2001b; Knights et al., 2000). This process can be
microbially catalyzed, and bacteria are generally more active at higher temperature.
Transitional Ecosystems
Transitional ecosystems exert important controls on watershed S budgets, especially in watersheds
that contain extensive wetland development. Sulfur storage in wetland soils provides an important
buffering system that restricts chronic S042 leaching to surface waters. Input-output studies of bogs in
Massachusetts (Hemond, 1980),Ontario (Urban and Bayley, 1986), and Minnesota (Urban and Eisenreich,
1988) suggested more than 50% retention of atmospheric S inputs. However, oxidation of S that was
previously stored in wetland soils can provide an important episodic source of S042 to downstream
surface waters. Thus, the presence of wetlands in a watershed can either temporarily increase or decrease
the flux of S042 to surface waters, and these differences are largely determined by changes in hydrology
and redox conditions in wetland soils. Overall, wetlands act as sinks for S because of microbial S042
reduction and sequestering of reduced S as sulfide minerals and organic S.
Changes in S flux that are controlled by processes in transitional ecosystems can have important
effects on surface water chemistry. For example, reduction of S042 in sediments by assimilatory and
dissimilatory processes is an important source of acid neutralizing capacity (ANC) to lakes having long
hydraulic residence time, and a likely source also to beaver ponds and wetlands. In-lake ANC production
is mostly due to S retention from microbial S042 reduction (Brezonik et al.; Schindler and Bayley, 1993;
Turner et al., 1990b). It is unlikely, however, that the changes in S flux caused by wetlands and ponds in a
watershed would be large enough to have any direct non-acidification effects on biota. More likely, the
major non-acidification effects of wetland influence on S cycling relate to changes in Hg methylation in
wetland soils. This is discussed in Section 3.4.1.4. Other changes can also occur, including enhanced
release of N and P from wetland soils.
Some of the organic S in wetlands can be converted to reduced S gasses, including dimethylsulfide
and hydrogen sulfide (under acidic conditions), and released to the atmosphere. Up to 30% of the
atmospheric deposition of S in remote areas may be derived from release of reduced S gasses from
wetlands (Nriagu et al., 1987). Thus, wetland processes can have important effects on local atmospheric S
deposition and trace gas emissions.
Aquatic Ecosystems
In aquatic ecosystems that are sensitive to acidification from atmospheric S deposition, S042 is
generally highly mobile within the ecosystem. Acid-sensitive streams tend to be relatively fast-flowing,
high-gradient, low-order streams that exhibit high S042 mobility. Acid sensitive lakes tend to be
relatively small, headwater lakes with short hydraulic residence times (weeks to months). In such streams
and lakes, most of the S042 contributed by inflowing ground and surface waters is directly flushed
through the ecosystem and emerges as outflow.
3-187
-------
However, larger streams, streams that flow through series of ponds (i.e., beaver ponds) or lakes,
and especially larger lakes, tend to have longer hydraulic residence, and provide opportunity for microbial
S reduction in sediments. This S reduction can have important effects on the concentration of S042 in
drainage water, and results in the generation of ANC. The importance of sediment reactions to the acid-
base chemistry of surface water depends mainly on the flux rate of material across the sediment-water
interface and the amount of time that water remains in contact with the sediment (Baker and Brezonik,
1988; Kelly et al., 1987; Turner et al., 1991). In some lakes having long water residence times, about half
of the input S042 is retained in lake sediments (Baker and Brezonik, 1988; Kelly et al., 1987).
S is an essential nutrient for algae and planktonic bacteria. Nevertheless, S concentration in most
lakes is well above the limiting concentration for algal productivity, and therefore biotic S uptake in the
water column is not a quantitatively important part of the S cycle in acid-sensitive lakes (Turner et al.,
1991).
3.4.1.3. Export of Sulfur
Terrestrial Ecosystems
In order for atmospherically deposited S to exert influence on drainage water, with the range of
associated environmental effects that can occur, it must be exported as S042 from the soil. If the
incoming S in atmospheric deposition is retained in the vegetation or soil compartments, it will not be
available to affect soil water or surface water downstream within the watershed. In areas of S deposition,
almost all deposited S moves into the soil and can then be exported from the terrestrial ecosystem or
adsorbed on soil. In most parts of the U.S., most deposited S is exported in drainage water. In much of the
southeastern U.S., however, S adsorption on soil substantially limits S export (See discussion in
Annex B).
Transitional Ecosystems
When saturated, wetland soils act as sinks for incoming S via S042 reduction. Sulfide is produced
through this process and sequestered in anoxic wetland sediments (Mitsch and Gosselink, 2000).
However, it has been observed that wetlands can act as sources of S042 to downstream drainage waters
during storm events that follow prolonged periods of drought (Dillon and LaZerte, 1992; Devito and Hill,
1999; Eimers and Dillon, 2002; Jeffries et al., 2002; Laudon et al., 2004; Mitchell et al., 2006). The
mechanism has been described as follows. S042 is produced through oxidative processes in wetland
sediments when they are exposed to atmospheric 02 as the water table falls during periods of drought.
This newly formed S042 is mobile, and therefore can be flushed from the wetland into streams or lakes
when the water table rises as more typical hydrologic conditions resume. This flush of S042 can result in
episodic acidification of downstream surface waters (Laudon et al., 2004) and potentially prolong the
chemical recovery of surface water ANC as S deposition declines (Aherne et al., 2006).
Much of the supporting research on this topic has been performed within the boreal watersheds of
Ontario, Canada (Dillon and LaZerte, 1992; Devito and Hill, 1997; Jeffries et al., 2002; Aherne et al.,
2004; Laudon et al., 2004). A Sphagnum-conifer wetland within the Plastic Lake watershed in Ontario
was determined to be a source of S042 to downstream drainage waters after extended periods of
summertime drought (Dillon and LaZerte, 1992). Comparisons of stream water chemistry were made
between the wetland inlet and the wetland outlet, which drains a watershed consisting entirely of upland
soils. The results showed little difference between S042 concentrations in the wetland inlet and outlet
during typical hydrologic conditions. However, S042 concentrations in the outlet increased by up to a
factor of 5 during storm events that followed extended periods of drought. This occurred during 4
separate years.
Most the study watersheds in the Plastic Lake region of Ontario, Canada have consistently exported
more S042 than was atmospherically deposited on an annual basis over an 18-year period (Eimers and
3-188
-------
Dillon, 2002). This observation suggests either the existence of an internal watershed S042 source, or an
underestimation of S deposition. It is possible that dry deposition is underestimated (Likens et al., 1990;
Edwards et al., 1999), but a variety of potential watershed sources of additional S have also been
proposed in areas that are sensitive to atmospheric S deposition, including:
¦ weathering of S-containing minerals (Baron et al., 1995)
¦ desorption of S042 previously adsorbed to soils when S deposition was higher (Driscoll et al.,
1995; (Mitchell etal., 1996)
¦ mineralization of S previously incorporated into organic matter (Driscoll et al., 1998)
¦ drought-related oxidation and release of S stored in wetlands and riparian soils (Dillon and
LaZerte, 1992; Dillon et al., 1997)
Underestimation of dry deposition was not considered to be a significant issue for the Canadian
study watersheds (Eimers and Dillon, 2002). Furthermore, mineral weathering is not considered a
significant source of S042 in that region, due to the low S content of the bedrock (Neary et al., 1987).
Reoxidation and mobilization of S stored in wetland sediments was considered the most likely
explanation for the observed higher S042 outputs for those watersheds that contain a significant
proportion of wetland. Other mechanisms, including increased soil S042 desorption and/or increased S
mineralization in response to decreased S deposition inputs, may explain the S042 input/output
imbalance observed in watersheds containing little or no wetland area (Alewell and Gehre, 1999; Eimers
and Dillon, 2002). Jeffries et al. (1995, 2002) determined that within the Turkey Lakes watershed in
western Ontario wetland, reoxidation and S042 remobilization mechanism can delay lake acidification
recovery by as much as 6 years.
Wetland S transformations have been incorporated into state-of-the-science modeling to better
describe climate-induced acidification effects on lake water chemistry (Aherne et al., 2004, 2006). A
wetland component to the MAGIC model was developed and tested for its ability to predict observed
stream water S042 fluxes from the Plastic Lake watershed (Aherne et al., 2004). This model was then
used to investigate acidification recovery under two different climate scenarios: an "average climate"
scenario consisting of long-term (most recent 20 years) monthly precipitation and runoff;and a "variable
climate" scenario that included sequential repetition of the measured monthly precipitation and runoff for
the preceding 20 years. The average climate scenario did not include any significant drought periods,
whereas the variable climate scenario included several periods of summer drought.
Model results under the average climate scenario suggested that chemical recovery of lake water
would occur, with ANC reaching 40 (ieq/L by 2020 and 50 j^ieq/L by 2080. However, the variable climate
scenario projected that recovery would be greatly reduced. ANC recovery by 2080 was estimated to only
reach 2.6 |_ieq/L. The authors acknowledged that reiterating the past 20 years of climate under the variable
climate scenario was somewhat arbitrary. Nevertheless, results suggested that climate effects on the
cycling of S can modify chemical recovery of lake water from acidification in watersheds that are
wetland-influenced.
Aquatic Ecosystems
Export of S from surface waters is controlled primarily by retention in sediments through microbial
S042 reduction. In-stream and in-lake biological demand for S is generally a very small component of the
S input levels in areas affected by atmospheric S deposition. Sulfur reduction can be an important process
regulating S export from aquatic ecosystems, mainly in waters that exhibit long hydraulic retention.
Sulfur reduction in lake and pond sediments can also be closely associated with Hg methylation.
Therefore, the dynamics of S storage and export can influence the bioavailability of Hg to fish,
piscivorous wildlife, and humans who consume large quantities of fish.
3-189
-------
3.4.1.4. Sulfur and Methylation of Mercury
Hg has long been established to be a potent neurological, reproductive, and developmental toxin
that accumulates at progressively higher concentrations in higher trophic levels (biomagnification). The
biogeochemical cycle of Hg is closely tied to that of S, and the presence of S042" in wetlands and lake
sediments is essential for entry of Hg into the food web. Hg is taken up by living organisms and
bioaccumulates in the MeHg form. For the protection of human health, the U.S. EPA set the fish tissue
criterion for MeHg at 0.3 jxg/g. This has resulted in 2,436 fish consumption advisories for Hg in 2004,
2,682 in 2005 and 3,080 in 2006. Forty-eight states, one territory, and two tribes have issued Hg
advisories. Eighty percent of all fish consumption advisories have been issued, at least in part, because of
Hg. Most of the new Hg advisories issued in 2005 and 2006 were in Wisconsin (293), Michigan (46),
New York (36) and Minnesota (32). In 2005, American Samoa, Kansas, Oklahoma and Utah started
issuing Hg advisories, and Iowa started in 2006. In 2006, atotal of 14,177,175 lake acres and 882,963
river miles were under advisory for Hg. As of July 2007, 23 states have issued statewide advisories for Hg
in freshwater lakes and/or rivers.
Examples of elevated tissue and blood Hg have also been reported in terrestrial animals, and a few
of those studies included comparisons between aquatic and terrestrial species. In a survey of four species
of turtles in and around a portion of river with a long-term history of industrial Hg pollution, Bergeron
et al. (2007) observed that species differences in blood Hg level varied with the proportion of the species'
diet that is aquatic. The species with the smallest aquatic component to its diet still had higher blood Hg at
contaminated sites than at non-contaminated sites. However, its blood Hg level at contaminated sites was
approximately 90% lower than the three other species. In non contaminated parts of the river, blood Hg
level was very low for all species, even though it was lower in the most terrestrial one than in the others.
A survey of songbirds on the same 24 km stretch of river (Cristol et al., 2008) found elevated Hg in the
blood of terrestrial birds in proximity to the river, but not at reference sites away from the river, or
upstream from the contaminated portion. In their survey of northeastern freshwater avian species, Evers
et al. (2005) observed that among birds living in ecosystems with a large aquatic component, blood, egg,
and tissue Hg levels were much lower in insectivorous species than in piscivorous ones. Rimmer et al.
(2005) sampled blood and feather samples of insectivorous passerines in their summer and winter habitat
in the northeastern U.S. and southeastern Canada, and Cuba, Haiti and the Dominican Republic. They
found elevated MeHg in birds without clear connection to aquatic food webs. However, the putative role
of S in Hg methylation in terrestrial ecosystems has not been explored.
SRB are the main agent of Hg methylation in transigional and aquatic environments, and changes
in S042 deposition have been shown to result in commensurate changes in both Hg methylation, and Hg
levels in fish.
Effects of Mercury in Aquatic Biota
Adverse effects of Hg, including behavioral, reproductive, neurochemical, and hormonal effects,
have been demonstrated in piscivorous mammals and birds (Scheuhammer et al., 2007; U.S. EPA,
1996b), and MeHg has been shown to be the form in which Hg accumulates in tissue of fish and
piscivorous species (Becker and Bigham, 1995; Bloom, 1992; Harris et al., 2003; Scheuhammer et al.,
2007). Exposure of fish and wildlife to Hg occurs primarily through the diet. Top predatory, especially
piscivorous, animals feeding on aquatic food chains are at greatest risk for Hg accumulation and toxicity
(Scheuhammer et al., 2007). Wildlife living in inland lake habitats tends to accumulate higher tissue
concentrations of Hg than those living in coastal habitats (Evers et al., 2005; Frederick et al., 2002) .
Available data suggest that numerous wild populations of fish, birds, and mammals experience
MeHg exposures that are high enough to cause substantial reproductive, behavioral or health impairment.
Reproduction is the component of response that appears to be most affected (Scheuhammer et al., 2007).
In fish, exposure to MeHg can affect growth, reproductive ability, morphological characteristics, and
feeding efficiency. Examples of studies documenting the effects of MeHg on fish include Friedmann et al.
3-190
-------
(1996), who investigated the effects of low-level (0.137 (.ig Hg/g) and high-level (0.987 (.ig Hg/g) dietary
Hg concentrations (as MeHg) on hatchery juvenile walleye (Stizostedion vitreum). These experimental Hg
exposures were chosen to reflect dietary Hg concentrations commonly encountered in North American
lakes and streams. Results showed impaired fish growth and impaired gonad development in males. Fjeld
et al. (1998) exposed grayling (Thymallus thymallus) embryos to varying concentrations of MeHg (0.16,
0.8, 4.0, and 20 (.ig Hg/L) during their first 10 days of development. This exposure resulted in body tissue
MeHg concentrations of 0.09, 0.27, 0.63, and 3.80 (.ig Hg/g respectively. Morphological deformities were
observed in fish exposed to the highest level of MeHg. Samson and Shenker (2000) also observed
morphological disturbance in zebrafish (Danio rerio) at embryonic MeHg exposure levels of 20 and
30 (.ig CH3HgCl/L. Other fish such as mummichog (Fundulus heteroclitus) and rainbow trout
0Oncorhynchns mykiss) have also been observed to suffer teratogenic effects such as cyclopia, tail
flexures, cardiac malformations, jaw deformities, twinning, and axial coiling from embryonic MeHg
exposure (Samson and Shenker, 2000). Fish survival and subsequent population status can be jeopardized
as a result of exposure to MeHg. Fathead minnows (Pimephales promelas) showed impaired feeding
efficiency after exposure to both 6.79 and 13.57 jj.g HgCWL (Grippo and Heath, 2003). Reduced feeding
efficiency and competitive ability was also observed in grayling exposed to 0.8 to 20 (.ig Hg/L as embryos
(Fjeld etal., 1998).
Role of Sulfur in the Biogeochemical Cycle of Mercury
The global cycle of Hg has atmospheric, aquatic, edaphic, and biotic components. In the
atmosphere, Hg is transported locally, regionally, and globally, depending on speciation. Both elemental
and oxidized forms are found in soil and aquatic environments, but the oxidized form is more prevalent.
MeHg is the form that is found in tissues (Figure 3-59). SRB are the main agent of Hg methylation in the
environment. Although Hg methylation in watersheds has been shown to occur through other processes,
their contribution to MeHg loads is negligible in comparison to that of SRB-mediated methylation.
Addition of S042 has been demonstrated to stimulate Hg methylation by SRB in studies spanning scales
from the culture of isolated bacteria, to the experimental amendment of entire lakes. Those studies have
included addition of S042 at rates corresponding to observed deposition.
surface
biomagnification
*
Hg
methylation
demethylation
Hg-organic/inorganic
complexes
methylation
Sulfate
Hg-S
complexes
reducing
bacteria
Hg cycle
Figure 3-59. Simplified cycle of mercury, showing the role of sulfur. Arrows are not proportional with
actual rates.
3-191
-------
SRB are commonly found in anoxic wetland and lake bottom sediments (Compeau and Bartha,
1985; Gilmour and Henry, 1991; 1992). Under increased S042 availability, their number and activity
increase. The mechanisms for Hg methylation, as mediated by SRB, have been discussed by Choi et al.
(1994), Ekstrom et al. (2003), and Ekstrom and Morel (2004). Abiotic mechanisms responsible for Hg
methylation have been discussed by Weber (1993), Hintelmann and Evans (1997), and Siciliano et al.
(2005). Studies demonstrating the response of SRB-mediated methylation to S042 in pure cultures
include King et al. (2000), and Benoit et al. (2001). This response has also been established in samples of
soil and sediments (Compeau and Bartha, 1985; Gilmour et al., 1992; Harmon et al., 2004), and in
experimental manipulations in wetlands and lakes (Benoit et al., 2003; Branfireun et al., 1999; 2001;
Frost et al., 1999; Harmon et al., 2004; Jeremiason et al., 2006; Watras et al., 2006; Wiener et al., 2006).
Evidence regarding the importance of SRB in MeHg production was provided by Compeau and
Bartha (1985), and Gilmour et al. (1992), who showed that MeHg production was substantially reduced
with addition of a known SRB inhibitor (Na2Mo04). This is also in agreement with observations of SRB-
mediated Hg methylation in salt marsh sediments (Compeau and Bartha, 1985). The work of Gilmour
et al. (1992) considered anoxic lake bottom sediments, rather than wetland sediments. However, anoxia is
also common in freshwater wetland sediments, where S042 addition has also been observed to enhance
Hg methylation (Branfireun et al., 1999; Harmon et al., 2004; Jeremiason et al., 2006). Accumulation of
sulfidic forms of S in sediments, also resulting from SRB activity, has been shown to diminish the
availability of S to SRBs, and thus net Hg methylation (Benoit et al., 1998; 1999a; 2001; Gilmour, 1998;
King et al., 2001a).
150
100
SO
0
Added S04 pM
O-O 0
' A—A 100
T"
10
15
Days
30 35
0}
X
a*
1 !
10 15
Days
20
30 36
Source: Gilmour et al. (1992).
Figure 3-60. (A) S042" and (B) MeHg concentrations as a function of time in sediment slurries made from
Quabbin Reservoir littoral sediments. Each delta point represents the average value from
three separate incubations and the associated standard error.
3-192
-------
Gilmour et al. (1992) investigated MeHg production within anoxic sediments of a reservoir located
in central Massachusetts. Elevated MeHg production with S042 addition was measured in both
experimental laboratory slurries (Gilmour et al., 1992) (Figure 3-60B) and intact sediment cores. The
background S042 concentration in the experimental sediment slurries was 60 |_icq/L. S042 additions of 0,
100, 200, and 400 (ieq/L were applied to these samples in the presence of 50 mg/L of Hg as HgCl2. The
rate of production and the final concentration of MeHg increased in proportion to the initial S042
concentration. Furthermore, S042 concentrations decreased during the experiment (Gilmour et al., 1992),
experiment (Figure 3-60A), suggesting that S042 reduction had occurred. MeHg production within
isolated lake bottom sediment cores was also enhanced across a gradient of S042 addition (3 to 1040
(imol sodium S042 [Na2S04]; Figure 3-61). Sediment MeHg production was most enhanced when S042
concentration was above about 60 (ieq/L, with increased production from a pre-treatment background
MeHg concentration of 0.26 ng/g to approximately 7.0 to 8.5 ng/g. These results suggest maximum
MeHg production at S042 concentrations between about 200 and 400 (ieq/L, although optimal conditions
for methylation are likely to vary with other factors that influence S042 reduction. S042 concentrations
in the range of these experiments (about 60 to 200 j^icq/L) are often found in waters affected by S
deposition in the U.S.
Interacting factors
Many studies have also shown an association between low lake water pH and high Hg
concentrations in fish (Driscoll et al., 1994a; Grieb et al., 1990; Kamman et al., 2004; Suns and Hitchin,
1990). Hrabik and Watras (2002) found that decreases in fish Hg concentration in an experimentally de-
acidified lake basin exceeded those in the reference lake basin by a factor of 2 over a 6-year period of
experimental de-acidification. The association between low pH and high Hg accumulation in fish suggests
a response of methylation to pH, but although SRB activity does respond to pH (Kelly et al., 2003),
quantification of the interactive effects of pH with S042 in the environment has only been tentative.
Other interacting factors, mainly Fe, P, and dissolved organic matter, have been identified, but very
incompletely quantified (Munthe, 2007; Watras and Morrison, 2008). Driscoll et al. (2007b) developed
indicators of Hg sensitivity using two stratified, random-probability surveys of northeastern lakes
combined with the survey data sets of Chen et al. (2005). This analysis showed that lakes with Hg levels
above the U.S. EPA criterion of 0.3 jj.g/g in yellow perch had significantly higher DOC, and lower pH,
ANC, and total P than lakes with fish Hg concentrations below 0.3 jj.g/g (Driscoll et al., 2007b). Based on
the probability surveys, they calculated that about 20% of lakes in the region had total P concentrations
above 30 j^ig/L and yellow perch Hg concentrations below 0.3 jj.g/g. In the remaining 80% of lakes, 75%
had yellow perch Hg concentrations exceeding 0.3 jj.g/g when surface water DOC levels exceeded 4.0 mg
C/L, a pH of less than 6.0, or an ANC of less than 100 (ieq/L. Most Hg in the water column of freshwaters
is bound to organic matter, either to DOC or to suspended particulate matter. Therefore, total Hg and
MeHg concentrations are often positively correlated with DOC in lake waters (Driscoll et al., 1994a;
Mierle and Ingram, 1991; U.S. EPA, 1996b). DOC, in turn, has an important influence on pH. Thus,
several interrelated factors seem to affect Hg loading in tissue. For example, Driscoll (1995) found one or
more yellow perch exceeding the 0.5 jj.g/g action level in 14 of 16 Adirondack study lakes despite wide
ranges in pH (to above 7) and ANC (to above 200 (ieq/L). Driscoll et al. (1994a) concluded that the most
obvious factor regulating the concentration and availability of both total Hg and MeHg in Adirondack
lakes is DOC. They found increased fish Hg concentrations with increasing DOC up to DOC
concentrations of about 8 mg/L, followed by lower concentrations in the highly dystrophic Rock Pond
(DOC = 26 mg/L). They hypothesized that DOC may bind with MeHg at very high DOC concentration,
limiting the bioavailability of the Hg. In addition, calculations made by Driscoll et al. (1995) with the Hg
Cycling Model suggested that increases in DOC result in increasing concentrations of Hg in biota, but
decreases in the bioconcentration factor of Hg in fish tissue. Because the transport of Hg to Adirondack
lakes appeared to be linked to DOC production from wetlands within the watersheds of the study lakes,
Driscoll et al. (1995) concluded that DOC is important in regulating Hg concentrations in the lakes, and
3-193
-------
ultimately the supply to fish. In a peatland experiment testing the effects of various sources of organic C,
Mitchell et al. (2008) demonstrated that while S042 is required for methylation of Hg, the addition of
some sources of C greatly enhanced the process. The combinations of C and S042 additions that
enhanced methylation in the experiment corresponded to the combinations present in MeHg 'hot spots'
within watersheds that include peatlands.
8 -
0
5
10
50
100
500 1000
pM SO/"
Figure 3-61. MeHg produced in sediment cores incubated two weeks under artificial lake water containing
3-1040 |jM Na2S04. Error bars represent standard error between two replicate cores. Data
from cores incubated under natural water are shown individually. The average MeHg
concentration in unamended Purgee sediments sampled in July was 0.26 ± 0.01 ng/g (n = 2).
Several researchers have suggested that the export of Hg from terrestrial watersheds to lakes may
be controlled in large part by the nature of watershed soils and the transport of naturally occurring organic
acids (Engstrom et al., 1994; Meili, 1991; Mierle, 1990; Mierle and Ingram, 1991). This suggestion is
based partly on the fact that dissolved organic matter strongly binds with Hg, and partly on the observed
positive correlation between Hg accumulation in lake sediments and the ratio of the watershed area to the
lake area in relatively undisturbed watersheds (r2 = 0.91; r2 = 0.91) (Engstrom et al., 1994). Engstrom
et al. (1994) concluded that Hg export from the terrestrial watershed to lake water may be explained by
factors regulating the export of fulvic and humic matter and by watershed area. They based this
conclusion on the close correlation between Hg concentration and humic matter in surface waters, the
observation that peak concentrations of both Hg and dissolved organic matter tend to occur during periods
of high runoff, and the experimental determination that Hg transport occurs primarily in upper soil
horizons.
Ecosystems Characteristics Conducive to Methylation
S deposition is most likely to result in enhanced Hg methylation in regions that receive relatively
high levels of atmospheric Hg and S deposition and that exhibit characteristics conducive to methylation.
These include low ANC and low pH surface waters, with large upstream or adjoining wetlands (Chen
et al., 2005; Scheuhammer and Blancher, 1994; Scheuhammer et al., 2007). Such sensitive ecosystems are
prevalent in portions of the northeastern U.S. and southeastern Canada, but ecosystems with high Hg
3-194
-------
methylation are present in other regions as well. Studies of Hg concentration in feathers, blood, and eggs
of the common loon (Gavia immer) indicate decreasing concentrations from west to east in this region
(Evers et al., 1998; 2003). This pattern is in general agreement with patterns of deposition of both Hg
and S.
Wetland environments have been shown to be significant areas of MeHg production and sources of
export to downstream receiving waters (St. Louis et al., 1994). Wetland MeHg production has been
measured at rates 26 to 79 times higher than in upland areas of a Canadian boreal forest (wetland: 1.84 to
5.55 mg/ha/yr; upland: 0.07 mg/ha/yr) (St. Louis et al., 1994). Watersheds containing 14.0% to 16.3%
wetland yielded 5 to 14 times more MeHg than upland catchments that lacked wetlands (St. Louis et al.,
1994). In the same region, St. Louis et al. (1996) found that all watersheds were net sinks for total Hg, but
that watersheds containing wetlands regularly exported MeHg (St. Louis et al., 1996). However, MeHg
export from these watersheds was not directly proportional to percent wetland coverage, indicating that
other variables are also involved in the major processes that regulate MeHg production and export. In
particular, the level of atmospheric Hg deposition and the acid-base chemistry of drainage water may be
important.
Branfireun et al. (1996) measured highest peat and pore water MeHg concentrations in wetland
areas that exhibited characteristics of a poor fen environment (i.e., interaction with nutrient-poor ground
water). St. Louis et al. (1996) observed that high water yield resulted in high MeHg export. Thus, the
proportion of upland to wetland land area within a watershed was not the only control on MeHg export,
but wetland type and annual water yield also played important roles (St. Louis et al., 1996).
As noted by Munthe (2007), multiple hydrological, chemical, and biological characteristics of
watersheds determine the movement of Hg between compartments. With regards to MeHg, however, the
chemical and biological characteristics of the lake compartment may be more critical: comparing two
remote lakes, one a seepage lake, and the other a drainage lake. Watras and Morrison (2008) found that
although wetland MeHg export was the dominant external source of MeHg to the drainage lake, in-lake
methylation remained four- to seven-fold greater than loading from the wetland. Likewise, Harris et al.
(2007) demonstrated, using traceable stable isotopes of Hg in a whole ecosystem experiment, that nearly
all of the increase in fish MeHg came from Hg deposited to the lake surface, with less than 1% of Hg
deposited to the watershed being exported to the lake, in any form.
Regardless, methylation of Hg occurs in anoxic sediments that contain a sufficient C source to
support S042 -reducing bacterial activity along with an adequate supply of S042 for SRB-mediated S042
reduction. These conditions are found in lake and pond bottom sediments (Gilmour et al., 1992),
freshwater wetland sediments (Branfireun et al., 1999; Harmon et al., 2004; Jeremiason et al., 2006), and
salt mash sediments (Compeau and Bartha, 1985). Such wetland systems, expected to exhibit high levels
of Hg methylation, can be found throughout the U.S.
In a 1998 preliminary national survey of 106 sites from 21 basins across the U.S., Krabbenhoft
et al. (1999) examined the relations of total Hg and MeHg in water, sediment and fish, and concluded that
wetland density was the single most important factor controlling MeHg production at the basin scale.
Four study basins along the east coast of the U.S. had the greatest methylation efficiency, while
nationwide, sub-basins characterized as mixed agriculture and forest cover types had the highest
methylation efficiency. A recent study of biological Hg hotspots in the northeastern U.S. and southeastern
Canada (Evers et al., 2007) analyzed more than 7,300 observations of Hg levels in seven species from
three major taxonomic groups to quantify the spatial heterogeneity in tissue Hg concentrations. Using
published effect thresholds for Hg tissue concentrations, they identified five known and nine possible
biological Hg hotspots. They reported that two of the biological hotspots, located in the Adirondack
Mountains of New York and south-central Nova Scotia, occur in areas with relatively low to moderate
atmospheric Hg deposition and high landscape sensitivity, as determined by the abundant forest and
wetland cover as well as the acidic surface water conditions (Evers et al., 2007). Using data collected by
the Northeastern Ecosystem Research Cooperative (NERC) initiative (Evers and Clair, 2005) to examine
the link between Hg deposition and biotic Hg, (Driscoll et al., 2007b) concluded that "forested regions
with a prevalence of wetland and unproductive surface waters," which are common in the northeastern
3-195
-------
U.S., "promote high concentrations of Hg in freshwater biota." In contrast, in a study of over 600
randomly selected streams and rivers throughout the western U.S., (Peterson et al., 2007) found little
relationship between fish tissue Hg concentrations and surface water pH, S042 . or DOC. They attributed
the lack of a relationship to the fact that low pH (<7) and high DOC systems were rare in the West. The
major factors controlling fish tissue Hg concentrations in western streams and rivers were fish size and
feeding group, not water chemistry. Likewise, a 1999 survey of high altitude western lakes with both low
Hg and low MeHg (Krabbenhoft et al., 2002) identified high pH and elevated rates of photo-
demethylation as the likely causes for low net methylation. Water clarity and high sunlight exposure were
cited as sources of enhanced photo degradation of MeHg.
Sulfur Deposition and MeHg in Fish
As shown by Harris and Rudd (2007), the response of fish MeHg to changes in Hg deposition can
occur on a time scale of less than a year. In their comprehensive synthesis of information on all elements
of the connection between environmental Hg loading, and Hg in fish, Munthe (2007) concluded that
several interacting factors are expected to affect the speed and magnitude of the changes in fish
contamination that result from changes in Hg loading. As indicated previously, numerous studies have
ascertained that S042 supply is a principal driver of MeHg production, and Hrabik and Watras (2002)
showed that decreased deposition of both Hg and S042 are followed by MeHg decrease in fish. Drevnick
et al. (2007), however, were able to establish an explicit linkage between S deposition and fish Hg, by
verifying that even in the absence of change in Hg deposition, changes in S deposition alone result in
commensurate changes in MeHg accumulation in fish.
3.4.1.5. Summary of S and Methylation of Mercury
The evidence is sufficient to infer a causal relationship between S deposition and increased
methylation of Hg, in aquatic environments where the value of other factors is within adequate range for
methylation. The main agent of Hg methylation is S042 -rcducing-bactcria. and experimental evidence
from laboratory to mesocosm scales has established that only inconsequential amounts of MeHg can be
produced in the absence of S042 . These experimental results are highly coherent with one another, and
with observational studies at larger scales. Changes in the amount of S042 present have been shown to be
followed by commensurate changes in MeHg, and mechanistic links have been established between
variation in S042 and variation in methylation of Hg.
Quantification of the relationship between S042 and methylation of Hg in natural settings has
proved difficult because of the presence of multiple interacting factors in aquatic environments where
S042 and Hg are present. The amount of MeHg produced has been shown to vary with 02 content,
temperature, pH, and supply of labile organic carbon. In some watersheds, such as high altitude lakes in
the Western U.S., no effect of changes in S042 deposition have been recorded on methylation of Hg,.
This is because one or several interacting factors were not present in the amounts required for methylation
to occur at more than inconsequential rates. Watersheds with conditions known to be conducive to Hg
methylation can be found in the northeastern U.S. and southeastern Canada, but significant biotic Hg
accumulation has been observed in other regions that have not been studied as extensively, and where a
different set of conditions may exist.
Mercury is a highly neurotoxic contaminant, and enters the food web in the methylated form.
MeHg is then concentrated in higher trophic levels, including fish eaten by humans, with undesirable
consequences for affected species, and for populations that consume large amounts of fish. Once MeHg is
present, other variables influence how much of it accumulates in fish. Current evidence indicates that
increased S deposition very likely results in MeHg accumulation in fish.
3-196
-------
3.4.1.6. S Nutrient Enrichment Case Study: Interactive Effects of S and Hg in Little
Rock Lake, Wl
Little Rock Lake is an 18-ha precipitation-dominated seepage lake located in a forested and
undisturbed catchment of north central Wisconsin. The extensive experimental work conducted at Little
Rock Lake was described by Hrabik and Watras (2002). The research at Little Rock Lake provides
considerable insight into the interactions of S and Hg in the lake, and also bioaccumulation of Hg in fish
in freshwater ecosystems.
In 1984, the lake was divided into two basins by placing an impermeable curtain across a narrow
lake section. One of the basins was experimentally acidified from pH 6.1 to 4.7 by mixing H2S04 into the
surface water over a period of 6 years (Watras and Frost, 1989). The other basin was left undisturbed to
serve as a reference. Beginning in 1990, the treated basin was left to de-acidify naturally.
Hg accumulation in yellow perch showed significant declines in fish in both the experimental and
reference basins between 1994 and 2000, commensurate with declines in atmospheric deposition of Hg.
Fish Hg concentrations in the experimental basin were 57% higher in 1994 than in 2000, whereas
concentrations were 36% higher in the reference basin (Hrabik and Watras, 2002). The authors
determined that half of the decrease in fish Hg concentration was attributable to lakewater de-acidification
and the other half was associated with regional declines in atmospheric Hg deposition. In the reference
basin, which had higher pH and exhibited a lower rate of de-acidification, 15% of the decrease in fish Hg
concentration was due to de-acidification (Hrabik and Watras, 2002).
These findings were consistent with the hypothesis that S042 and newly added Hg synergistically
contribute to enhanced bioaccumulation of Hg in fish. In subsequent analyses, Watras et al. (2006) found
that maximum MeHg concentrations in hypolimnetic waters were directly correlated with the S042
deficit (mean epilimnetic S042 concentration - minimum hypolimnetic S042 concentration) and they
observed a correlation between MeHg and lakewater S042 concentrations. The tracking of external loads
of Hg and S, and internal loads ofHg and MeHg suggested a tight biogeochemical connection among
atmospheric deposition, S042 reduction, and Hg methylation. However, these relationships did not fully
explain the observed large inter-annual variability in MeHg accumulation. The variability appeared to be
influenced by OC, terrestrial runoff and temperature.
The results from the Little Rock Lake acidification experiment suggest that S deposition plays an
important role in the accumulation and methylation of Hg in freshwater ecosystems, and that acid
deposition and Hg deposition have a disproportionately larger effect together than either would have
separately (Watras et al., 2006).
3.4.2. Direct Phytotoxic Effects of Gaseous N and S on Vegetation
This section is intended to provide a brief overview of the exposure and phytotoxic effects of
gaseous N and S compounds on vegetation. This recognizes that the major focus of this review is the
effect of acidifying deposition and N deposition on ecosystems. However, direct effects of gaseous N and
S could augment the effects of deposition on vegetation and effects of gaseous N and S may occur in
some areas.
The effects of gaseous pollutants such as S02, N02, NO, HN03 and 03 on vegetation have been
studied since the 1950s and 1960s. Methodologies have been developed to study these effects in the lab,
greenhouse, and in the field. The methodologies to study gaseous pollutants effects on vegetation have
been recently reviewed in the 2006 03 AQCD (U.S. EPA, 2006b). A thorough description of the
methodologies used to expose vegetation to gaseous pollutants can be found in Section AX9.1 of the 2006
03 AQCD (U.S. EPA, 2006b) and Section 9.2 in the 1993 N02 AQCD (U.S. EPA, 1993a).
Uptake of gaseous pollutants in a vascular plant canopy is a complex process involving adsorption
to surfaces (leaves, stems, and soil) and absorption into leaves. These pollutants penetrate into leaves
3-197
-------
primarily in gaseous form through the stomata, although there is evidence for limited pathways via the
cuticle. Pollutants must be transported from the bulk air to the leaf boundary layer to get to the stomata.
Although the transport of pollutants through a boundary layer into the stomata region is known to be
important, and even rate limiting in many cases of low wind velocity, its description has been defined
from aeronautical concepts and usually relates to smooth surfaces that are not typical of leaf-surface
morphology; however, it is nearly the only treatment available (Gates, 1968). Once through the boundary
layer, the gas must enter the leaf through the stomata. The entry of gases into a leaf is dependent upon the
physical and chemical processes of gas phase and surfaces as well as the stomatal aperture. The aperture
of the stomata is controlled largely by the prevailing environmental conditions, such as humidity,
temperature, and light intensity. When the stomata are closed, as occurs under dark or drought conditions,
resistance to gas uptake is very high and the plant has a very low degree of susceptibility to injury (Figure
3-62). The stomatal control of uptake of gaseous pollutants is described in more detail in AX9.2 of the
2006 03 AQCD (U.S. EPA, 2006b) and Section 9.3.1.5 of the 1993 NOx AQCD (U.S. EPA, 1993a). It
should be noted that unlike higher plants, mosses and lichens do not have a protective cuticle barrier to
gaseous pollutants, a major reason for their sensitivity to gaseous S and N.
Cuticle
Epidermis
Pallisade
Mesophyll
Spongy
Mesophyll
Epidermis
Cuticle
Light
4^
0->, NOy, SOv
j^C, = [C02]A
Guard Cell
H,0
Guard Cell
O3, NOx, SO^
Co= [C02]
Vascular
System
Figure 3-62. The microarchitecture of a dicot leaf. While details among species vary, the general overview
remains the same. Light that drives photosynthesis generally falls upon the upper (adaxial
leaf surface. CO2, SOx, NOx, and O3 gases generally enter through the stomata on the lower
(abaxial) leaf surface, while water vapor exits through the stomata (transpiration).
3.4.2.1. Direct Phytotoxic Effects of SO2 on Vegetation
It has been known since the early 1900s that exposure of plants to S02 can cause damage and death
(Wislicenus, 1914). The large sources of S02 were ore smelters. Sulfides in the ore were oxidized during
smelting and resulted in large releases of S02. Emissions from large ore smelters in the U.S. and Canada
resulted in large areas denuded of vegetation surrounding these facilities (Swain 1949; Thomas 1951).
Much of the damage to the vegetation was due to acute effects of high concentrations of S02. However, as
early as 1923 researchers recognized that S02 might reduce plant growth without acute symptoms of
foliar injury (Stoklasa, 1923). In the 1950s through the early 1980s, there was much research on the
effects of lower levels of S02 as well as the interaction with other pollutants such as O , and N02. Since
3-198
-------
then, there has been much less research on the effects of S02 on vegetation, especially in the U.S., due to
the decreasing ambient concentrations of S02. The effects of S02 on vegetation are summarized below.
Currently, S02 is the only criteria pollutant with a secondary NAAQS distinct from the primary
standard. This standard is to protect acute foliar injury resulting from S02 exposure. The standard is a 3-h
average of 0.50 ppm and was promulgated in 1970 to protect against acute foliar injury in vegetation. The
last AQCD for ecological effects of SOx was completed in 1982 and concluded that controlled
experiments and field observations supported retaining this secondary standard (U.S. EPA, 1971, 1982b,
1982c).
Acute foliar injury usually happens with hours of exposure, involves a rapid absorption of atoxic
dose and involves collapse or necrosis of plant tissues. Another type of visible injury is termed chronic
injury and is usually a result of variable S02 exposures over the growing season. After entering the leaf,
S02 is converted to sulfite and bisulfite, which may be oxidized to S042 . S042 is about 30 times less
toxic than sulfite and bisulfite. The conversion of sulfite and bisulfite to S042 results in net H production
in the cells. Kropff (1991) proposed that the appearance of S02-induced leaf injury was likely due to a
disturbance of intracellular pH regulation. Kropff (1991) pointed out several studies that the pH of
homogenates only shifted towards greater acidity when plants were lethally damaged from long-term S02
exposures (Grill, 1971; Jager and Klein, 1977; Thomas et al., 1944). The appearance of foliar injury can
vary significantly between species and growth conditions affecting stomatal conductance. Currently there
is not regular monitoring for S02 foliar injury effects in the U.S.
Besides foliar injury, long-term lower S02 concentrations can result in reduced photosynthesis,
growth, and yield of plants. These effects are cumulative over the season and are often not associated with
visible foliar injury. As with foliar injury, the effects of foliar injury vary among species and growing
environment. The 1982 S02 AQCD summarized the concentration-response information available at the
time (U.S. EPA, 1982b). Effects on growth and yield of vegetation were associated with increased S02
exposure concentration and time of exposure. However, that document concluded that more definitive
concentration-response studies were needed before useable exposure metrics could be identified. Because
of falling ambient S02 concentrations and focus on 03 vegetation effects research, few studies have
emerged to better inform a metric and levels of concern for effects of S02 on growth and productivity of
vegetation.
Since the 1982 S02 AQCD was published, several studies have investigated a number of different
effects of S02 effects on plants. Most recent research has been performed in areas of Europe where
ambient S02 concentrations are generally higher than in the U.S. A brief summary of some of the major
studies are presented in Table 3-28.
S02 is considered to be the primary factor causing the death of lichens in many urban and industrial
areas, with fruticose lichens being more susceptible to S02 than many foliose and crustose species
(Hutchinson et al., 1996). Damage caused to lichens in response to S02 exposure includes reduced
photosynthesis and respiration, damage to the algal component of the lichen, leakage of electrolytes,
inhibition of N fixation, reduced K+ absorption, and structural changes (Belnap et al., 1993; Farmer et al.,
1992; Hutchinson et al., 1996). Significant reductions in lichen photosynthesis have been measured at
concentrations as low as 91 ppb over 2-4 hours (Huebert, 1985; Sanz, 1992). Damage to the algal
component of the thallus is evidenced by its discoloration. The entire thallus dies soon after algal cells are
damaged (Hutchinson et al., 1996). At higher levels, S02 deactivates enzymes by chemical modification
leading to reduced metabolic activity and loss of membrane integrity (Zeigler, 1975, Nieboer et al., 1976).
It also binds to the central metal atoms of enzymes, adversely affecting membrane function and cell
osmolality. In addition, S02 competitively inhibits carbonate (HC03) and phosphate (H2P04) interactions
with enzymes (Hutchinson et al., 1996). Low pH increases the toxicity of S02 action (Farmer et al.,
1992). The toxic effects of atmospheric deposition of S02 are lessened when lichen are attached to a
substrate, typically bark or rock, having high pH or superior buffering capacity (Richardson and
Cameron, 2004). Van Herk (2001) evaluated relationships between bark pH and air pollution levels as two
significant variables affecting epiphytic lichen composition, and concluded that bark pH was the primary
factor regulating the distribution of acidophilic species in The Netherlands. In studies of unpolluted areas,
3-199
-------
differences in bark chemistry also affect the presence and distribution of epiphytes (Farmer et al., 1992).
Indirect effects on bark pH, caused by acidification and high S02 concentrations, also affect lichen
distribution (Farmer et al., 1992).
3.4.2.2. Direct Phytotoxic Effects of NO, NO2 and PAN
It is well known that in sufficient concentrations nitric oxide (NO) and N02 can have phytotoxic
effects on plants through decreasing photosynthesis and inducing visible foliar injury (U.S. EPA, 1993a).
The 1993 NOx AQCD concluded that concentrations of N02 or NO in the atmosphere are rarely high
enough to have phytotoxic effects on vegetation (U.S. EPA, 1993a). Since the 1993 NOx AQCD, very
little new research has been done on these phytotoxic effects to alter this conclusion. However, it is
known that these gases alter the N cycle in some ecosystems, especially in the western U.S., and
contributing N saturation (Bytnerowicz and Fenn, 1996; Fenn et al., 2003a). See Section 3.3 for a
discussion of the nutrient effects of N.
In general, NO and N02 enters leaves through stomata (Saxe, 1986). However, it has also been
shown that the leaf cuticle may be an important receptor for N02 and there is evidence of transport of NO
and N02 across isolated cuticles (Lendzian and Kerstiens, 1988). Several studies have demonstrated that
plant canopies can directly assimilate N in the form of N02, but canopy uptake of N02 is generally small
relative to total plant uptake (Hanson et al., 1989, Nussbaum et al., 1993; Ammann et al., 1999;
Segschneider et al., 1993; Vallano and Sparks, 2008; von Ballmoos et al., 1993). After entering the leaves,
N02 dissolves in the extracellular water of the sub-stomatal cavity to form HN02 and HN03, which then
dissociate to form nitrite, N03~, and protons (Bytnerowicz et al., 1998b). Both cell and tonoplast
membranes contain ATP-dependent H pumps and the tonoplast pumps are strongly inhibited by N03
(Bytnerowicz et al., 1998b). If extra protons are deposited in vacuoles of the plant cells during normal
cellular regulation, then additional acidity will occur in combination with additional N03 . This
combination can cause disruptions in cellular control (Taylor and MacLean, 1970). N03 and nitrite are
metabolized to amino acids and proteins through a series of enzymatic reactions mainly involving N03
and nitrite reductases (Amundson and MacLean, 1982). The effectiveness of plants to reduce N03 and
nitrite to amino acids and proteins determines the potential of the plant to detoxify NO and N02
(Wellburn, 1990). Reduction of N03 takes place outside of the chloroplast while the reduction of nitrite
is coupled with the light reactions of photosynthesis. Therefore, when leaves are exposed to NO and N02
in the dark, highly phytotoxic levels of nitrite accumulate and may lead to greater toxicity to NO and N02
at night (Amundson and MacLean, 1982). Exposure to NO produces both N03 and nitrite in the leaves,
but the rate of N03 accumulation is much slower than nitrite. Thus, plants exposed to high NO could be
at risk to elevated concentrations of nitrite (Wellburn, 1990). More detailed information on the cellular
effects of NO and N02 can be found in the 1993 NOx AQCD.
The functional relationship between ambient concentrations of NO orN02 and a specific plant
response, such as foliar injury or growth, is complex. Factors such as inherent rates of stomatal
conductance and detoxification mechanisms and external factors, including plant water status, light,
temperature, humidity, and the particular pollutant exposure regime, all affect the amount of a pollutant
needed to cause symptoms of foliar injury. Plant age and growing conditions, and experimental exposure
techniques also vary widely among reports of experimental exposures of plants to N02. An analysis
conducted in the 1993 NOx AQCD of over 50 peer-reviewed reports on the effects of N02 on foliar injury
indicated that plants are relatively resistant to N02, especially in comparison to foliar injury caused by
exposure to 03 (U.S. EPA, 1993a). With few exceptions, visible injury was not reported at concentrations
below 0.20 ppm, and these occurred when the cumulative duration of exposures extended to 100 hours or
longer. At 0.25 ppm, increased leaf abscission was reported on navel orange trees (Citrus sinensis L), but
only after exposures in excess of 1000 hours (Thompson et al., 1970). Green bean plants used as bio-
indicators of N02 injury in Israel developed foliar injury symptoms when ambient concentrations
3-200
-------
exceeded 0.5 ppm (Donagi and Goren, 1979). Only when concentrations exceeded 1 ppm did injury occur
on most plants in less than one day (U.S. EPA, 1993a).
Reductions in rates of photosynthesis have been recorded in experimental exposures of plants to
both NO and N02, but usually at concentrations significantly higher than would normally be encountered
in ambient air. For example, Sabaratnam et al. (1988) reported that soybean (Glycine max) exposed
7 hours/day for 5 days showed an increase in photosynthetic rates at a concentration of 0.2 ppm, but a
reduction in net photosynthesis at a concentration of 0.5 ppm. Short-term exposures of soybean to
0.6 ppm N02 for 2 to 3 hours also had no effect on net photosynthesis (Carlson, 1983). Most plants appear
to be more susceptible to NO than to N02, as shown by Saxe (1986), who exposed a variety of
horticultural plants raised in greenhouses (species of Hedera, Ficus, Hibiscus, Nephrolepis, and
Dieffenbachia) to both NO and N02. Saxe (1986) reported that reductions in net photosynthesis occurred
at doses of NO that were 22 times less than that for N02. However, these reductions in net photosynthesis
required concentrations as high as 1 ppm NO for 12 hours to elicit a response in these plants.
Hundreds of studies have been conducted on the effects of N02 on growth and yield of plants
mostly performed in the 1970s and 1980s. These studies varied widely in plant species, growing
conditions, exposure equipment, concentrations, durations, exposure regimes, and environmental
conditions during exposures. No clear dose-response relationships for exposure to N02 and reductions in
growth and/or yield of plants have emerged from these experiments. Readers are referred to the analysis
of over 100 studies conducted in the 1993 NOx AQCD. A few key studies are highlighted in this section.
Several plant species appear to be susceptible to reductions in growth by relatively low concentrations of
N02 (less than 0.2 ppm), particularly when exposed during low-light conditions. For example, nearly
continuous exposure to 0.1 ppm N02 for eight weeks significantly reduced growth of Kentucky blue grass
(Poa pratensis L) (Ashenden, 1979; Whitmore and Mansfield, 1983). Eight species of tree seedlings were
exposed to 0.1 ppm N02 for six hours/day for 28 days, resulting in reduced shoot or root growth in two
species, white ash (Fraxinus americana L.) and sweetgum (Liquidambar styraciflua L.), reduced height
growth in two clones of loblolly pine (Pinus taeda L.), and no effects on the other species (Kress and
Skelly, 1982). No effects of N02 at 0.1 ppm or lower were observed on numerous other species, including
potato (Solarium tuberosum L.), black poplar (Populus nigra L.), radish (Raphanus sativus L.), soybean,
or peas (Pisum sativum L.) (U.S. EPA, 1993a). No effects of N02 were observed on soybeans grown in
field plots subjected to a series of 10 episodic exposures averaging 0.4 ppm for 2.5 or 3 hours (Irving
et al., 1982). Numerous studies have reported negative effects on growth of a variety of plants exposed to
0.5 ppm N02 and above (U.S. EPA, 1993a), but these concentrations are unrealistically high relative to
current ambient levels of N02.
The 1993 NOx AQCD reviewed the extensive literature on the effects ofN02 in combination with
other gaseous air pollutants, particularly S02 and 03, and concluded that combinations of pollutants can
cause reductions in photosynthesis or foliar injury at concentrations lower than those associated with N02
acting alone. However, the plant responses occur at concentrations much higher than are found in ambient
air (U.S. EPA, 1993a). In addition, the presence of N02 in combination studies did not produce symptoms
different from those caused by the dominant pollutant, either S02 or 03, so that a plant response produced
by combinations of N02 with other air pollutants in the field would be difficult, if not impossible, to
distinguish from those of the other single pollutants (U.S. EPA, 1993a).
Since the 1993 NOx AQCD was completed, most new research on N02 exposure to vegetation has
taken place in Europe and other areas outside the U.S. For example, foliar N03 reductase activity was
increased in Norway spruce (Picea abies) trees growing near a highway with average exposures of about
0.027 ppm compared to trees growing 1300 meters away from the highway with N02 exposures less than
0.005 ppm (Ammann et al., 1995). This was consistent with other studies on Norway spruce in the field
and laboratory (von Ballmoos et al., 1993; Thoene et al., 1991). Muller et al. (1996) found that the uptake
rate of N03 by roots of Norway spruce seedlings was decreased by the exposure to 0.1 ppm of N02 for
48 hours. Similarly, soybean plants grown in Australia had decreased N03 uptake by roots and reduced
growth of plants exposed to 1.1 ppm of N02 for 7 days (Qiao and Murray, 1998). In a Swiss study, poplar
cuttings exposed to 0.1 ppm for of N02 for approximately 12 weeks resulted in decreased stomatal
3-201
-------
density and increased specific leaf weight, but did not result in other effects such as leaf injury or a
change in growth (Gunthardt-Goerg et al., 1996). However, N02 enhanced negative effects of ozone,
including leaf injury, on these poplars when the pollutants were applied in combination (Gunthardt-Goerg
et al., 1996).
Peroxyacetyl nitrate (PAN) is a well-known photochemical oxidant, often co-occurring with 03
during high photochemical episodes, which has been shown to cause injury to vegetation (See reviews by
Cape, 2003, 1997; Kleindienst, 1994 Smidt, 1994; Temple and Taylor, 1983). Acute foliar injury
symptoms resulting from exposure to PAN are generally characterized as a glazing, bronzing, or silvering
of the underside of the leaf surface; some sensitive plant species include spinach, Swiss chard, lettuces,
and tomatoes. Petunias have also been characterized as sensitive to PAN exposures and have been used as
bioindicators of in areas of Japan (Nouchi et al., 1984). Controlled experiments have also shown
significant negative effects on the net photosynthesis and growth of petunia (Petunia hybrida L.) and
kidney bean (Phaseolus vulgaris L.) after exposure of 30 ppb of PAN for four hours on each of three
alternate days (Izuta et al., 1993). As mentioned previously, it is known that oxides of N, including PAN,
could be altering the N cycle in some ecosystems, especially in the western U.S., and contributing
N saturation (Bytnerowicz and Fenn, 1996; Fenn et al., 2003a, see Section 3.3). However, PAN is a very
small component of N deposition in most areas of the U.S. Although PAN continues to persist as an
important component of photochemical pollutant episodes, there is little evidence in recent years
suggesting that PAN poses a significant risk to vegetation in the U.S.
3.4.2.3. Direct Phytotoxic Effects of HNO3
Relatively little is known about the direct effects of HN03 vapor on vegetation. It has been
established that HN03 has a very high deposition velocity compared to other pollutants and may be an
important source of N for plants (Hanson and Lindberg, 1991; Hanson and Garten, 1992; Vose and
Swank, 1990). This deposition could contribute to N saturation of some ecosystems close to sources of
photochemical smog (Fenn et al., 1998). For example, in mixed conifer forests of the Los Angeles basin
mountain ranges HN03 has been estimated to provide 60% of all dry deposited N (Bytnerowicz et al.,
1999).
Norby et al. (1989) reported that exposure of 75 ppb of HN03 for one day increased nitrate
reductase activity in red spruce foliage. In another study, foliar nitrate reductase activity was also
increased in California black oak (Quercus kelloggi), canyon live oak (Quercus chrysolepis) and
Pondersosa pine (Pinus ponderosa) seedlings exposed to HN03 concentrations of 65 to 80 ppb for 24
hours (Krywult and Bytnerowicz, 1997). Because the induction of nitrate reductase activity is a step in a
process leading to the formation of organic N compounds (amino acids), the nitrate from HN03 could
function as an alternative source of N for vegetation (Calanni et al., 1999). However, in plants under
stress, the reduction of nitrate to amino acids consumes energy needed for other metabolic processes.
At high ambient concentrations, HN03 can cause vegetation damage. Seedlings of Ponderosa pine
and California black oak subjected to short-term exposures from 50-250 ppb of HN03 vapor for 12 hours
showed deterioration of pine needle cuticle at 50 ppb in light (Bytnerowicz et al., 1998a). Oak leaves
appeared to be more resistant to HN03 vapor, however, with 12-h exposures in the dark at 200 ppb
producing damage to the epicuticular wax structure (Bytnerowicz et al., 1998a). The observed changes in
wax chemistry caused by HN03 and accompanying injury to the leaf cuticle (Bytnerowicz et al., 1998a)
may predispose plants to various environmental stresses such as drought, pathogens and other air
pollutants. Because elevated concentrations of HN03 and ozone co-occur in photochemical smog
(Solomon et al., 1988), synergistic interactions between the two pollutants are possible (Bytnerowicz
et al., 1998b). However, it should be noted that the experiments described above were observed at
relatively short-term exposures at above ambient concentrations of HN03. Long-term effects of lower air
concentrations that more approximate ambient HN03 should be investigated.
3-202
-------
It has been suspected that HN03 may have caused a dramatic decline in lichen species in the Los
Angles basin (Nash and Sigal, 1999). The suggestion was strengthened by transplant of Ramalina lichen
species from clean air habitats (Mount Palomar and San Nicolas Island) to analogous polluted habitats in
the Los Angeles basin and repeatedly observing death of the lichens over a few weeks in the summer
(Boonpragob and Nash, 1991). Associated with this death was massive accumulation of H4" and N03 by
the lichen thalli (Boonpragob et al., 1989). Recently, Riddell et al. (2008) exposed the healthy Ramalina
menziesii thalli to moderate (8-10 ppb) and high (10-14ppb) HN03 in month-long fumigations and
reported a significant decline in chlorophyll content and carbon exchange capacity compared to thalli in
control chambers. Thalli treated with HN03 showed visual signs of bleaching and by day 28 were clearly
damaged and dead. The damage may have occurred through several mechanisms including acidification
of pigments and cell membrane damage (Riddell et al., 2008). The authors concluded that Ramalina
menziesii has an unequivocally negative response to HN03 concentrations common to ambient summer
conditions in the Los Angeles air basin. They believed it was very likely that HN03 contributed to the
disappearance of this sensitive lichen species from the Los Angeles air basin, as well as other locations
with arid conditions with high deposition loads (Riddell et al., 2008).
3.4.2.4. Summary of Phytotoxic Effects of Gaseous Nitrogen and Sulfur on Vegetation
Sulfur Dioxide
The evidence is sufficient to infer a causal relationship between exposure to S02 and injury to
vegetation (Section 3.4.2.1). The current secondary standard for SO2 is a 3-h average of 0.50 ppm, which
is designed to protect against acute foliar injury in vegetation. There has been limited research on acute
foliar injury since the 1982 PM-SOx AQCD and there is no clear evidence of acute foliar injury below the
level of the current standard.
Effects on growth and yield of vegetation are associated with increased SO2 exposure concentration
and time of exposure. The 1982 PM-SOx AQCD concluded that more definitive concentration-response
studies were needed before useable exposure metrics could be identified. The few new studies published
since the 1982 PM-SOx AQCD continue to report associations between exposure to SO2 and reduced
vegetation growth. However, most these studies have been performed outside the U.S. and at levels well
above ambient concentrations observed in the U.S.
Nitrogen Oxide, Nitrogen Dioxide and PAN
The evidence is sufficient to infer a causal relationship between exposure to NO, NO2 and PAN and
injury to vegetation (Section 3.4.2.2). It is well known that in sufficient concentrations, NO, N02 and PAN
can have phytotoxic effects on plants through decreasing photosynthesis and induction of visible foliar
injury (U.S. EPA, 1993). However, the 1993 NOxAQCD concluded that concentrations of NO, N02and
PAN in the atmosphere are rarely high enough to have phytotoxic effects on vegetation (U.S. EPA, 1993).
Since the 1993 NOx AQCD, very little new research has been done on these phytotoxic effects at
concentrations currently observed in the U.S.
HNOs
The evidence is sufficient to infer a causal relationship between exposure to HNO3 and changes to
vegetation (Section 3.4.2.3). Experimental exposure of HN03 resulted in damage to the leaf cuticle of
pine and oak seedlings which may predispose those plants to other stressors such as drought, pathogens
and other air pollutants (Bytnerowicz et al., 1998a; Bytnerowicz et al., 1998b). However, these tree
seedling experiments used relatively short-term exposures at concentrations well above current ambient
conditions. In lichen studies, several lines of evidence, including transplant and controlled exposure
studies, indicate that past and current HN03 concentrations may be contributing to the decline in lichen
3-203
-------
species in the Los Angles basin (Boonpragob and Nash, 1991; Nash and Sigal, 1999; Riddell et al., 2008).
Current deposition of HN03 is contributing to N saturation of some ecosystems close to sources of
photochemical smog (Fenn et al., 1998) such as the mixed conifer forests of the Los Angeles basin
mountain (Bytnerowicz et al., 1999).
Table 3-28. Summary of recent studies of S02 exposure to plants.
Species
Exposure3
Endpoint(s)
Results
Reference
Scots pine (Pinus 0,50,100,150,155 ppb SO2 in growth
sylvestris L); chambers simulating natural weather in
Norway spruce Finland in early June. The SO2 concen-
(P/cea abies (L) trations represented the range of hourly
Karst.) SO2 concentrations in the vicinity of
industrial areas in Finland.
Concentrations of Exposure to SO2 (100 and 155 ppb) re-
carbohydrates and duced concentrations of glucose and
secondary compo- fructose and increased concentrations of
nents sucrose in pine needles. By contrast, one
spruce clone had more glucose and
fructose and less sucrose in needles
exposed to 100 ppb SO2, but in other
spruces, no changes in sugar concen-
trations were detected in different SO2
exposures (50-155 ppb). Exposure to SO2
had no effects on concentrations of
monoterpenes in pine or spruce needles.
Concentration of total resin acids was
significantly smaller in needles exposed to
the greatest concentration of SO2 (155
ppb), but no changes were detected in
other exposures (50-150 ppb) in either
tree species. Concentrations of palustric
and neoabietic acids were affected by SO2
in needles of pine (155 ppb SO2) and clo-
nal spruces (100 ppb SO2). Exposure to
SO2 did not affect foliar concentration of
total phenolics in pine and spruce seed-
lings. In exposure to 0,50,100 and 150
ppb SO2, total phenolic concentration of
spruce roots increased linearly with
elevated SO2 exposure level. By contrast,
one spruce clone had decreased
concentrations of phenolics in roots after
exposure to 155 ppb SO2.
Kainulainen
etal. (1995)
Scots pine (Pinus Mature trees growing at a polluted (32
sylvestris L); ppb SO2) and low pollution (1 ppb SO2)
Norway spruce sites in Finland. In addition, seedlings
(P/cea abies (L) were placed in the chambers and open-
Karst.) field plots in mid-Sept. 1991 and
fumigated 8 hours daily, 5 days a week
from 19 September to 15 Nov 1991 and
from 19 May to 12 Oct 1992. Mean
pollutant concentrations in the fumigated
chambers during the 8-h exposure
periods were 5-6 ppb SO2 and 7-8 ppb
NO2. The mean pollutant concentrations
not receiving the particular pollutant were
~2 ppb SO2 and 5 ppb NO2.
Response of needle Elevated concentrations of S were found
S and N concentra- in mature pine and spruce trees at pollu-
tions ted sites. The response of mature Scots
pine to SO2 differed from that of mature
Norway spruce. The greater increase in
the needle total S concentrations of pine
suggested more abundant stomatal up-
take of SO2 compared to spruce. Mature
pine was able to assimilate SO42" derived
from SO2 into organic S more effectively
than mature spruce at the high S and N
deposition sites, whereas both pine and
spruce seedlings accumulated S under
NO2+SO2 exposure.
Manninen
etal. (2000)
3-204
-------
Species
Exposure3
Endpoint(s)
Results
Reference
Mature red spruce Branches were fumigated in late summer
(Picea rubens Sarg.) of 1990 and 1991 in Canada. FourS02
treatment levels (0,100,200,400 ppb)
Net photosynthesis, Net photosynthesis and stomatal con-
stomatal conduc- ductance were found to decrease in direct
tance, visible foliar proportion with cumulative foliar SO2
injury absorption. Needle injury was observed in
sun branches exposed to 200 and 400
ppb of SO2 in 1990. Net photosynthesis
was depressed by SO2 regardless of
branch position. Foliage subjected to high
level SO2 did not recover from SO2
damage 1 year after treatment: needles
had fallen off twigs and twig length of new
foliage was reduced.
Meng et al.
(1994)
Scots pine (Pinus Open-air experiment, Finland- trees were Visible symptoms,
sylvestris L.)\ exposed to F, N and S pollutants
Norway spruce individually, or in mixtures, by spraying F
(Picea abies (L) and N compounds in aqueous solution
Karst.) and fumigating plants with gaseous SO2,
for 5 months in each of 3 consecutive
growing seasons. SO2 concentration
among the trees varied between 35 and
140 ppb SO2, depending on velocity and
direction of the wind. Exact
concentrations of SO2 were not reported.
pollutant concentra-
tions, ultrastructure
of seedlings
Visible injury symptoms were most pro-
nounced in combination exposures and
whenever F was applied. Visible symp-
toms correlated well with needle pollutant
concentrations. Exposure to F increased
needle F contents particularly when F was
applied with SO2 or NH4NO3. This
suggests that a reduction in N or SO2
emissions, in F polluted areas, could
improve the condition of conifers via
decreased accumulation of phytotoxic F in
the needles. Norway spruce needles
accumulated 2-10 times as much S and F
as those of Scots pine. In both species,
exposure to SO2 increased significantly
the amount of cytoplasmic vacuoles,
suggesting detoxification of excess
sulphate or low pH. All exposures
enhanced the accumulation of lipid bodies.
Both visible symptoms and ultrastructural
changes pointed to the more pronounced
sensitivity of Norway spruce compared to
Scots pine.
Wulffet al.
(1996)
European Beech
(Fagus sylvatica L),
Norway Spruce
(Picea abies (L)
Karst) European
Silver Fir (Abies alba
Mill.).
Weekly concentrations of SO2, (averaging Shoot length, leaf
3-42 ppb) and O3 (10-90 ppb) was surface area, dry
applied to trees in open-top chambers in weight
Hohenheim, Germany, for almost five
years.
Fumigation with SO2 alone caused
insignificant decreases of shoot length,
total dry weight, and needle surface of
spruce and fir. Fir trees fumigated with
SO2 in combination with O3 showed lower
rates of productivity compared to filtered
control treatments. Beech was not as
affected by SO2 than with O3 or SO2 + O3.
Billen et al.
(1990)
Norway spruce Weekly concentrations of SO2, (averaging Ectomycorrhizal SO2 resulted in higher percentages of
(Picea abies (L.) 3-42ppb) and O3 (10-90 ppb) was (EM) frequency, fine non-mycorrhizal short root tips, and
applied to trees in open-top chambers in root structure, decreased number of living short roots.
Hohenheim, Germany, for almost five distribution of short EM percentage decreased by 38% on SO2
years. roots exposed roots.
Blaschke
(1990)
European Beech
(Fagus sylvatica L),
Norway Spruce
(Picea abies (L.)
Karst.) European
Silver Fir (Abies alba
Mill.).
Weekly concentrations of SO2, (averaging Visible injury
3-42 ppb) and O3 (10-90 ppb) was
applied to trees in open-top chambers in
Hohenheim, Germany, for almost five
years
In Jan to Feb 1985, after long frost, SO2 Arndt et al.
treated fir showed development of tip (1990)
necrosis, showing SO2 inhibits frost resis-
tance. No clear visible effects were found
due to SO2 alone on beech or spruce.
3-205
-------
Species
Exposure3
Endpoint(s)
Results
Reference
European Beech
(Fagus sylvatica L.),
Norway Spruce
(P/cea abies (L)
Karst.) European
Silver Fir (Abies alba
Mill.).
Root samples taken in five-year Hohen-
heim Long Term Experiment in Germany
from 2 tree groups. Each group of trees
consisted of three younger (10-year-old)
and five older (13-year-old) trees. Weekly
concentrations of SO2, (averaging 3-42
ppb) and O3 (10-90 ppb) was applied to
trees in open-top chambers.
Fine root and my- In beech seedlings, SO2 and SO2 +O3 Wollmer and
corrhizae production resulted in reduced fine root production by Kottke
35% and 55%, respectively. SO2 had no (1990)
clear effect on fine root production in fir.
SO2 increased fine root production in
spruce by 31 %, but significantly reduced
relative frequency of mycorrhizae (-20).
Norway spruce One- and two-year old seedlings exposed
(P/cea abies L. to low levels of SO2 and O3 in open-top
Karst.) and fir seed- chambers in five year experiment 1983
lings (Abies aiba through 1988 in Hohenheim, Germany.
Mill.) SO2 concentrations averaged weekly be-
tween 3-42 ppb and O3 concentrations
were between 10-90 ppb.
Visible symptoms,
photosynthesis,
transpiration
The twigs did not exhibit any visible sign of Schweizer
injury due to SO2 treatments. Exposure of
fir to SO2 alone or in combination with O3
resulted in a significant decrease in
photosynthesis and transpiration. No
changes either in photosynthesis or in
transpiration were found in spruce under
fumigation with SO2 alone.
and Arndt
(1990)
Black sage (Salvia SO2 fumigation over 10 weeks, at 0, 50, Number of inflores- Decreased inflorescences were observed Westman
mellifera) CA sage-
brush (Artemisia
californica) Eastern
Mojave buckwheat
(Eriogonum
asciculatum) CA
brittlebush (Encelia
californica)
200, and 500 ppb, California
cences
at 50 ppb SO2 for black sage, and at 200 et al. (1985)
ppb SO2 for California sagebrush, Eastern
Mojave buckwheat, and California
brittlebush, with progressive declines as
SO2 concentration increased.
Timothy grass
(Phleum pratense)
Exposure to 120 ppb SO2 for 40 days. Leaf production, leaf Diminished leaf production and increased Mansfield
senescence, dry leaf senescence in seedlings exposed to
weight (LAR), leaf- 120 ppb SO2 at 35 days. Exposure to 120
area ratio, specific ppb SO2 in seedlings over 40 days
leaf area (SLA) resulted in a 62% reduction in the dry
weight of roots and 51 % reduction in the
dry weight of shoots, as well as a
significant decline in leaf-area ratio and
specific leaf area by the end of the
experiment.
and Jones
(1985)
Mixed native prairie
grassland
Exposed grasses to a control (~7 ppb)
and three elevated levels of SO2 (
64 ppb) over 5 year study.
Root and rhizome S Year-to-year S accumulation did not Laurenroth
-21,37, concentrations, bio- appear to occur over the 5-year course of and Milchu-
mass, primary pro- the treatment, though progressive nas, (1985)
ductivity, lichen increases in root and rhizome S concen-
cover, population trations were observed seasonally. No
significant negative effects on either
above-ground net primary productivity or
below-ground biomass dynamics in
grasses were observed, except a de-
crease in biomass for Bromus japonicus.
Lichen cover declined after 1 year of
exposure at the low treatment level.
Faba bean (Vicia Experiment done in 3 different years
faba L) (1986,1986,1988). Seasonal mean
elevated exposures of SO2 were 58 ppb
in 1985,22 ppb in 1986 and 26 ppb in
1988. Ambient concentrations were 6 ppb
in 1985,3 ppb in 1986 and 3 ppb in 1988.
Yield, leaf injury Exposure to elevated SO2 resulted in leaf Kropff (1990)
injury in all three years. SO2 exposure
reduced yield by 17% in 1985,7% in 1986
and 9% in 1988.
aConcentration, Duration (hours, days)
3-206
-------
Chapter 4. Summary and Conclusions
The previous chapters present the policy-relevant science pertaining to the emissions, atmospheric
transformation and transport, deposition and ecological effects of NOx and SOx. Ecological effects are
divided into broad categories of ecosystem types typically studied: terrestrial, wetlands, freshwater
aquatic, and estuarine aquatic. Several NOx and SOx chemical species were considered because of their
complex multi-phase and multi-species in both the atmosphere and the biosphere. For example, the
atmospheric chemistry of NOx and SOx would be incomplete if only gas-phase compounds were
considered; therefore, descriptions of current ambient concentrations and deposition amounts related to
the particulate forms of N and S are given in Chapter 2. Similarly, the roles of other atmospheric
pollutants including Hg, 03, NH3, and NH/, and their interactions with NOx and SOx in the atmosphere
and biosphere are also considered.
4.1. Source to Deposition
4.1.1. Chemical Families and Constituent Species
NOx is the name given to the family of chemical species containing oxidized N, chief among which
are NO, N02, HN03, and PAN in the gas phase. And because it has a prominent role in transporting N
from the atmosphere to the biosphere, particulate N03 is included in this ISA as well even though it is
not a member of the oxidized N family of species as typically defined. Some of these oxidized N species
are directly emitted; others are formed as secondary products from the emitted species.
Similarly, SOx is the name for the family of chemical species containing oxidized S, including SO,
S02, S03, and S20; however, of these gas-phase species, only S02 is present in concentrations relevant
for atmospheric chemistry and environmental exposures. In addition, and as was the case with the NOx
species, particulate S042 is included in this ISA because of its dominant role in transferring S species
from the atmosphere to the biosphere. Furthermore, this ISA includes extensive treatment of the reduced
N chemical species NH3 and NH4—together given the chemical family name NHX—because NHX can
play a crucial role controlling the transfer of total N and S to the biosphere on many levels of spatial
extent. The most salient points from the foregoing chapters are summarized below.
4.1.2. Transport and Transformation
Convective processes and small-scale turbulence will both transport pollutants up and down
throughout the PBL and the FT. Emitted NOx, SOx, NH3 and other pollutants can be transported vertically
by convection into upper part of the mixed layer on one day, and then transported overnight in a layer of
elevated mixing ratios like a nocturnal low-level jet or the conveyor belts that characterize flows around
frontal systems. Once pollutants are lofted to the middle and upper troposphere, they typically have a
much longer lifetime and, with the generally stronger winds at these altitudes, can be transported long
distances from their source regions. The length scale of this transport is highly variable owing to chemical
and meteorological conditions encountered along the transport path. Transport of NOx and SOx from the
boundary layer to the upper troposphere by convection, for example, usually dilutes high surface
concentrations and extends species lifetimes by several days. During this transport time, the emitted and
now chemically transformed pollutants can be entrained from aloft into the convective boundary layer
4-1
-------
downwind as it grows with increasing insolation. In this way, the transported, transformed pollutants can
be mixed back down to the surface.
CTMs are the prime tools for computing the emissions and interactions of NOx, SOx, NHxand
other pollutants; their transport and transformation, including production of secondary aerosols; the
evolution of particle size distributions; and the deposition of pollutants to the surface. CTMs are driven by
calculated pollutant emissions for primary species such as NOx, SOx, NH3, and primary PM, and by
computed meteorological fields produced by other numerical prediction models. Meteorological
quantities such as winds and temperatures are taken from operational analyses, reanalyses, or numerical
weather circulation models to determine both transport and emission components of the CTMs: higher
temperatures, for example, will substantially alter the rates and chemical forms of some emitted N
pollutants and the lofting of N and S pollutant plumes. Additional discussion of these processes and the
general performance of CTMs is provided in Section 2.8.
The emitted, transported, and transformed pollutants reach the earth's surface where they can exert
ecological effects largely through deposition. Wet and dry deposition are important removal processes for
pollutants on all scales and so are included in all CTMs.
Wet deposition results from the incorporation of atmospheric particles and gases into cloud droplets
and their subsequent precipitation as rain or snow, or from the scavenging of particles and gases by
raindrops or snowflakes as they fall. Receptor surface properties like vegetation leaf surfaces have little
effect on wet deposition, although leaves can retain liquid and solubilized particles containing S and N. In
terrain containing extensive vegetative canopies, any material deposited by precipitation to the upper
stratum of foliage is likely to be intercepted by several foliar surfaces before reaching the soil. This allows
such processes as foliar uptake, chemical transformation, and re-suspension into the atmosphere to occur.
Dry deposition is a complicated and poorly characterized process that appears to be controlled by a
highly varying complex of such variables as atmospheric stability, macro- and micro-surface roughness,
particle diameter, and receptor surface characteristics. The general approach for dry deposition used in
most CTMs is the resistance-in-series method described in Section 2.8.2. The range of particle sizes, the
diversity of canopy surfaces, and the variety of chemical constituents in airborne particulates have made it
difficult to estimate with precision dry particulate deposition totals for large expanses of the U.S.
Direct and indirect wet and dry deposition to specific locations like watersheds depend on air
pollutant concentrations in the airshed above the watershed, but the shape and extent of the airshed is
quite different from that of the watershed owing to the transport and transformation of emitted pollutants
described above. In a watershed, everything that falls in its area, by definition, flows into a single body of
water. An airshed, by contrast, is a theoretical concept that defines the source area containing the
emissions that contribute a given level, often 50 or 75%, to the deposition in a particular watershed or to a
given waterbody. Hence, airsheds are modeled domains containing the sources estimated to contribute a
given level of deposition from each pollutant of concern. The principal NOx airsheds and corresponding
watersheds for several regions in the eastern U.S. are shown in Section 2.8.4.
4.1.3. Emissions and Atmospheric Concentrations
Total anthropogenic NO and N02 emissions in the U.S. in 2002 were 23.19 Tg. Combustion
chemistry at EGUs contributed -22% of these total and transportation-related sources, -56%. Ambient
annual NOx concentrations have decreased -35% in the period 1990-2005 to current annual average
concentrations of -15 ppb.
Biogenic NOx sources are substantially smaller than anthropogenic ones and include biomass
burning, lightning, and soils. The NO and N20 emitted from soils as intermediate products from
denitrification can evolve either naturally or as stimulated by addition of N containing fertilizers to crops
and other soil management practices. N20, another member of the oxides of N family of compounds, is
also a contributor to total U.S. GHG emissions: -6.5% on a Tg C02e basis in 2005, and its U.S. emissions
4-2
-------
decreased -3% in the period 1990-2005, though there remains considerable interannual variation in this
value.
Concentrations of N02 in the CONUS from non-anthropogenic sources in the U.S. and elsewhere
in the world are <300 ppt over most of the CONUS and <100 ppt in the eastern U.S. on an annual average
basis. The 24-h ambient N02 concentrations in CMSAs where most of the regulatory monitors are located
and where most anthropogenic emissions originate were, on average, <20 ppb with a 99 percentile value
<50 ppb for the years 2003-2005. Annual-average N02 concentrations over the CONUS are calculated to
be <5 ppb for nearly all urban and rural and remote sites.
On a national scale, energy production at EGUs accounted for -66% of total S02 emissions in the
U.S. in 2001-2002; -5% of total S02 is emitted by transportation-related sources, with on-road vehicles
accounting for -40 % of the transportation fraction and off-road diesel and marine traffic together
accounting for the remainder. Ambient annual SOx concentrations have decreased -50% in the period
1990-2005 and now stand at -4 ppb for both aggregate annual and 24-h average concentrations nation-
wide.
Annual-average policy-relevant background S02 concentrations in the U.S. from uncontrolled
sources here and elsewhere in the world are <10 ppt over most of the CONUS, or <1% of observed S02
concentrations everywhere except areas in the Pacific Northwest where geogenic S02 sources are
particularly strong.
NH3 emissions are chiefly from livestock and from soils as stimulated by addition of N-containing
fertilizers to crops and other soil management practices. Confined animal feeding operations and other
intensified agricultural production methods over a period of many decades have resulted in greatly
increased volumes of animal wastes high in N; 30 to 70% of these wastes may be emitted as NH3. This
increase in NH3 emissions, and the consequent increase in NFL^ concentration and deposition, correlates
well with the local and regional increases in agricultural intensity. However, there remain no reliably
consistent estimates of national average NH3 concentrations owing to three complex issues: the high
spatial and temporal variability in NH3 emissions; the high uncertainty in the magnitude of those
emissions; and the lack of real-time, ambient level NH3 monitoring techniques. Nonetheless, U.S. national
NH3 emissions totals have been calculated taking into account these three drivers of uncertainty; for
2001-2002 that national NH3 emissions total was -4.08 Tg/yr.
4.1.4. Deposition
Increasing trends in urbanization, agricultural intensity, and industrial expansion during the
previous 100 years have produced a nearly 10-fold increase in N deposited from the atmosphere. NOx,
chiefly from fossil fuel combustion, often dominates total N pollution in the U.S. and comprises -50 to
75% of the total N atmospheric deposition.
For the period 2004-2006, the routine monitoring networks report the mean N deposition in the
U.S. was greatest in the Ohio River Valley, specifically in the states of Indiana and Ohio, with values as
high as 9.2 and 9.6 kg N/ha/yr, respectively. N deposition was lower in other parts of the East, including
the Southeast and in northern New England. In the central U.S., Kansas and Oklahoma reported the
highest deposition, 7.0 and 6.5 kg N/ha/yr, respectively.
N deposition primarily occurred in the form of wet N03 and NH/, followed with decreasing
amounts of dry HN03, dry NH/, and dry N03~. Although deposition in most areas of the U.S. occurred in
wet form, there were some exceptions, including parts of California where N deposition was primarily
dry. Data are very sparse for the central U.S. between the 100th meridian and the Mississippi River; but
where available, N deposition values there are lower than most of the eastern U.S., ranging from 4.1 to
5.3 kg N/ha/yr.
For the period 2004-2006, mean S deposition in the U.S. was greatest east of the Mississippi River
with the highest deposition amount, 21.3 kg S/ha/yr, in the Ohio River Valley where most recording
4-3
-------
stations reported 3 year averages >10 kg S/ha/yr. Numerous other stations in the East reported S
deposition >5 kg S/ha/yr. Total S deposition in the U.S. west of the 100th meridian is relatively low, with
all recording stations reporting less than 2 kg S/ha/yr and many reporting less than 1 kg S/ha/yr.
S was primarily deposited in the form of wet S042 followed in decreasing order by a smaller
proportion of dry S02 and a much smaller proportion of dry S042 . However, these S data in the western
U.S., like those for N deposition, are derived from networks with many fewer nodes in the West than in
the East.
4.1.5. Field Sampling and Analysis
The instrumentation deployed at present in the routine regulatory monitoring networks for
determination of gas-phase N02 and S02 concentrations is designed for determining compliance with the
current NAAQS. But in applications for determining environmental effects, all these methods have
important limitations which make them inadequate for fully characterizing the state of the atmosphere at
present; for correctly representing the complex heterogeneity of N and S deposition across the landscape;
and for realistically apportioning the contributions of reduced and oxidized forms of atmospheric N and S
in driving observed biological effects at a national scale.
Routine N02 measurements by chemilumenescnce difference (the FRM) are contaminated by
unknown and varying concentrations of higher-order oxidized N species, including gas-phase HN03,
important itself for N deposition to the biosphere and also as a precursor to pN03. Moreover, dry
deposition of NO, N02, and PANs is not at present estimated, but could be as much as 30% of total dry
oxidized N deposition in areas near strong NOx sources.
The present-day ambient annual average S02 concentrations are very near or even below the
operating limit of detection of most of the FRM monitors in the largest regulatory network. This produces
irresolvable uncertainty in these data which may be important for environmental effects from S
compounds, since they result in some cases from exposures at these current low concentrations.
Routine field sampling techniques for NH3 are at present limited to integrated values from several
days to one week because higher frequency semi-continuous methods are not yet sufficiently robust to
deploy in regulatory networks although passive NH3 samplers show excellent potential. Estimates for the
contribution of NH3 to the total N deposition budget range as high as 30% of total N, and are perhaps the
dominant source of reduced N.
Routine, regulatory, national-scale sampling and analysis for particulate-phase N03 . S042 . and
NH4+ are subject to positive and negative errors, chiefly from the loss or production of constituent species
on the surface of the filter used for the long time-integrated measurement.
The coverage of the networks is very thin over large expanses of the interior U.S. and especially so
west of the 100th meridian. This lack of monitored sites increases the likelihood that significant N and S
deposition is now occurring at current atmospheric concentrations where no measurements are available,
as predicted in numerical experiments with large-scale, first-principles models of atmospheric chemistry
and physics and deposition, and as measured at some few selected special sites.
4.2. Acidification
Acidifying deposition includes gases and particles derived from SOx, NOx, and NHX. The effects
of acidifying deposition on ecosystems have been well studied over the past several decades and
vulnerable areas have been identified for the U.S. The wealth of available data has led to the development
of robust ecological models used for predicting soil and surface water acidification. The principal factor
governing the sensitivity of terrestrial and aquatic ecosystems to acidifying deposition is geology.
4-4
-------
Watersheds of acid-sensitive lakes and streams have geologic formations with low base cation supply.
Other factors contribute to the sensitivity of soils and surface waters to acidifying deposition, including
topography, vegetation, soil chemistry, land use, and hydrologic flowpath. Regional and ecosystem
vulnerability to acidification results from sensitivity and exposure to acidifying deposition.
4.2.1. Terrestrial
In the 1982 PM-SOx AQCD foliar and root uptake pathways for SOx were described in detail, as
well as the role of S as a nutrient. Though small amounts of S02 may be beneficial, it was understood that
large amounts and high frequency of S02 exposure and S deposition can be detrimental in the long term.
At that time, there were no documented observations or measurements of changes in natural terrestrial
ecosystems that were directly attributed to acidic precipitation; however, changes still may have been
occurring.
The 1993 NOx AQCD documented few cases in which excessive atmospheric N deposition was
linked to soil acidification, although the process of soil acidification was already well understood. Since
the preparation of these assessments, direct links between NOx and SOx deposition and many adverse
effects associated with ecosystem loading have been reported.
4.2.1.1. Biogeochemistry and Chemical Effects
The evidence is sufficient to infer a causal relationship between acidifying deposition and changes in
biogeochemistry related to terrestrial ecosystems. The strongest evidence for a causal relationship comes
from studies of forested ecosystems, with supportive information on other plant communities, including
shrubs and lichens (Section 3.2.2.1.). Grasslands are likely less sensitive to acidification than forests. Soil
acidification occurs in response to inputs of sulfuric acid and nitric acid; the effect can be neutralized by
weathering or base cation exchange. Soil acidification is a natural process, but is often accelerated by
acidifying deposition. Acidifying deposition is important in decreasing concentrations of exchangeable
base cations in soils. Despite recent decreases in acidifying deposition, there are widespread observations
of ongoing soil acidification such as decreases in soil exchangeable base cations. The limited mobility of
anions associated with naturally derived acidity (organic acids and carbonic acid) controls the rate of base
cation leaching from soil under conditions of low atmospheric deposition of S and N. Because inputs of S
and N in acidifying deposition provide anions that are more mobile in the soil environment than anions of
naturally derived acids, these mineral acid anions can accelerate natural rates of base-cation leaching.
Nitrification is mediated by autotrophic bacteria that derive energy by oxidizing NH/ to N03 .
Nitrification produces acidity in the form of HN03 as a byproduct, which contributes to the acidification
of soils and surface waters.
There are three useful indicators of chemical changes and acidification effects on terrestrial
ecosystems, with consistency and coherence seen among multiple studies, including soil base saturation,
A1 concentration, and C:N Ratio (see Table 4-1).
¦ Soil base saturation is the concentration of exchangeable bases as a percent of the total soil cation
exchange capacity. Once base saturation decreases to a critical level (-15-20%), inputs of H2S04
and HN03 result in exchange of inorganic Al.
¦ Inorganic Al is toxic to some tree roots. Plants affected by high inorganic Al concentrations in
soil solution often have reduced root growth, which restricts the ability of the plant to take up
water and nutrients, especially calcium (Parker et al., 1989).
4-5
-------
¦ The C:N ratio of soil is used to indicate alterations to the N biogeochemical cycle. If the ratio
falls below about 20 to 25, nitrification is stimulated resulting in net nitrification and increased
acidity.
Table 4-1. Chemical indicators of acidification to terrestrial ecosystems.
Reference
Indicator
SOIL BASE SATURATION
Reuss (1983)
If base saturation is less than 15-20%, exchange ion chemistry is dominated by inorganic Al.
Cronan and Grigal
Base saturations below about 15% in the soil B-horizon could lead to effects from Al stress.
(1995)
Lawrence et al. (2005)
Base saturation decreases from 30% to 20% in the upper soil B-horizon showed decreases in diameter growth of
Norway spruce.
Bailey et al. (2004)
At Ca saturation less than 2% and Mg saturation less than 0.5% in the upper soil B-horizon, sugar maple mortality
was observed.
ALUMINUM CONCENTRATIONS
Johnson et al. (1991)
In soils with base saturation below about 20%, base cations reserves are so low that Al exchange dominates.
Joslin and Wolfe (1992)
Eagar et al. (1996)
There is a 50% risk of negative effects on tree growth if the molar ratio of Ca to Al in soil solution was 1.0.100% risk
Cronan and Grigal
for negative effects on growth at a molar ratio below 0.2.
(1995)
Johnson et al. (1994a;
Ca:AI ratios above 1.0 over the course of 4 years were found in a forest stand experiencing high mortality.
1994b)
DeWitt etal. (2001)
Ca:AI ratios below 0.5 in a Norway spruce stand showed reduced Mg concentrations in needles in the third year.
C:N RATIO
Aber et al. (2003)
Increased effects of nitrification occur only in soil with C:N ratio below about 20-25.
Ross et al. (2004)
4.2.1.2. Biological Effects
The evidence is sufficient to infer a causal relationship between acidifying deposition and changes in
terrestrial biota. The strongest evidence for a causal relationship comes from studies of terrestrial systems
exposed to elevated levels of acidifying deposition that show reduced plant health, reduced plant vigor,
and loss of terrestrial biodiversity. In multiple studies, consistent and coherent evidence shows that
acidifying deposition can affect terrestrial ecosystems by causing direct effects on plant foliage and
indirect effects associated with changes in soil chemistry (Section 3.2.2.3). Biological effects of
acidification on terrestrial ecosystems are generally attributable to A1 toxicity, decreased ability of plant
roots to take up nutrient cations and elevated leaching of Ca2+ from conifer needles. There are several
indicators of stress to terrestrial vegetation (see Table 3-3), including percent dieback of canopy trees,
dead tree basal area (as a percent), crown vigor index, and fine twig dieback.
4-6
-------
Species Level
¦ Changes in soil chemistry (e.g., depletion of soil base cations, A1 toxicity to tree roots, leaching
of base cations into drainage water) have contributed to high mortality rates and decreasing
growth trends of red spruce trees (Picea rubens) in some areas of the eastern U.S. over the past
three decades (see Red Spruce, Section 3.2.2.3).
¦ Acidifying deposition, in combination with other stressors, is a likely contributor to the decline of
sugar maple (Acer saccharum) trees that occur at higher elevation, in some portions of the eastern
U.S., on geologies dominated by sandstone or other base-poor substrate, and that have base-poor
soils (see Sugar Maple, Section 3.2.2.3).
¦ Lichens and bryophytes are among the first species affected by acidifying deposition in the
terrestrial ecosystem. Effects of S02 on lichens include reduced photosynthesis and respiration,
damage to the algal component of lichen, leakage of electrolytes, inhibition ofN fixation, reduced
potassium absorption, and structural changes.
¦ Data are insufficient to draw general conclusions for other species.
Community Level
¦ Species loss and reduced biodiversity of forests, shrubs, and meadow plant communities may
occur in response to acidifying deposition; however, such effects are likely more related to the
nutrient enrichment effects of N deposition.
4.2.1.3. Regional Vulnerability and Sensitivity
There has been no systematic national survey of terrestrial ecosystems to determine the extent and
distribution of terrestrial ecosystem sensitivity to the effects of acidifying deposition. However, one
preliminary national evaluation estimated that -15% of forest ecosystems in the U.S. exceeds the
estimated critical load of wet and dry deposition of S and N by >250 eq/ha/yr (McNulty et al., 2007).
Forests of the Adirondack Mountains of New York, Green Mountains of Vermont, White
Mountains of New Hampshire, the Allegheny Plateau of Pennsylvania, and high-elevation forest
ecosystems in the southern Appalachians are the regions most sensitive to terrestrial acidification effects
from acidifying deposition (Section 3.2.4.2). While studies show some recovery of surface waters, there
are widespread measurements of ongoing depletion of exchangeable base cations in forest soils in the
northeastern U.S. despite recent decreases in acidifying deposition.
4.2.2. Aquatic
In the 1982 PM-SOx AQCD, the evidence on acidifying deposition and its role in the acidification
of aquatic ecosystems was assessed. The most vulnerable regions were identified, including the
Adirondack Mountains of New York. Significant changes were reported in aquatic ecosystems with
increasing acidity, particularly as the pH decreases below -5.5. It was concluded that: changes in
community structure occur at all levels in the food web; bacterial decomposition is reduced and fungi that
feed on organic debris may become dominant in aquatic communities; organic matter accumulates
rapidly, tying up nutrients and limiting nutrient mineralization and cycling; phytoplankton productivity
may be reduced because of changes in nutrient cycling and increased acidity; biomass and total
productivity of benthic macrophytes and algae may increase, in part because of increased lake
transparency; andspecies diversity and total numbers of species of aquatic plants and animals (especially
invertebrates and fish species) are reduced, and acid-tolerant species predominate.
4-7
-------
In the 1993 N0X AQCD a much-expanded body of evidence was available on the role of N
deposition in the acidification of aquatic ecosystems. This was especially the case with respect to episodic
acidification, which is far more common than chronic acidification and has been well documented for
streams and lakes in the eastern U.S. The most well known examples are in the Adirondack and Catskill
Mountains of the Northeast, as well as in the Great Smoky Mountains of the Southeast. Instances of
episodic acidification were also reported in the western U.S. but to a much lesser extent than in the East.
4.2.2.1. Biogeochemistry and Chemical Effects
The evidence is sufficient to infer a causal relationship between acidifying deposition and changes in
biogeochemistry related to aquatic ecosystems. The strongest evidence for a causal relationship comes
from studies of changes in surface water chemistry including concentrations of S042 . N03 . sum and
surplus of base cations, ANC, inorganic Al, Ca, and surface water pH (see Section 3.2.2.1). Surface water
chemistry integrates the sum of upstream soil and water processes and reflects the results of watershed-
scale terrestrial effects of S and N deposition, including N saturation, forest decline, and soil acidification
(Stoddard et al., 2003). In many cases, surface water chemistry indicates the effects of acidification on
biotic species and communities found in fresh water ecosystems.
The status of surface water chemistry can be examined and reported as chronic chemistry or
episodic chemistry. Chronic chemistry refers to annual average conditions, which are often represented as
summer and fall chemistry for lakes, and as spring baseflow chemistry for streams. Episodic chemistry
refers to conditions during rainstorms or snowmelt when proportionately more drainage water is routed
through upper soil horizons, which tend to provide less neutralizing of atmospheric acidity as compared
with deeper soil horizons. Surface water chemistry has lower pH and ANC during storm runoff or
snowmelt than during baseflow conditions. One of the most important effects of acidifying deposition on
surface water chemistry is the short-term change in chemistry that is termed "episodic acidification."
Some streams may have chronic or average chemistry that is suitable for aquatic biota, but be subject to
occasional episodic acidification with lethal consequences. Episodic declines in pH and ANC are nearly
ubiquitous in drainage waters throughout the eastern U.S., caused partly by acidifying deposition and
partly by natural processes.
Acidification effects on aquatic biota are often evaluated using measures of either inorganic Al or
pH. ANC is also used because it is an indicator of acid-base status (although ANC does not relate directly
to the health of biota). The usefulness of ANC lies in the association between ANC and the surface water
constituents that directly contribute to or ameliorate acidity-related stress, in particular pH, Ca, S042 and
inorganic Al.
Sulfate, Nitrate, and Base Cations
Changes in water chemistry resulting from acidifying deposition typically include changes in
S042 . N03 . and base cation concentrations. Each plays an important role in the acid-base chemistry of
water; none, however, are directly toxic at concentrations commonly encountered in surface waters
(Table 4-2).
¦ Sulfate is the primary inorganic anion found in most acid-sensitive waters. Continued decreases
in S emissions should cause further decreases in S042 concentrations in surface waters.
However, the rate of decrease in surface water S042 concentrations may be delayed as
accumulated S leaches from watershed soils in some regions of the country, especially the
southern Appalachian Mountains.
4-8
-------
¦ The importance of N03 as an agent of acidification varies by region, but is particularly important
during periods of high hydrologic flow from soils to streams, such as those that occur during
snowmelt and rain events. The relationship between N deposition and surface water N03
concentration is complex and involves the terrestrial and aquatic cycling of N and other elements.
N03 contributes to the acidity of many lakes and streams in the eastern U.S. However, there is
no apparent relationship between recent trends in N deposition and trends in N03 concentrations
in these surface waters (in contrast to observed responses for S deposition and S042
concentrations). This suggests that the time scales ofN saturation may be longer than previously
considered (e.g., centuries, rather than decades). Nevertheless, long-term retention ofN deposited
in forested regions and consequent dampening of deposition effects on surface waters is unlikely
to continue (Aber et al., 2003).
¦ Decreases in base cation concentrations in eastern U.S. surface waters over the past two to three
decades are ubiquitous and are closely tied to trends in S042 concentrations. Rates of base cation
depletion have been similar to those for S042 plus N03 in most areas (Shenandoah National
Park is a notable exception). Decreasing trends in base cation concentrations do not necessarily
indicate further acidification or recovery of surface waters, but may indicate either lower base
cation leaching rates in soils or depletion of base cations from the soil system.
Acid Neutralizing Capacity, Aluminum, and pH
Acidification of surface water causes changes in ANC, Al concentration, and pH. Low pH and high
inorganic Al concentration can be directly toxic to aquatic biota (Section 3.2.3).
¦ ANC reflects the difference between base cations and anions of strong acids in solution; it is the
most widely used measure of acid sensitivity, acidification, and chemical recovery of surface
waters in response to changes in acidifying deposition. Acidic waters are defined as those having
ANC equal to or below zero. Waters with ANC of <50 j^ieq/L are considered "extremely acid-
sensitive" (Schindler, 1988), and are vulnerable to episodic acidification (DeWalle et al., 1987;
Eshleman, 1988). Lake and stream ANC values decreased throughout much of the 20th century in
a large number of acid-sensitive lakes and streams throughout the eastern U.S. Since -1990, the
ANC of many affected lakes and streams has increased slightly. The number of acidic surface
waters has decreased in some areas of the Northeast, but not in the central and southern
Appalachian Mountains.
¦ Dissolved inorganic Al is an important chemical indicator of the effects of acidifying deposition
on surface water because it is toxic to aquatic life and generally does not leach from soils in the
absence of acidification. When pH falls below approximately 5.5, inorganic Al generally
becomes a greater health risk to biota. Limited data suggest that acid-sensitive regions of the
northeastern U.S. have elevated inorganic Al concentrations in surface waters induced by years of
acidifying deposition, posing a threat to aquatic life. Concentrations have decreased slightly in
some surface waters in the northeastern U.S. during the last two decades in response to decreased
levels of acidifying deposition.
¦ The pH of freshwater streams and lakes is a common measure used to link acidification to
adverse effects on aquatic biota. Decreases in pH below values of 6.0 typically result in species
loss of benthic invertebrates, plankton species, and fish. A number of synoptic surveys indicated
loss of species diversity and absence of several fish species in the pH range of 5.0 to 5.5. If pH
decreases to lower values, there is a greater likelihood that more aquatic species could be lost
without replacement, resulting in decreased richness and diversity.
4-9
-------
Table 4-2. Chemical indicators of acidification in surface water.
Reference
Indicator
SULFATE
Driscoll et al.
Acidifying deposition at Hubbard Brook Experimental Forest in New Hampshire contributed to a nearly four-fold increase
(2001a)
in stream SO42" concentration between 1850 and 1970.
Stoddard et al.
Widespread decreasing trends in SO42" concentrations were documented by U.S. EPA during the period 1990-2000 in
(2003)
the eastern U.S. including New England lakes (1.77 peq/L/yr), Adirondack lakes (2.26 peq/L/yr), Appalachian streams
(2.27 [jeq/L/yr), and Upper Midwest lakes (3.36 peq/L/yr).
NITRATE
Driscoll and
NO3" concentrations in 20 Adirondack lakes in the early 1980 averaged 12% of SO42" concentrations.
Newton (1985)
Lovett et al.
Baseflow NO3" concentrations in 1994-97 were an average of 37% of SO42" concentrations in 39 Catskill streams.
(2000)
Murdoch and
During high-flows in Catskill streams NO3" concentrations periodically equaled or exceeded SO42" concentrations.
Stoddard (1993)
Webb et al.
Average concentrations of NO3" in most southeastern streams tend to be considerably less than SO42" concentrations.
(2004)
Cook etal. (1994)
Very high NO3" concentrations in streamwater were documented at high elevations in the Great Smoky Mountains in
North Carolina.
BASE CATIONS
Likens et al.
Approximately linear increasing relationship between concentrations of base cations and S042"+ NO3" concentrations in
(1996)
Hubbard Brook streams from 1964 to 1969, then a reversal in 1970 and a decreasing trend up to 1994.
Lawrence et al.
Decreasing concentrations of base cations at a rate that exceeded decreases in concentrations of S042"+ NO3" in Catskill
(1999)
Mountain streams from 1984 to 1997.
ACID NEUTRALIZING CAPACITY
Sullivan et al.
Model simulations suggest that none of the lakes in the Adirondack target lake population were chronically acidic or had
(2006b)
ANC less than 20 peq/L under preindustrial conditions. By 1980, there were hundreds of such lakes.
Stoddard et al.
Tendencies during the 1990s toward increasing surface water Gran ANC in all glaciated regions of the eastern U.S. (i.e.,
(2003)
New England, Adirondacks, Northern Appalachian Plateau) and Upper Midwest; and decreasing Gran ANC in the
Ridge/Blue Ridge province.
SURFACE WATER ALUMINUM
Gensemer and
Found that organically complex aluminum (organic Al) can occur in surface waters as a result of natural soil and
Playle (1999)
hydrologic processes, but this form of Al is not harmful to aquatic life.
Gensemer and
Demonstrated that inorganic Al has been found to be toxic to plant and animal species throughout the food web.
Playle (1999)
Baldigo et al.
20% mortality of young-of-the-year brook trout during a 30-day period with a median inorganic Al concentration of
(2007)
2 [jmol/L. 90% mortality occurs over 30 days with a median inorganic Al concentration of 4.0 pmol/L.
Lawrence et al.
49 of 195 streams (25%) in the western Adirondack region had inorganic Al concentrations above 2.0 |jM during August
(2007)
base flow.
4-10
-------
PH
Haines and Baker
pH values for biological effects have been summarized for a variety of aquatic organisms; common threshold values for
(1986)
pH are 5.0,5.5, and 6.0.
Baker et al.
(1990b)
Charles et al.
25 to 35% of the Adirondack lakes larger than 4 ha have been acidified since preindustrial time. An estimated 80% of the
(1989)
Adirondack lakes that had pH less than 5.2 in the mid-1980s were inferred to have experienced large declines in pH and
Sullivan et al.
ANC since the previous century. About 30 to 45% of the lakes with pH between 5.2 and 6.0 have also acidified.
(1990)
Cumming et al.
(1992; 1994)
Gbondo-
PnET-BGC modeling at Hubbard Brook estimated that past stream pH (circa 1850) was probably about 6.3, compared
Tugbawa et al.
with just above 5.0 in 2000.
(2002)
Stoddard et al.
An increase in the hydrogen ion concentration of Appalachian streams (0.08 peq/L/yr) and Upper Midwest lakes (0.01
(2003)
[jeq/L/yr) was reported. No trends were found in New England lakes or Appalachian streams in this study.
4.2.2.2. Biological Effects
The evidence is sufficient to infer a causal relationship between acidifying deposition and changes in
aquatic biota. The strongest evidence for a causal relationship comes from studies of aquatic systems
exposed to elevated levels of acidifying deposition that support fewer species of fishes,
macroinvertebrates, and diatoms (Section 3.2.3.3). Consistent and coherent evidence from multiple
species and studies shows that acidification can result in the loss of acid-sensitive species, and there is
evidence of a biological gradient in effects in that more species are lost with greater acidification.
Biological effects are linked to changes in water chemistry including ANC, inorganic Al, and pH.
Decreases in ANC and pH and increases in inorganic Al concentration contribute to declines in taxonomic
richness of zooplankton, macroinvertebrates, and fish. Chemical changes can occur over both long- and
short-term time scales, with additional effects on biological systems. Short-term (hours or days) episodic
changes in water chemistry can have biological effects, including reduced fish condition factor, changes
in species composition, and declines in aquatic species richness across multiple taxa, ecosystems and
regions.
Species Level
¦ High levels of acidification (to pH values below 5) virtually eliminate all mayflies, crustaceans,
and mollusks from some streams.
¦ In general, populations of salmonid fish are not found at pH levels less than 5.0, and smallmouth
bass (Micropterus dolomieu) populations are usually not found at pH values less than 5.5 to 5.2.
¦ Twenty percent mortality of young-of-year brook trout were documented during a 30-day period
with a median inorganic Al concentration of 2 (imol/L (Baldigo et al., 2007). It was estimated that
90% mortality would occur over 30 days with a median inorganic Al concentration of 4.0
(imol/L.
4-11
-------
Community Level
¦ Community-level effects were observed in the Adirondacks and Shenandoah National Park where
taxonomic richness is lower in lakes and streams having low ANC and pH.
¦ Decreases in pH and increases in inorganic A1 concentrations have reduced the species richness
of plankton, invertebrates, and fish in acid-affected surface waters.
¦ In the Adirondacks, a positive relationship exists between the pH and ANC in lakes and the
number of fish species present in those lakes. A number of synoptic surveys indicated suggested
loss of species diversity and absence of several sensitive fish species in the pH range of 5.0 to 6.0
(Section 3.2.4.4).
¦ In Shenandoah National Park streams, the fish species richness decreased with decreasing stream
ANC. On average, richness is lower by one fish species for every 21 j^icq/L decrease in ANC
(Section 3.2.4.5).
¦ Short-term episodes of acidification are particularly harmful to aquatic biota. Early life stages are
more sensitive to acidic conditions than the young-of-the-year, yearlings, and adults. Episodes are
most likely to affect biota if the water had pre-episode pH above 5.5 and minimum pH during the
episode of less than 5.0. Episodic acidification can have long-term adverse effects on fish
populations.
4.2.2.3 Regional Vulnerability and Sensitivity
The effects of acidifying deposition have been assessed by several national surveys since the
1980s, including the National Surface Water Survey and the National Stream Survey in the mid-1980s,
the Wadeable Streams Assessment (WSA) in 2004, the U.S. EPA Long-Term Monitoring program
beginning in 1983, and Temporally Integrated Monitoring of Ecosystems probability surveys beginning in
1991. These surveys indicate that acidifying deposition has acidified surface waters in the southwestern
Adirondacks, New England uplands, low-silica eastern Upper Midwest, forested Mid-Atlantic Highlands,
and Mid-Atlantic Coastal Plain (Section 3.2.4.2).
In the U.S., the Northeast and Mountainous West regions contain many of the surface waters most
sensitive to acidification. Levels of acidifying deposition in the West are low in most areas, acidic surface
waters are rare, and the extent of chronic surface water acidification that has occurred to date has been
very limited. However, episodic acidification does occur. In both the west and the northeast, the most
severe acidification of surface waters generally occurs during spring snowmelt. On average, spring ANC
values of acid-sensitive surface waters in New England, the Adirondacks, and the northern Appalachian
Plateau were on average 30 j^icq/L lower than summer values between 1990 and 2000. This implies that
lakes and streams in these regions would need on average to recover to chronic ANC values above
-30 (ieq/L or more before they could be expected not to experience acidic episodes (Stoddard et al.,
2003).
In 2004, the U.S. EPA conducted a national WSA survey and found that, overall, less than 1% of
the 1,020,000 km of stream in the target population was acidic due to acidifying deposition. No acidic
streams were observed in the Mountainous West, Xeric West, Upper Midwest, Northern Plains, Southern
Plains, or Temperate Plains ecoregions. Streams that were acidified from acidifying deposition were
found in the Northern Appalachians (2.8% of 96,100 km of stream), and the Southern Appalachians (1.8%
of 287,000 km). Very low ANC (0 to 25 j^ieq/L) streams, likely exposed to episodic acidification, were
found in the Northern Appalachians (2.7% of 96,100 km of stream), the Coastal Plain (6.3% of 119,000
km), and the Mountainous West (0.6% of 204,000 km). Stream surveys were not conducted in the
Adirondacks or New England.
It is important to address surface water recovery in response to reduced acidifying deposition over
the past few decades. The following summarizes recent regional trends in acidification recovery.
4-12
-------
About one-quarter to one-third of the lakes and streams that were chronically acidic during summer
in the 1980s were no longer chronically acidic in 2000. These improvements in water chemistry are
largely attributed to decreases in S deposition. Throughout the northeastern U.S., the concentration of
S042 in surface waters has decreased substantially in response to decreased emissions and atmospheric
deposition of S. Decreased S042 concentrations of a third or more in lakes and streams have been
commonly observed.
Data from U.S. EPA monitoring programs show that the following important changes in lake and
stream chemistry occurred during the 1990s in response to S and N emissions reductions: S042
concentration decreased as a percentage of total ion concentration in surface waters; ANC increased
modestly in three of the five regions included in surface water efforts; dissolved organic carbon and
associated natural organic acidity increased, perhaps toward more natural pre-disturbance concentrations,
as surface water acidity contributed from acidifying deposition decreased; and inorganic, and potentially
toxic, A1 concentrations appear to have decreased slightly in some sensitive aquatic systems.
Despite these improvements, some regions and specific locations remain sensitive to acidifying
deposition. For example, in the Adirondacks, the current rates of N and S deposition exceed the amount
that would allow recovery of the most acid-sensitive lakes. In the Shenandoah, past S042 has
accumulated in the soil and is slowly released from the soil into stream water, where it causes
acidification, making parts of this region sensitive to current loading. Numerical models specifically
calibrated to these locations and conditions suggest that the number of acidic streams will increase under
the current deposition rates.
4.2.3. Ecosystem Services
Acidification of ecosystems is primarily driven by NOx, NHX and SOx. Ecosystem services, as
defined by Hassan et al. (2005), are broadly grouped into four main categories (see Section 3.1.3). The
specific effects of acidification on ecosystem services may include:
¦ Supporting: altered nutrient cycling, decreased biodiversity, decline of productivity
¦ Provisioning: decline in the richness, abundance, and/or health of fish, other aquatic species and
some terrestrial trees
¦ Regulating: decline in water and soil quality
¦ Cultural: decline in forest aesthetics, fishing, ecotourism and cultural heritage values related to
ecosystem integrity and biodiversity
4.3. Nitrogen Nutrient Enrichment
NOx and NHX are the main contributors to N deposition across the U.S (see Chapter 2). Given the
complexity of the N cycle, a broadly applicable and well-tested predictive model of the ecological effects
of N deposition is not available. However, there is substantial empirical information for specific
ecosystems and endpoints about ecological and biogeochemical responses to N deposition. The most
commonly used experimental designs are N addition, N deposition gradient, and observational studies
that evaluate relationships between effects and changing pollution levels over time (see Table 4-4 and
Table 3-25). N addition experiments often use NH4N03 or (NEL^SC^ additions to simulate the chemical
species in atmospheric Nr deposition. Deposition gradient experiments often only measure oxidized and
reduced forms of N. Therefore, publications addressing N additions or deposition often do not include
4-13
-------
data on all components of Nr. Thus throughout this Section, causality determinations are made on effects
of N deposition that incorporate evidence related to various forms of N.
The different ecosystem types that occur across the U.S. have a broad range of sensitivity to N
deposition. In general, N deposition to terrestrial ecosystems causes accelerated growth rates in some
species, which may lead to altered competitive interactions among species and nutrient imbalances,
ultimately affecting biodiversity. The onset of these effects occurs with N deposition levels as low as 3 kg
N/ha/yr in sensitive terrestrial ecosystems. In aquatic ecosystems, N that is both leached from the soil and
directly deposited can pollute surface water. This causes alteration of the diatom community at levels as
low as 1.5 kg N/ha/yr in sensitive freshwater ecosystems. In estuarine and near coastal marine
ecosystems, total N loading, of which atmospheric deposition is a contributing factor, promotes
eutrophication. Eutrophication is a serious problem in many coastal areas of the U.S.
Factors that govern the sensitivity of terrestrial ecosystems to nutrient enrichment from N
deposition include the degree of N-limitation to plant growth, rates and form of N deposition, elevation,
climate, species composition, length of growing season, and soil N retention capacity. Critical N loads are
described for European ecosystems (see Section 3.3.7.1). The range of critical loads in Europe varies by
ecosystem type: terrestrial ecosystems are between 5-30 kg N/ha/yr; freshwater wetlands are between 5-
35 kg N/ha/yr; inland surface waters are between 5-20 kg N/ha/yr; coastal habitats are between 20-25 kg
N/ha/yr; and marine habitats are between 30-40 kg N/ha/yr. Less is known about the extent and
distribution of the terrestrial ecosystems in the U.S. that are most sensitive to the effects of nutrient
enrichment from atmospheric N deposition; however available data for quantified relationships between
loading and ecological effects are discussed in Section 3.3.7.2 and in Table 4-4. Table 4-4 summarizes
field studies of the ecological effects of N deposition and N addition. It includes 51 studies that were
conducted in the U.S. and 8 that were conducted in Europe and Asia since the last UNECE critical loads
assessment. The ecosystem types from the U.S. include coastal sage scrub, desert, sub-alpine forest,
coniferous forest, mixed hardwood forest, chaparral, oak savanna, grassland, freshwater lakes and
freshwater wetlands. Endpoints include lichen, mycorrhizae, herbaceous and woody vascular plant
species, algae and multiple biogeochemical indicators. Critical loads are available for few ecosystems and
endpoints in the U.S.
4.3.1. Terrestrial
The 1993 NOx AQCD concluded that N deposition may cause important effects on terrestrial
systems, and that the effects are often due to total N loading, not just that the oxidized forms. N deposited
to an N-deficient ecosystem is generally expected to increase growth. If N is deposited on an ecosystem
with adequate N or saturated with N, N03 leaching is expected to occur. Much of the information
presented in the 1993 NOx AQCD was based on results from studies of forest ecosystems. N saturation
was known to be more common in older forests. Disturbances such as fire and harvesting would push
ecosystems to a state more removed from a condition of N saturation. Fertilization was known to increase
growth in N-deficient forests in the short-term, but little was known about long-term effects of N
fertilization and the differential growth effects on various tree and herbaceous plant species. It was known
that plants do not necessarily benefit from added N. When N increases to the point that it is no longer
limiting, deficiencies of other nutrients can occur (Aber et al., 1989).
A few studies documented the deleterious effects of excessive N on tree growth and grassland
biodiversity. Alpine ecosystems were identified as particularly sensitive to N deposition. The studies
published since the 1993 NOx AQCD generally support its conclusions confirm findings of previous
studies, provide more information on the long-term effects of N deposition, and expand the knowledge of
effects to include more ecosystems and species.
4-14
-------
4.3.1.1. Biogeochemical Effects
N cycling
The evidence is sufficient to infer a causal relationship between N deposition and the alteration of
biogeochemical cycling of N in terrestrial ecosystems (Section 3.3.2.1). This is supported by numerous
observational, deposition gradient and field addition experiments. N deposition disrupts the nutrient
balance of ecosystems with numerous biogeochemical effects. The chemical indicators that are typically
measured include N03 leaching, C:N ratio, N mineralization, nitrification, denitrification, foliar N
concentration, and soil water N03 and NH/ concentrations. Note that N saturation does not need to
occur to cause adverse effects on terrestrial ecosystems. However, in some regions N saturation is a
plausible mechanism for net nitrification and associated N03 leaching in drainage water. Substantial
leaching of N03 from forest soils to stream water can acidify downstream waters (see Section 3.2) and
deplete soils of nutrient base cations, especially Ca and Mg (Likens et al., 1998).
¦ Two of the primary indicators of N enrichment in forested watersheds are the leaching of N03 in
soil drainage waters and the export of N03 in stream water, especially during the growing season
(Stoddard, 1994).
¦ There is consistent and coherent experimental evidence that N03 leaching can be induced by
chronic addition of N (Edwards et al., 2002b; Kahl et al., 1999; Kahl et al., 1993; Peteijohn et al.,
1996, Norton et al., 1999). Several N-exclusion studies in Europe demonstrated that decreases in
N deposition produced immediate reductions in N03 leaching from forest stands (Gundersen
et al., 1998; Quist et al., 1999).
¦ In upland forested areas in the U.S., most N received by atmospheric deposition is retained in
soil, and lesser amounts (7-16%) are retained in plant biomass (Nadelhoffer et al., 1999a).
Several different data compilations indicate consistent and coherent results that 80% to 100% of
N deposition is retained or denitrified within terrestrial ecosystems that receive less than about 8-
10 kg N/ha/yr (Aber et al., 2003; Dise and Wright, 1995; Kristensen et al., 2004; MacDonald
et al., 2002; Sullivan, 2000b).
¦ In the West, mixed conifer forests and chaparral watersheds with high smog exposure in the Los
Angeles Air Basin also are N-saturated and exhibit the highest stream water N03 concentrations
within wildlands in North America (Bytnerowicz and Fenn, 1996; Fenn et al., 1998). In the mixed
conifer forests of the Sierra Nevada and San Bernardino mountains, a critical load for increased
N03 leaching has been calculated to be 17 kg N/ha/yr. Several studies in the Rocky Mountains
indicate that the capacity of alpine catchments to sequester N is exceeded at levels greater than 5-
10 kg N/ha/yr (Baron et al., 1994; Williams and Tonnesen, 2000).
¦ Aber et al. (2003) found that surface water N03 concentrations exceeded 1 j^ieq/L in watersheds
receiving about 9 to 13 kg N/ha/yr of atmospheric N deposition. The lakes and streams found to
have high N03 concentration were those receiving N deposition above this range, but responses
were variable among those receiving high N deposition. Above this range, mean N03 export
increased linearly with increasing deposition at a rate of 0.85 kg N03 kg N/ha/yr for every 1 kg
N/ha/yr increase in deposition, although there was considerable variability in N retention among
watersheds at higher rates of deposition.
¦ Activities or disturbances such as logging or fire that export large quantities of N from the site
alter future N availability and site propensity to achieve N saturation (Chanasyk et al., 2003).
C cycling
The evidence is sufficient to infer a causal relationship between N deposition and the alteration of
biogeochemical cycling of C in terrestrial ecosystems (Section 3.3.3.1). The most extensive evidence on
4-15
-------
the interactions between N deposition and C cycling is available for forest ecosystems. Experimental N
addition studies show a range of responses in terms of tree mortality and productivity. In general,
moderate to high additions of N lead to either no significant change in growth rates or transient growth
increases (generally at deposition rates lower than 10 kg N/ha/yr), followed by increased mortality,
especially at higher rates of fertilization (see Section 3.3.3.1). This group of studies shows coherence in
effects, and indicates the presence of a biological gradient in responses with increasing N deposition.
Due to the complexity of interactions between the N and C cycling, the effects of N on C budgets
(quantified input and output of C to the ecosystem) are variable. Regional trends in NEP of forests have
been estimated through models based on gradient studies (Magnani et al., 2007). There have been
critiques of the method and the magnitude of these reported effects (Sutton et al., 2008b). N addition was
found to slightly increase ecosystem C in a meta-analysis that examined the effects of N fertilization
ranging from 25.5 to 200 kg N/ha/yr on forest ecosystem C content (see Section 3.3.3.1). In the western
U.S., atmospheric N deposition has been shown to cause increased litter accumulation and carbon storage
in above-ground woody biomass, which in turn may lead to increased susceptibility to more severe fires
(Fenn et al., 2003a).
Less is known regarding the effects of N deposition on C budgets of non-forest ecosystems. A
meta-analysis, including 16 observations from 9 publications, conducted to evaluate the relationship
between N addition ranging from 16 to 320 kg N/ha/yr and C sequestration of non-forest ecosystems
showed that N addition has no significant effect on net ecosystem exchange of non-forest ecosystems.
N deposition also affects the patterns of C allocation because most growth occurs above ground.
This increases the shoot-to-root ratio, which can be detrimental to the plant because of decreased
resistance to environmental stressors, such as drought and windthrow (Braun et al., 2003; Fangmeier
et al., 1994b; Krupa, 2003; Minnich et al., 1995).
N2O and ChUflux
The evidence is sufficient to infer a causal relationship between N deposition and the alteration of
biogeochemical flux of N2O in terrestrial ecosystems (Section 3.3.4.2). Terrestrial soil is the largest source
of N20, accounting for 60% of global emission (Johnson, 2002). In a meta-analysis of 80 observations of
terrestrial ecosystems that received chemical forms of N (NH/, N03 . NH4NO3, and urea) and addition
rates (10 to 562 kg N/ha/yr), N addition resulted in a two-fold increase in N20 emission. The response of
N20 emission to N addition for coniferous forest, deciduous forest and grasslands was significant (see
Section 3.3.4.2).
The evidence is sufficient to infer a causal relationship between N deposition and the alteration of
biogeochemical flux of CH4 in terrestrial ecosystems (Section 3.3.4.1). Non-flooded upland soil is the
largest biological sink for atmospheric CH4, consuming about 6% of the atmospheric CH4 (Le Mer and
Roger, 2001). A meta-analysis was performed on a data set of 41 observations including four forms of N
(NH4+, N03 . NH4NO3 and urea) and the addition rates ranging from 10 to 560 kg N/ha/yr. The results
indicated that N addition reduced CH4 uptake, but this inhibition was significant only for coniferous and
deciduous forests (see Section 3.3.4.1).
4.3.1.2. Species Richness, Composition and Biodiversity
The evidence is sufficient to infer a causal relationship between N deposition on the alteration of
species richness, species composition and biodiversity in terrestrial ecosystems. The most sensitive
terrestrial taxa are lichens. Empirical evidence indicates that lichens in the U.S. are adversely affected by
deposition levels as low as 3 kg N/ha/yr. Alpine ecosystems are also sensitive to N deposition, changes in
an individual species (Carex rupestris) were estimated to occur at deposition levels near 4 kg N/ha/yr and
modeling indicates that deposition levels near 10 kg N/ha/yr alter plant community assemblages. A
summary of N deposition effects are presented below, organized by ecosystem type.
4-16
-------
Lichens
There is consistent and coherent evidence indicating that lichen communities are affected by
current levels of N deposition. Sensitive lichens are frequently used as indicators of air pollution and
atmospheric deposition levels. In addition to being good subjects for biomonitoring, they constitute
important components of the forest ecosystem by contributing to biodiversity, regulating nutrient and
hydrological cycles, and providing habitat elements for wildlife (McCune and Geiser, 1997); see
Section 3.3.5).
Lichens that contain a cyanobacterial photobiont appear to be more sensitive to adverse effects
from atmospheric N deposition than most other lichens (Hallingback, 1991; Hallingback and Kellner,
1992). The decline of lichens containing cyanobacteria in parts of northern Europe has been associated
with N deposition in the range of 5 to 10 kg N/ha/yr (Bobbink et al., 1998). In the U.S., lichen species are
negatively affected by N inputs as low as 3 to 8 kg N/ha/yr (Fenn et al., 2003a).
In the San Bernardino Mountains, California, up to 50% of lichen species that occurred in the
region in the early 1900s have disappeared (Fenn et al., 2003a; Nash and Sigal, 1999). The critical load
has been calculated for lichen communities in mixed conifer forests in California at 3.1 kg N/ha/yr (Fenn
et al., 2008).
The Pacific Northwest retains widespread populations of pollution-sensitive lichens (Fenn et al.,
2003a). In this area, lichen communities are beginning to show evidence of changes in response to
increased N pollution, including decreased distribution of sensitive lichen taxa, and their replacement
with nitrophilous species (Geiser and Neitlich, 2007).
Alpine Plant Communities
Consistent and coherent evidence indicates that alpine plant communities are among the most
sensitive terrestrial communities to atmospheric N deposition. Factors that govern the sensitivity of alpine
tundra to N deposition include low rates of primary production, short growing season, low temperature,
and wide variation in moisture availability in the alpine environment (Bowman et al., 1993; 1994;
Bowman and Fisk, 2001; Fisk et al., 1998). Alpine herbaceous plants are generally considered N-limited
and changes in alpine plant productivity and species composition have been noted in response to
increased N inputs (Vitousek et al., 1997; Bowman et al., 2006). Alpine plant communities have also
developed under conditions of low nutrient supply, in part because soil-forming processes are poorly
developed, and this also contributes to their N-sensitivity.
The western U.S. contains extensive land areas that receive low levels of atmospheric N deposition,
interspersed with hot spots of relatively higher N deposition that typically occur downwind of large
metropolitan centers and agricultural areas. Some of these areas of higher N deposition occur at high
elevation. Results from several studies suggest that the capacity of Rocky Mountain alpine catchments to
sequester N is exceeded at input levels less than 10 kg N/ha/yr (Baron et al., 1994; Williams, 1999).
Changes in an individual species (Carex rupestris) were estimated to occur at deposition levels near
4 kg N/ha/yr. Changes in the plant community were estimated to occur at deposition levels near 10 kg
N/ha/yr. (Bowman et al., 2006). In comparison, critical loads for alpine plant communities in Europe are
between 5-15 kg N/ha/yr (Bobbink et al., 2003).
Grasslands
Consistent and coherent evidence for reduced biodiversity in response to N deposition is reported
for grasslands in the U.S. and Europe. Clark and Tilman (2008) evaluated the effects of chronic N
addition over 23 years in Minnesota prairie-like successional grasslands and in native savanna grassland
and found species numbers declined at the lowest addition level (10 kg N/ha/yr added to 6 kg N/ha/yr of
ambient deposition). The authors calculated the critical load as 5.3 kg N/ha/yr with an inverse prediction
interval of 1.3-9.8 kg N/ha/yr.
4-17
-------
Change in species composition in response to N deposition has been observed regardless of soil
type in European grasslands. Such effects have been found in calcareous, neutral, and acidic
environments, species-rich heaths, and montane-subalpine grasslands (Bobbink et al., 1992b; 1998;
Stevens, 2004).In a transect of 68 acid grasslands across Great Britain, covering the lower range of
ambient annual N deposition (5 to 35 kg N/ha/yr) chronic N deposition significantly reduced plant species
richness. Species richness declined as a linear function of the rate of inorganic N deposition, with a
reduction of one species per 4 m2 quadrant for every 2.5 kg N/ha/yr of chronic N deposition. The critical
loads for ten different types of grasslands in Europe ranged between 10 and 30 kg N/ha/yr, above which
changes in species composition were reported (Bobbink et al., 2003) (See Section 3.3.7.1).
In the San Francisco Bay area of California, which receives N deposition levels of 10 to
15 kg N/ha/yr, exotic nitrophilous grasses have displaced native grass species, likely due to greater N
availability from deposition and from the cessation of grazing, which previously exported N out of the
system (Fenn et al., 2003b).
Forests
Forests include overstory trees, understory herbaceous plants and mycorrhizae. There is very little
information on the effect of N deposition on the biodiversity of overstory trees within forests in the U.S.
This is due to the long life span and slow growth of trees, which makes such changes difficult to detect. A
study of the northern edge of the Great Plains (southern Canada), showed that increasing N deposition
over a range of 8 to 22 kg N/ha/yr to aspen-dominated and boreal forests increase forest expansion into
the grasslands (Kotchy and Wilson 2001). More is known concerning the effects of N deposition on
understory herbs, however most of the evidence is from Europe, where alteration of species composition
is known to occur over the gradient of N deposition ranging from 6 to 20 kg N/ha/yr for acid tolerant
species and a decline in the cover and abundance of ericaceous shrubs along a gradient from 3 to 12 kg
N/ha/yr (Gilliam, 2006a). Loss of mycorrhizal diversity was recorded for Alaskan coniferous forest over a
gradient of 1 to 20 kg N/ha/yr, and studies in oak savanna ecosystems in Minnesota show N addition
decreases mycorrhizal diversity (Avis et al., 2003).
Arid and Semi-arid Grasslands
Alteration to arid and semi-arid plant communities resulting from experimental N fertilization have
been reported in the Colorado Plateau, Joshua Tree National Park in California, and the coastal sage scrub
community (CSS) of Southern California.
Results from several lines of evidence showed increased biomass of non-native plant species over
native species; decreased soil moisture under some conditions; and increased fire risk where dense grasses
replaced shrub cover.
In some areas of the CSS of Southern California, dry N deposition may be upwards of
30 kg N/ha/yr (Bytnerowicz and Fenn, 1996). Native shrub and forb seedlings in this plant community are
unable to compete with dense stands of exotic grasses, and thus are gradually replaced by the grasses,
especially following disturbances such as fire (Cione et al., 2002; Eliason and Allen, 1997; Yoshida and
Allen, 2001). The CSS community in California has been declining in land area and in shrub density for
the past 60 years and is being replaced in many areas by Mediterranean annual grasses (Allen et al., 1998;
Padgett and Allen, 1999; Padgett et al., 1999}. N deposition is considered a possible cause or contributor
to this ecosystem alteration.
Egerton-Warburton and Allen (2000) discerned a shift in arbuscular mycorrhizal community
composition with decreased species richness and diversity along a deposition gradient (2 to 57 |_ig N/g as
soil N03 ). These shifts in mycorrhizal fungal communities may facilitate replacement of native plant
communities by Mediterranean annual grasslands in CSS.
A coherent body of evidence suggests that N deposition may be contributing to greater fuel loads
and thus altering the fire cycle in a variety of ecosystem types (Fenn, 2003b). Invasive grasses, which can
be favored by high N deposition, promote a rapid fire-cycle in many locations (D'Antonio and Vitousek,
4-18
-------
1992). The increased productivity of flammable understory grasses increases the spread of fire and has
been hypothesized as one mechanism for the recent conversion of CSS to grassland in California
(Minnich and Dezzani, 1998).
Deserts
Consistent and coherent evidence shows that N fertilization alters desert plant communities in the
Chihuahuan Desert, Jordana Basin, Mojave Desert, and the Great Basin. N additions stimulate plant
growth and cause the observed invasion of some exotic plant species and associated changes in ecosystem
function, especially where water supply is adequate. There is little evidence evaluating biological
responses across deposition gradients. However there are numerous field experiments that evaluate N
addition levels ranging from 10-100 kg N/ha/yr. This is within the range of deposition that occurs in the
U.S., which may be as high as 30-90 kg N/ha/yr downwind of major urban and agricultural areas (Fenn et
al., 2003). Increased grass biomass has also been associated with increased fire frequency in the Mojave
Desert (Brooks, 1999; Brooks and Esque, 2002; Brooks et al., 2004). This effect is most pronounced at
higher elevation, probably because the increased precipitation at higher elevation contributes to greater
grass productivity. In some cases, precipitation may be a more limiting factor than N to plant growth in
deserts. Increased N supply at lower elevation in arid lands can only increase productivity to the point at
which moisture limitation prevents additional growth. Fire was relatively rare in the Mojave Desert until
the past two decades, but now fire occurs frequently in areas that have experienced invasion of exotic
grasses (Brooks, 1999).
4.2.3. Transitional
Anaerobic conditions of waterlogged soils in wetlands result in slow decomposition of organic
matter and accelerated denitrification. N cycles of two types of wetland, ombrotrophic bogs and coastal
salt marshes, were discussed in 1993 NOx AQCD. Ombrotrophic bogs are generally considered the most
sensitive to atmospheric N deposition because they are nutrient poor, with a closed N cycle in which the
predominant source of N is rainfall. The 1993 NOx AQCD found that the three main ecological effects of
N deposition on wetland ecosystem are: increasing primary production; modifying microbial processes;
and reducing biodiversity. Recent studies support and extend the conclusions in the 1993 NOx AQCD,
especially with regard to the effects of N deposition on species diversity.
4.3.2.1. Biogeochemical Effects
The contribution of N deposition to total N load varies among wetland types. A wetland is more
vulnerable to N deposition as the relative contribution of N deposition to its total N load increases. For
example, in freshwater wetland ecosystems atmospheric deposition is the main source of N to the
ecosystem while N deposition is a minor contributor to N load in many coastal estuarine wetlands.
N Cycling
The evidence is sufficient to infer a causal relationship between N deposition and the alteration of
biogeochemical cycling of N (Section 3.3.2.2). N deposition contributes to total N load in wetlands. The
chemical indicators that are typically measured include N03 leaching, N mineralization, and
denitrification rates. N dynamics in wetland ecosystems are variable in time, among types of wetlands and
environmental factors, especially water availability (Howarth et al., 1996). A wetland can act as a source,
sink, or transformer of atmospherically deposited N (Devito et al., 1989) and these functions can vary
with season and with hydrological conditions. Vegetation type, physiography, local hydrology, and
4-19
-------
climate all play significant roles in determining source/sink N dynamics in wetlands (Arheimer and
Wittgren, 1994; Devito et al., 1989; Koerselman et al., 1993; Mitchell et al., 1996).
N mineralization has been shown to increase with N addition, and this can cause an increase in
wetland N export to adjacent surface water (Groffman, 1994). In general, leaching losses of N03 in water
derived directly from wetlands are often small because of N03 removal by denitrification. Elevated N
inputs to wetlands will often increase the rate of denitrification (Broderick et al., 1988; Cooper, 1990;
Dierberg and Brezonik, 1983). This decreases environmental effects associated with increased N supply
to soils and drainage waters; however it increases the emissions of greenhouse gases (e.g., N20) to the
atmosphere. Denitrification appears to be negligible in wetland environments that are typically nutrient
(including N) poor, such as some bogs and fens (Morris, 1991).
C Cycling
The evidence is sufficient to infer a causal relationship between N deposition and the alteration of
biogeochemical cycling of C. A meta-analysis that included wetlands with other non-forest ecosystems
indicated no effect ofN deposition on overall net ecosystem exchange of C (See Section 3.3.3). In other
words, any gain in C capture by photosynthesis was offset by ecosystem respiration and C leaching. There
were not enough studies to evaluate wetlands as a separate category. There is evidence that above and
below ground C exchange processes are affected by N deposition. In Sphagnum-dominated ombrotrophic
bogs, higher N deposition resulted in higher tissue N concentrations and greater NPP (Aldous, 2002), but
lower bulk density. A study of 23 ombrotrophic peatlands in Canada with deposition levels ranging from
2.7 to 8.1 kg N/ha/yr showed peat accumulation increases linearly with N deposition; however in recent
years this rate has begun to slow indicating limited capacity for N to stimulate accumulation (Turunen et
al., 2004). Soil respiration has been studied in European countries under a natural gradient of atmospheric
N deposition from 2 to 20 kg N/ha/yr. They found enhanced decomposition rates for material accumulated
under higher atmospheric N supplies resulted in higher carbon dioxide (C02).
In intertidal wetlands, primary production of plant species typically increases with N addition,
however most studies apply fertilizer treatments that are several orders of magnitude larger than
atmospheric deposition (Mendelssohn, 1979; Darby et al., 2008; Wigand et al., 2003, Tyler et al., 2007).
N fertilization experiments in salt marsh ecosystems show biomass stimulation from 6 to 413% with
application rates ranging from 7 to 3120 kg N/ha/yr (U.S. EPA, 1993a).
Increases in biomass linked to N deposition, have also increased evapotranspiration rates (Howes,
1986). This changed the soil water balance of water and may influence the direction of plant community
succession. Model results suggest 7 kg N/ha/yr is the threshold for an oligotrophic bog to become a
mesotrophic bog dominated by trees, as found in the 1993 NOx AQCD.
N2O and ChUflux
The evidence is sufficient to infer a causal relationship between N deposition and the alteration of N2O
flux in wetland ecosystems. Nineteen observations of the effects of different N forms (NH4+, N03 .
NH4NO3, and urea) and addition rates (15.4 to 300 kg N/ha/yr) on wetland N20 emissions were evaluated
in a meta-analysis (see Section 3.3.4.2; Mendelssohn, 1979). The results indicated that N addition
increased the production of N20 by about two-fold.
The evidence is sufficient to infer a causal relationship between N deposition and the alteration of CH4
flux in wetland ecosystems. Wetlands are generally net sources of CH4, but some wetlands can be net sinks
depending on environmental conditions such as drainage and vegetation (Crill et al., 1994; Saarnio et al.,
2003). A meta-analysis was performed on a data set of 17 observations to assess the effects of N additions
on CH4 fluxes (see Section 3.3.4.1). This data set included four forms of N (NH/, N03 . NH4NO3 and
urea) and the addition rates ranged from 30 to 240 N kg N/ha/yr (see Section 3.3.4.1). The results
indicated that N addition increased CH4 production from the wetlands, but had no significant effect on
CH4 uptake of wetlands.
4-20
-------
4.3.2.2. Biological Effects
The evidence is sufficient to infer a causal relationship between N deposition and the alteration of
species richness, species composition and biodiversity in wetland ecosystems (Section 3.3.5.2). Wetlands
contain a high number of rare plant species. Excess N deposition can cause shifts in wetland community
composition by altering competitive relationships among species, which potentially leads to effects such
as decreasing biodiversity, increasing non-native species establishment and increasing the risk of
extinction for sensitive and rare species.
Changes in plant species composition caused by elevated atmospheric N deposition haven been
demonstrated in Europe. Achermann and Bobbink (2003), see Table 3-24, evaluated the empirical
evidence linking N deposition to wetland species composition and biodiversity to develop the following
critical loads in Europe: raised and blanket bogs = 5-10 kg N/ha/yr; poor fens = 10-20 kg N/ha/yr; rich
fens 15-35 kg N/ha/yr; mountain rich fens 15-25 kg N/ha/yr; pioneer and low-mid salt marshes
30-40 kg N/ha/yr.
Some wetland species are adapted to low-N environments. High levels of atmospheric N deposition
increase the risk of decline and extinction of those sensitive species. In general, these include the genus
Isoetes sp., of which three species are federally endangered; insectivorous plants like the endangered
green pitcher Sarracenia oreophila; and the genus Sphagnum, of which there are 15 species listed as
endangered by eastern U.S. states. Roundleaf sundew (Drosera rotundifolia) is also susceptible to
elevated atmospheric N deposition (Redbo-Torstensson, 1994). This plant is native to, and broadly
distributed across, the U.S. and is listed as endangered in Illinois and Iowa, threatened in Tennessee, and
vulnerable in New York . In the U.S., Sarracenia purpurea can be used as a biological indicator of local N
deposition in some locations (Ellison and Gotelli, 2002). S. purpurea is a perennial pitcher plant native to
Canada and the eastern U.S. that grows in nutrient-poor peatlands and is sensitive to changes in N
availability. Based on the annual demographic rates, a non-stationary matrix model forecasted that the
extinction risk within the next 100 years increased substantially if N deposition rate increase (1-4.7%)
from the rate of 4.5-6.8 kg N/ha/yr (Gotelli and Ellison, 2002).
4.3.2.3. Regional Vulnerability and Sensitivity
Bogs are among the most sensitive wetland ecosystems to N deposition. In the U.S., peat-forming
bogs are most common in areas that were glaciated, especially in portions of the Northeast and Upper
Midwest (U.S. EPA, 1993a).
N input and output rates of fens are intermediate between bogs and coastal marshes. N deposition
could drastically change species composition, increase primary productivity and increase methane
emission in fens (Aerts and de Caluwe, 1999; Pauli et al., 2002).
Atmospheric N inputs contribute to eutrophication problems in coastal marshes at many locations.
However marine inputs of N are typically higher than direct atmospheric input. Models of sources of N to
wetland ecosystems are not yet available.
The effect of N deposition on wetland ecosystems depends on the fraction of rainfall in its total
water budget and the sensitivity to N deposition was suggested as bogs (70-100% rainfall) >fens (55-83%
rainfall) >intertidal wetlands (10-20% rainfall) (Morris, 1991).
4.3.3. Freshwater Aquatic
The 1993 NOx AQCD concluded that productivity of fresh water is usually limited by the
availability of phosphorus (P). However, it was noted that high inputs of P from anthropogenic sources
could lead to N limitation. The ratio of dissolved N to total phosphorus (molar basis) was used as an
4-21
-------
indicator for nutrient limitation, with values less than 2 indicating N limitation. The proportions of N
limited lakes showed a wide regional variation: Pacific Northwest (27.7%), Upper Midwest (19%),
Northeast (5%), and Southeast (2.5%). All sub regions of the West contain substantial numbers of N
limited lakes.
4.3.3.1. Biogeochemical Effects
N Cycling
The evidence is sufficient to infer a causal relationship between N deposition and the alteration of
biogeochemical cycling of N in freshwater aquatic ecosystems (Section 3.3.2.3). N deposition is the main
source ofN enrichment to headwater streams, lower order streams and high elevation lakes. The chemical
indicators that were studied included N03 and DIN concentration in surface waters as well as Chi a:total
P ratio. Elevated surface water N03 concentrations occur in both the eastern and western U.S. Bergstrom
and Jansson (2006) report a significant correlation between N deposition and lake biogeochemistry by
identifying a correlation between wet deposition and [DIN] and Chi a: Total P. Recent evidence provides
examples of lakes and streams that are limited by N and show signs of eutrophication in response to N
addition.
C Cycling
The evidence is sufficient to infer a causal relationship between N deposition and the alteration of
biogeochemical cycling of C in freshwater aquatic ecosystems (Section 3.3.3.3). If growth of the
autotrophic community of a freshwater stream is N-limited, then N addition will stimulate C-capture via
photosynthesis often altering the C cycle. Moreover, a freshwater lake or stream must be N-limited to be
sensitive to N-mediated eutrophication. Elser et al. (2008) in a meta-analysis of over 600 experiments
found that N-limitation occurs frequently in freshwater ecosystems, in contrast to the traditional
paradigm. There are also many examples of fresh waters that are N-limited or N and P co-limited (See
Annex C). Bergstrom and Jansson (2006) concluded that most lakes in the northern hemisphere are
limited by N in their natural state.
Numerous studies investigate the relationship between N concentration of freshwater and primary
productivity (reported as Chi a, NPP, or an index such as the lake chemistry ratio of DIN: TP) and
atmospheric N deposition. Typically N addition experiments of lake and stream bioassays in which N was
added to waters in field or laboratory to measure the response are conducted. A meta-analysis of
enrichment bioassays in 62 freshwater lakes of North America found algal growth enhancement from
N amendments to be common in slightly less than half the studies (Elser et al., 1990). Gradient studies of
undisturbed northern temperate, mountain, or boreal lakes, that generally receive low levels of
atmospheric N deposition found strong relationships between N-limitation and productivity where N
deposition was low, and P and N+P limitations where N deposition was higher (Fenn, 2003b; Bergstrom
et al., 2005; Bergstrom and Jansson, 2006). One such study in Sweden found, the lowest productivity at
sites where wetN deposition was about 1.3 kg N/ha/yr; increasing productivity occurred at deposition
levels greater than 2.2 kg N/ha/yr (Bergstrom et al., 2005).
4.3.3.2. Biological Effects
The evidence is sufficient to infer a causal relationship between N deposition and the alteration of
species richness, species composition and biodiversity in freshwater aquatic ecosystems (Section 3.3.5.3).
Increased N deposition can cause a shift in community composition and reduce algal biodiversity,
especially in sensitive oligotrophic lakes.
4-22
-------
In the West, a hindcasting exercise determined that the change in Rocky Mountain National Park
lake algae that occurred between 1850 and 1964 was associated with an increase in wet N deposition that
was only about 1.5 kg N/ha (Baron, 2006). Similar changes inferred from lake sediment cores of the
Beartooth Mountains of Wyoming also occurred at about 1.5 kg N/ha deposition (Saros et al., 2003).
Some freshwater algae are particularly sensitive to added nutrient N and experience shifts in
community composition and biodiversity with increased N deposition. For example, two species of
diatom (a group of algae), Asterionella formosa and Fragilaria crotonensis, now dominate the flora of at
least several alpine and montane Rocky Mountain lakes and sharp increases have occurred in Lake Tahoe
(Baron et al., 2000; Interlandi and Kilham, 1998; Saros et al., 2003; Saros et al., 2005; Wolfe et al.,
2001a; Wolfe et al., 2003). The timing of this shift has varied, with changes beginning in the 1950s in the
southern Rocky Mountains and in the 1970s or later in the central Rocky Mountains. These species are
opportunistic algae that have been observed to respond rapidly to disturbance and slight nutrient
enrichment in many parts of the world.
Extremely high N03 concentrations can have direct adverse effects on fish, invertebrates and
amphibians, but the concentrations required to elicit such effects are typically more than 30 times higher
than those that would commonly be attributable to atmospheric deposition. For example, mortality of
rainbow trout eggs and fry occurred after 30-day incubations in concentrations greater than 79 jj.g N/L;
adverse effects on amphibians and insects occur at even higher concentrations.
4.3.3.3. Regional Vulnerability and Sensitivity
Eutrophication effects on freshwater ecosystems from atmospheric deposition of N are most likely
to occur in lakes and streams that have low productivity, low nutrient levels and that are located in the
most undisturbed areas.
¦ In the western U.S., high-elevation lakes are considered the most sensitive aquatic ecosystems to
N deposition. Some examples include the Snowy Range in Wyoming, the Sierra Nevada
Mountains, and Lake Tahoe in California, and the Colorado Front Range.
¦ The most severe eutrophication from N deposition effects are expected downwind of major urban
and agricultural centers.
4.3.4. Estuarine Aquatic
The 1993 NOx AQCD concluded that the primary effect of N enrichment effect on aquatic
ecosystems is eutrophication of estuarine and near-coastal marine waters, which results in an increase of
algal biomass and changes in community composition. Recent studies generally support and expand upon
the conclusions of 1993 NOx AQCD. The data for estimating the contribution ofN deposition to the
nutrient budget of aquatic ecosystems were very sparse and mainly limited to the Chesapeake Bay before
1993. The contribution of N deposition to estuarine eutrophication is now better understood in the
Chesapeake Bay and other estuaries.
4.3.4.1. Biogeochemical Effects
A recent national assessment of eutrophic conditions in estuaries found that 65% of the assessed
systems had moderate to high overall eutrophic conditions and generally received the greatest N loads
from all sources, including atmospheric and land-based sources (Bricker et al., 2007). Estuarine and
coastal marine ecosystems experience a range of ecological problems associated with nutrient enrichment.
4-23
-------
Because the productivity of estuarine and near shore marine ecosystems is generally limited by the
availability of N, they are susceptible to eutrophication effect of N deposition.
N Cycling
The evidence is sufficient to infer a causal relationship between Nr deposition and the biogeochemical
cycling of N (Section 3.3.2.4). The N load from atmospheric deposition is estimated to comprise 10% to
40% of the total input of N to many coastal estuaries, and could be higher for some. Atmospheric N loads
to great waters and estuaries in the U.S. are estimated to range from 2 to 8% for Guadalupe Bay, TX on
the lowest end to -72% for the St. Catherines-Sapelo estuary, GA (Castro et al., 2003) on the highest. At
Chesapeake Bay, where N and S deposition and ecological effects have been extensively studied, total
atmospheric deposition of atmospheric N03 is estimated to contribute from 20% to 30% of total N and
14% of the NH4 loadings to the Bay.
Estimates of total N loadings to estuaries are computed using measurements of wet and dry N
deposition where these are available and interpolated with or without the use of air quality models. Direct
atmospheric inputs (directly to the water surface) of Nr to coastal waters are essentially equal to or greater
than those contained in riverine flow in the absence of deposition and may contribute from 20 to >50% of
N loadings to these systems: 11, 5.6, and 5.6 kg N/hafor the Northeast Atlantic coast of the U.S., the
Southeast Atlantic coast of the U.S., and the U.S. Eastern Gulf of Mexico, respectively.
It is unknown if current levels of atmospheric deposition alone are sufficient to cause
eutrophication. In general, estuaries tend to be N-limited (Elser et al., 2008), and many currently receive
high levels of N input from human activities to cause eutrophication (Howarth et al., 1996; Vitousek and
Howarth, 1991). The most widespread chemical indicators of eutrophication are submerged aquatic
vegetation, Chi a, algal blooms, macroalgae and dissolved 02.
C Cycling
The evidence is sufficient to infer a causal relationship between N deposition and alteration to the
biogeochemical cycling of C (Section 3.3.3.4). Estuaries and coastal waters tend to be N-limited and are
therefore inherently sensitive to increased atmospheric N loading (D'Elia et al., 1986; Howarth and
Marino, 2006; Eisner et al., 2007). This is at least partly because denitrification by microbes found in
estuarine and marine sediments releases much of the added N inputs back into the atmosphere (Vitousek
et al., 1997). However, other limiting factors occur in some locations and during some seasons. Levels of
N limitations are affected by seasonal patterns. N-limited conditions are likely to be found during the
peak of annual productivity in the summer.
Numerous studies evaluate the relationship between N loading, eutrophication and ecological
endpoints including Chi a concentration, macroalgal abundance, dissolved 02, nuisance or toxic and algal
blooms In the national estuary condition assessment, high Chi a concentration was the most widespread
documented indicator of eutrophication (Bricker et al., 2007) (see Figure 3-42).
Excess N inputs will affect the Si:N ratio in water. If the Si:N ratio decreases below about 1, the
marine food web structure would be expected to change, with decreasing diatom-to-zooplankton-to-
higher tropic level ratios and increasing abundance of flagellated algae.
4-24
-------
Table 4-3. Indicators of estuarine eutrophication.
Primary Symptom
Description
Chi a
Excess N input will stimulate primary productivity and Chi a concentration
indicator of phytoplankton biomass
Macroalgal Abundance
Macroalgal blooms were moderate or high for half of the nation's assessed
estuaries (Bricker et al., 2007). Macroalgal blooms can cause the loss of
important submerged aquatic vegetation by blocking sunlight.
Dissolved O2
Dissolved O2 concentration decreases with increasing algal abundance under
elevated N, because microbes consume O2 as they decompose dead algae.
Increased atmospheric N deposition could stimulate the development of hypoxic
or anoxic zones. The northern Gulf of Mexico is the largest documented zone of
hypoxic coastal water in U.S.
Nuisance/Toxic Algal Blooms
Excess N input can cause nuisance or toxic algal blooms, which release toxins
in the water that can poison aquatic animals and threaten human health. About
one third of the nation's assessed estuary systems exhibited a moderate or high
symptom expression for nuisance or toxic algae (Bricker et al., 2007).
Source: Brickeret al. (2007).
4.3.4.2. Biological Effects
The evidence is sufficient to infer a causal relationship between N deposition and the alteration of
species richness, species composition and biodiversity in estuarine ecosystems (Section 3.3.5.4).
Increased N deposition can cause shifts in community composition, reduced hypolimnetic DO, reduced
biodiversity, and mortality of submerged aquatic vegetation. The form of deposited N can significantly
affect phytoplankton community composition in estuarine and marine environments. Small diatoms are
more efficient in using N03 than NH/. Increasing NH/ deposition relative to N03 in the eastern U.S.
favors small diatoms at the expense of large diatoms. This alters the foundation of the food web.
Submerged aquatic vegetation is important to the quality of estuarine ecosystem habitats because it
provides habitat for a variety of aquatic organisms, absorbs excess nutrients, and traps sediments. Nutrient
enrichment is the major driving factor contributing to declines in submerged aquatic vegetation coverage.
The Mid-Atlantic region is the most heavily impacted area in terms of moderate or high loss of
submerged aquatic vegetation due to eutrophication. Indicators to assess the eutrophic condition of
estuarine and coastal waters are given in Table 4-3. Because estuaries receive N from multiple sources, it
is unknown if N deposition alone could cause estuary eutrophication.
4.3.4.3. Regional Vulnerability and Sensitivity
The most eutrophic estuaries were generally those that had large watershed-to-estuarine surface
area, high human population density, high rainfall and runoff, low dilution, and low flushing rates
(Bricker et al., 2007). The national estuary condition assessment conducted by Bricker et al. (2007) found
the most eutrophic estuaries occurred in the mid-Atlantic region and the estuaries with the lowest degree
of eutrophication were in the North Atlantic. Other regions had mixtures of low, moderate, and high
degree of eutrophication. The regional assessment results from the report of Bricker et al. (2007) are
summarized in Section 3.3.8.
The Chesapeake Bay is the largest estuary in the U.S. Its watershed covers 64,299 square miles and
the surface area of the bay and its major tributaries is 4,479 square miles (Pyzik et al., 2004). The
4-25
-------
Chesapeake Bay is perhaps the best-documented case study in the U.S. of the effects of human activities
on estuarine eutrophication. Recent studies (Boyer et al., 2002; Howarth, 2007) indicated that
atmospheric deposition makes a substantial contribution (about 25%) to the overall N budget of
Chesapeake Bay. Human disturbances, such as landscape changes, have exacerbated the negative impacts
of N deposition by reducing N removal and retention in the upper watershed region. Anthropogenic N
inputs have substantially altered the trophic condition of Chesapeake Bay over the last 50 to 100 years.
Signs of eutrophication in the bay include high algal production, low biodiversity, and large hypoxia and
anoxia zones. Submerged aquatic vegetation was once abundant in Chesapeake Bay, covering about
200,000 acres along the shallows and shorelines. Increased nutrient inputs caused submerged aquatic
vegetation declines since the mid-1960s, and had fallen to about 38,000 acres by 1984. Eutrophication has
been implicated in declines and disappearance of striped bass (Morone saxatilis) and blue crab
(Callinectes sapidus) in the Chesapeake Bay.
4.3.5. Ecosystem Services
Some valuation studies assess the effects of N enrichment from multiple sources (see Annex
F).There are no publications at this time which focus on the ecosystem services specifically affected by N
deposition. The evidence reviewed in this ISA illustrates that N deposition can affect ecosystem services
in the following categories (defined by Hassan et al., 2005):
¦ Supporting: nutrient cycling, biodiversity
¦ Provisioning: forest yields, fishing yields in estuaries
¦ Regulating: water quality, air quality, climate regulation (interactions with greenhouse gases C02,
N20, CH4), fire frequency and intensity, disease resistance
¦ Cultural: swimming, boating, recreation, biodiversity
4.4. Direct Phytotoxic Effects
4.4.1. Sulfur Dioxide
The evidence is sufficient to infer a causal relationship between exposure to SO2 and injury to
vegetation (Section 3.4.2.1). The current secondary standard for S02 is a 3-h average of 0.50 ppm, which
is designed to protect against acute foliar injury in vegetation. There has been limited research on acute
foliar injury since the 1982 PM-SOx AQCD and there is no clear evidence of acute foliar injury below the
level of the current standard.
Effects on growth and yield of vegetation are associated with increased S02 exposure concentration
and time of exposure. The 1982 PM-SOx AQCD concluded that more definitive concentration-response
studies were needed before useable exposure metrics could be identified. The few new studies published
since the 1982 PM-SOx AQCD continue to report associations between exposure to S02 and reduced
vegetation growth. However, most these studies have been performed outside the U.S. and at levels well
above ambient concentrations observed in the U.S.
4-26
-------
4.4.2. NO, N02 and PAN
The evidence is sufficient to infer a causal relationship between exposure to NO, NO2 and PAN and
injury to vegetation (Section 3.4.2.2). It is well known that in sufficient concentrations, NO, N02 and PAN
can have phytotoxic effects on plants through decreasing photosynthesis and induction of visible foliar
injury (U.S. EPA, 1993a). However, the 1993 NOxAQCD concluded that concentrations of NO, N02and
PAN in the atmosphere are rarely high enough to have phytotoxic effects on vegetation (U.S. EPA,
1993a). Since the 1993 NOx AQCD, very little new research has been done on these phytotoxic effects at
concentrations currently observed in the U.S.
4.4.3. HNOs
The evidence is sufficient to infer a causal relationship between exposure to HNO3 and changes to
vegetation (Section 3.4.2.3). Experimental exposure of HN03 resulted in damage to the leaf cuticle of
pine and oak seedlings, which may predispose those plants to other stressors such as drought, pathogens,
and other air pollutants (Bytnerowicz et al., 1998a, b). However, these tree seedling experiments used
relatively short-term exposures at concentrations well above current ambient conditions. In lichen studies,
several lines of evidence, including transplant and controlled exposure studies, indicate that past and
current HN03 concentrations may be contributing to the decline in lichen species in the Los Angles basin
(Boonpragob and Nash, 1991; Nash and Sigal, 1999; Riddell et al., 2008). Current deposition of HN03 is
contributing to N saturation of some ecosystems close to sources of photochemical smog (Fenn et al.,
1998) such as the mixed conifer forests of the Los Angeles basin mountain (Bytnerowicz et al., 1999).
4.5. Mercury Methylation
The evidence is sufficient to infer a causal relationship between S deposition and increased
methylation of Hg, in aquatic environments where the value of other factors is within adequate range for
methylation (Section 3.4.1.4). The main agent of Hg methylation is S042 -rcducing-bactcria. and
experimental evidence from laboratory to mesocosm scales has established that only inconsequential
amounts of MeHg can be produced in the absence of S042 . These experimental results are highly
coherent with one another, and with observational studies at larger scales. Changes in the amount of S042
present have been shown to be followed by commensurate changes in MeHg, and mechanistic links have
been established between variation in S042 and variation in methylation of Hg.
Quantification of the relationship between S042 and methylation of Hg in natural settings has
proved difficult because of the presence of multiple interacting factors in aquatic environments where
S042 and Hg are present. The amount of MeHg produced has been shown to vary with 02 content,
temperature, pH, and supply of labile organic carbon. In some watersheds, such as high altitude lakes in
the western U.S., where no effect of changes in S042 deposition have been recorded on methylation of
Hg. This is because one or several interacting factors were not present in the amounts required for
methylation to occur at more than inconsequential rates. Watersheds with conditions known to be
conducive to Hg methylation can be found in the northeastern U.S. and southeastern Canada, but
significant biotic Hg accumulation has been observed in other regions that have not been studied as
extensively, and where a different set of conditions may exist.
Hg is a highly neurotoxic contaminant, and enters the food web in the methylated form. MeHg is
then concentrated in higher trophic levels, including fish eaten by humans, with undesirable consequences
for affected species, and for populations that consume large amounts of fish. Once MeHg is present, other
4-27
-------
variables influence how much of it accumulates in fish. Current evidence indicates that increased S
deposition very likely results in MeHg accumulation in fish.
Table 4-4. Summary of N deposition levels and corresponding ecological effects.
Type of
Ecosystem
Ambient N
Deposition
(kg N/ha/yr)1
N Additions
(kg N/ha/yr)
Biological and Chemi- 0. . 0.. 0. . 0 n .
cal Effects Study Site Study Species Reference
U.S.
Coastal
Level = 9.8 to 35
60 kg N/ha/yr; as
N addition caused a shift in California
Egerton-
sage scrub
kg N/ha/yr
NH4NO3 between
arbuscular mycorrhizal
Warburton
Species = nitrate as
HNO3/NO3
Jan. and March
community composition
and Allen
1994,1995,1996
with decreased species
(2000)
in two
richness and diversity
Measure = referred to
30 kg N/ha/yr
promoting a shift from
Padgett etal. (1999)
applications
shrub to grasslands
Coastal Level = not reported
sage scrub Spedes = not reported
Measure = not reported
50 ml N solution
(10 |jg/g NH4NO3;
Or 50 |jg/g NO3";
Or 50 |jg/g NH4+)
every 2 weeks
from March to
May, 1997
No evidence that decline in Southern Cali- Artemisia californica
native coastal sage scrub fornia
and increase in exotic
grass is due to mycorrhizal
response to increased
NO3-
(native); Bromus
madritenis spp.
Rubens (exotic)
Yoshida and
Allen (2001)
Coastal Level = Up to 30
sage scrub kg N/ha/yr
Species = not reported
Measure = referred to
Bytnerowicz et al.
(1987); Fenn etal.
(2003a)
60 kg N/ha/yr; as
NH4NO3
Soil inoculum from high N
deposition site caused na-
tive shrub growth de-
pression likely due to
mycorrhizal fungi
response. Growth of exotic
Western River- Artemisia californica
side County
Multispecies
Reserve; Uni-
versity of Cali-
fornia River-
(native); Bromus
madritenis spp.
Rubens (exotic)
Siguenza
etal. (2006)
grass may be promoted by side Botanical
soil inoculum from high N
deposition site.
Gardens
Desert Level = 30 kg N/ha/yr
Species = not reported
Measure = referred to
Bytnerowicz etal. 1987
Two additions of
16 kg N/ha/yr, one
as NH4NO3 and
one as an NPK
treatment
N addition increased bio-
mass of non-native plants
by -54%, decreased
native species biomass by
about -39%
Mojave desert Creosote bush (Larrea Brooks
tridentate), invasive (2003)
grasses Bromus
madritensis spp.
Rubens, and Schis-
mus spp.; and the forb
Erodium cicutarium
1 Ambient N deposition information is divided into three categories: Level = deposition rate; Species = chemical species of N that were measured;
Measure = source of the deposition data reported (i.e. a monitoring network, reference to another publication, etc.)
2 Ambient N deposition information is divided into three categories: Level = deposition rate; Species = chemical species of N that were measured;
Measure = source of the deposition data reported (i.e. a monitoring network, reference to another publication, etc.)
4-28
-------
Type of
Ecosystem
Ambient N
Deposition
(kg N/ha/yr)1
N Additions
(kg N/ha/yr)
Biological and Chemi-
cal Effects
Study Site Study Species Reference
Desert Level = 1.71 to 2.45 kg
N/ha/yr
Species = NH4NO3;
Measure = field study
for 16 years with
network of six funnel
precipitation collectors
Long term experi-
ment: 100 kg
Long term: increased the
cover of warm season
N/ha/yr as granular grasses and decreased the
NH4NO3 to 10
plots 2x/year since
Dec. 1995;
Single season
experiment: 20
kg N/ha/yr to 40
plots once
cover of legumes;
Short term: increased N
lead to significant plant
community structure
change, especially in blue
and black grama grassland
patch types
Chihuahuan Blue and black grama Baez et al.
Desert (Bouteloua gracilis (2007)
and Bouteloua ero-
poda, respectively)
Forest
Level = 3.6 and 3.5 kg No addition
Compared to west side,
Eastern vs.
Englemann spruce-
Baron et al.
(sub-al-
N/ha/yr on east slope;
east side had lower C:N,
Western slope
subalpine fir forests
(2000)
pine)
1.1 kg N/ha/yr on west
slope
lignin:N, and higher N:Mg,
N:P, foliar [N], soil [N], N
of Continental
Divide in
Species = not reported
mineralization rates and
lake water [NO3].
Rocky
Mountains.
Measure = referred to
Williams etal. (1998);
NADP 1999;
Stottlemyer et al. (1997)
Forest
Level = 1-2 kg N/ha/yr No addition
East side had decreased
Eastern vs. Englemann spruce
Rueth and
(sub-al-
on west slope;
soil organic horizon C:N
Western slope
Baron
pine)
3-5kg N/ha/yr on east
and foliar C:N, and
of Continental
(2002)
slope
increased foliar N
Divide in
Species = not reported
concentration, foliar N:Mg,
Rocky
foliar N:P and potential net
Mountains.
Measure = unspecified
mineralization; Englemann
spruce forest biogeo-
chemistry altered
Forest
Level = 1.2 to 23 No addition
N saturation observed at
California-Los
Bytnerowicz
kg N/ha/yr-dry dep and
25-45 kg N/ha/yr of total
Angeles Air
and Fenn
0.8 to 45 kg N/ha/yr-wet
inorganic N deposition.
Basin
(1996)
dep
Where N saturation oc-
Species = NO3" and
curred, high NO3"
NH4+
concentrations in
streamwater, soil, leaves;
Measure = various
high NO emissions; high
citations
foliar N:P.
1 Ambient N deposition information is divided into three categories: Level = deposition rate; Species = chemical species of N that were measured;
Measure = source of the deposition data reported (i.e. a monitoring network, reference to another publication, etc.)
4-29
-------
Type of
Ecosystem
Ambient N
Deposition
(kg N/ha/yr)1
N Additions
(kg N/ha/yr)
Biological and Chemi-
cal Effects
Study Site Study Species Reference
Forest Level = 2 study sites:
(conifer- CP = 18.8 kg N/ha/yr
ous) BF= 2.9 kg N/ha/yr
Species = NO3" and
NH4+
Measure = throughfall
and fog water
No addition Site nearest to urban area
(Los Angeles) received
much more N deposition,
as well as other pollutants
(i.e. S deposition), and
received much more fog,
coinciding with much more
wet deposition of N in that
site
Ecosystem was N
saturated, as evidenced by
high streamwater NO3"
concentration, 151 and 65
[jeq/L at upper and lower
ends, respectively, of Devil
Canyon West Fork
San Bernar-
dino Moun-
tains, Califor-
nia
Coniferous forest
Fenn et al.
(2000)
Forest
(mixed-
chaparral,
hardwood,
coniferous)
Level = 11 to 40
kg N/ha/yr
Species = NO3" and
NH4+
Measure = throughfall
(Fenn, 2002a)
No addition DIN export was scale de-
pendent, with highest
export occurring in water-
sheds of —150—ha.
Differences attributed to
temporal asynchrony be-
tween N availability and
biological demand
San Bernar-
dino Moun-
tains, CA
Mixed forest- chap-
arral, hardwood, conif-
erous
Meixner and
Fenn (2004)
Forest
(conifer-
ous)
Level = 8 kg N/ha/yr
and 82 kg N/ha/yr
Species = not reported
Measure = referred to
Fenn et al. (2002a)
0,50,150
kg N/ha/yr
annually from 1996
to 2002
Tree mortality was 9%
higher and beetle activity
50 % higher for unfertilized
trees at the high deposition
site compared to the low
pollution site. Tree
mortality and beetle activity
increased 8% and 20%,
respectively under highest
N fertilization rates at the
low deposition site
San Bernar-
dino Moun-
tains, CA
Ponderosa pine
Jones et al.
(2004)
1 Ambient N deposition information is divided into three categories: Level = deposition rate; Species = chemical species of N that were measured;
Measure = source of the deposition data reported (i.e. a monitoring network, reference to another publication, etc.)
4-30
-------
Type of
Ecosystem
Ambient N
Deposition
(kg N/ha/yr)1
N Additions
(kg N/ha/yr)
Biological and Chemi-
cal Effects
Study Site Study Species Reference
Forest Level = 35 kg N/ha/yr No addition
(chaparral) Spedes = not reported
Measure = taking mean
of two methods (estima-
tion methods using data
in situ and published
data; NADP)
Continued high export (~3 San Dimas Ex- Chamise
kg N/ha/yr) of NO3" in perimental (Adenostoma
Meixner
etal. (2006)
stream water (15 year)
after prescribed burn indi-
cates that chaparral
ecosystem did not recover
N-retention capabilities
after disturbance
Forest (40 km
NW of LA),
San Gabriel
Mountains, CA
fasiculatum), Cean-
othus spp., live oak
(Quercus agrifblia)
Forest Level = 1.2 to 71.1 kg No addition
(conifer) N/ha/yr
Species = NO3" and
NH4+
Measure = throughfall
(Fenn and Poth, 2004)
Empirical critical load for
adverse impacts on lichen
at 3.1 kg N/ha/yr.
Enhanced NO3" leaching
calculated with N
deposition above 17
kg N/ha/yr. Lowered litter
C:N and increases foliar N
also observed at highly
polluted sites.
California
mixed conifer
forests
Lichens, Ponderosa
pine
Fenn et al.
(2008)
Forest Level = 20 to 35
kg N/ha/yr
Species = NO3"
Measure = field studies
and NuCM, a nutrient
cycling model (Fenn
etal., 1996)
No addition Areas with higher Southern Cali-
deposition had increased fornia
NO3" leaching, increased
soil acidity, and decreased
base cation saturation.
Fenn et al.
(2003)
Forest Level = 3.2 to 5.5 kg
N/ha/yr
Species = not reported
Measure = data from
Loch Vale, a
NADP/NTN monitoring
station and Campbell
etal. (2000)
25 kg N/ha/yr of
NH4NO3
N addition increased N
concentration in foliar and
organic soil horizon
Fraser Experi-
mental Forest,
Colorado
Spruce
Rueth et al.
(2003)
Forest Level = 1.7 kg N/ha/yr
Species = not reported
Measure = referred to
Stottlemyer and
Troendle 1992,
Stottlemyer etal. 1997
25 kg N/ha/yr of
NH4NO3
N addition doubled N
mineralization rates and
stimulated nitrification
Loch Vale
watershed,
Colorado
Old-growth spruce
Rueth et al.
(2003)
1 Ambient N deposition information is divided into three categories: Level = deposition rate; Species = chemical species of N that were measured;
Measure = source of the deposition data reported (i.e. a monitoring network, reference to another publication, etc.)
2 Ambient N deposition information is divided into three categories: Level = deposition rate; Species = chemical species of N that were measured;
Measure = source of the deposition data reported (i.e. a monitoring network, reference to another publication, etc.)
4-31
-------
Type of
Ecosystem
Ambient N
Deposition
(kg N/ha/yr)1
N Additions
(kg N/ha/yr)
Biological and Chemi-
cal Effects
Study Site Study Species Reference
Forest Level = 50 kg N/ha/yr 200 kg N/ha/yr in
(alpine) MeaSure referred to
Sievering etal. (1996);
Theodose and Bowman
(1997); Fisk et al.
(1998)
1993 and 1994;
No fertilizer in
1995;
100 kg N/ha/yr as
(NH4)2S04 in 1996
and 1997
Nutrient amend-
ments consisted of
a mixture a mixture
of(NH4)2)N03 and
(NH4)2S04forthe
N plots
N deposition increased
plant foliage productivity
but reduced species
richness. The reduction of
species is best explained
by changes in soil
chemistry that resulted di-
rectly or indirectly from N
additions
Niwot Ridge,
Colorado
Alpine tundra: sedge
Kobresia
myosuroides.
Acomastylis rossii,
Polygonum viviparurn
Trifolium. More mesic
tundra-A rossii and
Deschumnpsia caes-
pitosa. Snow bed-
D.caespitosa, Sib-
baldia rocurnbens,
Rifblium parryi
Seastedt
and Vaccaro
(2001)
Forest Level = 11.5 to 25.4 kg No addition
N/ha/yr
Species = NO3" and
NH4+
Measure = 16
IMPROVE monitoring
sites and 11 NADP/NTN
wet dep. sites
Concentrations of N in
lichen thallus were highest
at eastern and west-
ernmost sites where N
deposition was highest,
implicating both
agricultural (east) and
urban (west) sources;
Columbia
River Gorge,
OR/WA
Lichens
Fenn at al.
(2007).
Forest
Level = 5 to 8
30 kg N/ha/yr as
Arbuscular mycorrhizal
Northern
Maple-dominated
van Diepen
(maple-
kg N/ha/yr
NaNOs
fungal biomass, storage
Michigan
hardwood- Sugar
etal. (2007)
dominated
Species = not reported
Measure = NADP 2006
structures and lipid storage
maple (Acersaccha-
hardwood)
declined in response to N
addition
rum)
Forest Level = 3.3 to 12.7 kg No addition
N/ha/yr
Species = not reported
Measure = at NERC, 29
Jan 2003, compiled
data sets and used
them in stats model by
Ollinger etal., 1993, or
used published values
At deposition levels above Northeastern
approximately 7-10 kg U.S.
N/ha/yr, stream NO3" con-
centration increase with in-
creasing deposition
Soil C:N and nitrification
flux increased with N
deposition
N deposition did not alter
foliar chemistry
Aber et al.
(2003)
Forest Level = 5.4 kg N/ha/yr
Species = not reported
Measure = McNulty and
Aber (1993)
15.7 to 31.4 Forest trees in plots receiv- Vermont
kg N/ha/yr as ing <20 kg N/ha/yr had
NH4CI-N or high rate of growth initially
NaN03"N followed by a decline, and
forests trees in plots
receiving >25 kg N/ha/yr
showed moderate rates of
decline.
Red spruce
McNulty
etal. (1996)
1 Ambient N deposition information is divided into three categories: Level = deposition rate; Species = chemical species of N that were measured;
Measure = source of the deposition data reported (i.e. a monitoring network, reference to another publication, etc.)
4-32
-------
Type of
Ecosystem
Ambient N
Deposition
(kg N/ha/yr)1
N Additions
(kg N/ha/yr)
Biological and Chemi-
cal Effects Stud>'Sl,c
Study Species
Reference
Forest
(herba-
ceous
layer)
Level = 7 kg N/ha/yr
Species = not reported
Measure = Shepard
etal. (1989)
14 and 28
kg N/ha/yr as crys-
talline (NH4)2S04
N addition decreased Adirondack
herbaceous cover under Park, New
hardwoods York
Mixed hardwood -
American beech,
sugar maple, and
yellow birch
Hurd et al.
(1998).
Forest Level = 3.5-7.8
(ectomy- kg N/ha/yr in wet
corrhiza) N deposition
Species = not reported
Measure = NADP data
Greenhouse study:
3 rates of N appli-
cations (0,35,140
kg N/ha).
Ectomycorrhizal
abundance and richness
declined along increasing
N deposition transect
under pitch pine. The
decline in richness was
significantly correlated with
the N deposition rate. In
greenhouse study, pine
seedling biomass was in-
versely related to N
addition.
New Jersey Pitch pine
Pine Barrens
Dighton
etal. (2004).
Forest
Level = Wet plus dry 25.2 kg N/ha for 10 After 10 years of
Bear Brook, Sugar maple; red
deposition 600 eq ha/yr years,
for N and 900
equiv./ha/yr for S
Species: not reported
Measure =Kahl et al.
(1999)
treatment, basal area
increment of sugar maple
was enhanced 13 to 104%,
whereas red spruce was
not significantly affected.
The increase in sugar
maple radial growth was
attributed to a fertilization
effect from the (NH4)2S04
treatment
Maine
spruce
Elvir et al.
(2003)
Forest Level = 17 kg N/ha/yr
Species = not reported
Measure = referred to
Adams et al. (1993)
35.5 kg N/ha/yr
and
40.5 kg S/ha/yr
fertilized annually
(NH4)2S04
(17 years)
N addition enhanced
growth of black cherry and
yellow poplar during the
first 7 years, but reduced
growth of these species in
years 9 to 12, with no
change in red maple or
sweet birch
Fernow
Experimental
Forest, West
Virginia
Mixed hardwood- red
oak, red maple, tulip
poplar, black cherry,
sweet birch
DeWalle
etal. (2006)
1 Ambient N deposition information is divided into three categories: Level = deposition rate; Species = chemical species of N that were measured;
Measure = source of the deposition data reported (i.e. a monitoring network, reference to another publication, etc.)
4-33
-------
Type of
Ecosystem
Ambient N
Deposition
(kg N/ha/yr)1
N Additions
(kg N/ha/yr)
Biological and Chemi-
cal Effects
Study Site Study Species Reference
Forest
(mixed
hardwood)
Level = 17 kg N/ha/yr
Species = not reported
Measure = referred to
Adams et al. (1993)
35 kg N/ha/yr
(NH4)2S04
(16 years)
Possible declining growth
vigor in red maple, and to
lesser extent black cherry
and tulip poplar. Observed
interspecific differences in
growth and plant nutrition
responses suggest
eventual changes in
species composition under
increasing N saturation.
Fernow Mixed hardwood-red Mayetal.
Experimental oak, red maple, tulip (2005)
Forest, West poplar, black cherry,
Virginia sweet birch
Forest
(mixed
hardwood)
Level = 17 kg N/ha/yr
Species = not reported
Measure = referred to
Adams et al. (1993)
35 kg N/ha/yr
(NH4)2S04
(12 years)
N addition altered
response of N-processing
microbes to environmental
factors, becoming less
sensitive to seasonal
changes in soil moisture
and temperature
Fernow
Experi-
mentalExperi-
mental Forest,
West Virginia
Mixed hardwood- red
oak, red maple, tulip
poplar, black cherry,
sweet birch
Gilliam etal.
(2001).
Forest
(herba-
ceous
layer)
Level = 17 kg N/ha/yr
Species = not reported
Measure = referred to
Adams et al. (1993)
35 kg N/ha/yr
(NH4)2S04 (4
years)
Increased foliar N in over-
story tree species and
Viola rotundifblia and
decreased foliar Ca2+ and
Mg2+, in response to 4
years of treatment.
Nitrification rates were
equally high in soils of all
watersheds; Results
support earlier studies that
high amounts of ambient N
deposition brought about N
saturation on untreated
watersheds at the Fernow
Experimental Forest
Fernow
Experimental
Forest, West
Virginia
Viola rotundifolia
Michx
Gilliam etal.
(1996)
1 Ambient N deposition information is divided into three categories: Level = deposition rate; Species = chemical species of N that were measured;
Measure = source of the deposition data reported (i.e. a monitoring network, reference to another publication, etc.)
4-34
-------
Type of
Ecosystem
Ambient N
Deposition
(kg N/ha/yr)1
N Additions
(kg N/ha/yr)
Biological and Chemi-
cal Effects S,udl'Sl,e
Study Species
Reference
Forest
(herba-
ceous
layer)
Level = 17 kg N/ha/yr
Species = not reported
Measure = referred to
Adams et al. (1993)
35 kg N/ha/yr
(NH4)2S04 (6
years)
no significant impact on Fernow
the herbaceous layer Experimental
under hardwoods Forest, West
Virginia
Mixed hardwood
Gilliam etal.
(2006b)
Forest Level = 6.6 to
8 kg N/ha/yr
Measure = regional
extrapolation from
NADP sites. Ollinger
et al. (1993); and esti-
mates from Munger
etal. (1996).
50 and 150kg
N/ha/yr for
15 years
Mortality of red pine
reached 56% in 15 years
in the pine high N plot, and
biomass accumulation has
stopped altogether. The
high N hardwood stand
shows increased above-
ground NPP, but excess N
availability and a severe
drought in 1995 con-
tributed to mortality of 72%
of red maple trees by
2002. Species importance
and litterfall patterns were
altered in several plots
after 1995. Roots, foliage
and wood have diminished
as net sinks for added N,
re-emphasizing the role of
soils in N retention. DIN
was not detected in soil
water of high N hardwood
plot until the 15th year.
Losses of inorganic N
remain high in the high N
plots (higher in pines than
hardwoods) and low N
plots in the pine stand also
have measurable DIN
losses. Foliar and fine root
N concentrations are ele-
vated significantly
Harvard Forest Red pine (Pinus resi-
nosaAit.), black and
red oak (Quercus
velutina Q. rubra)
black birch (Betula
lenta), red maple
(Acer rubrum), Ameri-
can beech(Fagus
grandifblia) and black
cherry (Prunus serot-
ina)
Magill et al.
(2004)
1 Ambient N deposition information is divided into three categories: Level = deposition rate; Species = chemical species of N that were measured;
Measure = source of the deposition data reported (i.e. a monitoring network, reference to another publication, etc.)
4-35
-------
Ambient N
Type of
Ecosystem
Deposition
(kg N/ha/yr)1
N Additions
(kg N/ha/yr)
Biological and Chemi-
cal Effects
Study Site Study Species
Reference
Forest
Level = 6.6 to 8
50 and 150kg
Following 7 years of N
Harvard Forest Understory of red pine
Rainey et al.
(herba-
kg N/ha/yr
N/ha/yr
additions, density and bio-
(1999)
ceous
Measure = regional
extrapolation from
mass of herb layer species
layer)
had declined by 80% and
90%
NADP sites. Ollinger
et al. (1993); and esti-
mates from Munger
etal. (1996).
Forest
Level = 6.6 to 8
50 and 150kg
Wood production
Harvard Forest Red pine (Pinus resi-
Magill et al.
(mixed
kg N/ha/yr
N/ha/yr.
increased (hardwood) and
nosaAit.) stand and
(1997).
hardwood
and conif-
erous)
Measure = regional
extrapolation from
decreased (coniferous)
Foliar N increased 25%
mixed hardwood stand
NADP sites. Ollinger
et al. (1993); and esti-
mates from Munger
etal. (1996).
(hardwood) and 67% (coni-
ferous); NO3" leaching in-
creased continuously over
6-year study in coniferous
forest, but was unchanged
in hardwood forest, most
(85-99%) of added N was
retained, primarily in
recalcitrant soil pool. Of
the retained N, 50—83%
appears to be in the long-
term, recalcitrant soil pool.
1 Ambient N deposition information is divided into three categories: Level = deposition rate; Species = chemical species of N that were measured;
Measure = source of the deposition data reported (i.e. a monitoring network, reference to another publication, etc.)
4-36
-------
Type of
Ecosystem
Ambient N
Deposition
(kg N/ha/yr)1
N Additions
(kg N/ha/yr)
Biological and Chemi-
cal Effects Stud>'Sl,c
Study Species
Reference
Oak sa-
vanna
Level=5.3 kg N/ha/yr
Species = not reported
Measure = NADP 2003
54, and 170
kg N/ha/yr (16—yr
addition of
NH4NO3)
N addition decreased total Minnesota
ectomycorrhizal fungal
diversity by 50% and
changed species
composition.
Native oak savannah:
bur oak (Quercus
macrocarpa Michaux)
and pin oak (Q.ellip-
sold alls E.J. Hilf),
ectomycorrhizal fungi
Avis et al.
(2003)
Grassland Level = 6 kg N/ha/yr
Species = 58%NH4,
42% N03
Measure = CASTN35
for total inorganic (wet
plus dry) N deposition;
local wet deposition
from the on-site NADP
monitoring station
10,20, 34, 54 or
95 kg N/ha/yr at
3,5,5,7, and 9
years respectively
from 1982 to 2004
Reduced relative number
of species at every
deposition level. Species
numbers were reduced
more per unit of added N
at lower addition rates,
suggesting that chronic but
low-level N deposition may
have a greater impact on
diversity than previously
thought. Chronic (23 year)
N addition (10 kg N/ha/yr)
reduced plant species
numbers by 17% relative
to controls receiving
ambient.
Critical load calculated at
5.3 kg N/ha/yr with an
inverse prediction interval
of 1.3-9.8 kg N/ha/yr
Cedar Creek Species = rich mixture Clark and
Biological of native C4 grasses Tilman
Station, Minne- and forbs (full list at (2008)
sota http://www.cedar-
creek.umn.edu/)
Grassland Level = not reported
Species = not reported
Measure = not reported
0, 54.4, and 272
kg N/ha/yr
NH4NO3 added
twice a yr for 18
years
Most of the forbs were lost
from the high N plots, and
two grass species, P.
pratensis and A. repens,
dominated. Loss of plant
diversity in areas of high N.
N addition changed
composition of soil micro-
bial community; increased
bacterial and decreased
fungal fatty acid methyl
ester activity
Cedar Ck,
LTER, Minne-
sota
P. pratensis, A.
repens, and
Schizachyrium
scoparium
Bradley et
al. (2006)
1 Ambient N deposition information is divided into three categories: Level = deposition rate; Species = chemical species of N that were measured;
Measure = source of the deposition data reported (i.e. a monitoring network, reference to another publication, etc.)
4-37
-------
Type of
Ecosystem
Ambient N
Deposition
(kg N/ha/yr)1
N Additions
(kg N/ha/yr)
Biological and Chemi-
cal Effects
Study Site Study Species Reference
Grassland Level = 10-15
kg N/ha/yr in San Jose
grasslands; 4-6
kg N/ha/yr at peninsula
sites
Species = not reported
Measure = referred to
Blanchard et al., (1996)
No addition Dry N deposition from
smog contributes to grass
invasion. Soil N limits
grass invasion on
serpentinitic solids. Graz-
ing cattle select grasses
over forbs, and grazing
lead to a net export of N as
cattle are removed for
slaughter. Decreased
populations of the bay
checkerspot butterfly due
to invasion of grasses after
cattle grazing.
San Francisco Bay checkerspot but- Weiss
Bay area, terfly/serpentinitic (1999)
California grasslands
Grassland Level = 10-15
kg N/ha/yr
Species = not reported
Measure =referred to
Weiss 1999
No addition N deposition displaced na-
tive grass species by
exotic nitrophilous grasses.
Low levels of soil N
normally limit grass
invasion in serpentinitic
soils, but in ungrazed
areas with experimental N
fertilization or high N
deposition, the introduced
grasses crowd out many
native species.
San Francisco
Bay area,
California
Serpentinitic grass-
lands
Fenn et al.
(2003).
Grassland Level = not reported
Species = not reported
Measure = not reported
70 kg N/ha/yr di-
vided into a liquid
Ca(N03)2 pulse
with the first au-
tumn rains and a
time-release pellet
application (Osmo-
cote) in January of
each yr for 3
years.
After three years, N
addition suppressed plant
diversity, forb production,
and forb abundance in
association with enhanced
grass production
Jasper Ridge
Biological Pre-
serve, Califor-
nia
Avena barbata,
Bromushordeaceus,
Lolium multiflorum,
Avena fatua, and Bro-
mus diandrus, Anagal-
lis arvensis, Geranium
dissectum, Erodium
botrys, Vicia sativa,
Crepis vesicaria
Zavaleta
etal. (2003).
1 Ambient N deposition information is divided into three categories: Level = deposition rate; Species = chemical species of N that were measured;
Measure = source of the deposition data reported (i.e. a monitoring network, reference to another publication, etc.)
4-38
-------
Type of
Ecosystem
Ambient N
Deposition
(kg N/ha/yr)1
N Additions
(kg N/ha/yr)
Biological and Chemi-
cal Effects
Study Site Study Species Reference
Grassland
(alpine)
Level =6 kg N/ha/yr
Species = not reported
Measure = not reported
20,40,60
kg N/ha/yr for
8 years
N deposition caused
changes of plant species
composition within 3 years
of the initiation of the
experiment and were
significant at all levels of N
addition. Changes in net
nitrification were
detectable at levels above
20 kg N/ha/yr. N addition
increased NO3" leaching
and NO3" concentration in
soil water. Changes in the
plant community, based on
the first axis of a detrended
correspondence analysis,
were estimated to occur at
deposition levels near 10
kg N/ha/yr.
Colorado Front
Range
Alpine dry meadows
dominated by the
sedge Kobresia myo-
suroides, with a mix of
perennial forbs,
sedges, and grasses
making up the remain-
der of the community.
Bowman
et al.
(2006)
Grassland Level = 5 kg N/ha/yr
Measure =referred to
Sievering et al. (1996);
Theodose and Bowman
(1997); Fisk et al.
(1998)
200 kg N/ha/yr In
1993 and 1994;
No fertilizer in
1995;
100 kg N/ha/yr as
(NH4)2S04 in 1996
and 1997;
As a mixture of
NH4NO3 and
(NH4)2S04
Increased plant biomass Niwot Ridge,
and tissue N concentration Colorado
Alpine tundra: sedge
Kobresia
myosuroides.
Acomastylis rossii,
Polygonum viviparurn
Trifolium. More mesic
tundra-A
Rossii and
Deschumnpsia caes-
pitosa. Snow bed-
D. casepitosa, Sib-
baldia procurnbens,
Trifolium parryi
Seastedt
and Vaccaro
(2001)
Grassland Level = not reported
Species = not reported
Measure = not reported
0,10,40, 70, and N addition increased
100 kg N/ha/yr for
75 days
growth. Native species
gained more height at
every level of N availability
compared to exotics
Greenhouse Two N. American na- Lowe et al.
study tives (Blue grama, (2002)
western wheatgrass),
and Four exotics,
(cheatgrass, leafy
spurge, Canada this-
tle, Russian
knapweed)
1 Ambient N deposition information is divided into three categories: Level = deposition rate; Species = chemical species of N that were measured;
Measure = source of the deposition data reported (i.e. a monitoring network, reference to another publication, etc.)
4-39
-------
Type of
Ecosystem
Ambient N
Deposition
(kg N/ha/yr)1
N Additions
(kg N/ha/yr)
Biological and Chemi-
cal Effects
Study Site Study Species Reference
Grassland Level = 5 kg N/ha/yr
Measure =referred to
Sievering etal. (1996)
Theodose and Bowman
(1997); Fisk et al.
(1998)
25 kg N/ha/yr with
biweekly additions
of 2 g of NH4NO3
dissolved in deion-
ized water for 2
years
N addition caused a
community shift towards
greater dominance of
hairgrass in wet alpine
meadows.
Niwot Ridge,
Colorado
Alpine tundra: sedge
Kobresia
myosuroides.
Acomastylis rossii,
Polygonum viviparurn
Trifolium. More mesic
tundra-A
Rossii and
Deschumnpsia caes-
pitosa. Snow bed-
D.caespitosa, Sib-
baldia procurnbens,
Trifolium
Bowman
etal. (1995);
Burns,
(2004).
Lake (al-
pine)
No addition By analysis of a lake sedi-
ment cores for a 400 year
time interval, the study
shows a typical alpine lake
diatom flora, consisting
mainly of small Fragilaria
sensu lato species, domi-
nated until approximately
1995, at which time Fragi-
laria crotonensis and
Cyclotella bodanica var.
lemanica rapidly increased
to 30% each of the total
assemblage. The shifts
appear indicative of both
increased N loading to
these systems as well as
changes in thermal
stratification patterns
Beartooth
Lake, Minne-
sota
Diatoms: Fragilaria
sensu lato, Fragilaria
crotonensis, Cyclotella
bodanica var.
lemanica
Saros et al.
(2003)
Lake (al- Deposition values back No addition
pine) to 1900 (0.5 kg N/ha/yr)
were calculated using
19 years (1984-2003) of
measured values from
Loch Vale (Colorado,
USA; NADP site C098)
The mean wet N-
deposition values of 1.5 kg
N/ha/yr (1950-1964) cor-
responded with the
alteration of diatom
assemblages attributed to
N deposition in alpine
lakes in Rocky Mountain
National Park. This value
becomes the critical load
defining the threshold for
ecological change from eu-
trophication.
Loch Vale,
Colorado
Diatoms
Baron
(2006)
1 Ambient N deposition information is divided into three categories: Level = deposition rate; Species = chemical species of N that were measured;
Measure = source of the deposition data reported (i.e. a monitoring network, reference to another publication, etc.)
4-40
-------
Type of
Ecosystem
Ambient N
Deposition
(kg N/ha/yr)1
N Additions
(kg N/ha/yr)
Biological and Chemi-
cal Effects
Study Site Study Species Reference
Lake Level= 2.5 - 3.5
kg N/ha/yr at elevations
greater than 2,500m;
<2.5 kg N/ha/yr at ele-
vations less than
2,500m
Measure= NADP sites
in the five-state (WY,
CO, UT, AZ and NM)
region of the intermoun-
tain West for the period
1992-1996
Species=not reported
No addition The deposition of inorganic
N in wetfall increased at
the rate of 0.32 kg N/ha/yr
from 1984 to 1997 at the
Niwot Ridge NADP site;
Annual loading of NO3"
increased with elevation;
Current level of N
deposition in wetfall
decreased surface water
pH and ANC, and resulted
in episodic acidifycation
and NO3" leakage.
Green Lakes
Valley of the
Colorado Front
Range
Williams
and Ton-
nessen
(2000)
Lake
Level = 2 kg N/ha/yr
N additions in
Increasing N shifted
Rocky Moun-
Diatoms.Stephano-
Interlandi
Measure = NAPD site
semi-continuous
diatom species in alpine
tain , Colorado
discus minutulus,
and Kilham
from 1981 through 1996
laboratory bioas-
says of mixed
diatom assem-
blages,
lakes.
Stephunodiscus
niugarue, and
Cyclotella bodanica.
(1998)
Surface
Level = 8-10 kg N/ha/yr No addition
Chronic N deposition re-
Eastern U.S./
Stoddard
waters
Measure = NADP from
sulted in increased N
New England
(1994)
1982 to 1994
leaching.
and Adi-
rondack lakes
Surface
Level = 4.71 kg N/ha/yr No addition
A shift in from an N-limited
Niwot
Williams
waters
-200% increase in NO3"
system to an N-saturated
Ridge/Green
etal. (1996)
loading from wet
system. Many lakes having
Lakes, Colo-
deposition over the prior
(NO3") concentrations
rado Front
decade, increasing from
greater than 10 iequiv/L.
Range
8 kg N/ha/yr for 1985-
Increasing atmospheric
1987 to 16.5 kg N/ha/yr
deposition of N with
1990-1992
elevation is causing a
Species= not reported
change from N limitation to
P limitation in the highest-
Measure = NADP
elevation bristlecone pines
Wetland Level = 10 to 14
(freshwa- kg N/ha/yr
ter)
Hawley Bog: 2
treatments of 0.1
or 1.0 mg NH4-N/L
every 2 weeks
between June 1
and Sept 30 for
1998,1999, and
2000
Negative population
growth rate of pitcher plant
Hawley Bog
(MA)
Molly Bog (VA)
Sarracenia purpurea
Gotelli and
Ellison
(2002;
2006)
1 Ambient N deposition information is divided into three categories: Level = deposition rate; Species = chemical species of N that were measured;
Measure = source of the deposition data reported (i.e. a monitoring network, reference to another publication, etc.)
4-41
-------
Type of
Ecosystem
Ambient N
Deposition
(kg N/ha/yr)1
N Additions
(kg N/ha/yr)
Biological and Chemi-
cal Effects
Study Site Study Species Reference
Coastal-
Discontinuously
Large diatoms became
Marine phytoplankton Stolte et al.
Marine
diluted N limited
dominant when nitrate was
(1994)
cultures, which
supplied as the only N
were pulsed with
source once in 3 days.
N03_every 3 days;
Sinking rate of the nitrate
grown population was
higher (0.12 m/day) than
that of the ammonium
grown population (0
m/day).
Variable Variable
60-120 kg N/ha/yr
Soil CEC and temperature Variable
Variable Clark etal.
variable range
were the strongest contri-
(2007)
across 23 different
butors to multivariable
experiments
explanation of species
richness response to
experimental N addition in
23 studies throughout the
U.S. Greater plant species
loss was associated with
lower soil CEC, colder
temperature and larger
production increases
CHINA AND EUROPE
Forest Level = 3 kg N/ha/yr
Species = not reported
Measure = not reported
30 years of annual
additions of
NH4NO3 in 3 treat-
ments: N1 and N2
(34 and 68 kg
N/ha/yr) and 20
years of N3 (108
kg N/ha/yr), plus a
control
Stimulate stemwood
production in all levels of N den
addition until 7 years into
experiment; thereafter the
second and third
treatments (with medium
and high N addition)
decreesed stemwood pro-
duction and the first treat-
ment (the lowest amount of
N addition) continued to
increase stemwood pro-
duction throughout experi-
ment
Northern Swe- Scots pine
Hogberg
etal. (2006)
Forest (bo- Level = 3-12 kg N/ha/yr No addition
rea') Species = not reported
Measure = MATCH
model of N-deposition
for 1996; surveys in
permanent plots in the
National Forest
Inventory
N deposition decreased
the abundance and cover
of ericaceous shrubs
Sweden
Vaccinium species
(,Ericaceae)
Strengbom
etal. (2003)
1 Ambient N deposition information is divided into three categories: Level = deposition rate; Species = chemical species of N that were measured;
Measure = source of the deposition data reported (i.e. a monitoring network, reference to another publication, etc.)
4-42
-------
Type of
Ecosystem
Ambient N
Deposition
(kg N/ha/yr)1
N Additions
(kg N/ha/yr)
Biological and Chemi-
cal Effects
Study Site Study Species Reference
Forest Level= not reported 22 years of annual The early increase in
North Sweden Scot pine
Species= not reported
Measure=not reported
additions of
NH4NO3 in four
treatments: NO
(control); N1 (30-
60 kg N/ha/yr), N2
(60-120 kg
N/ha/yr) and N3
(90-180 kg
N/ha/yr)
growth rates was reduced
at the highest N level and
partly also at the middle
level. N retention in
ecosystem was high at the
lowest level of N addition,
while considerable N lost
from the sites at higher
levels, indicating N satura-
tion.
Tamm et al.
(1995)
Forest Level: wet deposition
from 1965-1990:
Norrliden (S:5.5-7.2kg
S/ha/yr; NH4-N: 1.3-1.9
kg N/ha/yr; NH3-N:1.6-
2.2 kg/ha/yr)
Ljssejbo (S:5.2-6.7kg
S/ha/yr; NH4-N: 1.2-1.9
kg N/ha/yr; NH3-N:1.4-
1.9 kg/ha/yr)
Species: S, NhUand
NH3
Measure=not reported
Norrliden: 60-
120kg N/ha/yr from
1971 to 1988
Lisseblo: 40-120kg
N/ha/yr from 1969
to 1988
N normally is a growth- Norrliden in
limiting factor in boreal northern Swe-
forest, but regular N den and Lis-
additions can induce boron selbo in south-
and magnesium deficiency, ern Sweden
and low internal concen-
trations in the trees, e.g.,
of potassium and phos-
phorus.
Scot pine
Tamm et al.
(1999)
Lake Wet DIN = 1.3-11 kg No addition
N/ha/yr
Total N = 1-4 and 9-18
kg N/ha/yr
Species = wet DIN and
total N
Measure = SEPAs use
of MATCH model
Increased lake concentra- Sweden
tions of inorganic N caused
P limitation in summer and
increased eutrophication
with increased lake algal
productivity
Bergstrom
etal. (2005)
Wetland Level = 0.2 - 0.3 g
N/m2/a (2 - 3 kg
N/ha/yr)
Species = NH4+and
NOs-
Measure = refers to
Ruoho-Airola etal. 1998
Cumulative addi-
tion of 3 g N/m2/a
(30 kg N/ha/yr) as
NH4NO3 on six
occasions during
growing season
Increased NH4NO3
affected comp. of moss
layer, specifically
decreasing Sphagnum
balticum, and caused de-
crease in litter and an in-
crease of a Vaccinium spe-
cies
Eastern Sphagnum sp., Erio- Saarnio
Finland phorumvaginatum, etal. (2003)
Carex pauciflora, Vac-
cinium oxycoccos,
Scheuchzeria palustris
1 Ambient N deposition information is divided into three categories: Level = deposition rate; Species = chemical species of N that were measured;
Measure = source of the deposition data reported (i.e. a monitoring network, reference to another publication, etc.)
4-43
-------
Type of
Ecosystem
Ambient N
Deposition
(kg N/ha/yr)1
N Additions
(kg N/ha/yr)
Biological and Chemi-
cal Effects
Study Site Study Species Reference
Wetland Level = not reported
Species = not reported
Measure = not reported
240 kg N/ha/yr as
NH4NO3 applied in
field experiment
where 3 of 6 plots
received treatment
every 2 weeks for
2 years
N addition increased
above ground biomass and
abundance oWeyeucia
angustifblia, stimulated
CO2 and CH4, and N2O
emissions from D. angusti-
fblia wetlands. N2O
emissions significantly
influenced by N addition.
Northeast Deyeucia angustifblia Zhang etal.
China (2007)
Wetland Level = 2 to 20 kg
N/ha/yr
Species = not reported
Measure = national
precipitation monitoring
programs of each
country and refer to the
3 years prior to peat
sampling in experiment
No addition Decomposition rates for
material accumulated
under higher atmospheric
N supplies resulted in
higher CO2 emissions and
dissolved organic carbon
release
Nine European
countries
Species like Sphag-
num
Bragazza
etal. (2006)
1 Ambient N deposition information is divided into three categories: Level = deposition rate; Species = chemical species of N that were measured;
Measure = source of the deposition data reported (i.e. a monitoring network, reference to another publication, etc.)
1 Ambient N deposition information is divided into three categories: Level = deposition rate; Species = chemical species of N that were measured;
Measure = source of the deposition data reported (i.e. a monitoring network, reference to another publication, etc.)
4-44
-------
Annex A. Ecosystem Monitoring
and Models
A.1. Introduction
A tremendous amount of research has been conducted in the U.S., and elsewhere, over the past
three decades on the ways in which atmospheric deposition of sulfur (S) and nitrogen (N) affect the
health, condition, and vitality of aquatic, transitional, and terrestrial ecosystems. Much of this work has
focused on developing a better understanding of acidification and nutrient enrichment processes. Some of
this work has been highly quantitative allowing researchers to determine key process rates in multiple
ecosystem compartments. Nevertheless, quantification of overall ecosystem response requires a higher
level of process rate aggregation. It is important to develop quantitative understanding of the extent of
past ecosystem effects in response to atmospheric S and N deposition, the extent to which conditions will
worsen or recover under continued or reduced deposition levels, and the sustained loads of deposition that
would be required to prevent further ecosystem damage and to allow damaged ecosystems to recover.
This kind of quantitative understanding cannot evolve directly out of process-based research. It requires
development of mathematical models that encode process knowledge and link it in such a way as to pro-
duce quantitative estimates of change in resource conditions over time in response to changes in the major
forcing functions, including atmospheric deposition, climate, and landscape disturbance. As described in
this Annex, many such models have been developed and used to estimate past and future changes in
ecosystem condition. Such models cannot be validated, per se, because environmental systems are never
closed and because important processes yield conflicting, often opposing, results. Therefore, a model can
produce the right answer for the wrong reason (Oreskes et al., 1994). Similarly, a particular process may
not be important at a particular site where a model is tested, but assume much greater importance
elsewhere. For these reasons, it is critical that environmental models be tested and confirmed at multiple
locations that exhibit differing conditions and pollutant loads before they are used as the foundation for
public policy (Sullivan, 2000a).
Some of the best data with which to test and confirm environmental models are derived from
long-term monitoring sites. These are locations where one or more attributes of a natural ecosystem
compartment (i.e., surface water, soil, plants) is periodically sampled and analyzed over a long period of
time. Such data are often especially valuable for sites, which experience rather large changes in one of the
forcing functions (often atmospheric deposition). This enables evaluation of the extent to which the model
accurately captures the dynamics of ecosystem response(s) that occur. Because many environmental
attributes undergo rather substantial intra- and inter-annual variability in response to climatic variation
and other changes, a long period of record is required before a monitoring data set can be used for
evaluation of ecosystem response or for model confirmation.
Long-term monitoring data provide not only data with which to test model projections, but also a
reality check on scientific understanding of damage and recovery processes. If observed (monitored)
changes are not in agreement with process understanding, it is possible or perhaps likely that one or more
key processes is not well understood or well formulated in the model.
This Annex summarizes the primary long-term monitoring sites and programs in the U.S., and the
principal mathematical models used to simulate environmental responses to atmospheric S and N
deposition. Quantitative data derived from the model projections and from trends analyses of the
monitoring data provide an important part of the foundation for evaluating the past, current, and future
effects of S and N deposition, and expected recovery as emissions levels decrease in the future.
A-1
-------
A.2. Ecosystem Monitoring
The effects of acidic and N nutrient deposition on ecosystems require long-term study. Changes in
ecosystems often occur gradually, and sustained monitoring of key variables provides the principal record
of change over time. Monitoring data are also useful for establishing a baseline of resource conditions and
determining if short-term events were unusual or extreme (Lovett et al., 2007). There are limited
monitoring programs and data to document ecosystem responses to changes in atmospheric deposition in
the past. It is often difficult to sustain funding for ecosystem monitoring, perhaps because results are
produced slowly and because results are seldom viewed as novel. Nevertheless, monitoring data provide
some of the best means for evaluating the completeness of the scientific knowledge base and for testing
how robust our projections of future conditions might be. This section describes some of the more
important and useful monitoring programs for evaluating the effects of N and S deposition on ecosystems
in the U.S.
There are long-term monitoring sites scattered throughout the U.S. where samples are periodically
collected and analyzed to determine the condition of aquatic, transitional, or terrestrial ecosystem
elements. Some have been in operation for only a short period of time; others have continued for decades.
None extend back far enough to have documented resource conditions before the advent of high levels of
atmospheric S and N deposition. Some of the monitoring sites exist as an individual entity, or small
collection of sites, often established primarily for research purposes. Despite the research focus, many of
these long-term research sites include collection of monitoring data. Other long-term monitoring sites
exist as part of large regional programs with a specific focus on long-term monitoring. The most
significant individual monitoring sites and networks are discussed below.
Lovett et al. (2007) reviewed the characteristics of successful environmental monitoring programs,
and argued that monitoring is a fundamental part of environmental science and policy. Their analysis
underscored the fact that environmental monitoring costs little relative to the value of the resources that it
protects and the policy that it informs. Monitoring data also have substantial added value because they
can be used for multiple purposes, including various research objectives.
Ecosystems also require long-term study because most changes occur slowly. When more rapid
change does occur, for example in response to an extreme event, a long-term record is needed to put the
effects of the extreme event into proper context.
A.2.1. Environmental Monitoring and Assessment Program
The EMAP began regional surveys of the nation's surface waters in 1991 with a survey of
northeastern U.S. lakes. Since then, EMAP and Regional-EMAP (REMAP) surveys have been conducted
on lakes and streams throughout the country. The objective of these EMAP surveys is to characterize
ecological condition across populations of surface waters. EMAP surveys are probability surveys where
sites are picked using a spatially balanced systematic randomized sample so that the results can be used to
make estimates of regional extent of condition (e.g., number of lakes, length of stream). EMAP sampling
typically consists of measures of aquatic biota (fish, macroinvertebrates, zooplankton, and periphyton),
water chemistry, and physical habitat.
Of particular interest with respect to acidic deposition effects were two EMAP surveys conducted
in the 1990s, the Northeastern Lake Survey and the Mid-Atlantic Highlands Assessment of streams
(MAHA). The Northeastern Lake Survey was conducted in summer from 1991 to 1994 and consisted of
345 randomly selected lakes in the states of New York, New Jersey, Vermont, New Hampshire, Maine,
Rhode Island, Connecticut, and Massachusetts (Whittier et al., 2002). To make more precise estimates of
the effects of acidic deposition, the sampling grid was intensified to increase the sample site density in the
Adirondacks and New England Uplands areas known to be susceptible to acidic deposition. The MAHA
study was conducted on 503 stream sites from 1993 to 1995 in the states of West Virginia, Virginia,
A-2
-------
Pennsylvania, Maryland, Delaware, and the Catskill Mountain region of New York (Herlihy et al., 2000).
Sampling was done during spring baseflow. Sample sites were restricted to first through third order
streams as depicted on the U.S. Geological Survey (USGS) 1:100,000 digital maps used in site selection.
To make more precise estimates of the effects of acidic deposition, the sampling grid was intensified to
increase the sample site density in the Blue Ridge, Appalachian Plateau, and Ridge section of the Valley
and Ridge ecoregions. Results from both of these surveys were used to develop and select the sampling
sites for the Temporally Integrated Monitoring of Ecosystems (TIME) program, which is described below.
A
i I
Acid-Sensitive Regions of the Eastern United States
Temporally Integrated Monitoring of Ecosystem (TIME) Sites
New England
Adirondack
Mountains V.
Northern
Appalachian Plateau
fN
y
Source: Stoddard et al. (2003).
Figure A-1. Location of acid-sensitive regions of the northern and eastern U.S. These are regions for
which statistical survey data are available in the 1990s, and locations of individual TIME sites
used in trend analysis.
A-3
-------
A.2.2. Surface Water Chemistry Monitoring
There are two surface water chemistry monitoring programs, administered by the U.S. EPA, that
are especially important to inform the assessment of aquatic ecosystem responses to changes in
atmospheric deposition. These are the TIME program (Stoddard et al., 2003) and the Long-term
Monitoring (LTM) program (Ford et al., 1993; Stoddard, 1998a). These efforts focus on portions of the
U.S. most affected by the acidifying influence of S and N deposition, including lakes in the Adirondack
Mountains of New York and in New England, and streams in the Northern Appalachian Plateau and Blue
Ridge in Virginia and West Virginia. Both projects are operated cooperatively with numerous
collaborators in state agencies, academic institutions and other federal agencies. The TIME and LTM
projects have slightly different objectives and structures, which are outlined below. Stoddard et al. (2003)
conducted a thorough trends analysis of the TIME and LTM data.
A.2.2.1. TIME Project
At the core of the TIME project is the concept of probability sampling, whereby each sampling site
is chosen statistically from a pre-defined target population. Collectively, the monitoring data collected at
the sites are representative of the target population of lakes or streams in each study (Figure A-l). The
target populations in these regions include lakes and streams likely to be responsive to changes in acidic
deposition, defined in terms of acid neutralizing capacity (ANC), which represents an estimate of the
ability of water to buffer acid. It can be either calculated (calculated ANC = sum of base cations - sum of
mineral acid anions, where all concentrations are in j^ieq/L) or titrated in the laboratory (Gran ANC).
Measurement of Gran ANC uses the Gran technique to find the inflection point in an acid-base titration of
a water sample (Gran, 1952). In the Northeast, the TIME target population consists of lakes with a Gran
ANC less than 100 (ieq/L. In the Mid-Atlantic, the target population is upland streams with Gran ANC
less than 100 (ieq/L. In both regions, the sample sites selected for future monitoring were selected from
the EMAP survey sites in the region (Section A.2.1) that met the TIME target population definition.
Each lake or stream is sampled annually (in summer for lakes; in spring for streams), and results
are extrapolated with known confidence to the target population(s) as a whole using the EMAP site
population expansion factors or weights (Larsen and Urquhart, 1993; Larsen et al., 1994; Stoddard et al.,
1996; Urquhart et al., 1998). TIME sites were selected using the methods developed by the EMAP
(Herlihy et al., 2000; Paulsen et al., 1991). The TIME project began sampling northeastern lakes in 1991.
Data from 43 Adirondack lakes can be extrapolated to the target population of low-ANC lakes in that
region. There are about 1,000 low-ANC Adirondack lakes, out of atotal population of 1830 lakes with
surface area greater than 1 ha. Data from 30 lakes (representing about 1,500 low-ANC lakes, out of a total
population of 6,800) form the basis for TIME monitoring in New England. Probability monitoring of
Mid-Atlantic streams began in 1993. Stoddard et al. (2003) analyzed data from 30 low-ANC streams in
the Northern Appalachian Plateau (representing about 24,000 km of low-ANC stream length out of atotal
stream length of 42,000 km).
The initial 1993-1995 EMAP-MAHA sample in the Mid-Atlantic was not dense enough to obtain
enough sites in the TIME target population in the Blue Ridge and Valley and Ridge ecoregions. In 1998,
another denser random sample was conducted in these ecoregions to identify more TIME sites. After
pooling TIME target sites taken from both MAHA and the 1998 survey, there are now 21 TIME sites in
the Blue Ridge and Ridge and Valley that can be used for trend detection in this aggregate ecoregion in
the Mid-Atlantic in addition to the Northern Appalachian Plateau ecoregion.
A-4
-------
A.2.2.2. Long-Term Monitoring Project
As a complement to the statistical lake and stream sampling in TIME, the LTM project samples a
subset of generally acid-sensitive lakes and streams that have long-term data, many dating back to the
early 1980s (Figure A-2). These sites are sampled 3 to 15 times per year. This information is used to
characterize how some of the most sensitive of aquatic systems in each region are responding to changing
deposition, as well as giving information on seasonal variation in water chemistry. In most regions, a
small number of higher-ANC (e.g., Gran ANC greater than 100 j^ieq/L) sites are also sampled, and help
separate temporal changes due to acidic deposition from those attributable to other disturbances (e.g.,
climate, land use change). Because of the availability of long-term records (more than two decades) at
many LTM sites, their trends can also be placed in a better historical context than those of the TIME sites,
where data are only available starting in the 1990s.
Monitored water chemistry variables include pH, ANC, major anions and cations, monomelic
aluminum (Al), silicon (Si), specific conductance, dissolved organic carbon (DOC), and dissolved
inorganic carbon (DIC). The field protocols, laboratory methods, and quality assurance procedures are
specific to each team of investigators. This information is contained in the cited publications of each
research group. The EMAP and TIME protocols and quality assurance methods are generally consistent
with those of the LTM cooperators. Details of LTM data from each region are given below.
Acid-Sensitive Regions of the Northern and Eastern United States
Long-Term Monitoring (LTM) Sites
New England
Upper
Midwesty
\ ; • *
T\,
( Northern "".
• 'Appalachian
and- •
Blite pidge.Provinces j
Source: Stoddard et ai. (2003).
Figure A-2. Location LTM sites used in the 2003 Surface Water report.
New England Lakes: The LTM project collects quarterly data from lakes in Maine (sampled by the
University of Maine; (Kahl et al., 1991; Kalil et al.. 1993) and Vermont (data collected by the Vennont
Department of Environmental Conservation; Stoddard and Kellogg, 1993; Stoddard et al., 1998a). Data
from 24 New England lakes were available for the trend analysis reported by Stoddard et al. (2003) for
the time period 1990 to 2000. In addition to quarterly samples, a subset of these lakes have outlet samples
A-5
-------
collected on a weekly basis during the snowmelt season; these data are used to characterize variation in
spring chemistry. Most New England LTM lakes have mean Gran ANC values ranging from -20 to
100 (ieq/L; two higher ANC lakes (Gran ANC between 100 and 200 |_icq/L) are also monitored.
Adirondack Lakes: The trend analysis of Stoddard et al. (2003) included data from 48 Adirondack
lakes, sampled monthly by the Adirondack Lake Survey Corporation (Driscoll and Van Dreason, 1993;
Driscoll and Paostek, 1995) a subset of these lakes are sampled weekly during spring snowmelt to help
characterize spring season variability. Sixteen of the lakes have been monitored since the early 1980s; the
others were added to the program in the 1990s. The Adirondack LTM dataset includes seepage and
drainage lakes, most with Gran ANC values in the range of-50 to 100 (ieq/L; three lakes with Gran ANC
between 100 j^icq/L and 200 j^ieq/L are also monitored.
Appalachian Plateau streams: Stream sampling in the Northern Appalachian Plateau is conducted
about 15 times per year, with the samples spread evenly between baseflow (e.g., summer and fall) and
high flow (e.g., spring) seasons. Data from four streams in the Catskill Mountains (collected by the U.S.
Geological Survey; (Murdoch and Stoddard, 1993) and five streams in Pennsylvania (collected by
Pennsylvania State University (DeWalle and Swistock, 1994a) were analyzed by Stoddard et al. (2003).
All of the Northern Appalachian LTM streams have mean Gran ANC values in the range -25 to 50 j^icq/L.
Upper Midwest lakes: Forty lakes in the Upper Midwest were originally included in the LTM
project, but funding in this region was terminated in 1995. The Wisconsin Department of Natural
Resources (funded by the Wisconsin Acid Deposition Research Council, the Wisconsin Utilities
Association, the Electric Power Research Institute and the Wisconsin Department of Natural Resources)
has continued limited sampling of a subset of these lakes, as well as carrying out additional sampling of
an independent subset of seepage lakes in the state. The data reported by Stoddard et al. (2003) included
16 lakes (both drainage and seepage) sampled quarterly (Webster et al., 1993) and 22 seepage lakes
sampled annually in the 1990s. All of the Upper Midwest LTM lakes exhibit mean Gran ANC values from
-30 to 80 (ieq/L.
Ridge/Blue Ridge streams: Data from the Ridge and Blue Ridge provinces consist of a large number
of streams sampled quarterly throughout the 1990s as part of the Virginia Trout Stream Sensitivity Study
(Webb et al., 1989), and a small number of streams sampled more intensively (as in the Northern
Appalachian Plateau). A total of 69 streams, all located in the Ridge section of the Ridge and Valley
province, or within the Blue Ridge province, and all within the state of Virginia, had sufficient data for the
trend analyses by Stoddard et al. (2003). The data are collected cooperatively with the University of
Virginia and the National Park Service. Mean Gran ANC values for the Ridge and Blue Ridge data range
from -15 to 200 (ieq/L, with 7 of the 69 sites exhibiting mean Gran ANC greater than 100 (ieq/L.FIA
field personnel collect soil data during the Phase 3 field season, which begins in early June and ends in
September. Soil samples are sent to the laboratory immediately after collection where they are stabilized
by air drying. Laboratory analyses are conducted throughout the fall and winter following the field
season. On-plot measurements include soil compaction and bare soil observations. Soil compaction, the
percentage of the soil surface exhibiting evidence of soil compaction as well as the type of compaction, is
measured by ocular estimation. The relative amount of bare soil is also estimated. Field measurements
related to erosion and compaction estimates are made on all four subplots on the Phase 3 field plot. Soil
samples are collected on FIA sample plots along soil sampling lines adjacent to subplots 2, 3, and 4. Soils
are collected if the soil sampling location is in a forested condition. A total of five samples are collected
on each plot (three forest floor, two mineral soil). The entire forest floor layer is sampled from a known
area after measuring the thickness of the litter and duff layers at the north, south, east, and west edges of a
12-inch diameter sampling frame. Only organic material that is
-------
In the lab, mineral soil samples collected from FIA plots are analyzed for a suite of physical and
chemical properties including:
¦ Bulk density, water content, and coarse fragment ( >2-mm) content
¦ pH (water and 0.01 M CaCl2)
¦ Total carbon
¦ Total inorganic carbon (carbonates) (pH >7.5 soils only)
¦ Total N
¦ Exchangeable cations (Na, K, Mg, Ca, Al, Mn)
¦ Extractable sulfur and trace metals (Sr, Ba, Mn, Ni, Cu, Zn, Cd, Pb)
¦ Extractable P (Bray 1 method for pH <6 soils, Olsen method for pH >6 soils)
Forest floor and litter samples are analyzed for:
¦ Bulk density and water content
¦ Total carbon
¦ Total N
Soil chemical and physical properties can be highly variable in the field and are expensive to
analyze. As a result, interpretation of soil chemical data is confounded by spatial variability within the
plot. In addition, depending upon the soil type, both the number of samples and the methods used in
collecting these samples may vary between plots, complicating compilation and estimation procedures.
Finally, soil samples reflect conditions only in the forest floor and upper 20 cm of the soil. In many
systems, the upper portion of the soil profile is likely to be more responsive to disturbance, providing a
useful index for monitoring changes in soil properties over time.
A.2.3. USGS Monitoring Programs
A.2.3.1. National Water Quality Assessment Program
The National Water Quality Assessment (NAWQA) Program was created in 1991 by the USGS to
assess the nation's water quality in 51 study units defined primarily by major drainage divides. These
units comprise approximately 50% of the conterminous U.S. Its major of the program are to determine the
condition of the nation's streams, rivers, and ground water; whether these conditions are changing over
time; and how these conditions are affected by natural features and human activities.
The major priority of the NAWQA Program since its inception has been on watersheds that have
experienced impacts from agriculture and various forms of development. The location of sites, sampling
frequency, and types of measurements taken, all reflect this priority. Each study unit runs on a 9-year
cycle, with approximately one-third of the study units beginning the cycle every 3 years. Each 9-year
cycle is comprised of 3 years of intensive data collection and 6 years of low-level assessment. Three types
of sampling sites are established within each study unit: integrator sites, indicator sites, and synoptic sites.
Integrator sites are located on major rivers at points that drain much or the entire study unit. Indicator
sites drain large fractions of the study unit that are representative of a particular landscape or land use
type. Some indicator sites are also located to evaluate point sources of water pollution, and some are
located downstream of undisturbed drainages to provide reference, or background conditions. Reference
indicator sites are generally located too low in the drainage basin for assessment of surface water
A-7
-------
acidification. Sites associated with synoptic studies are chosen for improving spatial resolution of data
collection within the study unit. The strategies for site selection, sampling, and analysis for synoptic sites
are issue-specific and keyed to hydrologic conditions, times, and places of specific interest for the
targeted water quality issue.
During the 3 year intensive sampling period, integrator and indicator sites are sampled multiple
times, both periodically and in association with high flows. The sampling approach for synoptic studies
varies depending on the issue of interest, but is usually done within the second or third year of the
intensive sampling period. During the six years of the low-level assessment, sampling usually involves
base-flow sampling of high priority integrator sites, and possibly some sampling of indicator sites.
Water quality measurements vary among study units, but usually include pathogens, nutrients
(including N and S), trace elements, pesticides, industrial organics, suspended sediment, salinity,
temperature, acidity, and dissolved 02.
NAWQA studies have resulted in over 1000 reports on an extensive list of water quality issues,
including freshwater and marine eutrophication associated with N pollution. None deal with acidification,
however. Program details and access to publications can be obtained at http://water.usgs.gov/nawqa/.
A.2.3.2. Hydrologic Benchmark Network
The Hydrologic Benchmark Network (HBN) was started in 1963 by the USGS and gradually grew
to include 57 river gauging stations and 1 lake-stage station in 39 states by 1990. Most of the stations
have been established at the outlet of watersheds that were virtually free of human activities, located in
places such as in national parks and forests, wilderness areas, or nature preserves. Streamflow was
initially monitored continuously at each station, and samples were collected every month for water-
quality analyses that included concentrations of nutrients and all major ions. The frequency of water
sampling at HBN stations was decreased to quarterly in 1986 because of budgetary restrictions. Sampling
was discontinued in October 1997, except for a small study in the eastern U.S. that focused on the initial
response of rivers to decreases in industrial emissions mandated by the CAAA of 1990
(http://www.epa. gov/oar/oaa caa.html/').
All HBN watersheds were evaluated in 2002 to determine whether upstream development had
made them unsuitable as reference watersheds. The 36 sites that best met the network criteria were
selected for continued streamflow monitoring, and water sampling was reinitiated at 15 of those 36 sites.
In 2003, 15 of the original HBN stations were equipped with refrigerated, automated samplers and
telemetry systems that allow program coordinators to monitor stream conditions and adjust sampling
frequency and capture unique stream conditions or special sampling needs. The automated sampling
system is designed to collect samples through a wide range of flow conditions and to transmit data by
satellite. About 25 water samples are collected annually at each HBN water quality station and
refrigerated on site until retrieved by field personnel who visit the sites regularly. The most recent trends
analysis was done by Clow and Mast (1999) to evaluate long-term trends in stream chemistry with respect
to the CAA. The program is further described in a fact sheet that can be found at:
http://nv.water.usgs.gov/pubs/fs/fs20053135/.
A.2.3.3. New York City Water Quality Network
The New York District of the U.S. Geological Survey operates a water quality network throughout
the water supply watershed for New York City, in the Catskill Mountains region. The purpose of the
network is to provide stream flow and water quality data at key locations within the watershed. There are
currently 34 sites throughout the network at which stream flow data are collected, and at thirteen of those
sites stream water quality data are also collected. The water quality network is composed of paired
"nodes" consisting of one or more "upper nodes" that provide water quality of undeveloped, forested
A-8
-------
watersheds, and "lower nodes" that provide downstream water quality data that may reflect some level of
development within the watershed.
Water quality sampling for this program began at the 13 sites in 1998-99. Water samples are
collected biweekly and during high flow for approximately 6 storms per year. All water samples are
analyzed for concentrations of nutrients and major ions. Because streams in this area are also affected by
acidic deposition, acid-neutralizing capacity and 3 forms of A1 are also measured. Further details on the
program are available at: http://nv.cf.er.usgs.gov/nyc/unoono.cfm.
A.2.3.4. Catskill Long-Term Monitoring Sites
Within the Catskill Mountains region of New York State, stream samples are collected and stream
flow is measured at three locations within the Neversink River basin, and at one site on Rondout Creek.
Water samples are collected biweekly and during most storms. These sites are currently part of the
U.S. EPA LTM program, but also are affiliated with other programs. Sampling at two of the four sites
began in the mid 1980s, whereas sampling at the remaining two sites began in 1991. The primary purpose
of these sites is to monitor effects of acidic deposition on stream chemistry. The full suite of analytes
needed to assess acidic deposition effects are measured on these water samples.
A.2.3.5. Buck Creek, New York
Stream flow and water chemistry are monitored at three locations within Buck Creek watershed, in
the western Adirondack Region of New York. Samples are collected biweekly and during most storms at
each location. Sampling began in 1998 at two sites and 2001 at the third site. The full suite of analytes
needed to assess acidic deposition effects is measured on these water samples. Measurements of ANC and
pH were also collected at one site weekly for the period of 1991 to 2001 (Lawrence et al., 2004). Recent
data from Buck Creek are presented in Lawrence et al. (2007). Buck Creek is the only stream within the
acidified region of the Adirondacks where base flow and storm samples are collected in conjunction with
flow monitoring.
A.2.4. NSF Long-Term Ecological Research Network
The Long-Term Ecological Research (LTER) program constitutes a loose network of 26 sites
(Table A-l), funded by the National Science Foundation (NSF). There is increasing concern over such
globally significant problems as loss of biodiversity, climate change, destruction of forests, depletion of
stratospheric ozone, regional air and water pollution, and soil erosion. The research conducted at the
various LTER sites has examined and continues to examine aspects of these problems and provides
scientific information which has been invaluable in the formation of public policy. Site locations and
research activities are summarized in (Table A-l). A few of the sites that have been used most extensively
for evaluation of long-term effects of N and sulfur deposition are discussed in greater detail below.
A-9
-------
Table A-1.
LTER site locations and basic site description information.
Site
Landsat WRS
Institutional
Affiliations
Principal Biome/
Main Communities
Research Topics
H.J.
Andrews
Experimental
Forest
(AND) 44.2,
-122.2
Path 46 Row 29;
Lat/Long: 44°14'N/
122°11'W
Oregon State University;
USDA Forest Service
Pacific Northwest
Research Station
Temperate coniferous forest.
Douglas-fir/western hemlock/
western red cedar; true fir and
mountain hemlock; streams
Successional changes in
ecosystems; forest-stream
interactions; population dynamics of
forest stands; patterns and rates of
decomposition; disturbance regimes
in forest landscapes
Arctic
Tundra
(ARC) 68.6,
-149.6
Path 73, Row 12;
Lat/Long: 68°38'N/
149°34'W
The Ecosystem Center,
Marine Biological
Laboratory; Universities
of Alaska,
Massachusetts,
Minnesota, Cincinnati,
and Kansas; Clarkson
University
Arctic tundra, lakes, streams.
Tussock tundra; heath tundra;
riverine willows; oligotrophic
lakes; headwater streams
Research topics: Movement of
nutrients from land to stream to
lake; changes due to anthropogenic
influences; controls of ecological
processes by nutrients and by
predation
Baltimore
Ecosystem
Study (BES)
39.1,-76.3
Path 15, Row 33;
Lat/Long 38°54" 04'
(N), 76°52" 04' (W)
Institute of Ecosystem
Studies; USDA Forest
Service, Johns Hopkins
University; University of
Maryland; Baltimore
County and College
Park; University of North
Carolina; Parks and
People Foundation; US
Geological Survey; Yale
University
Eastern deciduous forest/
Suburban Agriculture fringe,
urban parks, residential and
commercial patches, riparian
and stream habitats
Patch dynamics of built, social,
biological, and hydrological
components of the metropolitan
area; feedback's between social,
economic, and ecological
components of an urban ecosystem;
effect of infrastructure and
development on fluxes of nutrients,
energy, and water in upland, stream,
and coastal regions of metropolitan
Baltimore
Bonanza
Creek
Experimental
Forest (BNZ)
64.8, -148.0
Path 69, Row 15;
Lat/Long: 64°45'N/
148°00'W
University of Alaska;
Institute of Northern
Forestry, USDA Forest
Service, Pacific
Northwest Research
Station
Taiga. Areas of boreal forest
including permafrost-free
uplands and permafrost-
dominated north slopes and
lowlands; floodplain seres
Successional processes associated
with wildfire and floodplains;
facilitative and competitive
interactions among plant species
throughout succession; plant-
mediated changes in resource and
energy availability for decomposers;
herbivorous control of plant species
composition; hydrologic regime and
stream ecology
Cedar Creek
Natural
History Area
(CDR) 45.4,
-93.2
Path 27, Row 28;
Lat/Long: 45°24'N/
93°12'W
University of Minnesota
Eastern deciduous forest and
tallgrass prairie. Old fields; oak
savanna and forest, conifer
bog; lakes; pine forest; wetland
marsh and carr
Successional dynamics; primary
productivity and disturbance
patterns; nutrient budgets and
cycles; climatic variation and the
wetland/upland boundary; plant-
herbivore dynamics
A-10
-------
Central
Arizona -
Phoenix
(CAP) 33.5,
-11.2
Path 36, Row 37 and
Path 36, Row 36°
These have been
used by CAP although
Path 37, Row 37 is
centered more closely
in the City of Phoenix
Arizona State University
(Main and West)
Sonoran Desert scrub. Urban
parks, residential, interior
remnant desert patches,
commercial and industrial
patches, urban fringe,
regulated river and floodplain
(dry), effluent-dominated river
Interactions of ecological and socio-
economic systems in an urban
environment; influence of land use
change on ecological patterns and
processes; movement of nutrients
through highly manipulated, urban
flowpaths; interactions of introduced
and native species in urban
environment; millennium- and
century-scale geomorphic change in
landforms and interaction with
engineered landscapes
Coweeta
Hydrologic
Laboratory
(CWT) 35.0,
-83.5
Path 18, Row 36;
Lat/Long: 35°00'N/
83°30'W
University of Georgia;
USDA Forest Service,
Southeastern Forest
Experiment Station
Eastern deciduous forest.
Hardwood forests and white
pine plantations
Long-term dynamics of forest
ecosystems including forest
disturbance and stress along an
environmental gradient; stream
ecosystems along an environmental
gradient; and the riparian zone as a
regulator of terrestrial-aquatic
linkages
Harvard
Forest (HFR)
42.5, -72.2
Path 13, Row 30;
Lat/Long: 42°32'N/
72°10'W
Harvard University;
Universities of New
Hampshire and
Massachusetts; The
Ecosystem Center,
Marine Biological
Laboratory
Eastern deciduous forest.
Hardwood-white-pine-hemlock
forest; spruce swamp forest;
conifer plantations
Long-term climate change,
disturbance history and vegetation
dynamics; comparison of
community, population, and plant
architectural responses to human
and natural disturbance; forest-
atmosphere trace gas fluxes;
organic matter accumulation,
decomposition and mineralization;
element cycling, fine root dynamics
and forest microbiology
Hubbard
Brook
Experimental
Forest
(HBR) 43.9,
-71.8
Path 13, Row 29;
Lat/Long: 43°56'N/
71 °45'W
Yale, Cornell, and
Syracuse Universities;
Institute of Ecosystem
Studies; USDA Forest
Service, Northeastern
Forest Experiment
Station
Eastern deciduous forest.
Northern hardwood forests in
various developmental stages,
spruce-fir forests; streams and
lakes
Vegetation structure and production;
dynamics of detritus in terrestrial
and aquatic ecosystems;
atmosphere-terrestrial-aquatic
ecosystem linkages; heterotroph
population dynamics; effects of
human activities on ecosystems
Jornada Path 33, Row 37;
Experimental Lat/Long: 32°30'N/
Range (JRN) 106°45'W
32.5,-106.8
New Mexico State
University; USDAARS
Jornada Experimental
Range; Duke University;
NOAA, RTP, NC;
University of New
Mexico; Dartmouth
College, NH; Oregon
Graduate Center; Texas
Technological University;
SUNY Buffalo; University
of Keele, UK; Kings
College, London, UK;
U.S. EPA-EMAP, Las
Vegas, NV
Hot desert. Playa, piedmont,
and swale; bajada, basin,
mountain and swale shrubland;
mesquite dunes
Desertification; factors affecting
primary production; animal-induced
soil disturbances; direct and indirect
consumer effects; vertebrate and
invertebrate population dynamics;
grazing effects on ecosystem
structure and function; biodiversity
and ecosystem function; small
mammal effects on soil and
vegetation heterogeneity; soil
microbial processes; surface
hydrology; trace gas emissions from
soils; eolian processes
A-11
-------
W.K. Kellogg
Biological
Station
(KBS) 42.4,
-5.4
Path 21, Row 31;
Lat/Long: 85°24'W/
42°24'N
Michigan State
University, Michigan
Agricultural Experiment
Station
Row-crop agriculture.
Conventional and organic-
based corn-soybean-wheat
cultivation; perennial biomass
cultivation; native successional
communities
Ecological interactions underlying
the productivity and environmental
impact of production-level cropping
systems; patterns, causes, and
consequences of microbial, plant,
and insect diversity in agricultural
landscapes; gene transfer,
community dynamics,
biogeochemical fluxes
Konza
Prairie
Research
Natural
Area(KNZ)
39.1,-94.6
Path 28, Row 33;
Lat/Long: 39°05'N/
96°35'W
Kansas State University
Tallgrass prairie. Tallgrass
prairie; gallery forest; prairie
stream
Effects of fire, grazing and climatic
variability on ecological patterns and
processes in tallgrass prairie
ecosystems, use of remotely sensed
data and geographic information
systems to evaluate grassland
structure and dynamics
Luquillo
Experimental
Forest (LUQ)
18.3,-65.8
Path 4, Row 47 and
48; Lat/Long:
18018'N/65047'W
Center for Energy and
Environment Research,
University of Puerto
Rico; Institute of Tropical
Forestry, USDA Forest
Service, Southern
Experiment Station
Tropical rainforest. Tabonuco
forest; palo Colorado forest;
palm brake; dwarf forest and
montane streams
Patterns of and ecosystem
response to different patterns of
disturbance; land-stream
interactions; effect of management
on ecosystem properties; integration
of ecosystem models and
geographic information systems
McMurdo
Dry Valleys -
Antarctica
(MCM)
-78.0,
+165.0
Path 56, Row 116
Desert Research
Institute, Reno, Nevada;
U.S. Geological Survey,
Boulder, Colorado
Polar desert oases
Microbial ecosystem dynamics in
arid soils, ephemeral streams, and
closed basin lakes; resource and
environmental controls on terrestrial,
stream and lake ecosystems;
material transport between aquatic
and terrestrial ecosystems;
ecosystem response to greater
hydrologic flux driven by warming
climate
Niwot Ridge/
Green Lakes
Valley
(NWT) 40.1,
-105.6
Path 34, Row 32;
Lat/Long: 40°03'N/
105°37'W
Institute of Arctic and
Alpine Research,
University of Colorado
Alpine tundra; Fellfield;
meadow; herbaceous and
shrub tundras; cliffs and talus;
glacial lakes; streams and
wetlands
Patterns and controls of nutrient
cycling; trace gas dynamics, plant
primary productivity and species
composition; geomorphology, and
paleoecology
North
Temperate
Lakes (NTL)
46.0, -89.7
and 43.1,
89.4
Path 25, Row 28 and
Path 24, Row 30
Lat/Long: 46°00'N/
89°40'W and
89°24/43°06
Center for Limnology,
University of Wisconsin-
Madison, Wisconsin
Northern temperate lakes in
glacial landscapes in urban,
agricultural and forested
watersheds. Oligotrophia
dystrophic and eutrophic lakes;
temporary forest ponds; warm
and cold streams; sphagnum-
leatherleaf bog; conifer swamp;
mixed deciduous and
coniferous forests
Physical, chemical and biological
limnology; hydrology and
geochemistry; climate forcing;
producer and consumer ecology;
ecology of invasions; ecosystem
variability; lakescape and landscape
ecology
A-12
-------
Palmer
Path 219, Row 105;
University of California,
Polar marine. Coastal and
Oceanic-ice circulation and models;
Station
Lat/Long: 64°40'S/
Santa Barbara; Old
open ocean pelagic
sea-ice dynamics;
(PAL)
64°W
Dominion University
communities; seabird nesting
biological/physical interactions;
Antarctica
areas
effect of sea ice on primary
-64.7, -64.0
production, consumer populations
and apex predators; bio-optical
models of primary production;
spatial distribution and recruitment
in consumer populations; seabird
population dynamics and
reproductive ecology
Plum Island
Path 12, Row 30; Lat/
The Ecosystems Center,
Coastal estuary
Linkages between land and coastal
Sound (PIE)
Long: 42°40770°59'
Marine Biological
waters involving organic carbon and
42.67,
Site has the following
Laboratory; Universities
organic N inputs to estuarine
-70.99
X and Y bounds in
of South Carolina and
ecosystems from watersheds with
decimal coordinates:
New Hampshire;
various land covers and uses
X min =-71.22 X max
Massachusetts
=-70.75. Ymin =
Audubon; Wells, Maine,
42.50 Y max = 42.83.
NERRS
The total area is
approximately 37 km
x37 km or 1369 km2
Sevilleta
Path 33, Row 36;
University of New
Multiple: intersection
Landscape and organism population
National
To acquire entire site
area, Path 32, Row
36, Path 32, Row 37
and Path 33, Row 37
are also needed. Lat/
Mexico; U.S. Fish and
of subalpine mixed-conifer
dynamics in a biome tension zone;
Wildlife
Wildlife Service
forest/meadow, riparian
semiarid watershed ecology; climate
Refuge
cottonwood forest, dry
change; biospheric/atmospheric
(SEV) 34.3,
mountainland, grassland, cold
interactions; paleobotany/
-106.8
desert, hot desert. Conifer
archaeology; microbial role in gas
Long: 34° 19'/
106°62'W
savanna; creosote bush;
flux; and control of landscape
desert grassland; mesquite
heterogeneity; scale effects on
and sand dunes; Great Basin
shrub and shortgrass steppes;
tallgrass swales; riparian
communities
spatial and temporal variability
Shortgrass
Path 33, Row 32;
Colorado State
Floodplain; shrubland;
Soil water; above- and belowground
Steppe
Lat/Long: 40°49'N/
University; USDA Forest
saltmeadow
net primary production; plant
(SGS) 40.8,
104°46'W
Service; USDA
population and community dyna-
-104.8
Agricultural Research
Service
mics; effects of livestock grazing;
soil organic matter accumulation
and losses, soil nutrient dynamics;
and ecosystem recovery from
cultivation
Virginia
Path 14, Row 34;
University of Virginia
Coastal barrier islands. Sandy
Holocene barrier island geology; salt
Coast
Lat/Long: 37°30'N
intertidal; open beach;
marsh ecology, geology, and
Reserve
75°40'W
shrubthicket; mature pine
hydrology; ecology/ evolution of
(VCR) 37.5,
forest; salt marsh; estuary
insular vertebrates; primary/
-74.8
secondary succession; life-form
modeling of succession
A.2.4.1. Hubbard Brook Experimental Forest
The Hubbard Brook Ecosystem Study (HBES) at Hubbard Brook Experimental Forest (HBEF) is
the longest-running precipitation and stream chemistry (1963 to present) monitoring program in the U.S.
A-13
-------
(see http://www.hubbardbrook.org'). HBEF was established in 1955 as a major center for hydrologic
research in New England. The site is located within the boundaries of the White Mountain National Forest
in central New Hampshire. The 3138-ha, bowl-shaped valley has hilly terrain, ranging from 222 to 1015
m elevation. The HBES originated in 1960 with the intention of applying the small watershed approach to
the study of element fluxes and cycles. The goal of the study is to develop a better understanding of
ecological patterns and processes that characterize the northern forest in eastern North America, and its
response to both natural and human disturbances. In 1987, HBEF joined the NSF's LTER network
(http://www.lternet.edu'). Hubbard Brook is renowned for its long-term record of measurements,
landscape-scale experiments of whole watersheds, and the involvement of scientists from diverse
disciplines and institutions.
The HBEF is entirely forested, mainly with deciduous northern hardwoods: sugar maple (Acer
saccharum), beech (Fagus grandifolia), and yellow birch (Betula allegheniensis), and some white ash
(Fraxinus americana) on the lower and middle slopes. Other less abundant species include mountain
maple (Acer spicatum), striped maple (Acer pensylvanicum), and trembling aspen (Populus tremuloides).
Red spruce (Picea rubens), balsam fir (Abies balsamea), and white birch (Betula papyrifera var.
cordifolia) are abundant at higher elevations and on rock outcrops. Hemlock (Tsuga canadensis) is found
along the main Hubbard Brook. Pin cherry (Prunus pensylvanica), a shade intolerant species, dominates
all sites for the first decade following a major forest disturbance. Logging operations ending around
1915-1917 removed large portions of the conifers and better quality, accessible hardwoods. The present
second-growth forest is even-aged and composed of about 80 to 90% hardwoods and 10 to 20% conifers.
The HBEF is an oblong basin about 8 km long by 5 km wide. Hubbard Brook is the single major
stream draining the basin. Numerous smaller tributary streams of varying size drain into Hubbard Brook
including Watershed 6 (WS-6), which is the biogeochemical reference watershed.
One of the strengths of the HBES is the long-term monitoring program. Table A-2 lists the major
parameters included in the HBES long-term monitoring study. The monitoring data illustrate that short-
term observations can be misleading and that decades of monitoring may be required to detect real
changes in complex ecosystems. The long-term record at the HBEF provides: insight into ecosystem
function; empirical data for testing models and generating hypotheses; a record of extreme or unusual
events; and information that is relevant to regional national and global environmental issues.
Table A-2. Current long-term monitoring data sets developed through the Hubbard Brook Ecosystem
Study
PHYSICAUHYDR0L0GIC MONITORING
SOLUTION CHEMISTRY
Weekly bulk precipitation (6-10 stations)
Monthly soil solution WS5, WS6
Weekly stream at weirs of WS19
Monthly stream within WS5, WS6
Instantaneous streamflow (9 stations)
Daily precipitation (24 stations)
Class A weather station data
Weekly snow depth on snow courses
Daily soil temperature and moisture
ORGANISMS
Bird populations
Phytophagous insect populations WS2, WS4, WS5, WS6
Vegetation biomass, chemistry
AIR CHEMISTRY
(SO2, HNO3, particulates, ozone)
MIRROR LAKE
Instantaneous streamflow (3 inlets,outlet)
Daily precipitation (2 stations)
Weekly chemistry (3 inlets, outlet)
Bi-monthly limnology (temp, chemistry, plankton)
SOILS
Forest floor mass, chemistry (WS6, WS5; 5-yr intervals)
Chemical and physical properties from soil pits (WS5)
Chemical and physical properties from soil bags
A-14
-------
Some of the monitoring is done on experimentally manipulated watersheds. There are nine gauged
watersheds at the HBEF, four of which have been treated experimentally. A tenth ungauged watershed
was also treated. Table A-3 includes summary data on the various watersheds. Datasets for long-term
monitoring can be found at http://www.hubbardbrook .org/data/dataset search.php. The datasets most
often used to examine ecosystem response to ambient deposition of N and S are from WS-6 and Mirror
Lake, since they have not been experimentally manipulated. Watershed 6 is the biogeochemical reference
catchment at HBEF where monitoring began in June 1963. Measured stream chemistry parameters
include major anions and cations, pH, silica, dissolved organic and inorganic carbon, specific
conductance, dissolved 02, ANC, and PO4. Stream chemistry data can be accessed at
http://www.hubbardbrook.org/data/dataset .php ?id=8. The normal sampling interval for WS-6 is weekly,
with more frequent samples taken at times of increased discharge.
Table A-3. Study watersheds at HBEF.
ws
Area (ha)
Slope (°)
Aspect
Elevation (m)
Gauge Type
Initial Yr.
1
11.8
18.6
S22°E
488-747
V-notch weir
1956
2
15.6
18.5
S31°E
503-716
V-notch weir
1957
3
42.4
12.1
S23°W
527-732
V-notch weir
1957
4
36.1
15.6
S40°E
442-747
V-notch weir
1960
5
21.9
15.4
S24°E
488-762
V-notch weir, San Dimas flume
1962
6
13.2
15.8
S32°E
549-792
V-notch weir, San Dimas flume
1963
7
77.4
12.4
N16°W
619-899
V-notch weir, San Dimas flume
1965
8
59.4
14.0
N12°W
610-905
V-notch weir, San Dimas flume
1968
9
68.4
NE
685-910
V-notch weir
1995
10
12.1
SE
470-595
None
1970
Monitoring of streamflow and water chemistry has shown that the study watersheds have similar
characteristics. Within each watershed there are a variety of soils, vegetation, microtopographical
features, and micro-climate. Nevertheless, the composition of these variables seems to be similar from
watershed-to-watershed. Thus, the effects of experimental manipulations of watersheds can be adequately
evaluated by comparison with neighboring unmanipulated watersheds.
The most conspicuous streamflow characteristic is the seasonal shift from large volume of flow in
spring to very low flow in late summer and early autumn. These yearly highs and lows reflect seasonal
spring snowmelt that often occurs over a few days or weeks and the slow progressive decrease in flow
from the transpirational draft in summer, respectively. The numerous streams in the HBEF range from
small ephemeral channels that often dry up during summer to a large perennial 5th-order stream (Hubbard
Brook).
Mirror Lake is a 15-ha oligotrophic clearwater lake adjacent to HBEF. The lake normally mixes in
spring and fall, and is ice-covered from about December 1st to April 15th each year. Part of the drainage
A-15
-------
to the lake originates in the Experimental Forest. The lake water is dilute, slightly acidic, and quite clear,
with low productivity and low concentrations of nutrients in the water. Numerous studies have been
conducted on Mirror Lake since the mid-1960s, including extensive physical, chemical, biological, and
paleoecological research (Likens, 1985). Data are available since 1967 for lakewater concentrations at
discrete depths for base cations, pH, and dissolved 02. Ammonium, major anions, phosphate, and
dissolved silica have been measured routinely since 1970, although some data are available before these
dates for each solute. Other standard monitoring data include temperature and specific conductance at
each depth. Before 1990, not all records had complete solute arrays. Since 1990, DIC and ANC have also
been measured on a routine basis, although some prior data do exist for those parameters. The usual
sampling interval for Mirror Lake is four to six times each year, especially at times of maximum and
minimum thermal stratification. Data for Mirror Lake and inlet and outlet streams can be found at:
htto://www.hubbardbrook.org/data/dataset search .oho.
The soils, vegetation, and climate at the HBEF are characteristic of the northern hardwood forest
complex, which spans much of the north-central and northeastern U.S. and southeastern Canada.
Streamflow and stream chemistry reflect the landscape characteristics of the drainage area. Consequently,
results from the relatively small watersheds at the HBEF are to a first approximation representative of a
much larger regional area. During the scientific debate that occurred before passage of the CAAA of
1990, the trends in sulfate (S042 ) concentrations in streamwater and precipitation at HBEF were very
influential in convincing scientists and policy makers that decreasing S emissions would yield large
decreases in the concentration of S042 in precipitation and streamwater in the northeastern U.S. (Lovett
et al., 2007). Monitoring data collected since 1963 (Figure A-3) played a major role in development of the
watershed-ecosystem concept and methods for analyzing and understanding watershed biogeochemical
cycles (Bormann and Likens, 1967; Likens et al., 1978; Likens and Bormann, 1995; Lovett et al., 2007).
An extensive effort has been made to bring together some of the results of research that had been
done at Hubbard Brook over the last several decades. Over the duration of the HBES there have been six
books and more than 1,000 papers published. In addition, more than 500 abstracts were published and
more than 100 graduate theses completed. A complete list of titles is available at
htto://www.hubbardbrook.org/pubs/pub search.oho. To date, four synthesis volumes have been completed
(Bormann and Likens, 1979; Likens et al., 1977; Likens, 1985, 1992).
Figure A-3. Long-term record of SO42" concentration in streamwater and precipitation at Watershed 6 of
180
Streamwater
Precipitation
20 -
00
1960 1965 1970 1975 1980 1985 1890 1995 2000 2005
Source: Lovett et al. (2007); updated from Likens et al. (2002).
HBEF.
A-16
-------
A.2.4.2. Coweeta
The Coweeta LTER research program (http://coweeta.ecologv.uga.edu/') in North Carolina is based
in the eastern deciduous forest of the Blue Ridge Physiographic Province of the southern Appalachian
Mountains. The program entails long-term cooperation between the University of Georgia and the U.S.
Department of Agriculture (USDA) Forest Service Coweeta Hydrologic Laboratory. The research
program centers on the effects of disturbance and environmental gradients on biogeochemical cycling,
and the underlying watershed ecosystem processes that regulate and respond to those cycles. Coweeta
represents one of the longest continuous environmental studies of any North American landscape.
The research at Coweeta focuses largely on how water, soil, and forest resources respond to
management practices, natural disturbances, and the atmospheric environment. It also aims to identify
practices that mitigate impacts on these watershed resources. Current topics of emphasis include analyses
of long-term changes in hydrology, nutrient cycling, and productivity in response to management
practices and natural disturbances; assessment of prescribed burning effects on the forest environment;
interdisciplinary implementation of ecosystem management on the national forests; effects of climatic
change on productivity; impacts of atmospheric deposition on forest processes and ecosystems;
cumulative effects of land use practices on water quality; physiological studies of forest carbon balance
and competition; and biodiversity.
Investigators at the Coweeta Hydrologic Laboratory have recorded N dynamics of streams and
precipitation in mature mixed hardwood-covered watersheds since 1972. Research has been conducted on
responses to management practices such as clearcutting, selective cutting, conversion of native hardwood
to coniferous forest, and old-field succession. Reference watersheds were characterized as in a transition
phase between stage 0 and stage 1 of watershed N saturation. Evidence for stage 3 of N saturation, where
the watershed is a net source of N rather than a N sink, was found for the most disturbed watershed at
Coweeta.
The Coweeta Basin comprises 2185 hectares within the Blue Ridge geologic province in North
Carolina. The laboratory has been dedicated to forest hydrology research since its establishment in 1933.
Elevations range from 679 to 1592 m. More than 50 km of streams drain the area.
Coweeta is the first major mountain range contacted by air masses moving over the industrialized
Piedmont region to the south. Analyses of precipitation chemistry have shown the influence of both local
and regional activities on nutrient inputs to forest ecosystems.
Since Coweeta was established, 32 weirs have been installed on streams. Seventeen of these weirs
are currently operational. Stream gauging was initiated on most watersheds between 1934 and 1938, and
stream chemistry measurements date back to 1968.
Research has been conducted on eight mixed hardwood control areas and 13 catchments where
forest management prescriptions have been applied. Past treatments have included varying intensities of
cutting, ranging from light selection through clear-cutting; conversion of hardwoods to grass and
subsequent succession to hardwoods; multiple-use management; mountain farming; and the application of
herbicides and fertilizers.
Research and monitoring data from Coweeta has been extensively analyzed and reported in the
scientific literature. Example recent publications include Swank and Vose (1997), Grossman et al. (1998),
Schofield et al. (2001), and Scott and Helfman (2001).
Long-term changes in soils have been identified in reference and managed watersheds over two
decades (Knoepp and Swank, 1994). For example, changes in exchangeable soil cation content varied
with aspect: concentrations decreased on a north-facing slope but were stable on a south-facing slope. The
demonstrated impacts of forest management practices have varied considerably. Soils in a white pine
plantation showed stable C levels, but cations declined.
Commercial sawlog harvest resulted in large increases in soil C and cation concentrations, which
remained elevated for 17 years. Whole-tree harvest resulted in decreased soil C for the next 14 years.
Clearly soil response to harvest varies with type of harvest and site. Long-term studies like these have
A-17
-------
proven useful in guiding ecosystem management projects in the southern Appalachians (Meyer and
Swank, 1996).
A.2.4.3. Walker Branch
Walker Branch Watershed is located on the U.S. Department of Energy's Oak Ridge Reservation in
Tennessee. The 97.5 ha Walker Branch watershed has been the site of long-term, intensive environmental
studies since the late-1960s (see http://walkerbranch.ornl.gov/).
The forest soils are acidic, very cherty, infertile, and permeable. They are formed over dolomitic
bedrock, but retain little evidence of their carbonate parent material. The forest vegetation is primarily
oak-hickory with scattered pine on the ridges and mesophytic hardwoods in the valleys.
Initially, the research and monitoring of Walker Branch centered primarily on the geologic and
hydrologic processes that control the amounts and chemistry of water moving through the watershed. Past
projects have included:
¦ watershed hydrology and forest nutrient dynamics,
¦ forest micrometeorology,
¦ atmospheric deposition,
¦ International Biological Program Eastern Deciduous Forest Biome Project,
¦ trace element cycling and stream nutrient spiraling, and
¦ effects of acidic deposition on canopy processes and soil chemistry.
These projects have all contributed to a more complete understanding of how forest watersheds
function and have provided insights into the solution of energy-related problems associated with air
pollution, contaminant transport, and forest nutrient dynamics. Available long-term data at this site
include:
¦ Daily climate data
¦ Monthly climate data
¦ Precipitation
¦ Atmospheric deposition
¦ Stream discharge and annual runoff
¦ Stream chemistry
¦ Vegetation
A.2.5. Water, Energy, and Biogeochemical Budgets Program
The Water, Energy, and Biogeochemical Budgets (WEBB) Program was started in 1991 at five
small watersheds in the U.S. to examine water, energy, and biogeochemical fluxes and to determine the
effects of atmospheric deposition, climatic variables, and human influences on watershed processes. The
five sites are at Loch Vale, Colorado; Luquillo Experimental Forest, Puerto Rico; Panola Mountain,
Georgia; Sleepers River, Vermont; and Trout Lake, Wisconsin. These sites are supported, in part, by other
programs in the USGS, other Federal and State Agencies, and Universities. Two of these sites, Loch Vale
and Sleepers River, have been used extensively to evaluate the effects of atmospheric sulfur and N
deposition, and are described here. Each of those sites is also part of the LTER network.
A-18
-------
A.2.5.1. Sleepers River
The Sleepers River Research Watershed in northeastern Vermont was established by the
Agricultural Research Service (ARS) of the USDA in 1959 and is now operated jointly by the USGS and
the U.S. Army Cold Regions Research and Engineering Laboratory (CRREL), with collaboration from
several other Federal agencies and universities (see http://nh.water.usgs.gov/proiects/sleepers/index.htm').
The USGS uses hydrologic measurements and chemical and isotopic tracing techniques to determine how
water moves from the hillslope to the stream, and what processes cause chemical changes, including the
neutralization of acid rain. Research results provide insights on how pollutants move through ecosystems,
and how ecosystems may respond to climatic change.
The watershed is covered by 1 to 4 m of glacial till, a compacted fine silty material that formed
underneath glacial ice as it moved overland. The till was formed primarily from local bedrock, which is a
calcareous granulite/quartz-mica phyllite. About 60 to 80 cm of soil has developed in the till. Weathering
of calcite in the till and bedrock causes highly buffered streamflow, compared to most streams in New
England, and a nutrient-rich biological environment. Sleepers River is, therefore, an end member in
regional biogeochemical cycling studies (Hornbeck et al., 1997).
The Sleepers River area has reverted from a predominately cleared, agricultural landscape to a
mostly forested one. A Northern Hardwood forest, dominated by sugar maple, white ash, yellow birch,
and beech, with lesser amounts of red spruce and balsam fir, now covers two-thirds of the area; the
remaining open land is primarily pasture and hayfields. Dairy farming and logging are the primary human
enterprises in the watershed. The average annual temperature is 6 °C and the average annual precipitation
is 1.1m, 20% to 30% of which falls as snow.
Sleepers River has one of the longest historical hydrologic and climatologic data bases for a cold-
region area in the U.S., featuring measurements of precipitation and streamflow since 1959, snow depth
and corresponding water content since 1960, soil frost depth since 1984 (Shanley and Chalmers, 1999),
and ground-water levels since 1991. These and other measurements constitute a valuable resource for
hydrologic modeling and for the evaluation of climatic changes. Sampling site locations are shown in
Figure A-4.
A-19
-------
VERMONT
Study area
Montpelier
W-9
[forest]
W-3
W-2
(pastural)
Sriowmelt
research
station
T3
N, Danville
W-5
0 2 km
A Stream-gaging station i i
1 1
¦ Meteorological station o 2 mi
Figure A-4. Location of sampling stations in Sleepers River watershed, Vermont.
Recent research findings include the following:
¦ Precipitation is acidic, but streamflow is well-buffered from calcite weathering in till and
bedrock.
¦ Infiltrating snowmelt causes ground water to rise into the permeable soil zone, where it moves
rapidly downslope.
* Naturally occurring isotopic and chemical tracers indicate that "old" water dominates streamflow,
and that water acquires solutes from weathering and biogeochemical processes along both deep
and shallow flow paths.
* Nitrate (NO, ) in streamflow is supplied primarily by mineralization and nitrification in the soil,
rather than directly by the N content of precipitation.
The fate of N 03 in the forest ecosystem is being investigated by analysis of both the N and O
isotopes of the NOgT ion. The isotopic composition of NO?" in streamflow matches that of NO, produced
by mineralization and nitrification in the soil, indicating that streamflow NO;, is derived from the soil and
not from the rain or snowmelt that causes the high flow (Kendall et al., 1995). This finding suggests that
most incoming atmospheric N is incorporated at least temporarily in the soil where it is utilized by the
biota.
A-20
-------
A.2.5.2. Loch Vale
The Loch Vale Watershed is a 661-ha alpine/subalpine basin located in the south-central Rocky
Mountains, about 100 km northwest of Denver, Colorado. The basin is in a roadless area in Rocky
Mountain National Park and is accessed by a 5-km hike or ski from the trailhead near Bear Lake. The
western boundary of the basin is the Continental Divide; streams drain to the northeast. Basin elevations
range from 4192m(13,153 ft) at Taylor Peak to 3110 m (10,200 ft) at the outlet. There are two main
subbasins in Loch Vale: Andrews Creek drains the northern subbasin, and Icy Brook drains the southern
subbasin. These two creeks join above The Loch, which is the lowest of three lakes in the basin. Stream
gauges are operated on Andrews Creek, Icy Brook, and at The Loch outlet. Water chemistry monitoring
occurs at The Loch and on both inlet streams (see http://nh.water.usgs.gov/proiects/sleepers/index.htm').
Large glaciers that covered much of Rocky Mountain National Park during the late Pleistocene
sculpted the basin into characteristic glacial landforms, including steep U-shaped valleys, cirques, and
aretes. When the glaciers retreated about 12,500 years ago, they deposited till of varying thickness, which
is confined mostly to the forested, lower part of the basin. Smaller, more recent glacial advances left
younger till, talus, and rock deposits in the upper parts of the basin. The younger glacial and periglacial
deposits are largely unvegetated.
Available water chemistry data include major ions, nutrients, DIC, DOC, and, for selected samples,
a range of isotopes including 2H/H, 180/160,15N/14N and 180/160 in nitrate ion, 35S, 34S/32S, 87Sr/86Sr, 13C,
12C). Monitoring of precipitation and hydrology include the following elements:
¦ Precipitation: quantity — 3 sites continuous; chemistry, 1 site, biweekly.
¦ Stream discharge, 2 sites (Andrews Creek-Loch Vale, Icy Brook-Loch Vale).
¦ Stream chemistry, (Andrews Creek, Icy Brook-Loch Vale).
¦ Spring discharge, conductance, and temperature, 3 sites continuous.
¦ Spring water chemistry, 3 sites biweekly, 20-30 sites once during low-flow season.
¦ Soil lysimeters, 5 sites, biweekly to monthly during summer and fall.
¦ Snowpack amount and chemistry (depth, snow-water equivalent) basin-wide survey at maximum
accumulation, index sites biweekly to monthly.
¦ Selected microenvironment runoff, e.g., rock outcrop, talus fields, weekly to monthly.
¦ Meteorology: 3 sites (wind speed, wind direction, air temperature, incoming and outgoing
radiation, relative humidity), continuous.
¦ Gas flux (C02 and CH4) in wetland, forest, and talus soils, weekly to monthly; C02
concentrations in surface waters at 10-15 sites several times annually.
¦ Snowmelt lysimeter discharge and chemistry, monitored for three years, currently inactive.
Atmospheric deposition of N to Loch Vale is high compared to most other sites in the Rockies,
although considerably lower than most impacted sites in eastern North America and Europe. The
alpine/subalpine ecosystem at Loch Vale exhibits symptoms of advanced watershed N saturation,
indicating sensitivity to N deposition. Talus landscapes contribute substantially to N export in streamflow,
and soil microbial processes are important in cycling N, even in areas such as talus that have little soil
development. Research at this site indicates that N export is a function of both deposition and internal N-
cycling processes that are affected by variability in climate.
A-21
-------
A.2.6. Other Monitoring Programs
A.2.6.1. Bear Brook
The Bear Brook Watershed (BBW) is located in eastern Maine (44°52'15" latitude, 68°06'25"
longitude), approximately 60 km from the Atlantic coastline. The BBW is a paired watershed study
funded by the U.S. EPA since 1987 as part of The Watershed Manipulation Project (WMP) within the
National Acid Precipitation Assessment Program (NAPAP) (see http://hvdromodel.com/bbwm.htm;
htto://www.umaine.edu/DrSoils/bbwm/bbwm.html). As a long-term research watershed, the BBW
includes bench-scale, micro-site, plot, and whole watershed investigations. The major purposes of the
BBW project are to:
¦ Identify and quantify the major processes that control surface water acidity, with a major
emphasis on the role of excess S042 and nitrate provided via atmospheric deposition and
experimental application, and the rate of cation supply from chemical weathering and cation
desorption.
¦ Assess the quantitative and qualitative responses at the watershed level to different (both
increased and decreased) levels of acidic deposition.
¦ Evaluate the ability of existing models of water acidification to predict short- and long-term
chemical variations in surface water chemistry and to predict watershed soil responses to
increased and decreased loading of strong acids.
The watershed includes two first order streams: East Bear Brook (EBB) and West Bear Brook
(WBB). On each stream, a catchment outlet was selected and gauged so that both streams have about the
same catchment area (EBB = 10.7 ha and WBB = 10.2 ha). Since streams are close and face the same
slope direction, the watersheds are geographically similar and are appropriate for a paired watershed
studies. Both watersheds have a maximum discharge of about 0.01 mm/ha/s or 0.15 m3/s. Annual water
yield relative to incoming precipitation for WBB ranges from 68 to 77% and EBB ranges from 62-68%.
Stream channels in each watershed are well defined. Each stream bed is approximately 1 m wide at
the weir and water flows over exposed bedrock in places. Elsewhere, the streambeds are comprised of
boulders and gravel. Both streams have undergone intermittent dry periods during summer over the
course of the study. One V-notch weir was constructed on each of the streams during winter 1987-1988.
Mean discharge in each stream is about 0.13 cubic feet per second.
Sampling frequency at the weirs was every three weeks during the winter of 1986-1987 and at
least weekly thereafter. On the basis of sampling conducted before beginning the manipulation
experiment (1987-1989), the streams had the following characteristics: ANC, -5 to 90 j^ieq/L: air-
equilibrated pH, 4.7 to 7.2; specific conductance of approximately 26 (iS/cm; and DOC of 1 to 4 mg C/l.
Soils in the Bear Brook watersheds are primarily Spodosols. The average depth of the overburden
in the watersheds is 0.5 m, with a range of 0 to 5.2 m. Soil pH (0.01M CaCl2) values ranged from 2.9 in
the O horizon, to 3.9 in the B horizon, to 4.4 in the C horizon. The bedrock is primarily metamorphosed
and folded politic graded beds and quartzites, with granitic dikes. The surficial material is till.
The forest is comprised primarily of deciduous species with areas of conifers. Tree species include
American beech (Fagus grandifolia), birch (Betula sp.), maple (Acer sp.), red spruce (Picea rubens),
balsam fir (Abies balsamea), white pine (Pinus strobus), and hemlock (Tsuga canadensis). Coniferous
stands, which occupy approximately 17% of the total watershed area, occur more commonly in the upper,
steeper portions of the watersheds.
Although the Bear Brook project was intended as an experimental manipulation of West Bear
Brook, there is also great value in the long-term monitoring data collected at East Bear Brook, the non-
manipulated reference watershed. This two-decade long monitoring record provides information on the
A-22
-------
response of an acid-sensitive low-order stream in Maine to changes that have occurred in atmospheric
deposition since 1986.
Results of the Bear Brook project have been widely published (cf. Kahl et al., 1993b; Norton et al.,
1999a, 1999b, 1994).
A.2.6.2. Shenandoah Watershed Study
The Shenandoah Watershed Study (SWAS) program is a monitoring and research network focused
on low-order, high-gradient streams associated with public lands in western Virginia (see
htto ://swas .evsc.virginia.edu/'). The objectives of the program are to increase understanding of factors that
govern biogeochemical conditions and stressor-response relationships in forested mountain watersheds of
the central Appalachian region. Success in addressing these scientific and problem-oriented objectives has
been achieved through development of a data collection network that accounts for spatial gradients, as
well as temporal variation, in the chemical composition of the region's relatively undisturbed headwater
streams.
The program is notable for the length of the continuous data record that has been obtained,
including the longest-running record (28 years) of stream water composition and discharge in the National
Park System. The SWAS component of the program, which now includes 14 streams in Shenandoah
National Park, was initiated in 1979. The Virginia Trout Stream Sensitivity Study (VTSSS) component,
which now includes 51 streams in National Forests and other conservation lands, was initiated in 1987.
The distribution of SWAS-VTSSS study sites in relation to public lands is shown in Figure A-5
The SWAS-VTSSS program has been maintained as a cooperative effort involving the Department
of Environmental Sciences at the University of Virginia, the National Park Service, the U.S. EPA, the
USDA Forest Service, the U.S. Geological Survey, the Virginia Department of Game and Inland
Fisheries, and Trout Unlimited. The monitoring sites account for ecological variation among the region's
forested mountain watersheds with a data-collection strategy that represents: spatial variation through the
distribution of hydrochemical monitoring within a lithologic classification system; and temporal variation
through long-term data collection at fixed locations sampled at different frequencies.
A-23
-------
Shenandoah
National Park
George Washington and
Jefferson National Forests
Figure A-5. SWAS-VTSS Program Study Sites. The length of the stream water chemical composition
record for SWAS-VTSSS study watersheds is 20 to 28 years. SWAS sites are located in
Shenandoah National Park (shaded blue). VTSSS sites are mainly located in Virginia's
National Forests (shaded green).
The lithologic classification system includes 6 classes based on the physical and chemical
properties of bedrock formations in the region. ANC and concentrations of related acid-base constituents
in stream waters, as well as other biotic and abiotic properties of watersheds, differ among the lithologic
classes.
The SWAS-VTSSS data collection framework is most-well developed in the Blue Ridge Mountains
Province within Shenandoah National Park, where stream water composition data are collected seasonally
at 14 sites, weekly at 6 sites, and every four hours during episodic high-flow conditions at 3 sites with
continuous discharge gauging. Stream water composition data are collected on a seasonal basis at an
additional 51 sites located outside of the Park, in both the Blue Ridge Mountains and Ridge and Valley
Provinces.
Stream water samples collected through the SWAS and VTSSS programs are analyzed for ANC,
pH, and the major anions (S042 , nitrate, and chloride) and cations (calcium, magnesium, potassium, and
sodium) by methods appropriate for low-ionic strength natural waters. Both the SWAS and VTSSS
sample streams were selected based on geographic distribution, representation of the major bedrock types
underlying the mountain ridges in the region, and minimization of recent watershed disturbance. All but a
few of the sample streams currently support reproducing populations of native brook trout. All of the
sample streams supported brook trout populations historically.
Sustained data collection in a network constructed of intensively studied sites nested within a
geographically extensive set of less intensively studied sites has allowed detection and interpretation of
change that has occurred in a context of multiple time scales and stressors. Responses to multi-year
changes in acidic deposition have been reflected in long-term trends in quarterly concentrations of S042 ,
ANC, and other acid-base constituents of streams in the network. Expectations for southeastern
watersheds with soils that retain sulfur, for example, have been confirmed by the lack of regional
improvement in stream water quality following reductions in acidic deposition mandated by the CAA.
The acid-base chemistry of streams in the network also varies seasonally and on shorter time scales.
Weekly and higher-frequency automated stream water sampling during periods of high runoff have
supported the study of episodically more-acidic conditions, including the study of fish sensitivity with in-
stream bioassays and development of models to predict severity and recurrence intervals.
A-24
-------
By accounting for significant spatial gradients and temporal patterns in the region, the SWAS-
VTSSS hydrochemical data collection program provides a basis for both observing and interpreting
watershed-scale change, as well as an informed foundation for process-oriented research. Monitoring data
and research findings obtained through the SWAS-VTSSS program have contributed to increased
scientific understanding, as well as to policy formulation and implementation.
The mathematical model, Model of Acidification of Groundwater in Catchments (MAGIC), was
first calibrated using data obtained for White Oak Run, a SWAS-VTSSS study stream in Shenandoah
National Park. MAGIC is the most widely used acid-base chemistry model in the U.S. and Europe and the
principal model used by the National Acid Precipitation Assessment Program in the 1980s to estimate
future damage to lakes and streams in the eastern U.S. The MAGIC model has since been applied in a
number of regional assessments that relied extensively on stream water and soils data obtained through
the SWAS-VTSSS program. Among these are:
¦ The Southern Appalachian Mountain Initiative, a multi-state effort to evaluate alternative
approaches to solving regional air-pollution problems. MAGIC projections indicated that even
ambitious emission control strategies would not result in near-term recovery of the region's most
acidified surface waters - a consequence of base-cation depletion in soils exposed to decades of
acidic deposition.
¦ The Shenandoah Assessment, an assessment of acidification effects on aquatic systems in
Shenandoah National Park. MAGIC reconstructions indicated that Park streams associated with
base-poor bedrock lost about 70 j^ieq/L between 1900 and 1990. MAGIC projections indicated
that some streams may recover given prospective reductions in acidic deposition, but others will
not.
Data and findings provided through the SWAS-VTSSS program have also proven relevant to the
evaluation and implementation of national air pollution control policies. The SWAS-VTSSS program
provides data for the U.S. EPA long-term monitoring of surface water response to legislated reductions in
sulfur emissions. Whereas S042 concentrations in surface water declined during the 1990-2000 period
for four northeastern regions with sensitive surface waters, the SWAS-VTSSS study region, in contrast,
experienced increasing stream-water S042 concentrations and continuing acidification.
Recent publications that were based on analyses of SWAS-VTSSS data include Cosby et al. (1991),
Stoddard et al. (1993), Sullivan et al. (2003), and Webb et al. (2004).
A.2.6.3. Fernow
The Fernow Experimental Forest, established in 1934, is located just south of the city of Parsons in
the most mountainous region of West Virginia. It is surrounded by the Monongahela National Forest,
which comprises about 900,000 acres of rugged, hilly terrain. Most research at Fernow is focused on
improvement of forest management (see http://www.fs.fed.us/ne/parsons/fefhome.htm).
Scientists at Fernow are developing information and techniques for sustainably managing
hardwood forests in the central Appalachians. The mixed hardwood forest covers about 78% of West
Virginia and supplies important timber products, provides recreational opportunities, and supports a
diverse assemblage of wildlife and plant species.
The Fernow Experimental Forest was heavily logged between 1905 and 1911. The forest now
contains about 1900 ha of second- and third-growth Appalachian hardwood stands, which are
representative of average to better than average sites found on approximately 4 million ha of the forest
type in West Virginia and surrounding states. At the lowest elevations, the original forests consisted
mainly of hardwoods, with eastern hemlock (Tsuga canadensis [L.] Carr.) along stream bottoms and on
north slopes. Forests at the higher elevations were dominated by red spruce (Picea rubens Sarg.) and
hemlock. Small patches of pure spruce occurred on the tops of the mountains.
A-25
-------
Elevations in the Fernow range from 533 to 1112 m, with slopes of 10% to 60%. A rock layer
composed of fractured hard sandstone and shale underlies most of the Fernow. A majority of the soils are
of the Calvin and Dekalb series, which originated from these rocky materials (loamy-skeletal mixed
mesic Typic Dystrochrepts). On the southeastern part of the forest, Greenbrier limestone outcrops to
produce a midslope zone of limestone soil of the Belmont series (fine-loamy mixed mesic Typic
Hapludalfs). Almost all Fernow soils, including the sandstone, shale, and limestone soils, are well-
drained, medium textured loams and silt loams. Average soil depth is about 1 m, and average soil pH is
about 4.5.
A rainy, cool climate is typical on the Experimental Forest. Precipitation, which averages about 145
cm per year, is evenly distributed throughout the year. Mean annual temperature is about 9 °C, and the
length of the growing season is approximately 145 days.
The forest types and conditions today reflect the site qualities and past history of the area. Oaks
(Quercus spp.) are most common and are found on all sites along with American beech (Fagus
grandifolia Ehrh.) and sweet birth (Betula lenta [L].). Excellent sites in coves and on north slopes support
primarily northern red oak (Quercus rubra L.), sugar maple (Acer saccharum Marsh.), yellow-poplar
(Liriodendron tulipifera L.), black cherry (Prunus serotina Ehrh.), white ash (Fraxinus americana L.),
basswood (Tilia americana L.), cucumbertree (Magnolia acuminata L.), and beech. Fair sites on south
and east slopes usually support oak stands composed of red oak, white oak (Quercus alba L.), chestnut
oak (Quercus prinus L.), and scarlet oak (Quercus coccinea Muenchh.). Other fair site species include red
maple (Acer rubrum L.), sweet birch, black gum (Nyssa sylvatica Marsh.), sassafras (Sassafras albidum
Nutt.), and sourwood (Oxydendrum arboreum [L.] DC.). Good sites commonly support a mixture of
excellent and fair site species. Black locust (Robiniapseudoacacia L.), sweet birch, and Fraser magnolia
(Magnolia fraseri Walt.) are consistent but generally minor components of the forest on all sites.
American chestnut was a major forest component until it was eliminated by the chestnut blight.
The Fernow Experimental Forest encompasses practically the entire Elk Lick Run drainage, which
is about 5.8 km long and 3.5 km across at the widest point. Elk Lick Run has seven major tributaries
including Big Spring, which drains a headwater limestone formation. Headwater areas on two of these
tributaries have been gauged to show how forest management influences streamflow.
Research on the Fernow Experimental Forest by the Timber and Watershed Project scientists is
done in cooperation with the Monongahela National Forest, West Virginia University, Marshall
University, Pennsylvania State University, Virginia Tech, and the West Virginia Division of Natural
Resources.
Scientific studies on the Fernow have followed two lines of research, with considerable overlap.
Silvicultural research, focused mostly on mixed hardwood stands, addresses questions relating to
regenerating, growing, tending, and harvesting trees and stands. Watershed research has addressed some
of the more basic questions about water use by forests and forest hydrology, as well as critical issues
affecting roads, best management practices, and forest management effects on water and soil resources.
The Fernow also has been in the forefront of research on acidic deposition and N saturation. A whole-
watershed acidification study has been conducted since 1989. Recently, research on threatened and
endangered species has assumed a more prominent role, due to the presence of Indiana bat and running
buffalo clover on the Fernow.
A.2.6.4. National Ecological Observatory Network
The National Ecological Observatory Network (NEON) is a continental-scale research platform
that is primarily focused on discovering and understanding the impacts of climate change, land-use
change, and invasive species on ecology. It will also generate data that will be useful for assessing effects
of NOx and SOx deposition on ecosystems. NEON has not yet been implemented; it is described here
because it represents an ambitious monitoring program that is expected to be very useful in the near
future. NEON will gather long-term data on ecological responses of the biosphere to changes in land use
A-26
-------
and climate, and on feedbacks with the geosphere, hydrosphere, and atmosphere. NEON is proposed as a
national observatory, consisting of distributed sensor networks and experiments, linked by advanced
cyber infrastructure to record and archive ecological data for at least 30 years. Using standardized
protocols and an open data policy, NEON is intended to gather essential data for developing scientific
understanding and theory required to manage the nation's ecological challenges. The program description
is found at www.neoninc .org/.
A.3. Modeling
A.3.1. Principal Ecosystem Models Used in the U.S.
It is particularly difficult to study endpoints at the larger levels of biological organization (e.g., at
the population, community, biogeochemical, and ecosystem-level) with monitoring studies. Geographic
areas are larger, and timeframes are longer, rendering it difficult to obtain data in sufficient quantity to
detect impacts unless they are exceptionally severe. Therefore, the most common approach to study
endpoints at these scales is to develop and apply a model. Models may be calibrated using data from
monitoring, survey, or laboratory or field experiments and are useful tools in predicting larger-scale,
longer-term impacts. However, verifying the predictions and assessing the overall validity of the model
can be challenging.
Some of the most frequently used ecosystem models designed to quantify effects of atmospheric N
and S deposition are discussed below. It is important to note that the ecosystem models are parameterized
for specific areas and may not be readily applicable to other locations without significant re-
parameterization.
There are four principal models that are currently being used in the U.S. to assess the effects of S
and N deposition on terrestrial and freshwater aquatic ecosystems: MAGIC, NuCM, PnET/BGC, and
DayCent-Chem. Two models, SPARROW and WATERSN, are commonly used to evaluate N loading to
large river systems and to estuaries. These six models are briefly reviewed in the following sections. Each
review begins with a summary of the provenance and conceptual basis of the model and contains
references to some of the published applications. This is followed by a more detailed description of the
processes included in the model, the inputs required, and the output variables simulated by the model.
The ranges of process complexity, temporal resolution and spatial discretization represented in
these models are considerable. These ranges make comparative summaries of inputs, outputs, and
processes across the models problematic. The models are all currently in use because they are, in a sense,
complementary to each other, with each providing an approach or satisfying requirements unique to their
own structure and intended applications. As a result, there is no good way to develop satisfying
comparative equivalences among the components of the various structures. It is also beyond the scope of
this document to present the level of detail necessary to run any of the models. The descriptions below
must of necessity be brief. References to appropriate texts designed to provide more detail are given for
each model.
Following the discussion of the four models most frequently used in the U.S., there are brief
descriptions of the most important models of S and N deposition effects that are being used in Europe and
elsewhere.
A.3.1.1. MAGIC
The MAGIC model (Cosby et al., 1985a; 1985b; 1985c) is a mathematical model of soil and
surface water acidification in response to atmospheric deposition based on process-level information
A-27
-------
about acidification. MAGIC has been applied extensively in North America and Europe to both individual
sites and regional networks of sites, and has also been used in Asia, Africa and South America. The utility
of MAGIC for simulating a variety of water and soil acidification responses at the laboratory, plot,
hillslope, and catchment scales has been tested using long-term monitoring and experimental
manipulation data.
Figure A-6. Conceptual structure of the MAGIC model showing major pools and fluxes included in
MAGIC has been widely used in policy and assessment activities in the U.S. and in several
countries in Europe (e.g., Beier et al., 1995; Clair, 2004; Cosby, 1985b; Cosby et aL 1990, 1995, 1996;
Ferrier et aL 2001; Hornberger et al., 1989; Jenkins et al., 1990; Moldan et al, 1998; Sullivan et aL
1998, 2006b; Whitehead et aL 1988, 1997; Wright et al., 1994, 1998).
MAGIC Model Structure
MAGIC is a lumped-parameter model of intermediate complexity, developed to predict the long-
term effects of acidic deposition on surface water chemistry (see Figure A-6). The model simulates soil
solution chemistry and surface water chemistry to predict the monthly and annual average concentrations
of the major ions in these waters. MAGIC consists of: a section in which the concentrations of major ions
are assumed to be governed by simultaneous reactions involving S042 adsorption, cation exchange,
dissolution-precipitation- speciation of aluminum, and dissolution-speciation of inorganic carbon; and a
mass balance section in which the flux of major ions to and from the soil is assumed to be controlled by
atmospheric inputs, chemical weathering, net uptake and loss in biomass and losses to runoff. At the heart
of MAGIC is the size of the pool of exchangeable base cations in the soil. As the fluxes to and from this
pool change over time owing to changes in atmosphenc deposition, the chemical equilibria between soil
and soil solution shift to give changes in surface water chemistry. The degree and rate of change of
surface water acidity thus depend both on flux factors and the inherent characteristics of the affected soils.
Cation exchange is modeled using equilibrium (Gaines-Thomas) equations with selectivity
coefficients for each base cation and aluminum. S042 adsorption is represented by a Langmuir isotherm.
Aluminum dissolution and precipitation are assumed to be controlled by equilibrium with a solid phase of
Major Pools and Fluxes
Atmospheric
Deposition
Stream or Lake
Runoff
Weathering
Dissolved
Ions
simulation of effects of S and N deposition.
A-28
-------
aluminum trihydroxide. Aluminum speciation is calculated by considering hydrolysis reactions as well as
complexation with S042 , fluoride and dissolved organic compounds. Effects of carbon dioxide on pH and
on the speciation of inorganic carbon are computed from equilibrium equations. Organic acids are
represented in the model as tri-protic analogues. Weathering rates are assumed to be constant. Two
alternate mechanisms are offered for simulation of nitrate and ammonium in soils and water: either first
order equations representing net uptake and retention; or a set of equations and compartments describing
process-based N dynamics controlled by C and N pools and fluxes in the compartments.
Atmospheric deposition fluxes for the base cations and strong acid anions are required as inputs to
the model. These inputs are generally assumed to be uniform over the catchment. Atmospheric fluxes are
calculated from concentrations of the ions in precipitation and the rainfall volume into the catchment. The
atmospheric fluxes of the ions must be corrected for dry deposition of gas, particulates and aerosols and
for inputs in cloud/fog water. The volume discharge for the catchment must also be provided to the model.
In general, the model is implemented using average hydrologic conditions and meteorological conditions
in annual or seasonal simulations, i.e., mean annual or mean monthly deposition, precipitation and lake
discharge are used to drive the model. Values for soil and surface water temperature, partial pressure of
carbon dioxide and organic acid concentrations must also be provided at the appropriate temporal
resolution.
The MAGIC model can be implemented as a one- or two-soil representation of a catchment with or
without wetlands. Atmospheric deposition enters the soil compartment(s) and the equilibrium equations
are used to calculate soil water chemistry. The water is then routed to the stream compartment, and the
appropriate equilibrium equations are reapplied to calculate runoff chemistry. Input-output mass balance
equations are provided for base cations and strong acid anions, and charge balance is required for all ions
in each compartment (for complete details of the model see Cosby et al., 1985a, 1985b, 1985c; 2001).
For most applications, model outputs for 15 stream water variables are used. These variables
consist of the concentrations of 10 ions (H; Ca; Mg; Na; K; NH4; S042 : NO3; CI; and total inorganic Al),
the stream discharge (Q), stream pH, sum of base cation (SBC) concentrations (SBC = Ca + Mg + Na + K
+ NH4), sum of mineral acid anion (SAA) concentrations (SAA = CI + S042 + NO3) and the charge
balance acid neutralizing capacity (ANC = SBC - SAA). These variables are expressed in units of m/yr
(or m/mo) for Q, (imol/L for inorganic Al, and j^icq/L for all other variables. In addition, model output for
7 soil and soilwater variables are frequently used, the total base saturation and individual cation
saturations for Ca, Mg, Na, and K, the soilwater pH and the Ca/Al ratio in soil water.
The aggregated nature of the model requires that it be calibrated to observed data from a system
before it can be used to examine potential system response. Calibrations are based on volume weighted
mean annual or seasonal fluxes for a given period of observation. The length of the period of observation
used for calibration is not arbitrary. Model output will be more reliable if the annual flux estimates used in
calibration are based on a number of years rather than just one year. There is a lot of year-to-year
variability in atmospheric deposition and catchment runoff. Averaging over a number of years reduces the
likelihood that an "outlier" year (very dry, etc.) is used to specify the primary data on which model
forecasts are based. On the other hand, averaging over too long a period may remove important trends in
the data that need to be simulated by the model.
The calibration procedure requires that stream water quality, soil chemical and physical
characteristics, and atmospheric deposition data be available for each catchment. The water quality data
needed for calibration are the concentrations of the individual base cations (Ca, Mg, Na, and K) and acid
anions (CI, S042 , and N03) and the pH. The soil data used in the model include soil depth and bulk
density, soil pH, soil cation-exchange capacity, and exchangeable bases in the soil (Ca, Mg, Na, and K).
The atmospheric deposition inputs to the model must be estimates of total deposition, not just wet
deposition. In some instances, direct measurements of either atmospheric deposition or soil properties
may not be available for a given site with stream water data. In these cases, the required data can often be
estimated by: (a) assigning soil properties based on some landscape classification of the catchment; and
A-29
-------
(b) assigning deposition using model extrapolations from some national or regional atmospheric
deposition monitoring network.
Soil data for model calibration are usually derived as aerially averaged values of soil parameters
within a catchment. If soils data for a given location are vertically stratified, the soils data for the
individual soil horizons at that sampling site can be aggregated based on horizon, depth, and bulk density
to obtain single vertically aggregated values for the site, or the stratified data can be used directly in the
model.
Calibration of the model (and estimation of historical changes at the modeled sites) requires a
temporal sequence of historical anthropogenic deposition. Current understanding of ecosystem responses
to acidic deposition suggests that future ecosystem responses can be strongly conditioned by historical
acid loadings. Thus, as part of the model calibration process, the model should be constrained by some
measure of historical deposition to the site. However, such long-term, continuous historical deposition
data may not exist. The usual approach is to use historical emissions data as a surrogate for deposition.
The emissions for each year in the historical period can be normalized to emissions in a reference year (a
year for which observed deposition data are available). Using this scaled sequence of emissions, historical
deposition can be estimated by multiplying the total deposition estimated for each site in reference year
by the emissions scale factor for any year in the past to obtain deposition for that year.
A.3.1.2. NuCM
The current NuCM model is based on the original Integrated Lake Watershed Acidification Study
(ILWAS) model of the 1980s (cf. Chen et al., 1984; Gherini et al., 1985; Goldstein et al., 1984). NuCM
was developed as an extension to the ILWAS model by investigators in the Integrated Forest Study see
and the model code was written by Tetra-Tech, Inc. (Liu et al., 1991). NuCM was developed to explore
potential effects of atmospheric deposition, fertilization and harvesting in forest ecosystems. Because
NuCM was designed primarily for simulating the effects of atmospheric deposition on nutrient cycling
processes, its construction emphasizes soil and soil solution chemistry (Liu et al., 1991). As a stand-level
model, NuCM incorporates all major nutrient cycling processes (uptake, translocation, leaching,
weathering, organic matter decay, and accumulation). Vegetation is divided into leaf, bole and root
compartments for under- and overstory vegetation. NuCM simulates the cycling of N, P, K, Ca, Mg, Na,
and S based on expected optimal growth rates (input by the user and reduced in the event of nutrient
limitation), user-defined litterfall, weathering, N and S mineralization rates, soil minerals composition,
initial litter, soil organic matter pools, and C/N ratios.
The model has been calibrated for different vegetation types, including a loblolly pine (Pinus taeda
L.) stand at Duke University (Johnson et al., 1995), a mixed deciduous stand at Walker Branch (Johnson
et al., 1993) and a red spruce (Picea rubens Sarg.) stand in the Great Smoky Mountains (Johnson et al.,
1996). The NuCM model was used as part of the Southern Appalachian Mountain Assessment (Sullivan
et al., 2002a).
NuCM Model Structure
In NuCM, the ecosystem is represented as a series of vegetation and soil components. The
overstory consists of one generic conifer and one generic deciduous species of specified biomass and
nutrient concentration (foliage, branch, bole, roots). For mixed species stands, average values for biomass
and nutrient concentration by component must be used. NuCM also includes an understory, which can be
divided into canopy, bole, and roots. Maximum potential vegetative growth in the model is defined by the
user and is constrained in the model by the availability of nutrients and moisture. The forest floor is
simulated from litterfall inputs and litter decay. Litterfall mass inputs are defined by the user, and litter
decay is represented as a four stage process where: coarse litter decays to fine litter; fine litter decays to
A-30
-------
humus and cations; humus decays to organic acids, NH4+, S042 . H . and C02; and organic acids decay to
NH4+, S042 , H , and C02. Each stage is represented as a first-order equation.
The soil includes multiple layers (up to 10), and each layer can have different physical and
chemical characteristics. The user defines bulk density, cation exchange capacity, exchangeable cations,
adsorbed phosphate and S042 . and four soil minerals and their composition. These inputs define the
initial soil exchangeable/adsorbed pools and total pools. Initial total soil N pools are simulated from
litterfall and decay, as described above, and user-defined C/N ratios. Vegetation, litter, and soil pools
change over a simulation in response to growth, litterfall and decomposition, and nutrient fluxes via
deposition, leaching and weathering.
The processes that govern interactions among these pools include translocation, uptake, foliar
exudation and leaching, organic matter decay, nitrification, anion adsorption, cation exchange and mineral
weathering. Translocation, defined as the removal of nutrients from foliage before litterfall, is user-
specified. Maximum uptake is calculated from biomass and nutrient concentrations; actual uptake is equal
to this maximum value when sufficient nutrients are available and reduced when nutrients become
limiting. Reduced uptake first allows reduced nutrient concentrations in plant tissues, then causes a
reduction in growth. Foliar exudation and leaching rates are simulated as proportional to foliar
concentrations using user-defined coefficients.
Mineral weathering reactions are described in the model using rate expressions with dependencies
on the mass of mineral present and solution-phase hydrogen-ion concentration taken to a fractional power.
Cation exchange is represented by the Gapon equation. The model simulates atri-protic organic acid with
a fixed charge density. Nitrification is represented in the form of a Michaelis-Menton rate expression.
Phosphate adsorption is represented by a linear isotherm, and S042 adsorption is represented by a
Langmuir adsorption isotherm.
Climate inputs to the NuCM model are through input meteorological files (typically 1 to 5 years
long), which are repeated to generate long-term simulations. The meteorological files contain daily values
for precipitation quantity, maximum and minimum air temperature, cloud cover, dewpoint, atmospheric
pressure, and wind speed. Monthly soil temperature data are also required.
Precipitation is routed through the canopy and soil layers and evapotranspiration, deep seepage,
and lateral flow are simulated. The movement of water through the system is simulated using the
continuity equation, Darcy's equation for permeable media flow, and Manning's equation for free surface
flow. Percolation occurs between layers as a function of layer permeability's and differences in moisture
content. Nutrient pools associated with soil solution, the ion exchange complex, minerals, and soil organic
matter are all tracked explicitly by NuCM.
Wet deposition is calculated from precipitation amounts and user-input air quality files which
define precipitation concentrations on a monthly basis. Dry deposition is calculated from air
concentrations in the air quality files combined with user-defined deposition velocities and simulated leaf
areas. Leaching is calculated from soilwater percolation and simulated soil solution concentrations using
the soil chemical and biological algorithms defined above for each soil horizon.
The only processes in the NuCM model that are explicitly temperature-dependent are evaporation,
occurrence of precipitation as rainfall versus snowfall, snowpack melting, litter decay, and nitrification.
Temperature affects processes such as cation exchange, mineral weathering, and uptake only indirectly.
Precipitation effects are manifested strictly through the hydrologic simulations; none of the nutrient
processes are dependent explicitly upon moisture.
A.3.1.3. PnET-BGC
PnET-BGC is an integrated dynamic biogeochemical model that simulates chemical transforma-
tions of vegetation, soil and drainage water. The PnET-BGC model was formulated by linking two
submodels (vegetation and biogeochemical) to allow for the simultaneous simulation of major element
cycles in forest and interconnected aquatic ecosystems. The vegetation submodel is based on PnET-CN
A-31
-------
(Aber, 1992; Aber and Driscoll, 1997; Aber et al., 1997), a simple generalized model of monthly carbon,
water, and N balances that provides estimates of net primary productivity, N uptake, and water balances.
The biogeochemical submodel BGC (Gbondo-Tugbawa et al., 2001), expands PnET to include vegetation
and organic matter interactions of other elements (Ca2+, Mg2+, K+, Na+, Si, S, P, Al3+, CI . and F ). abiotic
soil processes, solution speciation, and surface water process.
PnET-BGC was initially developed for and applied to the northern hardwood forest ecosystem. The
model has been tested using vegetation, soil and water chemistry data from the Hubbard Brook
Experiment Forest (HBEF) (Gbondo-Tugbawa et al., 2001). The model has subsequently been applied to
intensively studied watersheds in the Adirondack and Catskill regions of New York and applied regionally
to the Adirondacks (Chen and Driscoll, 2005b) and northern New England (Chen and Driscoll, 2005a,
2005c). PnET-BGC has also been used to evaluate the effects of current and future atmospheric deposition
scenarios (Gbondo-Tugbawa, 2002a; Sullivan, 2006b).
PnET-BGC Model Structure
PnET/BGC simulates major biogeochemical processes, such as forest canopy element
transformations, hydrology, soil organic matter dynamics, N cycling, geochemical weathering, and
chemical equilibrium reactions in solid and solution phases, and allows for simulations of land
disturbance (see Figure A-7) (Gbondo-Tugbawa et al., 2001). The model uses mass transfer relationships
to describe weathering, canopy interactions and surface water processes. Chemical equilibrium
relationships describe anion adsorption, cation exchange and soil solution and surface water speciation.
Soil solution equilibrium reactions are described using the tableau approach (Morel and Hering, 1993). A
more detailed description of the model can be found in Gbondo-Tugbawa et al. (2001).
A-32
-------
Carbon/Nutrient (C/Nut)
Snow
vViilci
t 'I t
18
Foliar Canopy
Wood
C/Nut
a Nut
* Bud
C/Nut
Wood
Soil Solution
(Speciation)
Dead
Wood
>*\
Soil Organic Matter
Cation Exchange Sites
Anion Adsorption Sites
Surface Water
Processes
Processes Depicted:
1. Gross Photosynthesis
2. Foliar Respiration
3. Transfer to Mobil C
4 Growth and Maintenance
5. Allocation to Buds
6 Allocation to Fine Roots
7 Allocation to Wood
8 Foliar Production
9 Wood Production
10. Soil Respiration
11. Weathering Supply
12. Precipitation
13. Interception
14 Snow-Rain Partition
15. Snowmelt
16. Shallow Flow
17. Water Uptake
18. Transpiration
19. Deposition (Wet + Dry)
20 Foliar Nutrient Uptake
21. Foliar Exudation
22. Throughfall & Stemflow
23 Wood Litter
24 Root Litter
25 Foliar Litter
26- Wood Decay
27. Mineralization/Immobilization
28 Nutrient Uptake
29 Cation Exchange Reactions
30- Anion Adsorption Reactions
31 Drainage
Source: Gbondo-Tugbawa et al. (2001).
Figure A-7. The PnET-BGC model illustrating the compartments and flow paths of carbon and nutrients
(C/Nut) within the model.
The model operates on a monthly time step and is applied at the stand to small-watershed scale.
The process of photosynthesis, growth and productivity, litter production and decay, mineralization of
organic matter, immobilization, and nitrification in PnET have been described in Aber and Federer (1992)
and Aber et al. (1997). The BGC submodel uses the Gaines-Thomas formulation (White and Zelazny,
1986) to describe cation exchange reactions within the soil. The exchangeable cations considered in the
model include Ca2+, Mg2+, Na+, if, AT . K+, and NH4+. A pH-dependent adsorption isotherm is used to
describe the S042 adsorption process. Hie speciation of monomel ic aluminum is calculated in the model,
including both organic and inorganic forms. Organic acids are described using atriprotic analogue
(Driscoll et al., 1994b), and the total amount of organic acids is estimated as a certain fraction (based on
the charge density) of DOC. The model simulates ANC in surface waters as an analogue to ANC
measured by Gran plot analysis, by considering the contributions of DIC, organic anions and Al
complexes (Driscoll et al., 1994b).
The PnET/BGC model requires inputs of climate, wet and dry deposition chemistry, and
weathering data. Climate inputs consist of minimum and maximum air temperature, solar radiation, and
precipitation. The model uses a constant dry-to-wet deposition ratio by default, but a variable ratio can
also be applied (Chen and Driscoll, 2005b). The model inputs utilize canopy enhancement factors to
depict the increased dry deposition observed in coniferous and mixed forest stands compared to hardwood
forests. Deposition and weathering fluxes for all major elements are required as model inputs. Weathering
rates are assumed to remain constant over time.
Calibration of PnET-BGC is based on empirical relationships and observations. The model uses
historical reconstructions of climate, atmospheric deposition, and land disturbance to construct hindcasts
of the response of forests to past acidic deposition. The model can also be used to predict the response of
A-33
-------
acid-sensitive forest ecosystems to future changes in acidic deposition, for example in response to
controls on atmospheric emissions. A detailed description of the model, including a detailed uncertainty
analysis of parameter values, is available in Gbondo-Tugbawa et al. (2001).
A.3.1.4. DayCent-Chem
DayCent-Chem links two widely accepted and tested models, one of daily biogeochemistry for
forest, grassland, cropland, and savanna systems, DayCent (Parton et al., 1998), and the other of soil and
water geochemical equilibrium, PHREEQC (Parkhurst and Appelo, 1999). The linked
DayCent/PHREEQC model was created to capture the biogeochemical responses to atmospheric
deposition and to explicitly consider those biogeochemical influences on soil and surface water chemistry.
The linked model expands on DayCent's ability to simulate N, P, S, and C ecosystem dynamics by
incorporating the reactions of many other chemical species in surface water.
Hartman et al. (2007) et al. used DayCent-Chem to investigate how wet and dry deposition affect
biological assimilation, soil organic matter composition, ANC and pH of surface waters, and also Al
mobilization, soil base cation depletion, and base cation flux. Model results were tested against a long-
term data set available from Andrews Creek in Loch Vale Watershed, Rocky Mountain National Park,
Colorado.
DayCent-Chem Model Structure
DayCent is the daily time-step version of CENTURY, a non-spatial, lumped parameter model that
simulates C, N, P, S, and water dynamics in the soil-plant system at a monthly timestep over time scales
of centuries and millennia (Parton et al., 1994). CENTURY can represent a grassland, crop, forest, or
savanna system with parameters that describe the site-specific plant community and soil properties.
DayCent, the daily timestep version of CENTURY, adds layered soil temperature, a trace gas submodel, a
more detailed soil hydrology submodel, and explicitly represents inorganic N as either N03 or NH4+ (Del
Grosso et al., 2001; Kelly et al., 2000; Parton et al., 1998). DayCent 5 is an object-oriented model written
in the C++ programming language that implements a layered soil structure and algorithms to manage soil
layers. The model is initialized with an organic soil depth and up to 10 soil layers, where each layer has a
specified thickness, texture, bulk density, field capacity, wilting point, and saturated hydraulic
conductivity.
PHREEQC is a model based on equilibrium chemistry of aqueous solutions interacting with
minerals, gases, exchangers, and sorption surfaces. The model is written in the C programming language
and has an extensible chemical database. Version 2.7 of PHREEQC is used in the linked DayCent-Chem
model to compute aqueous speciation, ion-exchange equilibria, fixed-pressure gas-phase equilibria,
dissolution and precipitation of mineral phases to achieve equilibrium, and irreversible aqueous mineral
phase reactions. The aqueous model uses ion-association and Debye Huckel expressions. Ion-exchange
reactions are modeled with the Gaines-Thomas convention and equilibrium constants are derived from
Appelo and Postma (1993).
The DayCent-Chem model inputs are climate drivers consisting of daily precipitation, and mini-
mum and maximum air temperatures. The model also requires daily atmospheric wet deposition concen-
trations for precipitation species Ca2+, CI . K+, Mg2+, Na+, NH4+, N03 . S042 . and H and daily dry depo-
sition amounts or dry/wet ratios for all precipitation species. Initial conditions for model simulations in-
clude: initial snowpack water content and chemical composition; initial soil solution concentrations; and
initial exchangeable cations in each soil layer. Potential annual denudation rates for each mineral phase
that could be dissolved in the soil, groundwater, or stream solutions must also be provided.
DayCent-Chem implements a geochemical submodel of layered pools and properties that provides
information exchange, such as of water fluxes and solute concentrations, between the coupled models,
and calculates daily geochemical outputs. The geochemical submodel defines soil layers and a
A-34
-------
groundwater pool that correspond to those in Day-Cent 5's original soil class. Surface water
concentrations are computed in a two-step process where solutes are first transported, and then
PHREEQC undertakes solution reactions. At each timestep, the model updates exchangeable base cation
pools and soil solutions in each soil layer, along with groundwater and stream solutions.
DayCent 5 output includes daily evapotranspiration; soilwater content; outflow; inorganic and
organic C, N, P, and S stream fluxes; C, N, P, and S contents in soil and plant pools; net primary
production (NPP); nutrient uptake; trace gas flux; and heterotrophic respiration. In addition to standard
DayCent 5 outputs, at each daily timestep the model writes the solution chemistry for soil layers,
groundwater, and stream.
A.3.1.5. SPARROW
SPAtially Referenced Regressions on Watersheds (SPARROW) is a hybrid statistical/deterministic
model used to estimate pollutant sources and contaminant transport in surface waters. SPARROW can be
used to estimate pollutant loading to downstream receiving waters for a number of water quality
constituents. The model as constructed for evaluating N export to estuaries will be presented here.
SPARROW was first described by Smith et al. (1997) as a water quality model designed to reduce
problems with interpreting watershed data as a result of sparse sampling, network bias, and basin hetero-
geneity. SPARROW combines regression techniques and process information regarding contaminant
transport and retention in watershed and riverine systems. Literature values for watershed retention rates
are used; in-stream retention of N is estimated by a first-order decay function (Smith et al., 1997).
Others have developed similar regression models relating in-stream water quality measurements to
watershed nutrient sources and basin attributes (Howarth et al., 1996; Jaworski et al., 1997; Mueller et al.,
1997). These simple correlative models assume that contaminate sources and sinks are homogenously
distributed and do not make a distinction between watershed and in-stream loss processes. SPARROW is
distinct from these methods by incorporating spatial representation of basin attributes in the model. Model
correlations between basin attributes and water quality measurements are strengthened by incorporating
these spatial references (Alexander et al., 2001; Smith et al., 1997). Spatially referenced basin attributes
include land use, point and non-point N sources, temperature, soil permeability, and stream density,
among others. Figure A-8 shows the mathematical form of the SPARROW model (Preston and Brakebill,
1999).
A-35
-------
L,-2 ^,S,je(-'^el'«#)
n - f /. J a i
where
Lj = load in reach i;
n,N = source index where N is the total number of considered sources;
J( i ) = the set of all reaches upstream and including reach i, except
those containing or upstream of monitoring stations
upstream of reach i;
fi„ " estimated source parameter;
Sn.J = contaminant mass from source n in drainage to reach_/;
a = estimated vector of land-to-water delivery parameters;
Zj = land-surface characteristics associated with drainage to reach j;
3 = estimated vector of instream-Ioss parameters; and
Tjj = channel transport characteristics.
Source: Preston and Brakebill (1999).
Figure A-8. Mathematical form of the SPARROW model.
Smith et al. (1997) provided an example of SPARROW model development for application to the
conterminous U.S. Their exploratory model included five N sources and eight land surface characteristics
as potential factors that deliver N from land to water. In-stream decay coefficients for three stream size
classes were also tested for significance (Table A-4).
The final model resulted in the inclusion of each of the five N sources and three (temperature, soil
permeability, and stream density) of the eight land to water delivery factors. Parameter selection was
primarily based on statistical significance. Further discussion regarding the exclusion of precipitation and
irrigated land, both of which were determined to be significant, can be found in Smith et al. (1997).
Parameter estimates are evaluated for robustness through the use of bootstrap analysis.
The bootstrap procedure involves randomly selecting, with replacement, Mmonitored loads and
associated predictor variables from among the observations in the data set (M is the number of monitored
reaches in the reach network). Where a sampled observation has an upstream monitored load as one of its
predictors, the monitored value is used, regardless of whether the upstream station appears in the
bootstrap sample. Coefficient values are estimated from the bootstrap sample. The bootstrap process is
repeated 200 times, resulting in 200 estimates of each coefficient. From these estimates, the mean
coefficient value (called the bootstrap estimate), minimum confidence interval, and probability that the
estimated coefficient has the wrong sign are determined (Smith et al., 1997).
Spatial referencing in the model occurs in two ways: land surface polygons are mapped in
conjunction with nonpoint contaminant sources and the land-water delivery variables (temperature, soil
permeability, stream density, etc.); and the stream reach network is mapped along with point sources,
channel transport characteristics, and measured transport rates. The positive impacts of this spatial
referencing can be quantified by eliminating the channel decay coefficients from the model and creating a
new model with only the contaminant sources and land-water delivery variables in the original model
(Smith et al., 1997). Removing this spatial reference provided by the reach network results in a model
with significantly higher mean squared error and lower predictive capacity (Table A-5).
A-36
-------
Table A-4. Parameter estimates, probability levels, and regression results of parametric and bootstrap
regressions of total nitrogen at 414 national stream quality accounting network stations on
basin attributes, for the Chesapeake Bay total nitrogen SPARROW model.
Model
Parameters
Coefficient
Units3
Exploratory Model
Final Model
Parametric
Coefficient
P
Parametric
Coefficient
Parametric
P
Bootstrap
Coefficient
Lower Upper
90% Clb 90% Clb
Bootstrap
P
NITROGEN SOURCE B
Point sources
Dimensionless
0.4112
0.0004
0.3464
0.0049
0.4401
0.0864
0.8173
<0.005
Fertilizer
application
Dimensionless
2.798
0.0154
1.278
0.0022
1.433
0.6149
2.373
<0.005
Livestock waste
production
Dimensionless
1.340
0.1553
0.9723
0.0629
1.058
0.0859
1.919
0.005
Atmospheric
deposition
Dimensionless
3.334
0.2513
6.465
0.0033
6.555
3.270
9.323
<0.005
Nonagricultural
land
kg N/ha/yr
38.49
0.0154
14.67
0.0005
16.65
7.130
29.89
<0.005
LAND TO WATER DELIVERY A
Temperature
oF-1
0.0228
0.0001
0.0196
0.0001
0.0198
0.0117
0.0259
<0.005
Slopec
%
0.2034
0.2187
Soil permeability
h/cm
0.0295
0.0022
0.0442
0.0001
0.0447
0.0334
0.0572
<0.005
Stream density0
km-1
0.0205
0.0124
0.0215
0.0095
0.0243
-0.0003
0.0450
0.025
Wetlandd
Dimensionless
0.7177
0.2962
Irrigated lande
Dimensionless
1.101
0.0001
Precipitation'
cm
38.52
0.0057
Irrigated water
useg
cm
0.0772
0.3117
A-37
-------
Model Coefficient
Parameters Units3
Exploratory Model
Final Model
Parametric
Coefficient
P
Parametric
Coefficient
Parametric
P
Bootstrap
Coefficient
Lower Upper
90% Clb 90% Clb
Bootstrap
P
IN-STREAM DECAYH A
6 1 (Q <28.3 D-1
m3/s)
0.2917
0.0001
0.3758
0.0001
0.3842
0.2981
0.4768
<0.005
6 2 (28.3 m3/s 283 m3/s) D-1
0.0352
0.1794
0.0406
0.1321
0.0408
0.0176
0.0685
0.015
R2
0.8822
0.8743
Mean square error
0.4310
0.4543
Number of
observations
414
414
dependent variable (nitrogen transport) in kilograms per year. bMinimum bootstrap confidence intervals (CI). cVariable enters the model in reciprocal form. dRatio of wetland area to
total land area. eRatio of irrigated land area to total cropland area. fProduct of reciprocal precipitation and one minus the ratio of irrigated land area to total cropland area. aRatio of
irrigated land area to irrigated water use. hDecay coefficients fit separately for stream reaches with mean streamflow (Q) corresponding to indicated intervals. The streamflow interval
breakpoints of 28.3 and 282 m3/s correspond to 1000 and 10,000 ft3/s, respectively. Source: Smith et al. (1997).
SPARROW has also been applied to spatially identify N sources at the scale of the Chesapeake Bay
watershed (Preston and Brakebill, 1999). A similar set of N sources, land-to-water delivery parameters,
and in-stream loss rates to those used in Smith et al. (1997) were considered for this model (Preston and
Brakebill, 1999). Only estimates for parameters that were used in the final model are given in Table A-6.
The final model included five N sources, one land-to-water deliver parameter (soil permeability),
and four in-stream loss rates (including reservoir retention). Comparisons between predicted and observed
N loading provided an r2 value of 0.961 (Preston and Brakebill, 1999). Because the data that are used in
SPARROW are spatially referenced, model results can be mapped.
Table A-5. Effect of spatial referencing on measures of regression model performance for predicting
total N flux using the sparrow model.
Model Components
Mean Square Error
R2
Includes full spatial referencing (SPARROW)3
0.4544
0.8743
Excludes in-stream decay and reservoir retention
0.9659b
0.7307
A-38
-------
Table A-6. Parameter estimates, probability levels, and regression results for the Chesapeake Bay total N
SPARROW model
Explanatory Variables
Parameter Estimates
Probability Level
Nitrogen sources
P
Point Sources
1.496
<0.005
Urban area (acres)
7.008
0.010
Fertilizer application (Ib/yr)
0.2790
<0.005
Livestock waste production (Ib/yr)
0.3361
<0.005
Atmospheric deposition (Ib/yr)
1.024
<0.005
Land-to-water delivery
a
Temperature (EF)
Precipitation (in)
Avg slope (%)
Soil permeability (in/h)
0.0754
0.095
Stream density (l/mi)
Wetland (%)
Instream loss (days)3
6
T1 (Q #200 ft3/s)
0.7595
<0.005
T2 (200 ft3/s < Q #1,000 ft3/s)
0.3021
<0.005
T3 (Q> 1,000 ft3/s)
0.0669
<0.005
Tu (reservoir retention)
0.4145
<0.005
R-squared
0.961
Mean square error
0.1669
Number of observations
79
aT, travel time
Q, stream discharge
Ib/yr, pounds peryr
°F, degrees Fahrenheit
in/h, inches per hour
ft3/s, cubic feet per second
value not statistically significant
A.3.1.6. WATERSN
The Watershed Assessment Tool for Evaluating Reduction Strategies for Nitrogen (WATERSN)
model is a steady-state numerical N budgeting model that estimates the amount of N exported to rivers
and estuaries from forest, agricultural and urban land uses. The model is intended to provide an
understanding of the relative contribution of N export from these land uses to estuaries, and to evaluate N
A-39
-------
export reduction strategies that are specific to each land use type (Driscoll et al., 2007a). Figure A-9
shows a conceptual diagram of the N budgeting system used in WATERSN.
A detailed description of the original model calculations is provided in Castro et al. (2001).
Subsequent model applications (Castro et al., 2003; Castro and Driscoll, 2002; Driscoll et al., 2003a;
Whitall et al., 2004) have developed modifications to the approach originally described in Castro et al.
(2001).
WATERSN uses calculations described in Jordan and Weller (Jordan and Weller, 1996) to estimate
N inputs to the watershed/estuary system. Estimated anthropogenic sources of N inputs to the modeled
watershed/estuary system include: crop and lawn fertilizer application; biotic N fixation by leguminous
crops and pastures; atmospheric deposition of wet and dry inorganic N (NH4+, N03 ) net N import of food
for human consumption; and net N import of feed for livestock (Castro and Driscoll, 2002).
Agricultural Areas
N available for water-borne export to estuaries from agricultural lands is determined as the
difference between N inputs and outputs (Castro et al., 2001). Modeled N inputs to agricultural lands
consist of wet and dry atmospheric NH4+ and N03 deposition, N fertilization, biotic N fixation, and
livestock waste (Castro and Driscoll, 2002). Wet and dry deposition are derived from NADP and
CASTNet data. Average annual wet deposition rates of NH4+ and N03 are taken from NADP sites in or
near the study watersheds. Dry deposition ofNH4+ and N03 is calculated as an average of all CASTNet
sites nearest to the study watersheds. WATERSN assumes that dry deposition of both NH4+ and N03 to
the estuary surface is 25% less than dry deposition to the watershed (Castro and Driscoll, 2002). Meyers
et al. (2001) described the uncertainty of estimates of wet and dry deposition and considered it to be no
less than a factor of 2. Estimates of N fertilization are taken from agricultural census data. WATERSN
assumes that all fertilizer sold in a county is applied in that county. This is considered to be the most
certain N input to the model (±25%) (Castro and Driscoll, 2002). WATERSN estimates both non-
symbiotic and symbiotic N fixation for crops, pastures, hay fields and upland forests. Non-symbiotic rates
were taken from literature values for crops, orchards, upland forests, and non-wooded pastures
(Hendrickson, 1990; Stevenson, 1982; Woodmansee, 1978). Symbiotic rates ofN fixation are based on
type of legume, crop N harvest, N in unharvested portions of crops, soil N availability, and fertilization
rate. These estimates are less certain than for N fertilization, but are noted as being a relatively minor N
source in most of the study watersheds. Livestock waste was calculated as the difference between
livestock consumption of N in feed and production of N in meat, milk, and eggs for human consumption
(Jordan and Weller, 1996).
A-40
-------
N Imputs
Net Food Import
Atmospheric
Atmospheric
Atmospheric
NH/ and N03
NH4' and N03
NH/ and N03
Deposition
Deposition
Deposition
N Fixation
N Fixation
N Fertilization
Livestock Waste
Human Population
Point
Septic
Sources
Systems
1
Retention/Loss*
SflHL in Watersheds^!
Agriculture
Above Fall Lir»e Runoff
Below Fall Line Runoff
Upland Forest
£
-------
wastewater treatment facilities is used to estimate total N discharges from wastewater treatment plants
that do not have total N monitoring data available. Septic system output is determined by multiplying
watershed specific human per capita N excretion rates by the human population of the watershed.
WASTERSN assumes that 75% of this N is exported to the estuary (Castro and Driscoll, 2002). The soil
water assessment tool (SWAT) is used to estimate non-point source non-atmospheric total N runoff from
pervious and impervious urban lands. SWAT is a distributed parameter, continuous time model applicable
at the watershed scale. Required inputs to SWAT include climatic variables, soil properties, elevation,
vegetation information, and land use. SWAT is designed to predict land use and land management impacts
on water, sediment, and agricultural yields in large watersheds (Castro and Driscoll, 2002). The model
assumes that 75% of atmospheric N inputs to urban areas is exported to the estuary (Fisher and
Oppenheimer, 1991). Alternatively, this N export term can be modified.
Upland Forests
N inputs to forests are assumed to be in the form of atmospheric deposition and non-symbiotic N
fixation. Outputs from forests are estimated with a non-linear regression relationship between wet
deposition of inorganic N and stream water export of dissolved inorganic nitrogen (DIN) developed using
results from a multitude of forest watershed studies. Exported dissolved organic nitrogen (DON) was
assumed to be equal to 50% of the inorganic N export (Castro and Driscoll, 2002).
Watershed and In-Stream N Retention
Model validation efforts using measured N fluxes from the USGS National Stream Quality
Accounting Network (NASQAN) have shown that WATERSN tends to overestimate N export from
watersheds to estuaries (Castro and Driscoll, 2002). These differences are not unexpected since
WATERSN does not account for watershed and in-stream N sinks. Attempts have been made to improve
flux estimates by accounting for watershed and in-stream N retention (Castro and Driscoll, 2002; Castro
et al., 2001; 2003). A summary of the N retention rates applied to WATERSN in these studies is given in
Table A-7.
Castro and Driscoll (2002) assumed that 30% of the total N that entered rivers above the fall line
was lost during transport to the fall line and that inputs that enter the river below the fall line were not
attenuated because of the relatively short travel times to the estuary (See Table A-7). This 30% in-stream
N retention value represents the median retention value obtained in previous studies of northeastern U.S.
rivers (Castro et al., 2001) and falls within the range of retention values estimated by Howarth et al.
(1996) and Alexander et al. (2000). Castro and Driscoll (2002) also incorporated watershed N retention
fractions specific to individual land uses. They assumed that 60% of the excess N from agricultural land
and septic systems was lost (retained within the watershed) due to watershed processes. Support for this
value of N retention was given by several reports of riparian N removal rates from agricultural land,
ranging from about 50 to 90% (Jacobs and Gilliam, 1985; Jordan et al., 1993; Lowrance et al., 1983;
Peteijohn and Correll, 1984). After incorporating these assumptions, predicted fluxes closely matched
(r2 = 0.909) measured fluxes.
Table A-7. Summary of N retention rates used in recent WATERSN studies.
Study
Retention Type
% N Retention
Castro etal. (2001)
In Stream*
30%
Agriculture
50%
Castro and Driscoll (2002)
In Stream*
30%
A-42
-------
Agriculture
60%
Castro et al. (2003)
Agriculture
40%
Septic System
40%
In Stream
Adjusted until predicted N flux matched observed fluxes
*ln-stream N retention was only applied to lengths of river located above the "fall line." Below fall line N inputs to streams were assumed to not be attenuated due to the relatively short
travel time to the estuary. The "fall line" is defined as the boundary between the Piedmont and Coastal Plain physiographic provinces in the eastern U.S.
Source: Castro and Driscoll (2002)
Driscoll et al. (2003a) applied WATERSN to investigate anthropogenic N loading to estuaries in the
northeastern U.S. The objectives of the study were to apply WATERSN to quantify the inputs of Nr to the
region (Figure A-10), discuss the ecological effects of regional elevated anthropogenic Nr inputs, and
evaluate management options aimed at mitigating the effects of these elevated anthropogenic N inputs.
Modeled N reduction scenarios included reductions atmospheric N emissions, increased N removal
efficiencies of wastewater treatment plants, offshore pumping of wastewater, reductions in agricultural N
runoff to surface waters, and an integrated management scenario consisting of a combination of N
reductions from multiple sources. Other studies have applied WATERSN to address similar issues related
to N loading to estuaries in other regions of the U.S. (Castro et al., 2003; Whitall et al., 2004; Whitall and
Bricker, 2006).
(A
0) —-
£ ^
3
tf> CL
0) -o
o o>
V) w
3 a>
if
¦SJ B
O) 0)
O Z
£
30-
20-
M Sewage
~ Agriculture
¦I Urban
I 1 Forest
E"1 Atmospheric Deposition
^ ^
,v«;
art'
S&'
o^'
Source: Driscoll et al. (2003a).
Figure A-10. WATERSN model estimates of anthropogenic N inputs to the estuaries of the northeastern
U.S., in kilograms per hectare per year.
A-43
-------
A.3.2. Additional Effects Models Used Widely in Europe
The models of the effects of S and N deposition described below have been used primarily in
Europe. These descriptions are derived in part from the UNECE Convention of Long-Range
Transboundary Air Pollution Modelling and Mapping manual (Posch et al., 2003).
A.3.2.1. The Very Simple Dynamic Model
The Very Simple Dynamic (VSD) soil acidification model is frequently used in Europe to simulate
acidification effects in soils when observed data are sparse. It only includes weathering, cation exchange,
N immobilization processes, and a mass balance for cations, sulfur and N. It resembles the model
presented by Reuss (1980) which, however, did not consider N processes. In the VSD model, the various
ecosystem processes have been limited to a few key processes. Processes that are not taken into account
include canopy interactions; nutrient cycling processes; N fixation and NH4 adsorption; S042
transformations (adsorption, uptake, immobilization, and reduction); formation and protonation of organic
anions; and complexation of Al.
The VSD model consists of a set of mass balance equations, describing the soil input-output
relationships, and a set of equations describing the rate-limited and equilibrium soil processes. The soil
solution chemistry in VSD depends solely on the net element input from the atmosphere (deposition
minus net uptake minus net immobilization) and the geochemical interaction in the soil (C02 equilibria,
weathering of carbonates and silicates, and cation exchange). Soil interactions are described by simple
rate-limited (zero-order) reactions (e.g., uptake and silicate weathering) or by equilibrium reactions (e.g.,
cation exchange). It models the exchange of Al, H, and Ca + Mg + K with Gaines-Thomas or Gapon
equations.
Solute transport in VSD is described by assuming complete mixing of the element input within one
homogeneous soil compartment with a constant density and a fixed depth. Since VSD is a single layer soil
model neglecting vertical heterogeneity, it predicts the concentration of the soil water leaving this layer
(mostly the rootzone). The annual water flux percolating from this layer is taken as being equal to the
annual precipitation excess. The time step of the model is one year, and therefore seasonal variations are
not considered. A detailed description of the VSD model can be found in Posch and Reinds (2003).
A.3.2.2. SMART
The Simulation Model for Acidification's Regional Trends (SMART) model is similar to the VSD
model, but somewhat extended. It is described in De Vries et al. (1989) and Posch et al. (1993). As with
the VSD model, the SMART model consists of a set of mass balance equations, describing soil input-
output relationships, and a set of equations describing the rate-limited and equilibrium soil processes. It
includes most of the assumptions and simplifications given for the VSD model; and justifications for
them can be found in De Vries et al. (1989).
SMART models the exchange of Al, H, and divalent base cations using Gaines Thomas equations.
Additionally, S042 adsorption is modeled using a Langmuir equation (as in MAGIC) and organic acids
can be described as mono-, di-, or tri-protic. Furthermore, it does include a balance for carbonate and Al,
thus allowing application to a range of site conditions, from calcareous soils to completely acidified soils
that do not have an Al buffer left. Recently, a description of the complexation of aluminum with organic
acids has been included. The SMART model has been developed with regional applications in mind, and
an early example of an application to Europe can be found in De Vries et al. (1994).
A-44
-------
A.3.2.3. SAFE
The Soil Acidification in Forest Ecosystems (SAFE) model has been developed at the University of
Lund (WarfVinge et al., 1993) and a recent description of the model can be found in Alveteg and Sverdrup
(2002). The main differences between the SMART and MAGIC models are: (a) weathering of base
cations is not a model input, but it is modeled with the PROFILE (sub-)model, using soil mineralogy as
input (Warfvinge and Sverdrup, 1992) SAFE is oriented to soil profiles in which water is assumed to
move vertically through several soil layers (usually 4); and (c) Cation exchange between Al, H, and
(divalent) base cations is modeled with Gapon exchange reactions, and the exchange between soil matrix
and the soil solution is diffusion-limited.
The standard version of SAFE does not include S042 adsorption although a version, in which
S042 adsorption is dependent on S042 concentration and pH has recently been developed (Martinson
et al., 2003). The SAFE model has been applied to many sites and more recently also regional
applications have been carried out for Sweden (Alveteg and Sverdrup, 2002) and Switzerland (Kurz et al.,
1998).
A.3.3. Other Models
There are scores of models that can be useful in the context of developing a better understanding of
the ecological effects of atmospheric S and N deposition. The preceding sections have summarized a
relatively small number of models that are most commonly used for this purpose in the U.S. and Europe,
in particular those that contribute to substantive conclusions presented in the ISA. There are many other
models that are not covered in the discussion presented in this Annex. Several are highlighted in Table A-
8.
Table A-8. Some examples of models that could contribute to development of a better understanding of
the ecological efforts of atmospheric S and N deposition, but that are not explicitly addressed
in this annex.
Model Name Type1 Support2 Reference3 Notes
QUAL2K A S 1 QUAL2K is one dimensional river and stream water quality
model. QUAL2K assumes: that he channel is well-mixed
vertically and laterally; steady state hydraulics; diurnal water-
quality kinetics. QUAL2K addresses point and non-point loads,
BOD/DO, non-living particulate organic matter (detritus);
denitrification; sediment-water interactions; bottom algae; pH
(both alkalinity and total inorganic carbon).
WASP is a dynamic compartment-modeling program for aquatic
systems, including both the water column and the underlying
benthos. WASP allows the user to investigate 1, 2, and 3
dimensional systems, and a variety of pollutant types. The time
varying processes of advection, dispersion, point and diffuse
mass loading and boundary exchange are represented in the
model. WASP also can be linked with hydrodynamic and
sediment transport models that can provide flows, depths
velocities, temperature, salinity and sediment fluxes.
WASP7 Water Quality Analysis AS 2
Simulation Program
A-45
-------
Model Name Type1 Support2 Reference3
Notes
CE-QUAL- Water quality models
RIV1; CE- (river, reservoir, and
QUAL-R1; estuary/ coastal)
CE-QUAL-W2; supported by USACE
CE-QUAL-ICM
CE-QUAL-R1 is a one-dimensional, vertical reservoir model and
CE-QUAL-W2 is a two-dimensional (vertical and longitudinal),
laterally averaged, hydrodynamic and water quality model.
These two models are widely used by the Corps of Engineers,
other federal and state agencies, the private sector, and
agencies in other countries. CE-QUAL-RIV1 was developed for
highly unsteady flow conditions, such as storm water flows and
streams below peaking hydropower dams. CE-QUAL-ICM run in
a 2D mode. This approach has been used for large, shallow
waterways, harbors, and embayments.
RCA
Row Column AESOP
RCA evaluates the fate and transport of conventional and toxic
pollutants in surface waterbodies in one, two, or three
dimensions. RCA has been linked to various hydrodynamic
models. Subroutines have been developed to model coliforms,
pathogens, BOD/DO, simple and advanced eutrophication,
wetland systems, and toxic contaminants. A sediment nutrient
flux subroutine permits the coupling of the water column and
sediment bed.
WARMS Waterfowl Acidification A
Response Modeling
System
McNicol et al. WARMS includes an acidification model linked to fish and
(1995), waterfowl models. WARMS uses pH, area, dissolved organic
McNicol carbon, total P, and presence of fish to estimate preacidification,
(2002) present and eventual steady-state values for pH, fish presence
and waterfowl breeding parameters under proposed SO2
emission scenarios.
GT-MEL Georgia Tech I
hydrologic model and
the Multiple Element
Limitation model
GT-MEL is a spatially distributed, process-based ecohydrology
model that links a land surface hydrology model with a terrestrial
biogeochemistry model. GT-MEL differs from other available
ecohydrology models in its simplicity, flexibility, and theoretical
foundation. The coupled GT-MEL simulates the cycling and
transport of water and nutrients (C, N and P) within hillslopes
and watersheds. The model runs on a daily time step and can be
applied to user-defined landscape units that may vary in shape
and size (m2 to km2). Thus, GT-MEL can provide detailed spatial
and temporal information on nutrient acquisition and turnover in
plants and soils, and terrestrial flow pathways and discharge of
water and nutrients to surface waters. The same set of model
equations applies to any terrestrial ecosystem - agricultural
crops, forests, grasslands, wetlands, tundra, etc. GT-MEL
simulates the effects of multiple interacting stressors, including
changes in land use, land cover, climate, and atmospheric CO2
and N deposition.
ILWAS Integrated Lake- I
Watershed Acidification
Study
Gherini et al. ILWAS was developed to predict changes in surface water
(1985) acidity given changes in the acidity of precipitation and dry
deposition. The model routes precipitation through the forest
canopy, soil horizons, streams and lakes using mass balance
concepts and equations which relate flow to hydraulic gradients.
The physical-chemical processes which change the acid-base
characteristics of the water are simulated by rate (kinetic) and
equilibrium expressions and include mass transfers between
gas, liquid and solid phases.
A-46
-------
Model Name Type1 Support2 Reference3
Notes
THMB/IBIS I N 6, 7
THMB is a mechanistic simulation model of large river systems
that has been used recently in combination with a dynamic
terrestrial ecosystem model IBIS to quantify nitrate flux in the
Mississippi River Basin. The coupled models simulate time-
varying flow and storage of water and N in rivers, wetlands, and
reservoirs, based on major source inputs, subsurface drainage
and N leaching, topography, and precipitation and evaporation.
Evaluations of the model in the Mississippi basin indicated that
the model accurately simulated inter-annual variability in the
water and N budget from 1960 to 1994.
BIOME-BGC Biome-BGC is a multi- T N 8
biome generalization of
FOREST-BGC
Biome-BGC is a computer program that estimates fluxes and
storage of energy, water, carbon, and N for the vegetation and
soil components of terrestrial ecosystems. The primary model
purpose is to study global and regional interactions between
climate, disturbance, and biogeochemical cycles.
DNDC Denitrification- T N 9
decomposition model
DNDC was initially developed to quantifying nitrous oxide (N20)
emissions from agricultural soils in the U.S. The capability of the
model to simulate soil biogeochemistry also allows DNDC to
model emissions from other ecosystems through linkages with
vegetation models; the model can be applied from field site to
regional scales. The core of DNDC is a soil biogeochemistry
model.
EPIC Agricultural dynamic T N 11
simulation model
EPIC is a widely used dynamic simulation model that describes
the influence of agricultural management on crop productivity
and erosion. The model has been used in studies of climate
change, agricultural management and policy, and water-quality.
EPIC simulates N cycling processes in soils-including
mineralization, nitrification, immobilization, NH3 volatilization and
denitrification, runoff and subsurface leaching based on physical
principles and parameter values derived from extensive model
testing and specific field validation.
GLEAMS Groundwater Loading T N 12
Effects of Agricultural
Management Systems
GLEAMS was developed from both EPIC and CREAMS and
employs a more explicit description of soil water content. In
GLEAMS, the concentration of nitrate-N removed via
denitrification is a function of the factors describing the soil water
content, the soil temperature, and the organic C content. Under
this formulation, denitrification only occurs if the soil water
content is greater than a parameter related to the soil water
content at field capacity and saturation. The fraction of soil
nitrate- N lost to denitrification increases quickly as soil water
content increases beyond the field capacity. The EPIC and
GLEAMS method of simulating denitrification neglects
denitrification that may occur in anaerobic micro-zones when the
soil is not at field capacity or saturation.
Hole-in-the-pipe Hole-in-the-pipe T N Davidson The Hole-in-Pipe model relates the emissions of nitrous oxides
et al. (2000) to common soil processes. It regulates soil emissions of NO and
N2O at two levels: 1st, the rate of N cycling through ecosystems,
which is symbolized by the amount of N flowing through the
pipes, affects total emissions of NO and N2O; 2nd, soil water
content and perhaps other factors affect the ratio of N20:N0
emissions, symbolized by the relative sizes of the holes through
which nitric oxide and nitrous oxide "leak." Soil water content is
so important because it controls the transport of O2 into soil and
the transport of NO, N2O, and N2
A-47
-------
Model Name Type1 Support2 Reference3
Notes
MERLIN Model of Ecosystem T
Retention and Loss of
Inorganic Nitrogen
Cosby et al. MERLIN is a catchment-scale mass-balance model of linked
(1997), carbon and N cycling in ecosystems for simulating leaching
Kjonaas and losses of inorganic N. It considers linked biotic and abiotic
Wright (1998) processes affecting the cycling and storage of N. The model is
aggregated in space and time and contains compartments
intended to be observable and/or interpretable at the plot or
catchment scale. The structure of the model includes the
inorganic soil, a plant compartment, and two soil organic
compartments. Fluxes in and out of the ecosystem and between
compartments are regulated by atmospheric deposition,
hydrological discharge, plant uptake, litter production, wood
production, microbial immobilization, mineralization, nitrification,
and denitrification. N fluxes are controlled by carbon productivity,
the C:N ratios of organic compartments and inorganic N in soil
solution. Inputs include time series, constants, rates, source
terms, and soil characteristics (Cosby et al., 1997).
NLM Waquoit Bay Nitrogen T
Loading Model
RTI The Waquoit Bay Nitrogen Loading model estimates inputs from
International different N sources to defined land use categories and then
(2001) estimates losses of N in various compartments of the watershed
ecosystem, including the groundwater. This empirical N loading
model produces long-term average output. It is not currently
endorsed by a federal agency but has been published in peer-
reviewed journals. Most applications of the model have focused
on the Cape Cod area of Massachusetts. The empirical data for
this model are specific to that area and the model simulates N
transport exclusively in the subsurface (i.e., overland transport is
not considered).
Simple Mass
Balance
Method/Steady
State Mass
Balance
"Mass balance
approach"
Bhattacharya
etal. (2004),
Likens et al.
(1996),
Rodriguez
and Macias
(2006)
Examples of two mass balance approaches are:
Simple Mass Balance: This model is based on a balance of
inputs and outputs of N according to the equation:
Ndep + Nfix = Ni + Nu + Nad + Nde + Nfire + Neros + Nvol + Nle Where
the subscripts denote: dep (deposition); fix (fixation); i
(immobilization); u (uptake); ad (adsorption); de (denitrification);
fire (N loss during combustion); eras (erosion); vol (volatilization);
le (leaching).
Steady State Mass Balance: This method is the most commonly
used method for analysis of critical loads of acid deposition. Its
basic principle is based on identifying the long-term average
sources of acidity and alkalinity in order to determine the
maximum acid input that will balance the system at a
biogeochemical safe-limit. Several assumptions have been made
in the steady state calculations. First, it is assumed that ion
exchange is at steady state and there is no net change in base
saturation or no net transfer of ANC from soil solution to the ion
exchange matrix. It is assumed that for N there is no net
denitrification, adsorption or desorption and the N cycle is at
steady state. Sulfate is also assumed to be at steady state: no
sulfide oxidation, sulfate uptake, sulfate permanent fixation or
sulfate reduction are significant. Simple hydrology is assumed
where there is straight infiltration through the soil profile.
A-48
-------
Model Name Type1 Support2 Reference3
Notes
HSPF/LSPC Hydrological T* S 12,13
Simulation Program -
FORTRAN
HSPF simulates watershed hydrology and water quality for
conventional and toxic pollutants. HSPF incorporates watershed-
scale ARM and NPS models into a basin-scale analysis
framework that includes fate and transport in stream channels. It
is a comprehensive model of watershed hydrology and water
quality that allows the integrated simulation of land and soil
contaminant runoff processes with In-stream hydraulic and
sediment-chemical interactions. LSPC is a simplified version of
HSPF. Although LSPC was designed to provide a less data
intensive alternative to HSPF for modeling very large scale
watersheds, it can also be used to model smaller, more detailed
watersheds. The primary disadvantage of this simplified version
of HSPF is that the developers eliminated the atmospheric
deposition routines found in HSPF. For each model run, it
automatically generates comprehensive text-file output by
subwatershed for all land-layers, reaches, and simulated
modules, which can be expressed on hourly or daily intervals.
Output from LSPC has been linked to other model applications
such as EFDC, WASP, and CE-QUAL-W2.
PLOAD Pollutant Loading T* S 14
Model
PLOAD is part of U.S. EPA's BASINS (Better Assessment
Science Integrating Point and Nonpoint Sources) program and
estimates nonpoint pollution sources on an annual basis. PLOAD
can be combined with geographic information system (GIS)—
based data coverages to rapidly estimate N loading to the bay
using pass-through rates based on land uses from U.S. EPA
guidance documents, literature, or other studies. This model is
not precipitation driven; it does it include N speciation. However,
the model does capture differences in N transport for different
land uses in the watersheds.
SWAT Soil and Water T* S van Griensven SWAT is a public domain river basin scale model actively
Assessment Tool and Bauwens developed and primarily supported by the USDA (and included
(2001), 15 within U.S. EPA BASINS framework), that quantifies the impact
of land management practices in large, complex watersheds.
SWAT is a physically based model that applies to all land uses
and to include stormwater runoff in its calculations. The model
simulates NH3, nitrate, and organic N throughout the
waterbodies and vegetation in the modeled system. At this time
SWAT only accepts the nitrate concentration in the rain as the N
atmospheric component.
WARMF Watershed Analysis T* S 16
Risk Management
Framework
WARMF includes a GIS-based watershed model that calculates
daily runoff, shallow groundwater flow, hydrology and water
quality of a river basin. A river basin is divided into a network of
land catchments (including canopy and soil layers), stream
segments, and lake layers for hydrologic and water quality
simulations. Inputs include meteorology, air quality, point source,
reservoir release, and flow diversion data. WARMF also includes
two watershed approach modules for Consensus building and
TMDL calculation.
A-49
-------
Model Name Type1 Support2 Reference3
Notes
DRAINMOD T* N 17 DRAINMOD quantifies N losses and transport from agricultural
lands with shallow water tables where artificial drainage systems
are used. Watershed-scale versions of DRAINMOD have been
developed and evaluated based on data collected on a NC
coastal plain. DRAINMOD is based on water balances in the soil
and at the soil surface. It uses functional methods to quantify
infiltration, subsurface drainage, surface drainage,
evapotranspiration, seepage, freezing, thawing, snowmelt, and
seepage. The model predicts the water table depth and soil
water contents above the water table, drainage rates and the
other hydrologic components on an hourly and daily basis for
long periods of hydrologic record. Hydrologic predictions of the
model have been tested and found to be reliable for a wide
range of soil, crop, and climatological conditions.
INCA is a water and N mass balance simulation model; it
estimates the integrated effects of point and diffuse N sources on
stream nitrate and ammonium concentrations and loads and also
estimates the N loads related to processes in the plant/soil
system. It has been most commonly applied to watersheds within
the UK, but more recently has been modified for use in other
European watersheds. INCA quantifies plant uptake of nitrate
and ammonium, nitrification, denitrification, and mineralization
and immobilization within each land-use type and subcatchment.
Biogeochemical reactions are limited to the soil zone from which
water and N are leached to deeper groundwater.
INCA
T*
Wade et al.
(2005)
LWWM Linked T* N 18
Watershed/Waterbody
Model
The original release of the LWWM coupled the RUNOFF Block of
the U.S. EPA SWMM model (Version 4.21) with the U.S. EPA's
Water Quality Analysis Program (WASP5). All components were
accessed via a user-friendly operating shell. The LWWM
included a GIS interface based on Arc/INFO to automate the
reduction of spatial data within a watershed (i.e., land use and
soils) for input into the RUNOFF Block of SWMM. The LWWM
included pre-processors for inputting data into the RUNOFF
Block of SWMM, WASP5 (Eutro and Toxi), and two
hydrodynamic models associated with WASP5 (RIVMOD and
DYNHYD5), as well as a graphical post-processor for the review
of output from all model components. The post-processor was
also the means by which nonpoint source loading files from
RUNOFF were mapped to WASP segments.
ReNuM3 Regional Nutrient T* N 19
Management Model
ReNuMa is based on the Generalized Watershed Loading
Function (GWLF) model that has been used widely for purposes
such as TMDL development. ReNuMa improves on GWLF by
incorporating Net Anthropogenic N Inputs (NANI) accounting
system. The model now considers atmospheric deposition,
fertilizer application, septic system effluents, N fixation, and
denitrification.
RHESS T* N Boyeretal
(2006b)
RHESS has been used to explore N dynamics at the watershed
scale. RHESSys simulates the coupled effects of C, N, and
hydrological processes by coupling biogeochemical dynamics
from the BIOME_BGC and the NGAS model used in DAYCENT.
Streamflow is based on the implementation of variable source-
area concepts based on topography, quantifying routing of water
through the landscape from patch to patch using either a lumped
topographic approach adapted from TOPMODEL or a distributed
approach adapted from the DHSVM model
A-50
-------
Model Name Type1 Support2 Reference3
Notes
1Type: A = aquatic; I = integrated aquatic/terrestrial; T = terrestrial; T* = watershed
2 Support: S = currently supported by U.S. EPA; N = currently not supported by U.S. EPA
3Websites:
1: River and Stream Water Quality Model (QUAL2K); http://www.epa.gov/athens/wwqtsc/html/qual2k.html
2: U.S. EPA's WASP Website; http://epawasp.com/
3: U.S. Army Core of Engineers Environmental Laboratory - Water Quality Models; http://el.erdc.usace.army.mil/products.cfm?Topic=model&Type=watqual
4: Hydroqual: Row Column AESOP (RCA) Modeling Code Description and Technical Capabilities; http://www.hydroqual.com/pdf/RCA_Desc_doc.pdf
5: Woods Hole Marine Lab, Ecosystems Center MEL home page http://ecosystems.mbl.edu/Research/Models/mel/welcome.htmll
6: IBIS (Integrated Biosphere Simulator); http://water.usgs.gov/software/hspf.html
7: THMB (Terrestrial Hydrology Model with Biogeochemistry) - formerly HYDRA; http://www.sage.wisc.edu/download/HYDRA/hydra.html
8: Biome-BGC: Terrestrial Ecosystem Process Model, Version 4.1.1; http://www.daac.ornl.gov/MODELS/guides/biome-bgc_guide.html
9: Denitrification Modeling Workshop: Model Summary; http://marine.rutgers.edu/BGC/RCNsite/WS1/WS1models/DNDC-2.pdf
10: EPIC Fact Sheet; http://www.brc.tamus.edu/epic/epfact2004.htm
11: GLEAMS Y2K Update Website; http://www.tifton.uga.edu/sewrl/Gleams/gleams_y2k_update.htm
12: U.S. EPA's HSPF Website; http://www.epa.gov/ceampubl/swater/hspf/
13: USGS's Water Resources Applications Software: HSPF Website; http://water.usgs.gov/software/hspf.htm
14: U.S. EPA's Better Assessment Science Integrating Point & Nonpoint Sources (BASINS) Website; http://www.epa.gov/waterscience/basins/
15: Soil & Water Assessment Tool; http://www.brc.tamus.edu/swat/
16: US EPA's Watershed Analysis Risk Management Framework (WARMF) Website; http://www.epa.gov/athens/wwqtsc/html/warmf.html
17: DRAINMOD Download Website; http://www.bae.ncsu.edu/soil_water/drainmod/
18: Linked Watershed Waterbody Model at the Southwest Florida Water Management District; http://www.swfwmd.state.fl.us/software/lwwm.htm
19: Regional Nutrient Management (ReNuMa) at Cornell University College of Agriculture and Life Sciences;
http://www.eeb.cornell.edu/biogeo/nanc/usda/renuma.htm
A-51
-------
Annex B. Acidification Effects
B.1. Effects on Biogeochemical Processes along
Acidification Pathways
B.1.1. Atmospheric Deposition and Canopy Interaction
Inputs of N and S in wet, dry, and occult deposition first interact with the vegetative canopy. This
interaction can occur a few centimeters above the ground in some alpine or grassland ecosystems to over
100 m above the ground in some forest canopies. In the canopy, deposited pollutants (especially N) can be
taken up by the plants or by organisms that live within the canopy or on the leaf surface. Most of the
deposited S moves as throughfall to the soil where it can be temporarily, or permanently, adsorbed on the
soil. Sulfur that is not adsorbed on the soil moves readily into drainage water.
Earlier reviews (i.e., Hosker and Lindberg, 1982; Taylor et al., 1988) summarized information on
the deposition of N to vegetation surfaces and interactions between pollutant deposition and canopy and
leaf surfaces. Deposited N that is not taken up within the canopy then falls to the ground as throughfall,
where plants, bacteria, and fungi compete for it. This competition for deposited N has long been known to
play an important role in determining the extent to which N deposition will stimulate plant growth and the
degree to which added N is retained within the ecosystem (U.S. EPA, 1993a). The available surface area
of vegetation, onto which N gasses readily diffuse, has a significant effect on the dry deposition of N
(Heil and Bruggink, 1987). Coniferous forests tend to increase deposition rates (both dry and wet) relative
to deciduous forests, and landscape features such as elevation, aspect, and forest edge can play an
important role in creating high levels of variability in deposition rates in complex terrain (Weathers et al.,
2000).
B.1.2. Interactions with Soil
Air pollution is not the sole cause of soil acidity. High rates of soil acidification occur in low-
deposition regions of the western U.S. because of internal soil processes, including tree N uptake and
nitrification associated with extensive N fixation, for example on sites occupied by red alder trees (Alnus
rubra) (Johnson et al., 1991a). Acidic deposition is not a necessary condition for having acidic soils, as
evidenced by the common occurrence of acidic soils in unpolluted forests of the northwestern U.S. and
Alaska (Johnson et al., 1991a).
B.1.2.1. Sulfur Retention and Release
Soils in the U.S. that most effectively adsorb S042 occur south of the maximum extent of
glaciation that occurred during the most recent ice age. (Rochelle and Church, 1987). Sulfate adsorption is
strongly pH dependent, and a decrease in soil pH resulting from acidic deposition can enhance the ability
of soil to adsorb S042 (Fuller et al., 1987).
Considerable effort in the 1980s went into the computation of S budgets for watersheds and forest
plots, to evaluate S retention and release. These budgets were subject to complications from fluxes that
could not be measured directly, such as dry deposition and weathering, but generally indicated net S
B-1
-------
retention at sites south of the line of glaciation—a result attributed to net adsorption of S042 (Cappellato
et al., 1998; Rochelle et al., 1987). Through the 1990s little or no decrease in S042 concentrations
occurred in streams below the glaciation line, despite regional decreases in atmospheric deposition of S
(Webb et al., 2004). This lack of response has been generally attributed to the net release of adsorbed
S042 , resulting from a shift in equilibrium between the adsorbed and solution phases under conditions of
decreased atmospheric inputs of S042 . This interpretation is supported by a decrease in concentrations of
adsorbed S042 from 1982 to 1990 in a Piedmont soil in South Carolina that received decreasing levels of
S deposition during this period (Markewitz et al., 1998). This same soil also experienced an increase in
adsorbed S042 from 1962 to 1972 (Markewitz et al., 1998). The only published S budget more recent
than 1992 for an unglaciated site in the U.S. (Castro and Morgan, 2000) also suggests a net release of
S042 . This upland Maryland watershed released 1.6 times more S042 than measured in throughfall in
1996-97. Additional information was obtained in the German study of Martinson et al. (2005) in which a
"clean-roof' was used to exclude acidic deposition since 1989. Data collection enabled calibration of a
model that predicted elevated concentrations of desorbed S042 in soil water for at least several decades.
Although decreased levels of deposition are most likely resulting in net S042 desorption, limited research
is available on sulfate desorption over time periods relevant to the time scale of decreased levels of S
deposition (Johnson and Mitchell, 1998).
Numerous S budgets were also compiled in the 1980's for glaciated sites, and results generally
indicated that inputs approximately equaled outputs on an annual basis (Rochelle et al., 1987). Little or no
S retention at glaciated sites was attributed to relatively low S042 adsorption capacity in soils. Balanced
S budgets implied that decreases in atmospheric deposition of S would lead directly to decreases in S042
leaching, and the strong correlation between decreases in atmospheric deposition and decreases in S042
concentrations in surface waters is widely recognized as an indication of this direct linkage (Stoddard
et al., 2003). However, considerable evidence also indicates that S inputs in glaciated ecosystems do not
behave conservatively, but instead are cycled through microbial and plant biomass (Alewell and Gehre,
1999; David et al., 1987; Likens et al., 2002). As a result, large quantities of S are stored in organic forms
within the soil. David et al. (1987) found that annual S deposition (wet plus dry) at a site in the central
Adirondack region of New York was about 1% of the organic S pool in the soil. Houle et al. (2001)
estimated that annual S deposition at 11 sites in North America ranged from 1% to 13% of the organic S
pool in soil.
Courchesne et al. (2005) measured a downward trend in water-soluble S042 from 1993 to 2002 in
glaciated soils in Quebec, and attributed this response to net desorption of S042 rather than release of
organically associated S. However, during this period, deposition of S042 was essentially unchanged.
They attributed this discrepancy to a delay in the release of adsorbed S042 in response to a decrease in S
deposition over the previous decade. These authors did not provide a mechanism to explain how
desorption can continue under conditions of constant S042 inputs, however. On the basis of the abundant
evidence of biological S cycling, it seems more likely that the delay observed by Courchesne et al. (2005)
is the result of biological controls over the release of S.
Much of the organic S stored in soil is in carbon-bonded forms that are relatively unreactive, but
can be mineralized to S042 in oxic conditions, typically found in moderately well-drained to well-drained
soils (Johnson and Mitchell, 1998). Furthermore, strong correlations have been shown between levels of
atmospheric deposition of S and concentrations of S in soil (Driscoll et al., 2001b; Novak et al., 2001).
Long-term increases in concentrations of total S in soils that are at least partially attributable to increases
in organic S have also been documented (Knights et al., 2000; Lapenis et al., 2004). The study of Houle
et al. (2001) did not find a relation between these factors, however. A Swedish "clean-roof' study also
provides some insights into the role of organic S in possibly delaying recovery (Morth et al., 2005). After
9 years of pre-industrial levels of S deposition, the amount of S in runoff still exceeded inputs by 30%.
Most of the S in runoff was attributed to mineralization of organic S in the O horizon.
B-2
-------
B.1.2.2. Base Cation Depletion
Base cations are common in rocks and soils, but largely in forms that are unavailable to plants.
There is a pool of bioavailable base cations (termed exchangeable base cations) that are adsorbed to
negatively charged surfaces of soil particles. They can enter solution by exchanging with other dissolved
cations including acidic cations such as H or Al3+. Base cations in this pool are gradually leached from
the soil in drainage water, but are constantly resupplied through weathering. Weathering slowly breaks
down rocks and minerals, releasing base cations to the pool of adsorbed base cations in the soil. The
balance between base cation supply and base cation loss determines whether the pool of available base
cations is increasing or decreasing in size. Net forest growth can also potentially lower exchangeable base
cation concentrations through uptake of nutrient cations (Ca, Mg, and K), but these cations remain in the
terrestrial ecosystem and can become available in the future through mineralization or canopy leaching. It
has long been known that leaching of base cations by acidic deposition might deplete the soil of
exchangeable bases faster than they are resupplied (Cowling and Dochinger, 1980). However, base cation
depletion of soils had not been demonstrated at the time of the 1982 PM-SOx AQCD (U.S. EPA, 1982b).
Data that clearly showed soil base cation depletion in the U.S. did not become available until the
1990s, although decreases in exchangeable Ca2+ concentrations between the periods 1947 to 1950 and
1987 to 1988 had been identified in European soils through repeated sampling (Billett et al., 1990;
Falkengren-Grerup and Eriksson, 1990). In the only repeated sampling in the U.S. in which the original
soil sample pre-dated acidic deposition, Johnson et al. (1994b) documented a decrease in exchangeable
Ca2+ concentrations in both the O (combined Oa and Oe horizons) and B horizons from 1930 to 1984.
Richter et al. (1994) also observed Ca2+ depletion in the B horizon from 1960 to 1990, in repeated
sampling of Piedmont soil in South Carolina. The studies of Johnson et al. (1994b) and Richter et al.
(1994) acknowledged the potential role of acidic deposition in causing the loss of Ca2+, but focused on net
forest growth as the primary cause.
Through reanalysis of archived soils, Lawrence et al. (1995) measured decreases in concentrations
of exchangeable Ca2+ and acid-extractable Ca2+ in Oa horizons of spruce stands from 1969-70 to 1987-92
and presented relationships in soil chemistry that were not consistent with changes expected from
vegetation uptake effects, but that could be explained by acidic deposition. Drohan and Sharpe (1997)
also observed a decrease in Ca2+ concentrations in Oa and A horizons at 11 sites across Pennsylvania that
were sampled in 1957 or 1959 and again in 1993, although effects of vegetation and acidic deposition
were not distinguished.
A thorough soil re-sampling study in the U.S. was conducted by Bailey et al. (2003) in
northwestern Pennsylvania. Between 1967 and 1997, pronounced decreases, attributed largely to acidic
deposition, were measured in exchangeable Ca2+ and Mg2+ concentrations in Oa/A horizons and
throughout the B horizon. Courchesne et al. (2005) found higher concentrations of exchangeable Ca2+ in
the O horizon (combined Oe and Oa horizons) in 2002 than in 1994 at one of three sampling areas within
a 5.1 ha watershed, but no significant differences at the other two locations. No significant differences
were found for exchangeable Mg2+ at the three locations in the O horizon. In the upper 10 cm of the
B horizon, no significant differences were found in exchangeable Ca2+, but at two of three locations,
exchangeable Mg2+ concentrations were lower in 2002 than in 1994. The 8-yr interval between sampling
in this study is the shortest time in which changes in exchangeable base cations have been reported for
North American soils.
In a regionally designed assessment of changes in soil-exchange chemistry, Sullivan et al. (2006b)
found that base saturation and exchangeable Ca2+ concentrations in the Adirondack region of New York
had decreased in the upper 10 cm of the B horizon between the mid 1980s and 2003, in watersheds of
lakes with ANC less than 200 |_ieq/L. Soil chemistry in 36 lake watersheds in the mid 1980s was
compared to soil chemistry in 32 lake watersheds in 2003. Although this study did not involve repeated
sampling of the same sites, the comparison could be made on a regional basis because the sampling
locations were selected randomly in both the mid 1980s and in 2003, and a large and similar number of
sites were included in both samplings.
B-3
-------
In a widely cited article, Likens et al. (1996) used a watershed mass balance approach to estimate
changes in ecosystem Ca2+ pools at HBEF and found a sustained decrease in exchangeable Ca2+
concentrations from 1963 to 1993. The maximum depletion rate occurred in 1972, at the estimated peak
in acidic deposition levels. The dependence on Ca:Na ratios to estimate Ca2+ weathering fluxes in this
analysis adds uncertainty to the magnitude of changes reported for the exchangeable Ca2+ pool (Bailey
et al., 2003). Two additional mass balance studies used Sr isotopes to evaluate changes in soil Ca2+ pools
and fluxes. The study of Bailey et al. (1996) estimated substantial depletion rates in a watershed in the
White Mountains of New Hampshire. Miller et al. (1993) estimated that inputs from weathering and
atmospheric deposition approximately equaled leaching losses at a site in the Adirondack Mountains in
New York. The different findings in these two studies are related to differences in the mineralogical
composition of the respective soils. However, the Miller et al. (1993) study also estimated that 50 to 60%
of the Ca2+ in vegetation and the forest floor was derived from the atmosphere, despite the fact that the
weathering flux was estimated to be three times the rate of atmospheric inputs. This result suggests that
Ca2+ supply from weathering in the lower profile is not reaching the upper soil where most root activity
occurs, and that Ca2+ depletion has occurred in the upper soil.
The study ofYanai et al. (1999) investigated changes in Ca2+ and Mg2+ concentrations and content
in northeastern hardwood stands overtime intervals ranging from 10 to 21 years. The general conclusion
of this study was that little or no change in O horizon (Oi, Oe, and Oa horizons) exchange chemistry
occurred. However, a decrease in exchangeable Ca2+ concentrations in the Oa horizon was observed in
this study at the HBEF from 1978 to 1997, although no change was observed in this soil in exchangeable
Mg2+ concentrations, or Ca2+ or Mg2+ content in the Oa horizon. Results of this study were complicated
by high spatial variability and differences in field sampling techniques between the original collection and
the resampling (Yanai et al., 1999). Yanai et al. (2005) also found little difference over 15 years in
exchangeable and extractable Ca2+ and Mg2+ concentrations in Oa horizons at 6 sites, and in O horizons
(combined Oe and Oa horizons) at 13 sites, in hardwood stands in New Hampshire. In this study, it was
also estimated that a difference greater than 50% would be needed to be statistically detected due to a
large degree of spatial variability. Although most repeated sampling studies did identify decreases in
exchangeable base cations in the Oa or O horizon, the results ofYanai et al. (2005) indicated that this
change may not occur at all sites and may be difficult to detect in some soils due to inconsistencies in
identifying horizon separations during sampling.
Through direct and inferred evidence of Ca2+ depletion, and additional research on soil processes, a
detailed understanding of the mechanisms of Ca2+ depletion has developed over the past two decades.
Ulrich (1983) explained Ca2+ depletion as a three-stage process in which buffering of acidity in the
mineral soil is first accomplished by weathering of carbonates and other mineral forms that weather
relatively rapidly. Once these mineral forms are depleted, buffering is accomplished largely by cation
exchange, in which H+ is substituted for base cations and concentrations of exchangeable base cations
decrease. Once the buffering capacity provided by cation exchange is depleted, acid neutralization is
accomplished by weathering of crystalline minerals that contain large amounts of silicon (Si) and Al and
relatively small amounts of base cations. At this stage, Al is mobilized within the soil and exchangeable
Al concentrations increase. The shift in acid buffering from base cation exchange to alumino-silicate
weathering and exchangeable Al was documented in Russian soils sampled three times over 75 years
(Lawrence etal., 1995).
The effect of decreasing concentrations of exchangeable base cations on cation leaching in mineral
soil was shown in simulation modeling by Reuss (1983). Below a base saturation of 20%, leaching of
Ca2+ decreases substantially and becomes less sensitive to variations in acid inputs as base saturation
decreases further. This relationship was later shown experimentally by Lawrence et al. (1999). Samples
from the upper B-horizon in nearly all of the Adirondack lake watersheds sampled by Sullivan et al.
(2006b) had base saturation values less than 20%, as did soils at 11 sites in New York, Vermont, New
Hampshire, and Maine in a regional study of mature spruce-fir forests (David and Lawrence, 1996).
Exchangeable Ca2+ concentrations (expressed as a percentage of cation exchange capacity [CEC]) in the
regional spruce-fir study were weakly correlated with an estimate of the relative weathering potential of
B-4
-------
parent material in the upper 10 cm of the B horizon (r2 = 0.44). However, these factors were strongly
correlated in the Oa horizon (r2 = 0.92) (Lawrence and David, 1997). Because mineral weathering in the
B horizon is the primary source of soil Ca2+, a strong relationship between weathering potential and
exchangeable Ca2+ concentrations would be expected in this horizon. The weak correlation suggests that
concentrations of Ca2+ had decreased into the Al-buffering range sometime in the past. The parent
material signature in the Oa horizon was likely maintained through vegetative recycling—uptake of Ca2+
from the O and B horizons, followed by transport back into the O horizon in litterfall.
In summary, evidence from repeated sampling and studies of soil processes indicate that decreases
in exchangeable base cation concentrations in both Oa and B horizons are common and widespread in the
eastern U.S. Factors such as logging and net forest growth are likely to have contributed to this decrease
in varying degrees, but acidic deposition has played a major role (Huntington, 2000; Lawrence et al.,
1987). The magnitudes and rates at which Ca2+ depletion has occurred are less clear.
These base cation depletion issues relate directly to the chemical recovery potential of acidified
soils and surface waters. Replenishment of exchangeable base cation concentrations on soils will require
that inputs from weathering and atmospheric deposition exceed losses from leaching and vegetative
uptake. Inputs of Ca2+ from atmospheric deposition decreased sharply in the east through the 1980s
(Hedin et al., 1994), and have remained relatively stable since that time (http://nadp.sws.uiuc.edu/).
Atmospheric deposition of S042 currently remains several factors higher than that of Ca2+ even at sites
where S042 levels are relatively low (http: //nadp. sws .uiuc. edu/). so chemical recovery at current acidic
deposition levels will require inputs of base cations from weathering that are considerably greater than
inputs from the atmosphere.
Because of the importance of weathering to the base-cation status of soils, a great deal of effort has
been made to estimate in situ weathering flux with a variety of methods (Bailey et al., 2003; Likens et al.,
1996; Miller et al., 1993). The complexity and variability of factors that affect weathering flux rates, such
as soil mineralogy, particle surfaces, soil organic matter, moisture flux, and a host of other factors that are
difficult to quantify, add large uncertainties to weathering flux estimates. Weathering rates estimated in
geochemical models are generally assumed to be constant overtime, but lower weathering rates were
observed in a soil sampled in 1987 than in the same soil sampled and archived in 1949-50 (Zulla and
Billett, 1994). Further complexity in weathering flux rates results from the possible role of mycorrhizae in
penetrating silicate minerals to extract base cations while remaining isolated from the soil solution (Blum
et al., 2002; van Breemen et al., 2000). Lastly, as yet unidentified sources of base cations may exist in
forest soils. Bailey et al. (2003) found that elevated rates of Ca2+ loss from forest harvesting continued for
30 years after disturbance, but the source of the additional Ca2+ being lost could not be identified. Until
estimates of in situ weathering fluxes are better constrained and more data become available from
repeated soil sampling, predictions of recovery of exchangeable base cation concentrations will be highly
uncertain.
B.1.2.3. Aluminum Mobilization
Through the natural process of podzolization, dissolved organic acids derived from partially
decomposed organic matter in the O horizon move into the mineral soil where they weather soil particles
and release Al into solution. As soil solution moves deeper into the profile, acidity is neutralized and Al is
deposited as a secondary mineral or more likely as an organic Al (Al0) complex (De Coninck, 1980). The
limited mobility of organic anions results in retention of most Al within the mineral soil (often in the
Bh horizon). Complexation with dissolved organic matter can increase the mobility of Al within the soil
and lead to transport of organic Al into surface waters from shallow soils that are high in organic matter
(Lawrence et al., 1987).
Increased concentrations of exchangeable Al in the mineral soil have been identified through
repeated sampling in the U.S. and Europe over periods ranging from 17 years to 41 years in studies by
Billet et al. (1990), Bailey et al. (2005), Falkengren-Grerup and Eriksson (1990), and Lawrence et al.
B-5
-------
(1995). In areas of Europe with excessively high acidic deposition levels, evidence of A1 depletion in the
mineral soil has also been found (Lapenis et al., 2004; Mulder et al., 1989), but A1 depletion has not been
documented in the U.S.
Increases in exchangeable Al concentrations in the O horizon have been documented over periods
from 17 to 30 years (Bailey et al., 2005; Drohan and Sharpe, 1997; Lawrence et al., 1995), although the
study ofYanai et al. (2005) did not find consistent changes in Oa horizons over 15 years.
Numerous papers have evaluated solubility controls on Al in both the mineral soil and the
O horizon. These papers have commonly related Al solubility to gibbsite (Al(OH)3) or a gibbsite-like
mineral to determine if inorganic Al concentrations could be predicted from gibbsite solubility constants
and pH (Cronan and Goldstein, 1989; David and Driscoll, 1984; Johnson et al., 1981; Lawrence and
David, 1997). These efforts have shown that inorganic Al concentrations are often undersaturated with
respect to gibbsite and do not support Al-trihydroxide as the primary control in natural systems. Gibbsite
solubility should therefore be considered a useful point of reference in evaluating Al-solubility rather a
mineral form that is an important control of Al solubility in natural systems.
Through the 1990s, evidence accumulated to indicate that secondary Al in the mineral soil is in a
form associated with organic matter, and in some soils, imogolite (Berggren and Mulder, 1995; Dahlgren
and Walker, 1993; Mulder and Stein, 1994; Simonsson, 1998; Skyllberg, 1999). Organic matter also plays
a major role in controlling Al solubility in O horizons. This interaction has been described by Cronan
et al. (1986) in O horizons through the bound Al ratio, which reflects the equivalents of adsorbed Al per
mol of carboxyl groups (Cronan et al., 1986). Tipping et al. (1995) described Al solubility on organic and
mineral soil horizons through equilibrium humic ion binding. Each of these approaches has had success in
describing dissolved Al concentrations in organic soils as a function of pH through formulations that rely
on concentrations of solid-phase organic bound Al. Further work has shown these relationships to be
specific to the particular horizon, and the pool sizes of Al and humic substances (Lofts et al., 2001).
However, inputs of acidity may alter concentrations of solid-phase organic bound Al (Lawrence and
David, 1997). Changes in atmospheric deposition levels may therefore shift these relationships over time
as soils further acidify or recover.
B.1.2.4. Soil Acidification
In the B horizon of soils north of the maximum extent of glaciation, CEC is largely derived from
organic matter, whereas in older southern soils the surface charge of highly weathered clay minerals is the
primary source of CEC. The CEC derived from organic matter is pH-dependent. Decreases in pH result in
a decreases in CEC. In both cases, the CEC of the B horizon is much lower than in organic-rich
surface horizons (Oa or A horizons). Less acidity from organic matter and a limited capacity for buffering
due to low CEC makes the B horizon more susceptible to a lowering of pH from acidic deposition, and
decreases in pH lower the CEC, further reducing the acid-buffering capacity from cation exchange. Two
studies in the U.S. have provided measurements to assess changes in soil pH in the B horizon from acidic
deposition. Bailey et al. (2005) found lower pH values in the upper B horizon in northwestern
Pennsylvania soils in 1967 than in 1997, at 50 cm depth (p <0.001) and at 100 cm depth (p <0.001),
which were largely attributable to acidic deposition. Markewitz et al. (1998) also found pronounced
decreases in soil pH down to 60 cm in highly weathered Piedmont soils from 1962 to 1990. The latter
study was conducted in a former cotton field in which loblolly pines were planted in 1956-57. Forest
regrowth undoubtedly played a large role in the soil pH changes that were measured, but atmospheric
deposition was estimated to account for 38% of the H inputs during the 28 years that elapsed between
measurements.
Other studies in Europe have found similar decreases in soil pH of the B horizon that could be
attributed, at least in part, to acidic deposition. These include the study of Lawrence et al. (1995) in
northwestern Russia, which documented decreases in soil pH in the B horizon down to 90 cm, from 1926
to 1964, and further decreases from 1964 to 2001. Acidic deposition was identified as the probable
B-6
-------
primary cause of decreasing pH in this study. The study of Lawrence et al. (1995) also observed a
decrease in CEC in this soil, as did a previous study of Russian soils (Lapenis et al., 2004). The decrease
in pH was likely to have contributed to the decreased CEC of these soils, but a more important factor may
have been a decrease in organic carbon concentrations that was also measured. To our knowledge, data to
assess possible changes in CEC in soils in the U.S. has not become available, but change in CEC has
implications for recovery potential of soils from acidic deposition effects (Sullivan et al., 2006a).
Increased CEC driven by increases in pH could foster soil recovery by increasing the opportunity for
adsorption of base cations, as soil solution becomes less acidic. Decreases in soil organic matter driven by
climate and/or vegetation changes, such as those seen in Russian soils, would result in a decrease in acid-
buffering capacity through cation-exchange. There are currently no data in the U.S. that indicate increases
in soil pH associated with recent declines in acidic deposition levels. These data limitations make future
projections of recovery of soil pH highly uncertain.
B.1.2.5. Nitrogen Saturation
Severe symptoms of N saturation, have been observed in high-elevation, nonaggrading spruce-fir
ecosystems in the Appalachian Mountains, as well as in the eastern hardwood watersheds at Fernow
Experimental Forest near Parsons, WV and throughout the northeastern U.S. Mixed conifer forests and
chaparral watersheds with high smog exposure in the Los Angeles Air Basin also are N-saturated and
exhibit the highest stream water N03 concentrations documented within wildlands in North America
(Bytnerowicz and Fenn, 1996; Fenn et al., 1998).
Some examples of N-saturated forests in North America, including estimated inputs and outputs,
are shown in Table B-l (Fenn et al., 1998). The Harvard Forest hardwood stand in western Massachusetts
absorbed >900 kg N/ha without significant N03 leaching during an 8-yr N amendment study (Fenn
et al., 1998). In contrast, N03 leaching losses were high at the Harvard Forest pine sites. In the 8-yr
experimental study, N03 leaching was observed in the pine stand after the first year (1989) in the high-N
application plots, and further increases were observed in 1995 and 1996. The hardwood stand did not
show significant increases in N03 leaching until 1996. The differences in response of the pine and
hardwood stands indicate that the mosaic of community types across the landscape must be considered
when evaluating landscape-scale responses to N deposition (Magill et al., 2000).
Utilization of N in the terrestrial ecosystem is accomplished through complex interactions between
plants and microbes that are not fully understood (Schimel and Bennett, 2004). Long-term N retention is
largely accomplished by incorporation ofN into soil organic matter through biological assimilation (Aber
et al., 1998), and to a lesser extent by abiotic processes that are not well understood (Dail et al., 2001).
The forms in which N is assimilated by plants and microbes are determined by availability, as described
in Schimel and Bennett (2004). In the most N-limited ecosystems, competition between plants and
microbes is high and N is assimilated primarily in depolymerized organic forms, resulting in low
mineralization rates and minimal buildup of inorganic N in the soil. Increased availability of N increases
the mineralization rate, which enhances competition between plants and microbes for available NH4+
produced by mineralization. Further increase in the availability of N (for example by high levels of
atmospheric N deposition) lessens competition for NH4+ between plants and microbes and leads to
increased production of N03 by autotrophic nitrifying bacteria. Some of this N03 can be taken up by
plants and microbes, but because much of the N demand is satisfied by NH4+ under these conditions,
N03 tends to be mobile within the soil, enabling it to leach to drainage water. Based on the definitions of
Aber et al. (1989, 1998) and Stoddard (1994), the first stage of N saturation is reached when competition
between plants and microbes for NH4+ has decreased to the point that net nitrification occurs.
Substantial leaching of N03 from forest soils to streamwater can acidify downstream waters
(Webb et al., 1995), eutrophy estuaries and marine waters (Fisher and Oppenheimer, 1991), and deplete
soils of nutrient base cations, especially Ca2+ and Mg2+ (Likens et al., 1998). Considerable evidence is
available to link N deposition to acidification of soils. Much of this evidence comes from the northeastern
B-7
-------
U.S., where increased accumulation of N in soil is suggested by a strong positive correlation between
atmospheric deposition levels and total N concentration in the Oa horizon, at sites in New York, Vermont,
New Hampshire, and Maine (Driscoll et al., 2001b). Further evidence that atmospheric deposition has
increased availability of N in soil is shown by a strong negative correlation between atmospheric
deposition levels and the C:N ratio of the Oa horizon in this region (Aber et al., 2003). If the C:N ratio
falls below about 25, nitrification is stimulated, resulting in elevated N03 in surface waters (Aber et al.,
2003). Similar results were found in Europe, where a C:N ratio of 24 was identified as the critical level
below which nitrification occurred (Emmett et al., 1998).
Analyses have been conducted in the northeastern U.S. and Europe to examine the relationships
between N deposition and N03 leaching to surface waters. The relationship between measured wet
deposition of N and streamwater output of N03 was evaluated by Driscoll et al. (1989) for sites in North
America (mostly eastern areas), and augmented by Stoddard (1994). The resulting data showed a pattern
of N leaching at wet inputs greater than approximately 5.6 kg N/ha/yr. Stoddard (1994) presented a
geographical analysis of patterns of watershed loss of N throughout the northeastern U.S. He identified
approximately 100 surface water sites in the region with sufficiently intensive data to determine their N
status. Sites were coded according to their presumed stage ofN retention, and sites ranged from Stage 0
(background condition) through Stage 2 (chronic effects). The geographic pattern in watershed N
retention depicted by Stoddard (1994) followed the geographic pattern of N deposition. Sites in the
Adirondack and Catskill Mountains in New York, where N deposition was about 11 to 13 kg N/ha/yr,
were typically identified as Stage 1 (episodic effects) or Stage 2. Sites in Maine, where N deposition was
about half as high, were nearly all Stage 0. Sites in New Hampshire and Vermont, which received
intermediate levels of N deposition, were identified as primarily Stage 0, with some Stage 1 sites. Based
on this analysis, a reasonable threshold of N deposition for transforming a northeastern site from the
"natural" Stage 0 condition to Stage 1 would correspond to the deposition levels found throughout New
Hampshire and Vermont, approximately 8 kg N/ha/yr. This is in agreement with the interpretation by
Driscoll et al. (1989), which would probably correspond to total N inputs near 8 tolO kg N/ha/yr. This is
probably the approximate level at which episodic aquatic effects of N deposition would become apparent
in many watersheds of the eastern U.S.
Analysis of data from surveys of N outputs from 65 forested plots and catchments throughout
Europe were conducted by Dise and Wright (1995) and Tietema and Beier (1995). Below the throughfall
inputs of about 10 kg N/ha/yr, there was very little N leaching at any of the study sites. At throughfall
inputs greater than 25 kg N/ha/yr, the study catchments consistently leached high concentrations of
inorganic N. At intermediate deposition values (10 to 25 kg N/ha/yr), Dise and Wright (1995) observed a
broad range of watershed responses. Nitrogen output was most highly correlated with N input (r2 = 0.69),
but also significantly correlated with S input, soil pH, percent slope, bedrock type, and latitude. A
combination of N input (positive correlation) and soil pH (negative correlation) explained 87% of the
variation in N output at the study sites (Dise and Wright, 1995).
The threshold level of atmospheric deposition that causes release of N03 to surface waters was
identified by Aber et al. (2003) as approximately 7 kg N/ha/yr for the northeastern U.S. In watersheds
receiving N deposition above this level, concentrations of N 03 in surface waters were positively
correlated with atmospheric deposition, whereas most watersheds with deposition less than 7 kg/ha/yr had
little or no N03 (undetectable at most sites) in their surface waters (Aber et al., 2003). The threshold
value of 7 kg/ha/yr was based on atmospheric deposition levels for the base of forested watersheds. When
scaled to include higher deposition levels expected at upper elevations this value was estimated to equal
about 10 kg/ha/yr, similar to the European estimate of Dise et al. (1998).
The common deposition threshold for release of N03 to surface waters in forested watersheds
found in the northeastern U.S. and Europe represents an important advance in relating N inputs to
ecosystem effects, but a considerable amount of variability in ecosystem response has also been
demonstrated. Lovett et al. (2000) found that 39 watersheds in the Catskill region of New York State
retained from 49% to 90% of atmospheric N inputs. Castro and Morgan (2000) showed that N03 export
B-8
-------
from watersheds in eastern North America can range from nearly 0 to over 400 eq/ha/yr in watersheds
that receive similar levels of inorganic N in wet deposition in the range of 400 to 500 eq/ha/yr.
Experimental additions of N to plots and watersheds have also demonstrated variations in terrestrial
retention of N. Additions of N (approximately twice ambient deposition) to hardwood watersheds in
Maine (25 kg N/ha/yr) and West Virginia (35.5 kg N/ha/yr), which were releasing N03 to surface waters
before the additions, resulted in substantial increases in N03 concentrations in soil water and stream
water within the first treatment year (Kahl et al., 1993; Peteijohn et al., 1996) Additions of 25 kg N/ha/yr
to spruce plots in Vermont (ambient bulk deposition 5.4 kg N/ha/yr), in which net nitrification did not
occur before treatment, triggered net nitrification in the second year of treatment, whereas nitrification
was not triggered until the third year in plots receiving 19.8 kg N/ha/yr (McNulty et al., 1996). Similar
results to these were seen in two studies from Colorado. Additions of 25 kg N/ha/yr to old-growth spruce
plots in Loch Vale watershed (ambient bulk deposition ~4-5 kg N/ha/yr) doubled N mineralization rates
and stimulated nitrification, while the addition of the same amount to plots receiving ambient bulk
deposition of -2.0 kg N/ha/yr in Fraser Experimental Forest elicited no microbial response but
significantly increased foliar and organic soil horizon N (Rueth et al., 2003). A comparison study of old-
growth spruce plots across a depositional gradient in Colorado found mineralization rates to be higher
where N deposition ranged from 3 to 5 kg N/ha/yr than where N deposition ranged from 1 to 2 kg
N/ha/yr, with measurable nitrification rates at sites with the highest deposition amounts (Rueth and Baron,
2002). In marked contrast to these results, concentrations of N03 plus NH4+ were not detected until the
seventh year in hardwood plots in Harvard Forest, which received additions of 150 kg N/ha/yr (Magill
and Aber, 2004). Concentrations of (N03 + NH4+) in hardwood plots receiving 50 kg N/ha/yr were not
yet detectable in the 15th year of treatments. The same treatments were applied to red pine (Pinus
resinosa) plots, which exhibited elevated concentrations of (N03 + NH4+) in soil water after 1 year of
150 kg N/ha/yr doses, and after 5 years of 50 kg N/ha/yr doses.
In general, deciduous forest stands in the eastern U.S. have not progressed toward N-saturation as
rapidly or as far as spruce-fir stands. Deciduous forests may have a greater capacity for N retention than
coniferous forests. In addition, deciduous forests tend to be located at lower elevation and receive lower
atmospheric inputs ofN. Many deciduous forests have higher rates ofN uptake and greater N requirement
than spruce-fir forests. Decreased growth and increased mortality have more commonly been observed in
high-elevation coniferous stands than in lower elevation hardwood forests, and these differences have
been partially attributed to excess inputs ofN (Aber et al., 1998). Indeed, many of the lower elevation
deciduous stands are N-deficient and are therefore likely to benefit (i.e., grow faster), at least up to a
point, with increased inputs ofN.
There are examples ofN saturation in lower-elevation eastern forests, especially in West Virginia.
For example, progressive increases in streamwater N03 and Ca2+ concentrations were measured at the
Fernow Experimental Forest in the 1970s and 1980s (Adams et al., 1997, 2000; Edwards and Helvey,
1991; Peteijohn et al., 1996). This watershed has received higher N deposition (average throughfall input
of 22 kg/ha/yr ofN in the 1980s) than is typical for low-elevation areas of the eastern U.S., however
(Eagar et al., 1996), and this may help to explain the observed N saturation.
Varying responses to N additions reflect differences in N status of the treatment sites. These
variations have most often been attributed to disturbance history, dating back a century or more (Goodale
and Aber, 2001). Sites which have undergone disturbances that cause loss of soil N, such as logging, fire,
and agriculture, tend to be most effective at retaining atmospheric and experimental inputs ofN. Nitrogen
retention capability often decreases with stand age, which suggests that older forests are more susceptible
than younger forests to becoming N-saturated (Hedin et al., 1995). Aber et al. (1998) surmised that land
use history may be more important than cumulative atmospheric deposition ofN in determining the N
status of a forest ecosystem.
Although considerable progress has been made in understanding the factors that control N
retention, efforts to quantify net N retention through known processes have not been fully successful.
Assimilation ofN by mycorrhizae followed by exudation as dissolved organic matter was proposed by
B-9
-------
Aber et al. (1998) as a possible explanation for unaccounted conversion of inorganic N into soil organic
matter. However, Frey et al. (2004) found that elevated N inputs reduced active mycorrhizal biomass,
fungal diversity and fungal:bacterial biomass ratios. These results suggested a decreased role for
mycorrhizae in fixation of N under elevated N inputs.
Abiotic transformation of inorganic N into soil organic matter has also been proposed as a possible
mechanism to explain high rates of N retention in soil, and some evidence has been presented to support
this possibility. Dail et al. (2001) observed retention of 15N03~ and 15N02 in sterile soil, but the method of
sterilization may have increased dissolved organic carbon (DOC) concentrations and artificially increased
the opportunity for formation of soluble organic N compounds. Davidson et al. (2003) developed the
ferrous wheel hypothesis to explain incorporation of inorganic N into organic matter. The hypothesized
mechanism involves conversion of N03 to N02 through oxidation of Fe2+. Testing of this hypothesis in
situ was not found in the literature, but the small amount of Fe2+ that typically occurs in the forest floor,
where presumably much of the conversion to organic N occurs, may limit the importance of this pathway.
Fitzhugh et al. (2003) showed that N02 produced in the first step of nitrification may be directly
converted to soluble organic N rather than becoming fully oxidized to N03 . However, concentrations of
introduced 15N02 in this experiment were several orders of magnitude higher than that normally seen in
forest soils. Therefore, the evidence at this time for abiotic retention of N is not fully convincing, and the
importance of this process requires further research.
In addition to our limited understanding of N retention mechanisms, there is no direct information
on ecosystem recovery from N saturation in the U.S. This may be at least partly because atmospheric
deposition of N has been relatively stable in the eastern U.S. over the past two to three decades. An
important source of information on N recovery responses has been provided by the European NITREX
study, which reduced ambient N deposition for 5 years with roofs constructed over experimental plots in
Germany and The Netherlands. At the German site, deposition was reduced from approximately 38 kg
N/ha/yr to levels that varied from 10 to 20 kg N/ha/yr. At the Dutch site, deposition was reduced from
45 kg N/ha/yr to levels that varied from 1 to 10 kg N/ha/yr. At both of these sites, deposition levels before
the experiment were approximately three to four times greater than the highest deposition levels
commonly found in the eastern U.S., whereas after the reduction, levels at the Dutch site fell within the
range of deposition in the eastern U.S. over the past two decades, and values at the German site were
somewhat higher than this range (Emmett et al., 1998; Ollinger et al., 1993). The decrease in ambient N
inputs resulted in a marked decrease in N outputs at each site within 2 to 3 years. The responses at the two
sites were somewhat different, however. At the Dutch site, outputs ofN exceeded inputs both before and
after experimental reduction of inputs. At the German site, inputs exceeded outputs before and after
reduction of inputs, but outputs were more similar to inputs after the reduction. At both sites, outputs after
the reduction in deposition remained two to three times higher than outputs commonly measured in the
eastern U.S.
Thus, atmospheric deposition of N has increased N availability in soils, which has led to increased
nitrification and associated acidification of soil and soil water. The N retention capacity of soils is
strongly dependant on land-use history, however, so the relationships between N deposition and
ecosystem N status and percent of terrestrial retention are variable. In general, however, atmospheric
deposition of 10 kg N/ha/yr or higher is required for appreciable amounts of N03 to leach to surface
waters in the eastern U.S. and northern Europe. Future projections of chemical recovery from N-driven
acidification are uncertain because retention mechanisms are not fully understood, and there are only
limited data on recovery responses. European experiments that reduced inputs of N found decreased
outputs of N within 2 to 3 years, which indicates a relatively rapid response to decreased deposition
levels. However, these studies are difficult to directly apply to the U.S. because deposition levels were
much higher at the European sites before the experiment, and the 5-yearr duration of the experiments
only demonstrated recovery to levels of N saturation that are higher than the more heavily affected sites in
the eastern U.S.
High concentrations of lake or strcamwatcr N03 . indicative of ecosystem saturation, have been
found at a variety of locations throughout the U.S., including the San Bernardino and San Gabriel
B-10
-------
Mountains within the Los Angeles Air Basin (Fenn et al., 1996), the Front Range of Colorado (Baron
et al., 1994; Williams et al., 1996a, 1996b), the Allegheny Mountains of West Virginia (Gilliam et al.,
1996), the Catskill Mountains of New York (Murdoch and Stoddard, 1992; Stoddard, 1994), the
Adirondack Mountains of New York (Wigington et al., 1996), and the Great Smoky Mountains in
Tennessee (Cook et al., 1994). All of these regions, except Colorado, received relatively high (more than
about 10 kg N/ha/yr) atmospheric deposition of oxidized N throughout the 1980s and 1990s. In contrast,
the Front Range of Colorado receives less than about 4 or 5 kg N/ha/yr of total (wet plus dry) deposition
(Sullivan et al., 2005), less than half of the total N deposition received at these other locations. The cause
of N-saturation at high-elevation western watersheds that receive low to moderate levels of atmospheric
deposition has been a subject of debate. High concentrations of N03 in surface waters in the western
U.S. are not widespread. Nitrate concentrations during the fall sampling season were low in most western
lakes sampled in the Western Lakes Survey (WLS). Only 24 sampled lakes were found to have N03
concentrations greater than 10 j^icq/L. Of those, 19 lakes were situated at high elevation, most above 3,000
m. Cold temperatures in such lakes undoubtedly play an important role in maintaining high N03
concentrations by limiting biological uptake processes. The high N03 concentrations are likely to affect
acid-base chemistry only where ANC is low. Eight lakes showed high N03 (>10 (ieq/L) and low ANC
(<50 (ieq/L), all of which occurred at elevations higher than 3,100 m. Four were located in Colorado, two
in Wyoming, and one each in California and Utah. In all cases, pH was above 6.5 and ANC was greater
than or equal to 15 (ieq/L. Such lakes are sensitive to episodic pulses of N03 acidity; such pulses have
been reported from Colorado Front Range lakes (Williams and Tonnessen, 2000). Episodic acidification
of western lakes could be important biologically.
In the Uinta Mountains of Utah and the Bighorn Mountains of central Wyoming, 19% of the lakes
included within the WLS had N03 >10 (ieq/L. This suggests that N deposition in these areas may have
exceeded the capability of these systems to assimilate N. It is unknown if these concentrations of N03
represent effects from anthropogenic sources or if this constituted a natural condition associated with
inhibited N03 assimilation in cold alpine environments.
Williams et al. (1996a; 1996b) contended that N-saturation is occurring throughout high-elevation
catchments of the Colorado Front Range. Many lakes in the Colorado Front Range have chronic N03
concentrations greater than 10 j^ieq/L and concentrations during snowmelt are frequently much higher, due
at least in part to leaching from tundra, exposed bedrock, and talus areas. Although biological N demand
may be high in subalpine forests, uptake is limited in alpine areas by large N inputs from snowmelt, steep
watershed gradients, rapid water flushing, extensive areas having little or no soil development, and
limitations on the growth of phytoplankton in some alpine lakes by factors other than N (e.g., phosphorus,
temperature) (Baron et al., 1994).
B.1.2.6. Nitrate Leaching
Nitrate leaching losses from soils to drainage waters are governed by a complex suite of ecosystem
processes in addition to N inputs from atmospheric deposition. In particular, mineralization and
nitrification processes play important roles in regulating the quantity of, and temporal variability in, the
concentration of N03 in soil solution, and consequently leaching losses from the rooting zone (Johnson
et al., 1991b, 1991c; Joslin et al., 1987; Reuss and Johnson, 1985)(Johnson et al., 1991b; 1991c). Thus,
N03 leaching is mostly under biological control and typically shows pronounced seasonal variability
(Van Miegroet and Johnson, 1993). Peak concentrations of N03 in soil solution appear to be largely
responsible for the potentially toxic peaks in Al concentration that sometimes occur in soil solution,
although S042 may also play a role by serving to elevate chronic Al concentrations (Eagar et al., 1996).
High leaching of N03 in soil water and streamwater draining high-elevation spruce-fir forests has
been documented in numerous studies in the Southern Appalachian Mountain region (Joslin et al., 1992;
Joslin and Wolfe, 1992b, 1994; Nodvin et al., 1995; Van Miegroet et al., 1992). This high N03 leaching
has been attributed to a combination of high N deposition, low N uptake by forest vegetation, and
B-11
-------
inherently high N release from soils. Forest age is another major factor-affecting uptake, with mature
forests requiring minimal N for new growth and, hence, often exhibiting higher N03 leaching that
younger, faster growing stands (Goodale and Aber, 2001). Old-growth red spruce stands in the Southern
Appalachians have been demonstrated to have significantly slower growth rates than stands younger than
120 years (Smith and Nicholas, 1999). The latter feature is associated with low C:N ratios in mineral soil,
high N mineralization potential and high nitrification (Eagar et al., 1996; Joslin et al., 1992).
In most terrestrial ecosystems in the U.S., N is strongly retained and there is limited mobility of
N03 . Exceptions to this pattern tend to occur in spatially limited regions that receive high levels of total
N deposition (higher than about 10 to 20 kg N/ha/yr) and in alpine and subalpine environments that have
little soil or vegetative development over substantial portions of the watersheds.
B.1.3. Interactions with Transitional Ecosystems
B.1.3.1. Sulfur Storage and Release in Transitional Ecosystems
Although S is generally mobile in upland soils in most parts of the U.S., wetlands act as both
sources and sinks of atmospherically deposited S. Wetlands retain and release S in response to variations
in hydrology, which in turn affect oxidation and the reduction process in wetland soils. Ito et al. (2005)
evaluated the influence of land cover types on S042 fluxes in Adirondack lake watersheds. They found
that S042 concentration in drainage water decreased in association with increased wetland area within the
lake watershed (adj. r2 = 0.58, p > 0.001). They attributed this observed pattern to dissimilatory S042
reduction in anaerobic wetland soils.
Sulfur storage in wetland soils to some degree prevents or delays the acidification of downstream
surface waters with mineral acidity. However, the water table in wetland areas typically drops during
drought conditions, and this allows development of aerobic conditions in surface wetland soils. Under
aerobic conditions, stored S is re-oxidized to S042 , which can then be rapidly mobilized under high-flow
conditions that occur in response to rainfall or snowmelt. This can cause substantial episodic pulses of
acidity in surface waters that receive drainage water from wetlands. Thus, wetlands buffer downstream
receiving waters against chronic acidity to some degree, but can be an important source of periodic
episodes of more extreme acidity.
B.1.3.2. Organic Acidity in Transitional Ecosystems and Downstream Surface Waters
Organic acids in fresh water originate from the degradation of biomass in upland areas, wetlands,
near-stream riparian zones, the water column, and stream and lake sediments (Hemond, 1994). The
watersheds of surface waters that have high concentrations of organic matter (DOC >about 400 |_iM) often
contain extensive wetlands and/or extensive organic-rich riparian areas (Hemond, 1990; Sullivan, 2000a).
Organic acids contributed by wetlands to downstream drainage waters can influence surface water
acid-base chemistry, particularly in dilute waters having moderate to high (greater than about 400 (.iM)
DOC concentrations. Organic acids in surface waters include a mixture of functional groups having both
strong and weak acid character. Some lakes and streams are naturally acidic as a consequence of organic
acids contributed to solution by wetlands. The presence of organic acids also provides buffering to
minimize pH change in response to changes in the amount of S042 and N03 derived from acidic
deposition.
There are many lakes and streams that are chronically acidic or low in ANC mainly due to the
presence of organic acids. In many cases, the principal source of these organic acids is the wetlands
within the watershed. The NAPAP (Sullivan et al., 1992) concluded that about one-fourth of all acidic
lakes and streams surveyed in the National Surface Water Survey (NSWS) (Kaufmann et al., 1988;
B-12
-------
Linthurst et al., 1986a) were acidic largely as a consequence of organic acids. A survey of 1400 lakes in
the Adirondack Mountains by the Adirondack Lake Survey Corporation (ALSC) (Kretser et al., 1989),
which included many small lakes and ponds (1 to 4 ha) having relatively high DOC, revealed that about
38% of the lakes had pH <5 due to the presence of organic acids, and that organic acids generally
depressed the pH of Adirondack lakes by 0.5 to 2.5 pH units in the ANC range of 0 to 50 j^ieq/L (Baker
etal., 1990a).
Specification of the acid-base character of water high in DOC is somewhat uncertain. Attempts
have been made to describe the acid-base behavior of organic acids using a single H+ dissociation
constant (pKa), despite the fact that organic acids in natural waters are made up of a complex mixture of
acidic functional groups. A portion (perhaps one-third) of the acidity in organic acids is quite strong, with
some ionization occurring at pH values well below 4.0 (Driscoll et al., 1994b; Hemond, 1994). A number
of modeling approaches have been used to estimate the acidity of organic acids in fresh waters, often as
simple organic acid analogs having different pKa values (Oliver et al., 1983; Perdue et al., 1984; Driscoll
etal., 1994b).
The importance of naturally occurring organic acids as agents of surface water acidification was
reinforced by a modeling study (Sullivan and Eilers, 1996) that showed that inclusion of organic acids in
the Model of Acidification of Groundwater in Catchments (MAGIC) had a substantial effect on model
predictions of surface water pH, even in waters where DOC concentrations were only moderate. MAGIC
hindcasts of pre-industrial lakewater pH of Adirondack lakes showed poor agreement with diatom
inferences of pre-industrial pH when organic acids were not considered in the MAGIC model (Sullivan
and Eilers, 1996). Revised MAGIC hindcasts of pre-industrial lakewater pH that included an organic acid
representation (Driscoll et al., 1994b) showed considerably closer agreement with diatom inferences
(Figure B-l). The mean difference between MAGIC and diatom estimates of pre-industrial pH was
reduced from 0.6 pH units to 0.2 pH units when organic acids were included in the model, and the
agreement for individual lakes improved by up to a full pH unit (Sullivan et al., 1996a).
Rosenqvist (1978) and Krug et al. (1985) hypothesized that a significant component of the mobile
acid anions contributed from atmospheric deposition (e.g., S042 , N03 ) replace organic anions that were
previously present in solution. Under this anion substitution hypothesis, the net result of acidic deposition
is not so much an increase in cations (including potentially toxic H and Aln+) as much as an exchange of
S042 and N03 anions for organic anions, with little or no change in ANC and pH. This hypothesis has
received some support from paleolimnological studies, which suggested historic decreases in DOC
concentrations during the period of lakewater acidification in the 1900s (Davis et al., 1985; Dixit et al.,
2001; Kingston and Birks, 1990). Other studies have found a decrease in organic acidity which was at
least partly attributable to the extent of organic acid protonation. David et al. (1999) measured a decrease
in organic anion concentrations in stream water in response to the experimental whole-watershed
acidification experiment at the Bear Brook Watershed in Maine. Wright et al. (1993) concluded that ANC
increases in a small watershed in Norway, where rates of acidic deposition were experimentally reduced,
were limited by the increasing role of organic acids that accompanied decreasing acid deposition load.
Complexation of organic acids by metals (Aimer et al., 1974; Cronan and Aiken, 1985; Dickson,
1978; Lind and Hem, 1975) and pH-dependent changes in dissociation of organic acids (Oliver et al.,
1983; Wright et al., 1988a) are probably important components of the organic acidity response. Loss of
DOC in response to acidic deposition can also cause a shift in Al species composition towards lesser
complexation with organic ligands. Such a shift from AL0 to Al; increases toxicity of the Al to aquatic
biota (Baker and Schofield, 1982). Changes in pH can alter the charge density of organic solutes and thus
influence organic contributions to acidity (Wright et al., 1988a, 1988b). David et al. (1999) found that the
charge density of organic acids decreased by about 1 (ieq/L/mg C at West Bear Brook in Maine, in
response to 6 years of experimental acidification, probably due to greater protonation of organic acid
anions at the lower pH. Similar results were reported by Lydersen et al. (1996) at Lake Skjervatjern in
Norway. Values of the organic acid charge density in the ALSC lakes in the Adirondack Mountains
B-13
-------
increased with increasing pH between pH values of 5.0 to 7.0 due to the presence of weakly acidic
functional groups (Driscoll et al., 1994b).
Organic Acids Not Included
M %¦
5 5.5 6 6.5 7
Diatom-inferred Pre-industrial pH
_ 7-5-
a
| 7.
I
I
£L
6.5 .
6
5.5 -
5-
4.5.
Organic Acids Not Included
3^i
T
4,5
n 1 r~
5 5.5 6 6.5 7
Diatom-inferred Pre-industrial pH
7,5
Source: Sullivan et al. (1996).
Figure B-1. MAGIC model hindcast estimates of pre-industrial pH versus diatom-inferred pH for 33
statistically selected Adirondack lakes: a) without including organic acid representation in
the MAGIC simulations, and b) including a triprotic organic acid analog model in the MAGIC
simulations.
Hedin et al. (1990) artificially acidified a small, moderately high-DOC (725 (.iM C) stream with
H2S04 at HBEF in New Hampshire. Streamwater pH (4.4) was near the range of reported average pKa
values for organic acids, suggesting that the capacity of organic acids to buffer mineral acidity should be
high. The acid loading rate was adjusted to achieve an increased streamwater S042 concentrations of
150 (ieq/L at the downstream sampling point 108 m below the point of acid addition. Adjustments were
made for dilution by soil water or inflow from small tributaries. Although streamwater DOC did not
change significantly, the concentration of organic anions (as calculated from the charge balance)
decreased by 17 (ieq/L. Thus, the overall capacity of organic anions to neutralize mineral acid inputs
offset about 11% of the added acid (Hedin et al., 1990). This experiment only considered interactions
between mineral acid and organic matter within the stream. Any additional buffering that may have been
provided within the terrestrial catchment was not represented in the experimental design. Also, any
B-14
-------
possible catchment-mediated influences of the experimental acidification on organic acid properties or
terrestrial DOC mobilization were excluded from the experiment because the acid was not applied to the
catchment soils.
Regional Sulfate Trends in LTM Network
Regional Nitrate Trends in LTM Network
100
CD
O
CO
O
70
0
60
LL
V
50
3
40
h
O
30
20
10
0
New England Lakes
AaironaacK LaKes
Appalachian Streams
Upper Midwest Lakes
Ridge and Blue Ridge Streams
f /
z 1
fi (
/'/
- s
100
90
80
c
70
50
3
40
t-
j
O
30
20
10
0
New England Lakes
Adirondacks Lakes
Appalachian Streams
ZL
Upper Michwest Lakes
Ridge/Blue Ridge Streams
/ / Jl
/ / J]
-8 -6 -4 -2
Slope of Trend (peq/L/yr)
-2 -1 0
Slope of Trend (peq/L/yr)
100
90
80
70
60
0.
a
50
3
40
fc
0
30
20
10
0
Regional ANC Trends in LTM Network
1 1
f
New England Lakes
Appalachian Streams
Ridge and Blueridge Streams
100
90
80
70
60
50
40
30
20
10
0
Regional Hydrogen Ion Trends in LTM Network
New England Lakes
Adirondack Lakes
Appalachian Streams
Upper Midwest Lakes
Ridge and Blue Ridge Streams
J
/
J
-2 0 2 4 6
Slope of Trend (peq/L/yr)
-1.0 -0.5 0.0
Slope of Trend (peq/L/yr)
100
90
80
70
60
50
40
30
20
10
0
Regional [Ca2+ + Mg2+] Trends in LTM Network
—
Mew England Lakes
Adirondack Lakes
Appalachian Streams
Upper Midwest Lakes
?idge/Blue Ridge Streams
-6 -4 -2 0 2 4
Slope of Trend (peq/L/yr)
100
on
yu
80
70
0)
u
Q-
60
a)
50
~5
40
E
3
O
30
20
10
0
Regional DOC Trends in LTM Network
New England Lakes
Adirondack Lakes
Appalachian Streams
Upper Midwest Lakes
-0.4
-0.2 0.0 0.2 0.4
Slope of Trend (mg/L/yr)
0.6
Source: Stoddard et al. (2003).
Figure B-2. Cumlative frequency diagram (distribution) of slopes for SO42", NO3", Gran ANC, hydrogen
ion, [Ca2+ + Mg2+], and DOC concentrations in LTM surface water monitoring sites, by region,
for the period 1990-2000. The Ridge/Blue Ridge Province did not have sufficient DOC data to
allow trend analysis.
B-15
-------
Results of a resurvey of 485 Norwegian lakes sampled in both 1986 and 1995 provided evidence in
support of an increase in organic acid anion concentrations in association with decreased lakewater S042
concentration (Skjelkvale et al., 1998). The organic acid anion concentration increased by an amount
equal to between 9% and 15% of the decrease in S042 concentration in the four regions of the country
most heavily affected by the decrease in S deposition during the intervening 10 year period. Lakewater
S042 concentrations decreased by 9 |_icq/L (western and northern Norway) to 20 to 21 |_ieq/L (eastern and
southern Norway). Only in mid-Norway, where average S042 concentration decreased by only 6 (ieq/L,
did the organic acid anion concentration remain unchanged between 1986 and 1995 (Skjelkvale et al.,
1998).
Recent monitoring data have shown that DOC and organic acid anion concentrations in many lakes
and streams in the U.S. have increased in association with decreased S deposition. It is likely that a high
percentage of this DOC originates from wetland soils within the monitored watersheds. This result
appears to be partly responsible for the limited lakewater ANC and pH recovery that has occurred at many
locations. The response of surface waters to changes in acidic deposition has included a general increase
in surface water DOC (Figure B-2). All regions of the eastern U.S. analyzed by Stoddard et al. (2003) that
had sufficient DOC data for analysis exhibited increases in DOC concentrations during the 1990s. All
regional trends were significant with the exception of the Northern Appalachian Plateau, the region with
the lowest median DOC concentration. The median increase in DOC of 0.05 mg/L/yr reported by
Stoddard et al. (2003) corresponds to an overall increase of about 10% across study regions, similar to
trends reported elsewhere in the northern hemisphere (Evans and Monteith, 2002; Skjelkvale et al., 2001).
This suggests a common cause. Both climate warming and decreasing acidic deposition are possible
causal agents.
B.2. Factors That Determine Ecosystem Sensitivity
B.2.1. Transitional Ecosystems
B.2.1.1. Wetlands and Peatlands
Wetlands and peatlands often contain highly acidic soils. Their acidity is mainly attributable to the
presence of large quantities of naturally occurring organic materials. Fulvic and humic acids, formed
during the breakdown of organic matter, contribute substantial organic acidity to soil and surface waters
in wetland and peatland environments. In the case of ombotrophic bogs and poor fens, there is also a
scarcity of base cations, which would serve to buffer both organic and mineral acidity.
Because wetland and peatland vegetative communities are adapted to high levels of natural organic
acidity, it is unlikely that S or N deposition would cause any acidification-related effects at levels of
acidic deposition commonly found in the U.S. Nevertheless, wetlands are closely tied to a number of
important biogeochemical processes that regulate watershed response to acidic deposition. The major
interactions are described below.
High concentrations of DOC in brownwater lakes and streams are often due to the influence of
wetlands on hydrography within the watershed. This presence of high concentrations (higher than about
500 (.iM) of DOC can substantially reduce the pH and ANC of surface waters, buffer those waters against
pH changes in response to added mineral acidity, and form stable complexes with dissolved Al, thereby
reducing its toxicity to aquatic life. Therefore, the response of surface waters to acidic deposition is
strongly influenced by the extent of upstream and shoreline wetland development.
B-16
-------
Wetlands also serve as a (sometimes-temporary) sink for atmospheric S and N. Chemical reduction
reactions and biological uptake contribute to S and N storage in wetland soils. Oxidation during drought
periods, when water levels recede, followed by flushing from wetland to downstream surface water
during subsequent storm flow, can cause substantial pulses of mineral acidity in downstream receiving
waters. On a chronic basis, the concentration of S042 (and associated acidity) in surface water can be
substantially lower as a consequence of dissimulatory S reduction in upslope wetlands. On an episodic
basis, wetlands can contribute to wide fluctuations in downstream surface water acid-base chemistry.
Such fluctuations can include pulses of acidity that may be toxic to aquatic biota.
Wetlands provide anaerobic substrate for S-reducing bacteria. These bacteria are also partly
responsible for the increased rate of mercury (Hg) methylation that is known to occur in wetlands. As a
consequence, fish in lakes drained by wetlands often have much higher concentrations of tissue methyl
mercury, as compared with fish in lakes that lack watershed wetlands (Driscoll et al., 2007b).
B.2.1.2. Ponds
The factors that determine the sensitivity of ponds to acidification from acidic deposition are
generally similar to those that determine the sensitivity of lakes (discussed in the following section). In
general, however, ponds and small lakes tend to exhibit low ANC and pH at a greater frequency than do
larger lakes (Sullivan et al., 1990b). This pattern is mainly a consequence of the higher concentrations of
DOC frequently found in ponds as compared with larger lakes. In addition, because larger bodies of water
tend to have larger watersheds, there is a greater likelihood that they will integrate conditions across a
broader landscape, increasing the possibility of receiving at least a moderate level of base cation supply
(Sullivan et al., 1990b). Thus, where lakes are acid-sensitive, it is likely that ponds are also acid-sensitive.
However, synoptic databases of pond acid-base chemistry are generally not available.
B.2.2. Streams and Lakes
Acidic deposition that falls as precipitation directly on the lake surface may eventually be
neutralized by in-lake reduction processes which are controlled in part by hydraulic residence time.
Natural hydrologic events also alter acidification and neutralization processes during snowmelt and
change flowpaths during extended droughts (Webster et al., 1990).
Leaching of base cations by acidic deposition can deplete the soil of exchangeable bases. The
importance of this response has recently been widely recognized because most watersheds are not
exhibiting much ANC and pH recovery of drainage water in response to recent large decreases in S
deposition. This limited recovery can be at least partially attributed to decreased base cation
concentrations in surface water. This understanding of the base cation response has developed slowly.
During the 1980s, the generally accepted paradigm of watershed response to acidic deposition was
analogous to a large-scale titration of ANC (Henriksen, 1984). Atmospheric inputs of acidic anions were
believed to result in movement of those anions through soils into drainage waters with near proportional
loss of surface water ANC. This view was modified by Henriksen (1984), who suggested that a modest
component of the added S042 (up to a maximum of about 40%) could be charge-balanced by increased
mobilization of base cations from soils, and the remaining 60% to 100% of the added S042 resulted in
loss of ANC in surface waters. During the latter part of the 1980s, it became increasingly clear that a
larger component (>40%) of the added S042 was in fact neutralized by base cation release in most cases
and the ANC (and therefore also pH) of surface waters typically did not change as much as was earlier
believed. This understanding developed in large part from paleoecological studies (Charles et al., 1990b;
Sullivan et al., 1990b), which indicated that past changes in lakewater pH and ANC had been small
relative to estimated increases in lakewater S042 concentrations since pre-industrial times (Sullivan,
2000a). The belief that changes in acidic deposition were accompanied mainly by changes in ANC and
B-17
-------
pH has been replaced by the realization that changes in S042 were accompanied mainly by changes in
base cations. Thus, surface waters have not been acidified as much by historical deposition as was earlier
believed. Furthermore, surface water ANC and pH should not be expected to show substantial chemical
recovery upon reduced emissions and deposition of S and N. The magnitude of the base cation response
has clearly limited the extent of surface water acidification caused by acidic deposition. However, this
same response has contributed to base cation deficiencies in some soils, with associated adverse terrestrial
effects.
B.2.3. Other Types of Ecosystems
There has been little work on the rates of atmospheric deposition to urban ecosystems despite
extensive data on concentrations and chemical reactions of air pollutants in cities (U.S. EPA, 2004).
Nevertheless, urban ecosystems are often subjected to large rates of deposition of anthropogenic
pollutants (Lovett et al., 2000a). Decades of research on urban air quality indicate that cities are often
important sources of emissions of NOx, SOx, and dust. Urban N deposition may affect nutrient cycles and
soil acid-base chemistry in vegetated areas in and around cities, but such possible effects have not been
studied sufficiently to draw conclusions about sensitivities or effects.
To determine the patterns of atmospheric deposition and throughfall in the vicinity of a large city,
Lovett et al. (2000a) measured bulk deposition, oak forest throughfall, and particulate dust at sites along a
transect within and to the north of New York City. They found that throughfall N was twice as high in the
urban areas compared with suburban and rural areas. Most of the urban dry deposition of N03 was from
gaseous NOx. Because there is limited biological uptake of throughfall N in an urban setting, it is
believed that a relatively high (but unknown) percentage of N deposited to the urban landscape leaches to
surface waters. Aquatic effects associated with N leaching from urban environments would be expected to
be most pronounced near coastal cities. This is because atmospheric deposition to near-coastal urban
environments can provide a substantial N load to estuaries and near shore oceanic environments, which
tend to be N-limited. See further discussion in ISA Section 3.3.2.4.
B.3. Distribution and Extent of Ecosystem Effects
B.3.1. Terrestrial Ecosystems
Coniferous forests, with soils that are naturally more acidic, generally have lower pH and base
saturation than soils in deciduous forests (Fernandez et al., 2003). In a paired watershed study at Bear
Brook Watershed in Maine, one watershed with mixed coniferous and deciduous species received
(NH4)2S04 corresponding to about 25 kg N/ha/yr and 29 kg S/ha/yr. After a decade of experimental
acidification, the treated watershed had 66 kg/ha/yr less exchangeable Ca2+ and 27 kg/ha/yr less
exchangeable Mg2+ than the untreated watershed (Fernandez et al., 2003). Soils under conifers (red
spruce, balsam fir [Abies balsa me a \. hemlock [Tsuga canadensis]) appeared to be more sensitive to
acidification than those under hardwoods (American beech [Fagus grandifolia], yellow birch [Betula
alleghaniensis], sugar, and red maples [Acer rubrum]). The hardwoods demonstrated no significant
effects from (NH4)2S04 addition. Differences in response to acid treatment among vegetation covers were
most pronounced in upper soil (O horizon and upper 5 cm of the B horizon). The study did not distinguish
between effects from NH4+ versus S042 additions.
B-18
-------
Kozlowski (1985) suggested that plants and soils act as sinks for SO2 deposition at low exposures
of 1 to 4 |_ig/m3. with no discernible effects on ecosystem structure at those levels. Shugart and
McLaughlin (1985) cautioned that forest responses to SO2 and other stressors are strongly controlled by
the successional dynamics of impacted forests. Thus, efforts to better understand and quantify forest
dynamics and development will be paramount to predicting chronic pollution effects.
Results of N fertilization studies have been used to infer the response of forests to atmospheric N
deposition. Such studies were reviewed by Johnson (1991) and U.S. EPA (1993a), illustrating that forests
can respond differently to periodic large pulsed fertilizer inputs, as compared with steady, low-level
inputs from atmospheric deposition. For example, multiple or continuous inputs of N may stimulate
populations of nitrifying bacteria (U.S. EPA, 1993a). This might be expected to modify the competitive
interactions between trees and microbes and affect both the forest growth response and the extent of N03
leaching and associated acidification.
In the southern Appalachian Mountains, acidification sensitivity has been evaluated for two
common tree species: red spruce (sensitive) and loblolly pine (Pinus taeda; insensitive). Dendro-
chronological analyses of tree cores collected for permanent plots in the Great Smoky Mountains National
Park (37 trees cores from low elevation [-1500 m]; 35 tree cores from high elevation sites [-2000 m]),
demonstrated a positive correlation between temporal and spatial trends in red spruce growth and acidic
deposition, with a greater response in trees on ridges than in draws. Ridges are naturally more acidified,
receive higher levels of acidic deposition, and have shallower soils with lower base saturation (Webster
et al., 2004).
Loblolly pine seems to have low susceptibility to adverse effects from acidic deposition. A
simulated acid addition experiment showed no significant effect of acidification on foliar nutrition in
loblolly pine seedlings, at application levels of 21 to 26 kg/ha SO4-S and 8 to 10 kg/ha N03 N (Baker
et al., 1994). Loblolly pines grown on old agricultural fields showed signs of N deficiency over 25 years
of growth despite atmospheric deposition of 5 to 10 kg N/ha/yr (Richter et al., 2000).
In the northeastern U.S., two species of coniferous tree (red spruce and red pine) have been shown
to be sensitive to acidification. Aber et al. (2003) reported a decrease in C:N ratio from about 35 to about
25 along an increasing N deposition gradient of 3 to 12 kg N/ha/yr across the Northeast. At the Harvard
Forest LTER site, at chronic experimental N addition levels of 50 and 150 kg N/ha/yr, Magill et al. (2004)
found 31% and 54% decreases, respectively, in red pine growth after 15 years of chronic N application.
No additive effect of S was seen after 11 years of a combined N and S treatment, with an application of
74 kg S/ha/yr and 50 kg N/ha/yr. There were no significant differences in baseline measurements between
the low N and combined N and S treatments.
Recent evidence indicates that mortality in red spruce in the southern Appalachian Mountains is not
abnormal when compared to historical rates, and that Fraser fir stands killed by the balsam woolly adelgid
(Adelges piceae) are largely being replaced by vigorous re-growth of young stands of that species (Van
Miegroet et al., 2007). To what extent spruce or fir mortality in the southern Appalachian Mountains will
be replaced with a species mix similar to that existing before the mortality remains to be seen.
At the Fernow Experimental Forest in West Virginia, (NH4)2S04 inputs of 54 kg N/ha/yr and 61 kg
S/ha/yr (application plus ambient atmospheric deposition), each about three times the ambient deposition
level, were applied to one watershed for 4 years. Few differences in soil and forest floor chemistry were
found in response to the N addition, although pH was significantly lower in the treatment watershed,
corresponding to increased base-flow concentrations of N03 and Ca2+ (Gilliam et al., 1996).
Deciduous forests show variable responses to acidification depending on the tree species present.
Along an increasing N deposition gradient in the northeastern U.S., from 4.2 to 11.1 kg N/ha/yr, Lovett
and Rueth (1999) found a twofold increase in mineralization in soils of sugar maple stands, but no
significant relationship between increased deposition and mineralization in American beech stands. This
difference might be attributable to the lower litter quality in beech stands. Thus, sugar maple appears to be
more susceptible to effects of increasing deposition and concomitant soil acidification from either direct
leaching of N03 or enhanced nitrification. For northeastern hardwoods, Aber et al. (2003) found a
B-19
-------
decrease in C:N ratio from 24 to 17 over a deposition gradient of 3 to 12 kg N/ha/yr. This decrease was
similar but less steep than the decrease seen in conifers.
Across an 800 km pollution gradient (3 to 11 kg S04 -S/ha/yr; 2 to 4 kg N03 -N/ha/yr) in northern
hardwood forests, with maples dominant, Pregitzer et al. (1992) found a 200 to 300 jj.g/g increase in foliar
S, and litter fall S content ranged from 872 to 1356 jj.g/g. While foliar N did not change across the
gradient, litterfall N was correlated with changing deposition. Pregitzer and Burton assert that their data
did not suggest a causal link between acid deposition and forest decline. Decline would be impossible to
document in the short 5-year time frame of their study. They did, however, assert that their results
supported the plausibility of altered tree nutrition across large geographic regions due to atmospheric
deposition.
B.3.2. Transitional Ecosystems
Wetlands are common in many areas that contain acid-sensitive surface waters. For example,
wetlands constitute about 14% of the land surface in the Oswegatchie/Black River watershed in the
southwestern Adirondack Mountains (Ito et al., 2005), one of the regions of the U.S. most affected by
surface water acidification from acidic deposition. There are no studies, however, that have documented
the extent or magnitude of acidification effects of S and N deposition on wetland ecosystems in the U.S.
The topic of acidification effects on wetlands is not well represented in the literature, and therefore
the distribution of ecosystem effects for these systems is not presented. Because levels of natural organic
acidity tend to be high in wetland soils and water, it is not likely that such ecosystems are affected by the
levels of acidic deposition commonly encountered in the U.S. It is more likely that atmospheric
deposition affects wetlands via nutrient N enrichment pathways. (See Discussion in ISA Sections 3.3.2.2
and 3.3.5.2) Gorham et al. (1987) hypothesized that acidic deposition to mineral-poor fens might cause
depletion of exchangeable base cations and decreased pH of soil water. This mechanism was suggested as
a possible cause of transition from mineral-poor fen to Sphagnum bog. Such an effect has not been
observed in response to acidic deposition at levels found in the U.S.
Synoptic surveys of ponded waters generally are restricted to lakes larger than 1 ha, 4 ha, or 10 ha.
Reasons for this limitation are varied and can include the perception that larger lakes are more important,
the failure of regional topographic map coverages to include the smaller lakes and ponds, and the fact that
smaller lakes tend to be much more numerous than larger lakes within the major lake districts of the U.S.
In general, if the larger lakes in a given region are sensitive to acidification, the smaller ponds would also
be expected to be sensitive. In most cases, data to demonstrate this are not available.
Ponds that have been observed to be sensitive to acidic deposition have been found in the
Adirondack Mountains in New York (Kretser et al., 1989) and the Mount Zirkel Wilderness Area located
in the Colorado Rockies (Campbell et al., 2004). Acid-sensitive ponds are likely to be found elsewhere as
well.
B.3.3. Aquatic Ecosystems
In most regions of the U.S., most lakes and streams are not highly sensitive to existing levels of
ambient air pollution. In addition, air pollution levels are generally decreasing in many parts of the
country, especially in the eastern U.S., in response to federal and state air pollution control regulations.
Therefore, the highly sensitive, and acidified, systems tend to be restricted to a relatively small percentage
of the overall aquatic resource base. There are exceptions to this generalization, such as for example in
Monongahela National Forest, WV, where a high percentage of the streams are acid-sensitive and highly
acid-affected (Sullivan et al., 2002b). Similarly, a high percentage of Adirondack lakes (and presumably
B-20
-------
also streams) are acid-sensitive and have been acidified by atmospheric deposition of S and N (Driscoll
etal., 1991).
Studies to assess relationships between atmospheric deposition loading (N and S) and the estimated
or expected extent, magnitude, and timing of aquatic acidification effects (NAPAP, 1998; U.S. EPA,
1995a; Van Sickle and Church, 1995) often employ a "weight of evidence" evaluation of the relationships
between deposition and effects, as followed by NAPAP in the Integrated Assessment (IA) (NAPAP,
1991).
Several kinds of evidence were used in the IA to assess the extent and magnitude of acidification in
sensitive regions of the U.S. These included:
¦ results of watershed simulation models that projected past or future chemical changes in response
to changes in S deposition
¦ empirical biological dose/response models
¦ improved relationships between surface water chemistry and ambient acidic deposition
¦ trend analyses of long-term monitoring chemical data in regions that have experienced large
recent changes in acidic deposition levels
¦ paleolimnological reconstructions of past water chemistry using fossil remains of algae deposited
in lake sediments
¦ results from whole-watershed or whole-lake acidification or deacidification field experiments
Evidence from each type of study contributes to understanding of the quantitative importance of
acidification and neutralization processes and their effects on the chemistry and biology of affected
ecosystems.
B.3.3.1. Status of Surface Waters - Regional Overview
In the NSS, DOC concentrations were much higher in lowland coastal streams, compared with
inland streams. National Stream Survey data also supported the hypothesis that atmospheric sources and
watershed retention of S control regional patterns in streamwater S042 concentrations. Most NSS
watersheds retained the vast majority of the total N loading from wet deposition. The 1986 data
suggested, however, that some atmospherically deposited N may have been reaching streams in the
northern Appalachians (Kaufmann et al., 1991).
In the 1990 NAPAP State of Science/Technology (SOS/T) summary, Baker et al. (1991a) identified
six high interest subpopulations that accounted for most of the U.S. surface waters that were acidic with
acidic deposition having been identified as the likely source (Figure B-3; Table B-2):
¦ Southwestern Adirondacks
¦ New England Uplands
¦ Low Si Eastern Upper Midwest
¦ Forested Mid-Atlantic Highlands
¦ Mid-Atlantic Coastal Plain
¦ Northern Florida Highlands
Stream data for the NSS was unavailable for three of these high-interest areas: the Adirondacks,
New England, and Upper Midwest. The national WSA data indicated that acidic streams in the Upper
Midwest are likely to be rare but there are acidic streams in the Adirondacks/New England region.
Specific areas of interest within the other three high-interest regions are described below.
B-21
-------
In addition to the large water chemistry databases developed by the U.S. EPA, there are also some
important supplemental databases in some regions. For example, based on results of lake surveys
conducted during the 1980s, about 70% of the known acidic lakes in Maine were either seepage type or
high elevation (Kahl et aL 1991). The Maine seepage lake dataset includes 120 of the estimated 150 lakes
in Maine that meet the following criteria: located in sand and gravel mapped by the USGS or Maine
Geological Surveys; depth at least 1 in; and area at least 0.4 ha (1 ac). Sampling was conducted in 1986—
87 and 1998-2000, and included at least one fall index sample for each lake. There were 87 lakes with
Gran ANC less than 100 jjcq/L.
Hie Maine high elevation lake dataset includes 90 lakes above 600 m elevation. Sampling was
conducted during the periods 1986-88 and 1997-99. The study included the vast majority of Maine lakes
that are at least 1 m deep and at least 0.4 ha (1 ac) in area. There were 64 lakes with Gran ANC less than
100 ^.leq/L.
LOW SILICA
EASTERN UPPER MIDWEST
southiest
AO IRQH0ACKS
D
Figure B-3. Location and percentage of acidic surface waters in U.S. high-interest subpopulations with
respect to acidic deposition effects. Estimates are for the upstream reach ends in the NSS
data. Population estimates in the Mid-Atlantic Highlands apply only to the forested
watersheds, while estimates in the eastern Upper Midwest apply only to low-silica (> 1 mg/L)
lakes. Figure taken from Figure 9-106 in the 1990 NAPAP SOS/T report.
B.3.3.2. Recent Changes in Surface Water Chemistry
Surface water acid-base chemistry monitoring throughout the eastern U.S. occurs primarily in two
EPA programs: the Temporally Integrated Monitoring of Ecosystems (TIME) project (Stoddard, 1990)
and the Long-Term Monitoring (LTM) project (Ford et aL 1993; Stoddard et al.. 1998a). Both projects
are operated in cooperation with numerous state agencies, academic institutions and other federal
agencies. Each is described below.
Hie regions represented by the LTM and TIME monitoring programs (Table B-3) are estimated to
contain 95% of the lakes and 84% of the streams in the U.S. that have been anthropogenically acidified by
acidic deposition . The Adirondacks had a large proportion of acidic surface waters (14%) in the NSWS;
from 1984 to 1987, the ALSC sampled 1,469 Adirondack lakes greater than 0.5 ha in size and estimated
B-22
-------
that many more (26%) were acidic (Driscoll et al., 1991). The higher percentage of acidic lakes in the
ALSC sample was due to inclusion of smaller lakes and ponds (1 to 4 ha in area), many of which were
acidic as a consequence of naturally occurring organic acids (Sullivan et al., 1990a). The proportions of
lakes estimated by NSWS to be acidic were smaller in New England and the Upper Midwest (5% and 3%,
respectively), but because of the large numbers of lakes in these regions, there were several hundred
acidic waters in each of these two regions.
The Valley and Ridge Province and Northern Appalachian Plateau had 5% and 6% acidic sites,
respectively. The only potentially acid-sensitive region in the eastern U.S. not assessed in the Stoddard
et al. (2003) report was Florida, where the high proportion of naturally acidic lakes, and a lack of long-
term monitoring data, make assessment of the effects of acidic deposition problematic (Stoddard et al.,
2003).
The TIME project is structured as a probability sampling. Each site is chosen statistically to be
representative of a target population. In the Northeast (New England and Adirondacks), this target
population consists of lakes with Gran ANC <100 j^icq/L, which are those likely to be most responsive to
changes in acidic deposition. In the Mid-Atlantic, the target population is upland streams with ANC
<100 (ieq/L. Each lake or stream is sampled annually, and results are extrapolated to the target
populations (Larsen and Urquhart, 1993; Larsen et al., 1994; Stoddard et al., 1996; Urquhart et al., 1998).
The TIME project began sampling northeastern lakes in 1991. Data from 43 Adirondack lakes can be
extrapolated to the target population of about 1,000 lakes having ANC <100 (ieq/L, out of a total
population of 1,830 lakes with surface area >1 ha. Data from 30 lakes representing about 1,500 lakes
having ANC <100 (ieq/L, out of a total population of 6,800 lakes, are included in the TIME program in
New England.
As a compliment to lake and stream sampling in the statistical populations of lakes in TIME, the
LTM project samples a subset of sensitive lakes and streams with long-term data, many dating back to the
early 1980s. Each LTM site is sampled 3 to 15 times per year, and the resulting data are used to
characterize the response of the most sensitive aquatic systems in each region to changing levels of acidic
deposition. In most regions, a small number of higher ANC (e.g., Gran ANC >100 j^icq/L) sites are also
sampled. Because of the long-term records at most LTM sites, their trends can also be placed in a better
historical context than those of TIME sites, where data are only available from the 1990s. Monitoring
results from the LTM project have been widely published (DeWalle and Swistock, 1994; Driscoll et al.,
1995; Driscoll and Van Dreason, 1993; Kahl et al., 1991, 1993; Murdoch and Stoddard, 1993; Stoddard
et al., 1998; Stoddard and Kellogg, 1993; Webster et al., 1993). Overall results were summarized by
Stoddard and Kellog (1993).
Monitoring data from the LTM and TIME projects were used to evaluate recent changes in lake and
stream chemistry from 1990 to 2000 in many of the sensitive areas of the eastern U.S. including New
England, the Adirondack Mountains, the Northern Appalachian Plateau, the Ridge and Blue Ridge
provinces of Virginia, and the Upper Midwest (Stoddard et al., 2003). There are also substantial numbers
of acid-sensitive streams in the Blue Ridge Province in North Carolina and portions of South Carolina
and Tennessee that have been affected by acidic deposition but that were not included in this analysis. In
general, the results of the TIME/LTM data analysis suggest that about one-quarter to one-third of the
lakes and streams that were chronically acidic in the 1980s were no longer chronically acidic in the year
2000. However many still had low ANC and were potentially susceptible to episodic acidification
(Stoddard et al., 2003).
Stoddard et al. (2003) found little evidence of regional change in the acidity status of lakes in New
England or streams in the Ridge/Blue Ridge regions. Furthermore, none of the study regions showed an
increase in the number of chronically acidic waters, even though there was a decline in base cation
concentrations and a likely increase in natural organic acidity (Stoddard et al., 2003). An important caveat
in this analysis is that changes in Gran ANC used in the analysis were based on the median change of all
sites in a region (Table B-4). However, as shown in (Table B-5), the rates of ANC increase were generally
more rapid in chronically acidic lakes with ANC less than 0 (ieq/L and streams with ANC between 0 and
B-23
-------
25 (ieq/L. If acidic sites are recovering more rapidly than the population of sites as a whole, then the
estimates of change in the number of acidic lakes and streams presented would be conservative.
While general estimates for large regions are useful in providing a broad picture of the extent and
status of surface water acidity, specific results from studies within those regions can help isolate trends
and determine the specific mechanisms that contribute to change. The following sections report on the
current status, past acidification, and potential future conditions for lakes and streams in acid sensitive
areas of the Northeast, Southeast, Upper Midwest, and Western U.S.
B.3.4. Regional Assessments
B.3.4.1. Northeastern Surface Waters
Current Status
The Adirondacks and New England are two of the most acid sensitive and intensively studied
regions in the Northeast. The glaciated soils and location downwind from emissions sources have made
these areas the subject of intense scientific study over the past four decades. Most of this research has
focused on lake ecosystems, though important stream studies have been undertaken at specific research
sites and more regional stream survey work is being conducted. As discussed below, the surface water
chemistry in these areas integrates atmospheric deposition, local geology, and upland watershed
characteristics.
Available surface water datasets for Adirondack lakes include TIME, EMAP, and ALSC, each of
which is useful for documenting chemical status and recent chemical changes. Population estimates from
the TIME dataset suggest that 13.0% of Adirondack lakes (238 lakes) were chronically acidic in the early
1990s during baseflow conditions in the summer. By applying an approximate rate of change in Gran
ANC of +0.8 (ieq/L/yrto these estimates (based on trend slopes for TIME and LTM data, (Table B-6),
Stoddard et al. (2003) projected that approximately 8.1% of the population (149 Adirondack lakes)
remained chronically acidic in 2000. This finding suggests that roughly 38% of lakes in the Adirondacks
that were chronically acidic in the early 1990s were not chronically acidic a decade later. Certain caveats
need to be included with the results of this analysis, however. Summertime baseflow sampling reflects the
least acidic conditions experienced throughout the year. In addition, LTM trends, which are based on year-
round sampling, may not be representative of trends in the summer-only sampling of the TIME program,
and the rate of change determined through the TIME program was not controlled for possible differences
in flow conditions between the two sample periods. Lastly, the ANC value of 0 (ieq/L/yr used to define
acidic waters has been shown to be below the level needed to protect aquatic ecosystems in the
Adirondack region (Baldigo et al., 2007; Lawrence et al., 2007).
A study by Driscoll et al. (2001b) used EMAP data from 1991 to 1994 to evaluate the extent of
acidic lakes in the Adirondacks for that period. The EMAP survey is a probability based survey
representative of lakes with surface area greater than 1 ha (1,812 lakes). The survey was conducted during
low-flow summer conditions, and the results therefore likely reflect the highest ANC values for the year.
Results from the survey indicate that 10% of the population of Adirondack lakes were chronically acidic
(ANC values of less than 0) and 31% were sensitive to episodic acidification (ANC values between 0 and
50) during the study period (Driscoll et al., 2001b).
The ALSC conducted a comprehensive survey of Adirondack lakes greater than 0.2 ha in surface
area between 1984 and 1987 (Kretser et al., 1989). Of the 1,489 lakes surveyed, 24% had summer pH
values below 5.0, 27% were chronically acidic (ANC <0), and an additional 21% were probably
susceptible to episodic acidification (ANC between 0 and 50) (Driscoll et al., 2007a).
For the New England region, the TIME population data indicates that 5.6% of the population (386
lakes) in New England exhibited Gran ANC <0 j^icq/L during the period of 1991 to 1994. This result is
B-24
-------
similar to the EMAP findings which indicate that 5% of lakes in New England and in the eastern Catskill
region of New York had ANC values less than 0 j^icq/L. The EMAP analysis also estimated that an
additional 10% of the population had low ANC values, between 0 and 50 (ieq/L, and were probably
sensitive to episodic acidification (Driscoll et al., 2001b).
Both TIME and LTM data from the New England region indicate that only a small increase in Gran
ANC had occurred during the reported monitoring period (+0.3 (ieq/L/yr). As a result, it is estimated that
the proportion of chronically acidic lakes decreased only 0.1% from 5.6% to 5.5% over the previous 10
years (Table B-4) (Stoddard et al., 2003).
State surveys within New England provide additional information on the variation in lakewater
chemistry across the region. In Maine, approximately 100 clearwater lakes in that state have been
classified as acidic, based on surveys conducted by the U.S. EPA and the University of Maine (Kahl et al.,
1999). An estimated 13% of the high-elevation lakes in Maine are acidic, compared to less than 1% of
Maine lakes (>4 ha) represented in EPA's Eastern Lakes Survey (ELS) (Kahl et al., 1991; Linthurst et al.,
1986a). Most acidic lakes in Maine are either seepage lakes located in sand and gravel deposits, or high-
elevation lakes located above 600 m elevation. Roughly 60% of the acidic lakes are seepage lakes. The
acid-sensitive seepage lakes are located in mapped sand and gravel deposits, are at least 1 m deep, and are
at least 0.4 ha (1 acre) in surface area. About 45 of the 150 lakes of this type in Maine are acidic (Kahl
et al., 1999).
Whereas lakes in the Adirondacks and New England have been intensively studied, there are no
published data which describe the status of streamwater acid-base chemistry at a regional scale, except for
the Catskill Mountains.
In the absence of regional streamwater studies, insights can be gained from site-specific long-term
studies in the region. The HBEF has one of the longest continuous records of precipitation and
streamwater chemistry in the U.S. Compared to model hindcast approximations, current conditions at
HBEF indicate that soil percent base saturation has decreased in response to acidic deposition and
because of accumulation of nutrient cations by forest vegetation. Further, acidic deposition has
contributed to a nearly fourfold increase in stream S042 concentration; a decrease in ANC from positive
to negative values; a decrease in stream pH to 5.0; and an increase in stream Al, largely occurring as the
inorganic form which has been shown by Lawrence at al. (2007) to be an unequivocal indication of the
effects of acidic deposition. Driscoll et al. (2001b) estimated that roughly 6% of lakes and streams in the
Northeast are considered more sensitive to acidic deposition than the stream monitored at HBEF (Driscoll
et al., 2001b).
Past Acidification
There are limited surface water data that directly document historic conditions and the response to
atmospheric deposition since the time of the Industrial Revolution (Charles, 1991). To address this gap,
scientists use sediment cores from lakes and detailed computer models to try to reconstruct past
conditions as well as understand the mechanisms that contribute to changing conditions.
Paleolimnological Studies
Paleolimnological studies use the remains of diatoms and other algae embedded in lake sediments
to reconstruct historical water chemistry. In the Adirondack Mountains and northern New England, both
diatom and chrysophyte algal remains in sediment cores have been used to evaluate patterns of past
acidification in a large number of lakes.
Major findings of the Paleoecological Investigation of Recent Lakewater Acidification (PIRLA)-I
and PIRLA-II research programs in the Adirondack Mountains suggested that: Adirondack lakes had not
acidified as much since pre-industrial times as had been widely believed before 1990; many Adirondack
lakes with ambient pH greater than 6.0 had not acidified historically, whereas many of the lakes having
pH less than about 6.0 had acidified; many of the lakes having high pH and ANC had increased in pH and
ANC since the previous century; and the average F-factor for acid-sensitive Adirondack lakes was near
B-25
-------
0.8 (Charles et al., 1990b; Sullivan et al., 1990). The results of these studies had major effects on
scientific understanding of the extent to which lakes had acidified in response to acidic deposition. The
view of surface water acidification as a large-scale titration of ANC (Henriksen, 1980, 1984) was replaced
by the realization that base cation concentrations typically changed more than ANC in response to acidic
deposition. This realization also modified expectations for chemical recovery of surface waters as acidic
deposition levels have decreased (Sullivan, 2000).
Diatom and chrysophyte reconstructions of pH and ANC for a statistically selected group of
Adirondack lakes suggested that about 25% to 35% of the Adirondack lakes that are larger than 4 ha had
acidified since preindustrial time (Cumming et al., 1992). Low-ANC lakes of the southwestern
Adirondacks acidified the most, probably due to low initial buffering capacity and high rainfall and
deposition of S and N in that area. Cumming et al. (1992) estimated that 80% of the Adirondack lakes that
had ambient pH >5.2 had experienced large declines in pH and ANC since the previous century, and that
30% to 45% of the lakes with pH between 5.2 and 6.0 had also acidified.
Cumming et al. (1994) reported the results of chrysophyte inferences of pH in recently deposited
lake sediments to assess acidification timing for 20 low-ANC Adirondack lakes. Lakes that were inferred
to have acidified since about 1900 tended to be small, high-elevation lakes with lower inferred pre-
industrial pH than the group of study lakes as a whole. These were probably the most acid-sensitive and
were the first to acidify with increasing acidic deposition. Husar and Sullivan (1991) estimated that S
deposition was about 4 kg S/ha/yr at that time. These lakes are located in the high peaks area and in the
southwestern portion of Adirondack Park. A second category of acidification response included high-
elevation lakes that were historically low in pH (<5.5) but that acidified further beginning about 1900.
The third identified type of response included lakes with pre-industrial pH in the range of about 5.7 to 6.3
that started to acidify around 1900 but showed their greatest pH change around 1930 to 1950. The final
category included lakes that were not inferred to have acidified. They had pre-industrial pH around 6.0
and are located at lower elevation.
Davis et al. (1994) conducted paleolimnological studies of 12 lakes in northern New England that
were low in pH and ANC. Past logging, forest fire, and vegetation composition in the watersheds were
estimated from oral and written historical information, aerial photographs, and tree ring analyses. Lake
sediment cores were collected and analyzed for pollen, diatoms, and chemistry to reconstruct past lake
conditions for several hundred years. All 12 lakes were historically low in pH and ANC, with diatom-
inferred pre-industrial ANC of-12 to 31 (ieq/L. The inferred pH and ANC values of the lakes were
relatively stable throughout the one to three centuries of sediment record before watershed disturbance by
Euro-Americans. From the early 19th into the 20th century, however, all of the lakes showed increased
diatom-inferred pH changes of about 0.05 to 0.6 pH units and increased diatom-inferred ANC of about 5
to 40 (ieq/L. Most of these inferred increases in pH and ANC coincided with watershed logging. For all
study lakes, recovery to pre-logging acid-base lake chemistry conditions was followed by continued
decline in pH by 0.05 to 0.44 pH units and ANC by up to 26 (ieq/L, probably because of acidic
deposition. The 12-lake mean inferred decreases in pH and ANC in response to acidic deposition were
0.24 pH units and 14 (ieq/L, respectively (Davis et al., 1994).
Modeling Studies
The most extensive regional modeling study that provides estimates of past acidification of
Adirondack lakes is that of Sullivan et al. (2006b). They modeled past changes in the acid-base chemistry
of 70 Adirondack lake watersheds, including 44 that were statistically selected to be representative of the
approximately 1,320 lake watersheds in the Adirondacks that have lakes larger than 1 ha and deeper than
1 m and that have ANC > 200 j^ieq/L. Model hindcasts were constructed using both the MAGIC and
PnET-BGC models.
Based on MAGIC model outputs extrapolated to the regional Adirondack lake population,
maximum past acidification occurred by about 1980 or 1990, with median ANC of the population under
investigation of about 61 |_icq/L (reduced from a median of 92 j^ieq/L estimated for the pre-industrial
B-26
-------
period). By 1990, 10% of the population target lakes had decreased in ANC to below -16 j^ieq/L and 25%
had ANC <28 j^ieq/L. Percentile values in 2000 illustrated limited chemical recovery (3 to 5 j^icq/L)
compared with simulated values in 1980 and 1990.
The MAGIC model simulations suggest that none of the target lakes were chronically acidic (had
ANC <0 (ieq/L) under pre-industrial conditions, but that by 1980 there were about 204 acidic Adirondack
lakes. That number decreased by an estimated 14% between 1980 and 2000. Similarly, the MAGIC model
simulations suggested that there were no Adirondack lakes having ANC <20 j^ieq/L in 1850, but by 1990
there were 263 such lakes. Many lakes (N = 191) were estimated to have had pre-industrial ANC below
50 |_ieq/L. and this estimate increased threefold by 1990, followed by a decrease to 399 lakes in 2000.
PnET-BGC model simulations generated output generally similar to results provided by MAGIC
model simulations. PnET-BGC simulations suggested that lakewater S042 . N03 . and base cation
concentrations under pre-industrial conditions were much lower than current values. In 1850, simulated
S042 concentrations in all study lakes were less than 25 j^icq/L, and the median value was about
15 (ieq/L. By 1980, the median simulated S042 concentration had increased more than sixfold to about
100 (ieq/L. Simulated lake N03 concentrations also increased markedly during that time, with the
median value increasing from about 4 j^icq/L in 1850 to 12 j^ieq/L in 1980. Simulated increases in the sum
of divalent base cation concentrations were less than for S042 concentrations, with the median value
increasing from 93 j^ieq/L in 1850 to 140 |_icq/L in 1980. This large change in S042 + N03 relative to the
change in the sum of base cation concentrations was the major mechanism driving the decreases in ANC
and pH associated with historical increases in acidic deposition.
Simulated lakewater ANC and pH and soil base saturation decreased from pre-industrial conditions
to recent times. Results from PnET-BGC suggested that the median Adirondack lake, from among the
estimated 1,320 lakes in the population larger than 1 ha that had measured recent ANC <200 |_icq/L. had
pre-industrial ANC near 80 (ieq/L; an estimated 10% of the lake population had pre-industrial ANC
<41 (ieq/L; and one-fourth had pre-industrial ANC <64 (ieq/L. Percentiles for the year 2000 suggest
decreases in S042 , N03 , and sum of base cations, and small increases in ANC since 1990 for lower-
ANC lakes.
Results from PnET-BGC suggest that none of the lakes in the Adirondack population had pre-
industrial ANC below 20 |_icq/L. By 1990, there were 289 lakes having ANC <20 |_icq/L and 217
chronically acidic (ANC > 0 j^ieq/L) lakes according to PnET-BGC simulations. There were 202 lakes in
the population simulated to have had pre-industrial ANC below 50 (ieq/L, and this number increased 2.8
times by 1980 under the PnET-BGC simulations.
PnET-BGC has also been used to characterize past conditions at streams within the HBEF.
Gbondo-Tugbawa et al. (2002) used relationships between current emissions and deposition, and
estimates of past emissions to reconstruct historical deposition conditions. Their analysis also considered
land disturbances such as logging in 1918-1920 and hurricane damage in 1938. Using this approach, they
estimated that past soil base saturation in the mineral soil (circa 1850) was approximately 20%, stream
S042 concentration was approximately 10 (ieq/L, stream ANC was about 40 (ieq/L, stream pH was
approximately 6.3, and stream Al; concentration was below 1 (imol/L (Driscoll et al., 2001b).
Recent Trends
Sulfur deposition has contributed to chronic soil and surface water acidification in the eastern U.S.
to a much greater extent than has N deposition. Nitrate concentrations in acid-sensitive drainage waters in
the eastern U.S. are generally much lower than S042 concentrations.
The concentration of S042 in precipitation has declined for about the past three decades throughout
the northeastern U.S., in response to decreased atmospheric emissions and deposition. EPA's LTM
Program has been collecting monitoring data since the early 1980s for many lakes and streams in acid-
sensitive areas of the U.S., including the Northeast. These data allow evaluation of trends and variability
in key components of lake and streamwater chemistry before, during, and subsequent to Title IV
implementation. Throughout the northeastern U.S., the concentration of S042 in surface waters has
B-27
-------
decreased substantially (Figure B-4) in response to decreased emissions and atmospheric deposition of S.
Decreased concentrations of S042 in lakes and streams of a third, or more, have been commonly
observed.
Lakewater S042 concentrations have decreased steadily in Adirondack lakes, at least since 1978
(Driscoll et al., 1995; Stoddard et al., 2003). Initially, there was not a systematic increase in lakewater pH
or ANC in response to the decreased S042 concentrations. Rather, the decline in lakewater S042 during
the 1980s was charge-balanced by a nearly equivalent decrease in concentrations of base cations in many
of the low-ANC lakes (Driscoll et al., 1995). Similar findings were reported by Stoddard and Kellog
(1993) for lakes in Vermont. F-factors for the nine ALTM lakes that showed significant declines in both
the sum of base cations (SBC) and (S042 + N03 ) concentration ranged from 0.55 to greater than 1.0,
with a mean of 0.93 (Driscoll et al., 1995). These high F-factor values for chemical recovery from
acidification were similar to results of historical acidification obtained by Sullivan et al. (1990), based on
diatom reconstructions of historical change for 33 Adirondack lakes.
Trend analysis results for the period 1982 to 1994 were reported by Stoddard et al. (1998) for 36
lakes in the northeastern U.S. having ANC > 100 j^ieq/L. Trend statistics among sites were combined
through a meta-analytical technique to determine whether the combined results from multiple sites had
more significance than the individual Seasonal Kendall Test statistics. All lakes showed significant
declining trends in S042 concentration (A S042 = -1.7 (ieq/L/yr; p > 0.001). Lakewater ANC responses
were regionally variable. Lakes in New England showed evidence of ANC recovery (A ANC =
0.8 (ieq/L/yr; p > 0.001), whereas Adirondack lakes exhibited either no trend or further acidification,
largely because of declines in base cation concentrations.
The observed changes in the concentration of N03 in some surface waters have likely been due to
a variety of factors, including climate. During the 1980s, N03 concentration increased in many surface
waters in the Adirondack and Catskill Mountains in New York (Driscoll and Van Dreason, 1993; Murdoch
and Stoddard, 1993). There was concern that northeastern forests were becoming N-saturated, leading to
increased N03 leaching from forest soils throughout the region. Such a response could negate the
benefits of decreased S042 concentrations in lake and stream waters. However, this trend was reversed in
about 1990, and the reversal could not be attributed to a change in N deposition. Nitrate leaching through
soils to drainage waters is the result of a complex set of biological and hydrological processes. Key
components include N uptake by plants and microbes, transformations between the various forms of
inorganic and organic N, and local precipitation patterns. Most of the major processes are influenced by
climatic factors, including temperature, moisture, and snowpack development. Therefore, N03
concentrations in surface waters respond to many factors in addition to N deposition and can be difficult
to predict. It is likely that monitoring programs of several decades or longer will be needed to separate
trends in N03 leaching from climatic variability in forested watersheds (Driscoll et al, 1995).
Monitoring data collected during the 1990s in the LTM and TIME projects illustrated that most
regions included in the monitoring efforts showed large declines in S042 concentrations in surface waters
over the 10 years of monitoring, with rates of change ranging from 1.5 to 3 (ieq/L/yr (Figure B-5). These
declines in lake and stream S042 concentration were considered consistent with observed declines in S
wet deposition. Surface water N03 concentrations also decreased, but only in the two regions that had
the highest ambient surface water N03 concentrations (Adirondacks and Northern Appalachian Plateau),
but were relatively unchanged in regions with lower concentrations. DOC increased in each region over
time. This finding suggests an increase in the contribution of natural organic acidity, which would
partially offset the expected chemical recovery from decreased acidic deposition.
B-28
-------
200
Sulfate concentration in Lake and Stream Water (|jeq/L)
Lake Rondaxe. NY Biscuit Brook, PA
200
100
200
1Jan81 1Jars88 1 Jan91
Long Pond, ME
100
0
200
100
1Jan81
Uari96
1Jan81 1Jan86 1Jan91
Lake Elbert, CO
1Jan96
1Jan86
1Jan91
1Jan96
100
-i 1 1 r
26Nov81 22Aug84 19May8712Feb90 8Nov92 5Aug95
200
Vandercook Lake, Wl
100
1Jan81
200
1Jan86 1Jan91
Grout Pond, VT
1Jan96
100
0
1Jan81
1Jan86
1Jan91
1Jan96
Figure B-4.
Measured concentration of S042~ in selected representative lakes and streams in six regions
of the U.S. during the past approximately 15 years. Data were taken from EPA's Long-Term
Monitoring (LTM) program.
ANC increased in the Adirondacks at a rate of about +1 (ieq/L/yr, despite a decline in surface water
base cation concentrations (Ca2+ + Mg2+. The decline in base cations offset some of the decline in S042 .
B-29
-------
and thus limited the increase in ANC or pH that occurred in response to lower S042 concentrations.
Surface water ANC and pH increased significantly in the 1990s; Alj concentrations declined slightly.
Regional surface water ANC did not change significantly in New England (Stoddard et al., 2003).
Moderate increases in surface water ANC during the 1990s reduced the estimated number of acidic
lakes and stream segments in the northeastern U.S. Stoddard et al. (2003) estimated that there were 150
Adirondack lakes in the year 2000 that had ANC less than 0 (8.1% of the lake population), compared to
13% (240 lakes) in the early 1990s.
Lakewater SO : concentrations in the most acid-sensitive Maine lakes declined by about 12% to
22% during the period 1982 to 1998 (Kahl, 1999). Only in the seepage lakes, however, was there
evidence of a small decline in lakewater acidity during that period (Table B-7).
Regional Trends, 1990-2000
{in lakes and streams)
ANC (peq/L/yr)
Hydrogen Ion (peq/L/yr)
Base Cations (peq/L/yr)
DOC (mg/L/yr)
Aluminum (|jeq/Uyr)
¦3
¦2
2
¦1
0
1
Slope of Trend
New England Lakes
Adirondack Lakes
Northern Appalachian Streams
Upper Midwest Lakes
Ridge and Blue Ridge Streams
Source: Stoddard et al. (2003).
Figure B-5, Summary of regional trends in surface water chemistry from 1990 through 2000 in regions
covered by the Stoddard et al. (2003) report.
However, evidence for reductions in lakewater ANC in seepage lakes from the mid-1980s to 1998
were based on a comparison of only two sampling points, which may have been influenced by climatic
variation. Therefore, the conclusion of decreasing acidity of seepage lakes is considered preliminary. The
seepage lakes are generally hydrologically isolated from their surrounding soil environment. They
therefore did not show a clear decreasing trend in base cation concentrations, as has been found in
drainage lakes throughout the Northeast. The high-elevation lakes, in contrast, showed small declines in
lakewater acidity during the 1980s, but that trend slowed or reversed in the 1990s (Kahl, 1999). Both the
seepage and high elevation lakes showed increased DOC concentrations of 10% to 20%, generally by
about 0.5 to 1.0 mg/L. The increase in dissolved organic matter would be expected to limit the extent of
ANC and pH recovery that would otherwise accompany the observed decreases in SOa2 concentration.
Whereas HQs- concentrations decreased during the 1990s in many lake chemistry datasets (Stoddard
et al., 2003), the high-elevation lakes in Maine continued to show high concentrations.
B-30
-------
The reference stream of the Bear Brook Watershed Study (East Bear Brook) has the longest
continuous, high-frequency data record of stream chemistry in Maine. Sulfate and N03 concentration
have both declined substantially since 1987. Base cations declined by an almost equivalent amount, and
the increase in ANC has been limited (Kahl, 1999).
Long-term stream water data from the HBEF reveal a number of changes that are consistent with
trends in lakes and streams across Europe and eastern North America (Evans and Monteith, 2002;
Stoddard et al., 2003, 1999). Stream water draining the HBEF reference watershed (Watershed 6) had a
32% decline in annual volume-weighted concentrations of S042 (-1.1 j^icq/L) between 1963 and 2000
(Driscoll et al., 2007a). This decrease in stream S042 concentration corresponds to both decreases in
atmospheric emissions of SO2 and to bulk precipitation concentrations of S042 (Likens et al., 2001). In
addition, there has been a long-term decrease in stream concentrations of N03 that is not correlated with
a commensurate change in emissions of NOx or in bulk deposition ofN03 . The long-term declines in
stream concentrations of strong acids (SO42 + N03 : -1.9 (ieq/L/yr) have resulted in small but significant
increases in pH, from 4.8 to 5.0 (Driscoll et al., 2007a). Streams at HBEF remain acidic compared to
background conditions, when stream pH was estimated to be approximately 6.0 (Driscoll et al., 2007).
The increase in stream pH has been limited because of marked concurrent decreases in the sum of base
cations (-1.6 (ieq/L/yr; Driscoll et al., 2001b).
Future Projections
MAGIC model simulations were conducted for the NAPAP IAto forecast the response of lakes and
streams in the eastern U.S. to S deposition. Results were reported by NAPAP (1991), Sullivan et al.
(1992), and Turner et al. (1992). The projected median change in lakewater or stream water ANC during
50-year simulations was similar among regions, except in the Southern Blue Ridge and Mid-Atlantic
Highlands, where acidification was delayed due to S adsorption on watershed soils. MAGIC projected
relatively small future loss of ANC in most northeastern watersheds under continued constant deposition.
These modeled changes were due to a simulated slight depletion of the supply of base cations from soils
(Turner et al., 1992).
On average, each kg/ha/yr change in S deposition was projected by MAGIC to cause a 3 to 4 j^icq/L
median change in surface water ANC. Such projected changes in ANC, while considerably smaller than
was generally thought to occur in the 1980s, nevertheless suggested widespread sensitivity of surface
water ANC to changes in S deposition throughout the northeastern U.S. (Sullivan, 2000).
Since 1990, adjustments have been made to the MAGIC model and its application method in
response to model testing using paleolimnological data (Sullivan et al., 1992, 1996a) and the results of
acidification and deacidification experiments (Cosby et al., 1995, 1996; Norton et al., 1982) and empirical
studies (Sullivan and Cosby, 1998). The net effect has been that the model projects somewhat less
sensitivity of Adirondack lakes to change in S deposition as compared with the version of MAGIC
applied in 1990 (Sullivan and Cosby, 1998).
Model projections of future acid-base chemistry under three scenarios of future atmospheric
emissions controls were presented by Sullivan et al. (2006a) and Zhai et al. (2008) for lakes in the
Adirondack Mountains to evaluate the extent to which lakes might be expected to continue to increase in
ANC in the future. Estimated levels of S deposition at one representative watershed are shown in Figure
B-6 for the hindcast period and in the future under the three emissions control scenarios. Model
simulations for 44 statistically selected Adirondack lakes using the MAGIC and PnET-BGC models were
extrapolated to the regional lake population. Cumulative distribution frequencies of ANC response
projected by MAGIC are shown in Figure B-7 for the past (1850), peak acidification period (1990), and
future (2100). Results for the future are given for each of the scenarios.
Results suggested that the ongoing trend of increasing lakewater ANC for the most acid-sensitive
lakes would not continue under future emissions and deposition levels anticipated as of 2003 (Base Case
Scenario). The numbers of Adirondack lakes having ANC below 20 and below 50 |_icq/L were projected to
increase between 2000 and 2100 under that scenario, and the number of chronically acidic Adirondack
B-31
-------
lakes (i.e., ANC less than 0) was projected to stabilize at the level reached in 2000. This projected reversal
of chemical recovery of acid-sensitive lakes was due to a continuing decline in the simulated pool of
exchangeable base cations in watershed soils.
Simulations suggested that re-acidification might be prevented with further reductions in emissions
and deposition.
Chen and Driscoll (2005b) applied the PnET-BGC model to 44 EMAP lake watersheds in the
Adirondacks. PnET-BGC was used to predict the acid-base chemistry of soils and surface waters, and to
assess the fisheries status during pre-industrial conditions (-1850) and under three future acidic
deposition scenarios.
25
m
i
1
I
1
H
1850
1900
1950
2000
2050
Year
Source: Sullivan et al. (2006b).
Figure B-6. Estimated time series of S deposition at one example watershed in the SW Adirondack
Mountains used by Sullivan et al. (2006b) as input to the MAGIC model for projecting past and
future changes in lakewater chemistry attributable to acidic deposition. Future deposition
estimates were based on three emissions control scenarios (Base Case, Moderate Additional
Controls, Aggressive Additional Controls).
B-32
-------
MAGIC Model Estimates of ANC Distribution
Adirondack Lakes with ANC < 200 peq/L
1850
1990
2100-Aggresive
2100-Base
— — — 2100-Moderate
Predicted ANC (|Jeq/L)
Source: Sullivan et al. (2006b).
Simulated cumulative frequency distributions of lakewater ANC at three points in time for the
population of Adirondack lakes.
Figure B-7.
Model hindcasts using PnET-BGC indicated that acidic deposition has greatly altered surface
waters and soils in the Adirondacks over the past 150 years, and that some ecosystems are continuing to
acidify despite decreases in S deposition. The model was applied to three future emissions scenarios: base
case, moderate emissions reductions, and aggressive emissions reductions. A case study for Indian Lake
in the Adirondacks illustrated that larger reductions in deposition caused greater decreases in S042 and
base cation concentrations in stream water and greater recovery in pH and ANC. Within the full
population of lake-watersheds, some showed decreasing ANC and pH values from 1990 to 2050 even
under the moderate and aggressive reduction scenarios. By 2050 to 2100, however, nearly all lakes
experienced increasing ANC and pH. The rate of soil base saturation regeneration increased very slowly
over the modeled time period, compared to changes in surface water chemistry. For 95% of the lake-
watersheds studied, simulated soil base saturation remained below 20% in 2100 under all emissions
scenarios.
There are few streams in the northeastern U.S. for which future acid-base chemistry status has been
modeled. One notable exception is the modeling conducted for streams at the HBEF and in the Catskill
Mountains. Calculations performed by Driscoll et al. (2003a) using the PnET-BGC model suggested that
"aggressive reductions in N emissions alone will not result in marked improvements in the acid-base
status of forest streams." For example, in response to an aggressive utility emissions control scenario
(hypothesized 75% reduction in utility N emissions beyond the CAAA), the ANC values of Watershed 6
at HBEF in New Hampshire and Biscuit Brook in the Catskill Mountains in New York were only
projected to increase by 1 and 2 (ieq/L, respectively, by the year 2050 (Driscoll et al., 2003a). Projected
changes in water chemistry in response to differing levels of N deposition were small in comparison with
model projections of variations resulting from climatic factors (Aber, 1997; Driscoll et al., 2003a).
B-33
-------
B.3.4.2. Southeastern Surface Waters
The two regions in the Southeast that were identified by Charles (1991) as containing low-ANC
surface waters are the Appalachian Mountains and Northern Florida. The Appalachian Mountain region
contains many streams that have low ANC, and it receives one of the highest rates of acidic deposition in
the U.S. (Herlihy et al., 1993). Streamwater acid-base chemistry has been extensively studied in this
region (Church et al., 1992; Herlihy et al., 1993; Sullivan et al., 2002, 2003; Van Sickle and Church,
1995).
Northern Florida contains the highest percentage of acidic lakes of any lake population in the U.S.
(Linthurst et al., 1986a, b). Most lakes in Florida are located in marine sands overlying carbonate bedrock
and the Floridan aquifer, a series of limestone and dolomite formations that underlies most of Florida.
Most of the acidic and low-ANC lakes are located in the Panhandle and north central lake districts.
The current status, past acidification and recent trends in surface waters chemistry for both the
Appalachian Mountains and northern Florida are discussed below.
Appalachian Mountains
Current Status
One of the most important processes affecting watershed acid-neutralization throughout much of
the Southeast is S-adsorption on soil. If S adsorption on soil is high, relatively high levels of S deposition
have little or no effect on stream acid-base chemistry, at least in the short-term. However, this
S-adsorption capacity can become depleted over time under continued S deposition, and this causes a
delayed acidification response.
Sulfur-adsorption varies by physiographic province. It is highest in the soils of the Southern Blue
Ridge, where typically about half of the incoming S is retained. Adsorption is lower in the Valley and
Ridge watersheds and especially in the Appalachian Plateau (Herlihy et al., 1993). In general, S
adsorption is higher in the southern portions of the Appalachian Mountain region.
The Mid-Atlantic Highlands consists of the portions of the Blue Ridge Mountains, Ridge and
Valley, and Appalachian Plateau ecoregions between the Virginia-North Carolina border and the Catskill
Mountains in southeastern New York. Acid mine drainage (AMD) is a major source of acidity to streams
in the Mid-Atlantic Highlands but in many cases is easy to identify due to the high concentrations of
S042 in the streams that are influenced by AMD (Herlihy et al., 1991). Acidic and low-ANC streams
affected by AMD were removed before analyses of acid-base chemistry population statistics.
Streams in the Appalachian Mountain portion of the mid-Atlantic region receive some of the largest
acidic deposition loadings of any region of the U.S. A compilation of survey data from the mid-
Appalachians yields a consistent picture of the acid-base status of streams. Acidic streams, and streams
with very low ANC, are almost all located in small (watershed area <20 km2), upland, forested
catchments in areas of base-poor bedrock. Acidic surface waters in this region are nearly always found in
forested watersheds because the thin soils and steep slopes that make these watersheds unsuitable for
agriculture and other development also contribute to their sensitivity to acidic deposition (Baker et al.,
1991b).
In the subpopulation of upland forested streams, which comprises about half of the total stream
population in the mid-Appalachian area, data from various local surveys showed that 5% to 20% of the
streams were acidic, and about 25 to 50% had ANC <50 j^ieq/L (Herlihy et al., 1993). NSS estimates for
the whole region showed that there were 2330 km of acidic streams and 7500 km of streams with ANC
<50 (ieq/L. In these forested reaches, 12% of the upstream reach ends were acidic and 17% had pH >5.5
(Table B-4). Sulfate from atmospheric deposition was the dominant source of acid anions in acidic mid-
Appalachian streams. In these acidic streams, the low pH (median 4.9) and high levels of Al j (median
B-34
-------
129 jj.g/L) leached through soils by acidic deposition were considered to be sufficiently high to cause
damage to aquatic biota. Acidic streams in this subpopulation typically had low DOC (mean 1.5 mg/L).
Localized studies have clearly shown that streamwater ANC is closely related to bedrock
mineralogy (Herlihy et al., 1993). Sullivan et al. (2007a) delimited a high-interest area for streamwater
acidification sensitivity within the Southern Appalachian Mountain region (Virginia/West Virginia to
Georgia) based on geological classification and elevation. It covered only 28% of the region, and yet
included almost all known acidic and low ANC (<20 |_icq/L) streams, based on evaluation of about 1,000
streams for which water chemistry data were available. They found that the vast majority of low ANC
sample streams were underlain by the siliceous geologic sensitivity class, which is represented by such
lithologies as sandstone and quartzite. Low ANC streamwater throughout the region was also found to be
associated with a number of watershed features in addition to lithology and elevation, including
ecoregion, physiographic province, soil type, forest type and watershed area.
Sulfate mass balance analyses indicated that, because of watershed S042 retention, soils and
surface waters of the region have not yet realized the full effects of elevated S deposition. On average,
based on NSS data, sites in the Blue Ridge Mountains retain 35% of incoming S042 from atmospheric
deposition. The amount of S042 retention was strongly related to ecoregion in the order, Piedmont >Blue
Ridge and Ridge & Valley >Appalachian Plateau.
Herlihy et al. (1993) believed that the observed differences were due to the effects of cumulative
loadings from atmospheric S deposition and not due to inherent ecoregional differences in the soils. They
also concluded that S retention will likely continue to decrease in the future, resulting in further losses of
stream ANC.
Both mineral acids and organic acids play important roles in the acid-base status of streams in the
Mid-Atlantic Coastal Plain (Baker et al., 1991b). Acidic streams in the New Jersey Pine Barrens (Table B-
2) are largely inorganically dominated, but most likely they were naturally organically acidic in the past.
It is uncertain what effect the addition of inorganic acids from acidic deposition has had on these low
ionic strength colored systems. Over half the streams in the Pine Barrens included in the NSS were acidic,
and virtually all (96%) had ANC less than 50 j^icq/L. Human disturbances in the Pine Barrens often result
in the alkalinization of streams (increases in pH and ANC) that alter the natural Pine Barrens aquatic
biota. Outside of the Pine Barrens in the NSS, the remainder of the acidic streams in the Coastal Plain
were all high DOC organically dominated systems. Low DOC (<4 mg/L), acidic streams have been
observed, however, in other Mid-Atlantic Coastal Plain surveys.
The Virginia Trout Stream Sensitivity Study (VTSSS) surveyed streamwater chemistry for 344
(-80%) of the native brook trout (Salvelinus fontinalis) streams in western Virginia. About half of the
streams included in the VTSSS had ANC <50 (ieq/L. In contrast, the NSS (Kaufmann et al., 1988) data
for western Virginia suggested that only about 15% of the streams in the NSS target population had ANC
<50 (ieq/L. These differences may reflect the smaller watershed size, more mountainous topography, and
generally more inert bedrock of the VTSSS watersheds, as compared with the overall NSS stream
population.
In the Appalachian Plateau of West Virginia there are two wilderness areas located in close
proximity in an area of base-poor bedrock — Dolly Sods and Otter Creek Wilderness Areas. Most streams
draining these wilderness areas are acidic or low in ANC.
In the Great Smoky Mountains National Park in North Carolina and Tennessee, Cook et al. (1994)
reported high N03 concentrations (-100 |_icq/L) in upland streams which were correlated with elevation
and forest stand age. The old growth sites at higher elevation showed the highest N03 concentrations.
This pattern could have been due to the higher rates of N deposition and flashier hydrology at high
elevation, and also decreased N uptake by trees in older forest stands. High N deposition at these sites has
likely contributed to both chronic and episodic acidification of streamwater (Flum and Nodvin, 1995;
Nodvinetal., 1995).
B-35
-------
Recent Trends
Population estimates from TIME surveys of streams in the Northern Appalachian Plateau region
suggested that 5014 km of streams (11.8% of the stream length) were acidic in 1993-94, but that only
3600 km of streams (8.5% of the stream population) remained acidic in this region in 2000. The
approximate rate of estimated change in Gran ANC in the region (Table B-5) was +0.79 (ieq/L/yr. On this
basis, Stoddard et al. (2003) estimated that roughly 3600 km of stream (8.5%) remained acidic 10 years
later. This represents about a 28% decrease in acidic stream length over the preceding decade.
Future Projections
Model projections of future changes in acid-base chemistry of streams in the southeastern U.S.
were presented by Sullivan et al. (2002, 2003, 2005). In the eight-state Southern Appalachian Mountains
region, Sullivan et al. (2005) modeled future effects of atmospheric S and N deposition on aquatic
resources. Modeling was conducted with the MAGIC model for 40 to 50 sites within each of three
physiographic provinces, stratified by stream water ANC class. Simulations were based on assumed
constant future atmospheric deposition at 1995 levels and on three regional strategies of emissions
controls provided by the Southern Appalachian Mountains Initiative (SAMI), based on the Urban to
Regional Multiscale One-Atmosphere model (Odman et al., 2002).
The NSS statistical frame (Kaufmann et al., 1991) was used to estimate the number and percentage
of stream reaches in the region that were projected to change their chemistry in response to the emissions
control strategies. There was a small decline in the estimated length of projected acidic (ANC > 0)
streams in 2040 from the least to the most restrictive emissions control strategy, but there was little
difference in projected stream length in the other ANC classes as a consequence of adopting one or
another strategy. However, projections of continued future acidification were substantially larger under a
scenario in which S and N deposition were held constant into the future at 1995 levels. Turner et al.
(1992) also reported MAGIC model simulation results that suggested substantial acidification
(-20 |_ieq/L) of aquatic systems would occur in the southeastern U.S. if deposition remained constant at
1985 levels. Those model analyses were conducted as part of the National Acid Precipitation Assessment
Program (NAPAP, 1991).
The SAMI emissions control strategies used in the modeling represented air regulatory
requirements being implemented at the time of SAMI's formation, expected reductions under recent
federal regulatory actions, and additional emissions controls applied to all emissions sectors in the eight
SAMI states. The spatial variability of these emissions controls resulted in varying estimated future
changes in S and N deposition at different locations within the SAMI region. The SAMI strategies were
designated A2, Bl, and B3. A2 is the reference strategy that represented SAMI's best estimates for air
emission controls under regulations for which implementation strategies were relatively certain at the
time of the study (about the year 2000). Emissions reductions under the A2 strategy included the acid rain
controls under Title IV of the 1990 Amendments to the CAAA, the 1-h O3 standard, NOx reductions
required under EPA's call for revised State Implementation Plans (SIPs), and several highway vehicle and
fuel reductions. The A2 strategy was applied for all eastern states and focused on the utility and highway
vehicle sectors. The Bl and B3 strategies assumed progressively larger emissions reductions, targeted
only to the eight SAMI states but covering all emissions sectors.
Streams exhibited a broad range of response to the cumulative S deposition loadings received to
date and the large simulated decreases in S deposition in the future under the emissions control strategies
(Figure B-8). Some streams showed modeled stream water S042 concentrations increasing in the future,
even while S deposition was reduced by more than two-thirds. These were mostly sites that had relatively
low S042 concentrations in 1995 (> about 50 j^ieq/L) because of S adsorption on soils. They generally
showed simulated future acidification, which was most pronounced for the A2 strategy. Other streams
were simulated to show relatively large decreases in future stream water S042 concentrations and
concurrent increases in ANC in response to the strategies, with progressively larger changes from the A2
to the B3 strategy. These tended to be streams that had relatively high concentrations of S042
B-36
-------
(>50 (ieq/L) in 1995, suggesting that they were closer to steady state with respect to S inputs and outputs.
Some streams were projected to exhibit future decreases in both S042 and NO;, concentrations but
nevertheless to continue to acidify. This response was attributed by Sullivan et al. (2004) to large
simulated decreases in base cation concentrations at these sites due to soil base cation depletion.
Legend
Recovery
{25 ueq/L)
Miles
Source: Sullivan et al. (2004)
Figure B-8. Map showing simulated changes in streamwater ANC from 1995 to 2040 in response to the
SAMIA2 emmissions control strategy, representing existing emissions control regulations.
Most simulated changes in stream water ANC from 1995 to 2040 were rather modest (Table B-8),
given the very large estimates of decreased S deposition. Few modeled streams showed projected change
in ANC of more than about 20 (ieq/L. Some of the largest changes were simulated for some of the streams
that were most acidic in 1995. For such streams, however, even relatively large increases in ANC would
still result in negative ANC stream water, and therefore little biological benefit would be expected from
the simulated improvement in chemistry. The model results suggested, however, that benefits would
continue to accrue well beyond 2040 for all strategies, even if deposition was held constant at 2040 levels
into the future.
B-37
-------
Florida Surface Waters
Current Status
According to the ELS survey conducted in 1984, 75% of the Florida Panhandle lakes were acidic at
that time, as were 26% of the lakes in the northern peninsula. Most of the acidic lakes were clearwater
(DOC <400 (imol) seepage lakes in which the dominant acid anions were CI and S042 . Most of the
acidic and low-ANC lakes are located in the Panhandle and north central lake districts. In these areas, the
Floridan aquifer is separated from overlying sand deposits by a confining layer called the Hawthorne
formation. The major lake districts are located in karst terrain, where lakes formed through dissolution of
the underlying limestone followed by movement of surficial deposits into solution cavities (Arrington and
Lindquist, 1987).
Flow of water from most of the lakes is downward, recharging the Floridan aquifer. Lake stage
varies in response to long-term trends in precipitation, and perhaps in response to groundwater
withdrawals. ANC generation in most lakes that have been studied appears to be due primarily to in-lake
S042 and N03 reduction (Baker et al, 1988; Pollman and Canfield, 1991). Retention of S042 by
watershed soils may also be important. Lakes can be highly alkaline where groundwater interacts with the
deeper aquifer. Lakes with hydrologic contributions from shallow aquifers in highly weathered sands can
be quite acidic and may be sensitive to acidic deposition.
DOC concentrations are high in many Florida lakes, but organic anions are less important than
S042 in most low-ANC and acidic lakes (Pollman and Canfield, 1991). Aluminum concentrations tend to
be very low in Florida lakes despite the high lakewater acidity because most of the Aln+ is removed from
soil solution by precipitation and ion exchange reactions within 75 cm depths (Graetz et al., 1985), and
relatively little Aln+ is transported in groundwater to lake waters.
Baker et al. (1988) reported that retention of inorganic N is nearly 100% in most Florida lakes.
ANC generation from S042 retention may approach 100 j^icq/L in some Florida lakes (Pollman and
Canfield, 1991). These in-lake processes are important for generating ANC. Base cation deposition and
NH4+ assimilation can also influence the acid-base status of lakes in Florida.
The Northern Florida Highlands high interest area identified by Baker et al. (1991a) consists of the
northern portion (north of 29°N latitude) of the Central Lake District and the Florida Panhandle (Figure
B-3). Acidic streams were located in the Florida Panhandle and were mildly acidic (mean pH 5.0) and
extremely dilute, with very low sea salt-corrected SBC (mean 21 j^ieq/L) and sea salt corrected S042
concentrations (mean 16 j^ieq/L). One-fourth of these acidic Panhandle streams were organic-dominated
but the remaining sites all had DOC <2 mg/L and were inorganically dominated. Inorganic monomeric Al
concentrations in these acidic streams were very low (mean 11 (ig/L). In these low DOC, low ANC
Panhandle streams, it was suggested that the degree of S042 and N03 retention was an important control
on streamwater ANC (Baker et al., 1991a).
Past Acidification
Considerable research has been conducted on past acidification in Florida lakes. Historical analyses
of lake chemistry (Battoe and Lowe, 1992), inferred historical deposition (Hendry and Brezonik, 1984;
Husar and Sullivan, 1991), and paleolimnological reconstructions of lake pH (Sweets, 1992; Sweets et al.,
1990) suggest evidence that some Florida lakes have acidified in response to acidic deposition. However,
the role of acid deposition in lakewater acidification is not entirely clear (Pollman and Canfield, 1991),
and the interpretation is complicated by regional and local changes in land use and hydrology (Sullivan,
2000).
An alternative explanation for the apparent acidification of some lakes is the regional decline in the
potentiometric surface of the groundwater (Sweets et al., 1990). Large groundwater withdrawals of the
Floridan aquifer for residential and agricultural purposes may have reduced groundwater inflow of base
cations into seepage lakes, and therefore caused less buffering of acidity. Other land use changes may
B-38
-------
have increased lake pH by providing inputs of fertilizer, which would increase lake productivity.
Paleolimnological evidence of this effect was provided by Brenner and Binford (1988) and Deevey et al.
(1986).
It is likely that lake chemistry in Florida has been heavily effected by land use changes. More than
half of the Florida lakes included in the ELS showed evidence of disturbance based on deviations from
expected geochemistry (Pollman and Canfield, 1991). Such effects complicate efforts to determine the
role of acidic deposition in controlling lakewater acid-base chemistry.
Diatom-inferred pH reconstructions of lakewater chemistry of six seepage lakes in Florida were
calculated as part of the PIRLA-I project and reported by Sweets et al. (1990). An additional 10 seepage
lakes were cored as part of PIRLA-II, and results of those analyses were reported by Sweets (1992).
Paleolimnological study lakes are located in the Panhandle, the Trail Ridge Lake District, and Ocala
National Forest, generally in terraces of highly weathered loose sand that were deposited on top of the
clay-confining layer.
Of the six lakes analyzed in PIRLA-I, two (Barco and Suggs) were inferred to have acidified since
1950 (Sweets et al., 1990). The timing of the onset of inferred acidification correlated with increases in
S02 emissions and S deposition, which increased consistently between about 1945 and 1985 (Husar and
Sullivan, 1991).
Of the 16 Florida seepage lakes studied in the PIRLA-II projects, 5 were located in or near the Trail
Ridge Lake District, and all showed diatom-inferred acidification of at least 0.2 pH units (Sweets, 1992).
Lakes located in the Panhandle region and Ocala National Forest generally did not show evidence of
recent acidification. The exception was Lake Five-O, which was inferred to have decreased by 2 pH units.
However, the diatom data suggested that this pH decline was associated with a sudden change in
chemistry, probably caused by a catastrophic disturbance such as sinkhole development, rather than by
acidification from atmospheric deposition (Pollman and Sweets, 1990; Sweets, 1992).
B.3.4.3. Upper Midwest
The Upper Midwest contains numerous lakes created by glaciation. The region has little
topographic relief and with its deep glacial overburden, it also has little or no exposed bedrock. Acid-
sensitive surface waters in the Upper Midwest are mainly seepage lakes (Eilers, 1983). Most drainage
lakes and some of the seepage lakes in the Upper Midwest region receive substantial inflow from
groundwater, which is generally high in base cation concentrations from dissolution of carbonate and
silicate minerals. Relatively high concentrations of base cations in these lakes make them insensitive to
acidification from acidic deposition. The seepage lakes that have low base cation concentrations, and that
are therefore more acid-sensitive, generally receive most of their water input from precipitation directly
on the lake surface (Baker et al., 1991b).
Current Status
Based on the ELS survey, the Upper Midwest has a large population of low ANC lakes, but
relatively few chronically acidic (ANC > 0) lakes (Linthurst et al., 1986a; 1986b). Acidic lakes in the
Upper Midwest are primarily small, shallow, seepage lakes that have low concentrations of base cations
and Al and moderate S042 concentrations. Organic anions, estimated by both the Oliver et al. (1983)
method and the anion deficit, tend to be less than half the measured S042 concentrations in the acidic
lakes (Eilers et al., 1988), but much higher in many of the drainage lakes that are less sensitive to
acidification from acidic deposition.
Groundwater flow-through lakes in the Upper Midwest can be identified on the basis of having Si
concentration greater than about 1 mg/L (Baker et al., 1991b). They generally have high pH and ANC,
due to groundwater inputs of base cations (e.g., (Baker et al., 1991b). Based on results from the ELS
survey, only 6% of these lakes had ANC >50 j^ieq/L and none were acidic. Groundwater recharge lakes
B-39
-------
(those having Si concentration less than 1 mg/L) constituted 71% of the seepage lakes in the Upper
Midwest, and were more frequently low pH and ANC. Five percent were acidic and 9% had pH >5.5.
Nearly 90% of Upper Midwestern lakes that had ANC > 50 |_icq/L were in this category (Baker et al.,
1991b). Such lakes tend to be susceptible to acidification from acidic deposition.
Sullivan (2000) summarized patterns in lakewater chemistry across the Upper Midwest from the
ELS survey. Lakewater pH, ANC, base cations, and DOC all decreased from west to east across the
region. Lakewater S042 concentrations did not show a comparable change, despite a substantial increase
in wet S042 deposition from Wisconsin to Michigan. Cook and Jager (1991) attributed the absence of a
more pronounced gradient in lakewater S042 concentration across the region to watershed sources of S in
Minnesota and high anion retention in seepage lakes, which predominate in the eastern portion of the
region. The retention of S042 by dissimilatory reduction is generally high for seepage lakes because of
their long hydraulic retention times (xw). For example, an Upper Midwestern seepage lake with mean
depth of 3 m and hydraulic retention time of 7.5 years would be expected to lose about 50 (ieq/L of S042
from the water column by this process (Cook and Jager, 1991).
Lakewater concentrations of inorganic N reported by the ELS were low throughout the Upper
Midwest. In addition, snowmelt would not be expected to provide any significant N03 influx to lakes in
the Upper Midwest because most snowmelt infiltrates the soil before reaching the drainage lakes, and
because snowmelt input of N into seepage lakes would be limited mainly to the snow on the lake surface
and immediate near-shore environment. Aluminum concentrations are far lower in the Upper Midwest
than in lakes of similar pH in the Northeast.
Wetlands are common throughout the Upper Midwest. They contribute to high production of
organic matter which is reflected in high DOC concentrations in many lakes. Despite the abundant
wetlands, S042 is the dominant anion in the low-ANC (> 50 j^ieq/L) groundwater recharge seepage lakes.
Base cation production is the dominant ion-enrichment process in most Upper Midwestern lakes.
Even in low-ANC groundwater-recharge seepage lakes, base cation production accounts for 72% to 86%
of total ANC production (Cook and Jager, 1991).
Past Acidification
Space-for-time substitution analysis was used to infer the general levels of past change in lake
water acid-base chemistry in the Upper Midwest. Such an analysis assumes that study lakes were
generally similar in acid-base chemistry before the onset of acidic deposition and that the only substantial
driver of recent change in acid-base chemistry has been the level of acidic deposition. Across an
increasing S depositional gradient from eastern Minnesota eastward to eastern Michigan, ANC expressed
as (HCO3 - H ) decreased and the ratio S042 to SBC increased in the groundwater recharge seepage
lakes. In Michigan and Wisconsin, many lakes had S042 >SBC, indicating that the acidity was due to
high S042 relative to SBC concentration. There were also many lakes that had high concentrations of
DOC, and organic acidity probably accounted for many of these lakes having ANC <0. The spatial pattern
in (HCO3 - H ) could not be attributed to DOC, which generally showed a decreasing trend with
increasing acidic deposition.
The concentration of lakewater (Ca2+ + Mg2+) also decreased with increasing acidic deposition,
probably due to lower levels of base cation deposition and greater amounts of precipitation in the eastern
portion of the region. Atmospheric deposition is an important source of base cations for groundwater
recharge seepage lakes because of minimal groundwater inputs. In the eastern portion of the region, such
lakes are more sensitive to pH and ANC depression in response to either elevated S042 or DOC. The
spatial patterns for low ANC groundwater recharge lakes in the Upper Midwest are consistent with the
following hypotheses (Sullivan et al, 1990a; Sullivan, 1990, 2000a):
1. Sensitivity to mineral and organic acidity increased from west to east because of decreasing
lakewater base cation concentrations, and this may have been due, in part, to changes in base
cation deposition and precipitation volume along this gradient.
B-40
-------
2. High concentrations of DOC were responsible for the acidic conditions in some of the lakes, and
DOC may have decreased in response to acidic deposition.
3. Many of the lakes in eastern Michigan, and some in Wisconsin, were acidic because of high S042
relative to base cation concentration, and had probably been acidified by acidic deposition.
Diatom-inferred pH reconstructions were completed for 15 lakes in the Upper Midwest region, and
summarized by Kingston et al. (1990) and Cook and Jager (1991). Four lakes, all of which had measured
pH <5.7, showed a diatom-inferred pH decline of 0.2 to 0.5 pH units during the preceding 50 to 100
years. Diatom-inferred pH increased in one lake by 0.2 pH units. No change was inferred for the other 10
lakes, including 4 lakes with pH >6.0. No major, recent, regional acidification was indicated by the
diatom-inferred pH reconstructions. Inferred changes in most lakes were small, and were no greater
during the industrial period than during the pre-industrial period (Sullivan, 2000a).
Although diatom data suggested that some Upper Midwestern lakes may have acidified since pre-
industrial times, there is little paleolimnological evidence indicating substantial widespread acidification
in this region (Cook and Jager, 1991; Kingston and Birks, 1990). Land use changes and other human
disturbances of Upper Midwestern lakes and their watersheds have probably exerted more influence on
the acid-base chemistry of lakes than has acidic deposition (Kingston and Birks, 1990; Sullivan et al,
1990a; Sullivan, 1990, 2000a). The portion of the region most likely to have experienced acidification
from acidic deposition is the Upper Peninsula of Michigan, where acidic seepage lakes are particularly
numerous (Baker et al., 1991b); acidic deposition is highest for the region, and the |S042 |/| SBC | ratio is
commonly >1.0. The percentage of acidic lakes in the eastern portion of the Upper Peninsula of Michigan
(east of longitude 87°) was estimated to be 18% to 19% in 1984 (Eilers et al., 1988; Schnoor et al., 1986).
Recent Trends
Regional trend values for long-term monitoring lakes during the period 1990 to 2000 suggested
that S042 declined in lakewater by 3.63 (ieq/L/yr, whereas lakewater N03 concentrations were relatively
constant. The large decrease in S042 concentration was mainly balanced by a large decrease in base
cation concentrations (-1.42 (ieq/L/yr) and an increase in ANC (+1.07 (ieq/L/yr). All of these trends were
significant at p <0.01 (Stoddard et al., 2003). In the Upper Midwest, an estimated 80 of 251 lakes that
were acidic in the mid-1980s were no longer acidic in 2000. This change may be due to reduced levels of
S deposition (Stoddard et al., 2003).
B.3.4.4. West
Portions of the mountainous West contain large areas of exposed bedrock, with little soil or
vegetative cover to neutralize acidic inputs. This is particularly true of alpine regions of the Sierra
Nevada, northern Washington Cascades, the Idaho batholith, and portions of the Rocky Mountains in
Wyoming and Colorado. However, the percentage of exposed bedrock in a watershed does not always
indicate acid-sensitivity. If the bedrock contains even small deposits of calcareous minerals or if physical
weathering such as that caused by glaciers causes a high production of base cations within the watershed
(Drever and Hurcomb, 1986), surface waters may be alkaline, and are not sensitive to acidification from
acidic deposition.
The areas that are sensitive to adverse effects from acidic deposition in the western U.S. form two
nearly continuous ranges, the Sierra Nevada, which extends through most of the length of California, and
the Cascade Mountains, which extend from northern California to northern Washington (see Figure B-9.).
The sensitivity of the Rocky Mountains varies widely because the ranges are discontinuous with highly
variable geological composition. For that reason, assessments of the sensitivity of Rocky Mountain
aquatic resources to acidification should be specific to individual ranges (Turk and Spahr, 1991).
B-41
-------
ftorlh 4C
Washington Selkirk Cabinet
Cascades Mountains Mountains
1
Olympic
Law's Rance
O'UGrioot
Range \
AmcoMta-PinUat
Low!ends
Mountains
Oeartooth
. Uplift
nitris
Uigho/n
fliVi"
utom
Yellowstone
t'laieau
Cascades
Salmon River
Mountains
Sawiootn
Mountains
Van (re
Medicine Row
flange
Marge
0^7
Wasatch
Mountains
front
Range.
Uinta
A,chy . up!»I
Sta/rn
'ievfila
San Juan
Mountains
Klamath
Mountains
/
lv California
Cascades
S* watch Uph/t
^ Sangrc Da Cttsto
' Uplift
Source: Landers et al. (1987).
Figure B-9, Major geomorphic units and locations of lakes sampled in the Western Lake Survey. Those
areas known to contain sensitive lake resources are shaded with cross-hatching.
The NAPAP SOS/T Reports and the IA (NAPAP, 1991) provided only a cursory treatment of
aquatic effects issues in the West, largely because it was well known that atmospheric deposition of S and
N were generally low compared to highly affected areas in the East (Sullivan. 2000a) and because results
from the WLS (Landers et al., 1987) indicated that there were virtually no acidic (ANC > 0) lakes in the
West. NAPAP (1991) indicated, however, that high-elevation areas of the West contained many of the
watersheds most sensitive to the potential effects of acidic deposition.
Because of the proximity of western urban and industnal pollution sources to individual mountain
ranges, it is important to consider emissions in the immediate vicinity of sensitive resources as well as
regional emissions. Atmospheric deposition in the far western ranges (i.e., Sierra Nevada, Cascade
Mountains) is largely influenced by local emissions, particularly emission sources to the west (upwind) of
sensitive resources. In the Rocky Mountains, deposition chemistry is often influenced by a more complex
collection of emission sources. For example, in the Mt. Zirkel Wilderness of northwestern Colorado,
S042 and NO.i in the snow appeared to originate largely from sources in the Yampa Valley, about 75 km
to the west (Turk, 1992). Rocky Mountain National Park is affected by emissions from the Front Range to
the southeast.
B-42
-------
I
z
Jl
A
-90 -60 -30 0 30 60
-30 0 30 60 -90 -60 -30 0 30 60
(J) O
01 0
ID ~
-90-60 -30 0 30
L
-90 -60 -30 0 30 60 -90 -60 -30 0 30
JL.
-90 -60 -30 0
Jt
,k
-90 -60 -30 0 30 60 -90 -GO -30 0 30 60
Projected Percent Change in Total Deposition from 1995 to 2040
Source: Sullivan et al. (2004).
Figure B-10. Estimated percent changes in the total deposition of sulfur, reduced nitrogen, and nitrate-
nitrogen at MAGIC modeling sites from 1995 to 2040 under each of the emissions control
strategies
The acid base chemistry of lake and stream waters in Rocky Mountain National Park appears to be
primarily a function of the interactions among several key parameters and associated processes:
atmospheric deposition, bedrock geology, the depth and composition of surficial deposits and associated
hydrologic flowpaths, and the occurrence of soils, tundra, and forest vegetation (Sullivan, 2000a).
Potential biological effects of acidic deposition on lakes in the Rocky Mountains are primarily attributable
to acidification from high NO;, concentration. In general, such effects tend to be episodic, rather than
chronic. Highest NO;, concentrations in both precipitation and surface waters are found above timberline
in Colorado, where biological activity, and therefore NO; uptake, by terrestrial and aquatic biota is
lowest.
Current Status
The available information on acid-base chemistry of surface waters in the West is based mostly on
synoptic data from the WLS (Landers et al., 1987) and some more localized studies. Acid anion
concentrations in most western lakes are low during fall, but can be higher during snowmelt (Williams
and Melack, 1991). Available data from intensive study sites in the West (e.g., Loch Vale, CO, Emerald
Lake Basin, CA, and the Glacier Lakes Watershed, WY) suggest that episodic depression of stream pH
may be more pronounced than for lakes. However, there are no available systematic regional stream
chemistry data with which to assess regional sensitivity of streams to acidic deposition.
In most western lakes concentrations of S042 are low, although watershed sources of S are
substantial in some cases (Table B-9). Turk and Spahr (1991) presented a conceptual model for expected
S042 distributions in lakewaters in the West that can be used as an aid in identifying the proportion of
watersheds with significant watershed sources of S. Considering that atmospheric sources can account for
generally <30 j^icq/L of S042 in the West, it appears that many lakes, particularly in Colorado, receive
variable amounts of watershed S (Sullivan, 2000a) (Figure B-10).
B-43
-------
Nitrate concentrations in most WLS lakes were near 0 during fall (Landers et al., 1987), although
fall N03 concentrations were high in some cases (Table B-10). For example, nearly one fourth of the
lakes in northwest Wyoming had N03 >5 j^ieq/L and almost 10% had N03 >10 j^ieq/L (Table B-10). In
both the Sierra Nevada and Colorado Rockies subregions, about 10% of the lakes had fall N03
concentrations above 5 j^icq/L (Table B-10).
It is important to note that even low to moderate concentrations of N03 in western lakes might be
significant in view of: the low base cation concentrations in many lakes; potential for continuing N
deposition to eventually exhaust natural assimilative capabilities; and the fact that these distributions are
based on fall data. Time-intensive discharge and chemical data for two alpine streams in Loch Vale
watershed identified strong seasonal control on streamwater N03 concentrations (Campbell et al., 1995).
In spite of the paucity of soil cover, the chemical composition of streams is regulated much as in typical
forested watersheds. Soils and other shallow groundwater matrices such as boulder fields are more
important in controlling surface water chemistry than their abundance would indicate. Spring streamwater
N03 concentrations ranged to 40 (ieq/L, compared with summer minimum values near 10 (ieq/L. Elution
of acidic waters from snowpack along with dilution of base cations originating in shallow groundwater
caused episodes of decreased ANC in alpine streams (Campbell et al., 1995). A subalpine stream in the
same watershed similarly displayed decreased ANC and elevated monomeric Al concentrations during
snowmelt over 2 years of intensive sampling, related to elevated concentrations of S042 , N03 . and DOC
(Denning et al., 1991). Limited data collected during snowmelt suggest that spring concentrations could
be several times higher than samples collected during the fall (e.g., Reuss et al., 1995).
The surface water chemistry data for the West indicate that the Sierra Nevada and Cascade
Mountains constitute the mountain ranges with the greatest number of sensitive resources (Table B-9 and
Table B-10). Lakes in the Sierra Nevada are especially sensitive to effects from acidic deposition because
of the predominance of granitic bedrock, thin acidic soils, large amounts of precipitation, coniferous
vegetation, and dilute nature of the lakes (Melack et al., 1985; Melack and Stoddard, 1991). Surface
waters in this region are among the poorly buffered surface waters in the U.S. (Landers et al., 1987;
Melack and Stoddard, 1991). The hydrologic cycle is dominated by the annual accumulation and melting
of a dilute, mildly acidic (pH about 5.5) snowpack.
During the 1980s, an Integrated Watershed Study (IWS) was conducted at seven lakes in the Sierra
Nevada, including Emerald Lake and surrounding watersheds (-3,000 m elevation) to determine the
effects of acidification on surface waters (Tonnessen, 1991). Three lakes (Lost, Pear, and Emerald) had
volume-weighted mean ANC in the range of 15 to 30 j^icq/L. Moderate ANC waters (Topaz, Spuller, and
Marble Fork) exhibited mean ANC in the range of 30 to 50 j^ieq/L. Crystal and Ruby Lakes had mean
annual ANC >50 j^ieq/L.
Many Cascade and Rocky Mountain lakes are also highly sensitive to potential acidic deposition
effects (Nelson, 1991; Turk and Spahr, 1991). It does not appear that chronic acidification has occurred to
any significant degree, although episodic acidification has been reported for lakes in the Colorado Front
Range (Williams and Tonnessen, 2000). The data that would be needed for determining the extent and
magnitude of episodic acidification have not been collected to a sufficient degree in acid-sensitive areas
of the West to support regional assessment of episodic acidification (Sullivan, 2000a).
Along the eastern edge of the Continental Divide in Colorado and southeastern Wyoming,
Musselman et al. (1996) conducted a synoptic survey of surface water chemistry in the mountainous areas
that are exposed to relatively high (by western standards) deposition of N. A total of 267 high-elevation
lakes situated in watersheds having a high percentage of exposed bedrock or glaciated landscape were
selected for sampling. None of the lakes were chronically acidic (ANC <0), although several had ANC
<10 (ieq/L, and more than 10% of the lakes had ANC <50 |_ieq/L. The WLS data for lakes in Colorado
and Wyoming demonstrate that surface waters in this area had fall concentrations of N03 in the range of
10 to 30 (ieq/L, and likely had substantially higher N03 concentrations during spring.
The weight of evidence suggests many high-elevation lakes in the West receive N deposition
sufficiently high to cause chronic N03 leaching, and likely some degree of associated chronic and
B-44
-------
episodic acidification. However, existing data are insufficient to make a conclusive determination
(Sullivan, 2000a).
Past Acidification
The limited paleolimnological data available for lakes in the western U.S. suggest that widespread
chronic acidification probably has not occurred. Some lakes may have experienced recent pH declines,
but the magnitude of such changes has likely been small (Sullivan, 2000a).
In the Sierra Nevada, paleolimnological reconstructions of lakewater pH and ANC were calculated
by Holmes et al. (1989) at 24 depth intervals at Emerald Lake, for the period 1825 to the present.
Significant trends were not found for either pH or ANC, and the authors concluded that Emerald Lake had
not been acidified by acidic deposition. Whiting et al. (1989) completed paleolimnological analyses of
three additional lakes in the Sierra Nevada. Eastern Book Lake (pH = 7.06) showed evidence of both
long-term alkalization (-0.3 pH units over the past 200 years) and pH fluctuations since 1970. Lake 45
(pH = 5.16) may have acidified slightly (-0.2 pH units) over the last 60 years. Lake Harriet (pH = 6.52)
showed no significant change.
In Rocky Mountain National Park, Colorado, Baron et al. (1994) investigated metal stratigraphy,
diatom stratigraphy, and inferred pH profiles of four subalpine lakes. They found no evidence of historical
influence on pH attributable to atmospheric deposition. Other paleolimnological studies of Rocky
Mountain lakes report similar results: metals (primarily lead) exhibit temporal dynamics related to the
increase and decline of precious metal mining in the region, but these are asynchronous with other metal
or biological indicators of acidification (Wolfe et al., 2003). Both the study by Wolfe et al. (2003) and a
study by Saros et al. (2003) showed no evidence of acidification of lake waters over time, but increasing
evidence of eutrophication from atmospheric N deposition (see Annex C).
DayCent-Chem, a model that simulates the daily dynamics of plant production, soil organic matter,
cation exchange, mineral weathering, elution, stream discharge, and stream solute concentrations, was
able to recreate daily stream chemistry dynamics over 13 years for an alpine watershed in the Colorado
Front Range (Hartman et al., 2007). Using the model to hindcast stream chemical dynamics back to 1900
revealed changes in simulated pH coincident with maximum SO2 emissions in the late 1960s and early
1970s. Model simulations suggested annual mean pH values decreased to 5.6 to 5.8 during the years of
maximum regional SO2 emissions, and have since recovered to circumneutral values. Simulated ANC
responded to both SO2 and NOx emissions, decreasing to annual values of 20 to 25 j^icq/L during years of
highest SO2 or NOx emissions compared with current mean annual ANC values near 50 j^ieq/L Hartman.
Recent Trends
Limited monitoring data are available on recent trends in surface water chemistry in the western
regions and are mostly limited to the recent past and a number of reconnaissance studies (Melack and
Stoddard, 1991; Nelson, 1991; Turk and Spahr, 1991). Existing information on recent trends in surface
water chemistry since the 1980s suggests that conditions vary widely across the West. Parts of Colorado,
Wyoming and the western Cascades showed decreased ANC, while Emerald Lake experienced reduced
N03 concentrations.
Turk et al. (1993) reported the results of 5 years of monitoring for ten lakes in the Mt. Zirkel and
Weminuche Wilderness areas in Colorado. Based on lake concentrations of S042 and CI and on wet
deposition concentrations of S042 , N03 . and H . Turk and Spahr (1991) concluded that low-ANC lakes
had lost no more than 5 j^icq/L ANC in the Bitterroot Range of the Northern Rocky Mountains, 12 j^ieq/L
ANC in the Wind River Range of Wyoming, and 10 (ieq/L ANC in the Front Range of Colorado. It is
likely that the actual ANC losses had been much less than these estimates (Sullivan, 2000a).
B-45
-------
Future Projections
The DayCent-Chem model was used to project a timeline to acidification for an alpine watershed
of Rocky Mountain National Park (Hartman et al., 2007). At current levels of N deposition of 4 to 6 kg
N/ha/yr, acidification does not occur over 48 years of simulation, but increasing deposition amounts lead
to first episodic acidification overtime at deposition of 7.0 to 7.5 kg N/ha/yr. MAGIC model simulation
results suggested that a sustained N deposition load of 12.2 kg N/ha/yr would be required over a period of
50 years to cause chronic acidification of the Andrews Creek watershed in Rocky Mountain National Park
(Sullivan et al., 2005).
B.3.4.5. Temporal Variability in Water Chemistry
Water chemistry changes on both intra-annual and inter-annual time scales in response to changes
in environmental conditions. Because of this variability, many years of data are required to establish the
existence of trends in surface water chemistry. Assignment of causality to changes that are found to occur
is even more difficult.
Temporal variability in surface water and soil solution chemistry, and patterns in nutrient uptake by
terrestrial and aquatic biota, influence acidification processes and pathways. Thus, conditions are
constantly changing in response to episodic, seasonal, and inter-annual cycles and processes. In particular,
climatic fluctuations that govern the amount and timing of precipitation inputs, snowmelt, vegetative
growth, depth to groundwater tables, and evapoconcentration of solutes influence soil and surface water
chemistry and the interactions between pollution stress and sensitive aquatic and terrestrial biological
receptors.
Decreases in pH with increases in flow are nearly ubiquitous in drainage waters throughout the
U.S. (Wigington et al., 1991). Chemical changes during episodes are controlled in part by acidic
deposition and in part by natural processes, including dilution of base cation concentrations, nitrification,
flushing of organic acids from terrestrial to aquatic systems, and the neutral salt effect. Episodic
acidification pulses may last for hours to weeks, and sometimes result in depletion of ANC in acid-
sensitive streams and lakes to negative values and concomitant increases in Al; in solution to toxic levels.
During episodes, which are driven by rainstorms and/or snowmelt events, both discharge
(streamflow volume per unit time) and water chemistry change, sometimes dramatically. This is important
because streams may in some cases exhibit chronic chemistry that is still suitable for aquatic biota, but
nevertheless experience occasional episodic acidification with lethal consequences (Wigington et al.,
1993).
The most important factor governing watershed sensitivity to episodic acidification is the pathways
followed by snowmelt water and stormflow water through the watershed. These pathways determine the
extent of acid neutralization provided by the soils and bedrock in that watershed. High-elevation
watersheds with steep topography, extensive areas of exposed bedrock, deep snowpack accumulation, and
shallow, base-poor soils tend to be most sensitive to episodic acidification.
Rainfall and snowmelt typically pass through the soil profile before reach a stream channel. The
typical soil profile in acid-sensitive watersheds has lowest pH in upper organic soil horizons, increasing
down the profile to higher pH at depth. Drainage water chemistry during baseflow conditions is generally
reflective of conditions in the lower soil horizons and the subsoil. During high flows during snowmelt or
rainfall events, however, flow-routing favors water flowpaths through upper horizons. During such
events, drainage water chemistry, therefore, typically reflects the lower pH, higher organic content, and
lower ANC of these upper soil horizons (Sullivan, 2000a). As such, storm flow and snowmelt are often
associated with episodes of extreme surface water acidity due to an increase in the proportion of flow
derived from water that has moved laterally through the surface soil without infiltration to deeper
soil horizons (Wigington et al., 1991).
B-46
-------
The routing of water as it flows through a watershed determines the degree of contact with
acidifying or neutralizing materials and therefore influences (along with soils and bedrock characteristics)
the amount of episodic acidification that occurs. In any given watershed, surface water ANC may vary in
time depending upon the proportion of the flow that has contact with deep versus shallow soil horizons;
the more subsurface contact, the higher the surface water ANC (Turner et al., 1991). This can be
attributed in part to higher base saturation and (in some watersheds) greater S042 adsorption capacity in
subsurface soils. It may also relate to the accumulation in the upper soil horizons of acidic material
derived from atmospheric deposition and decay processes (Lynch and Corbett, 1989; Turner et al., 1991).
Episodic acidification is often the limiting condition for aquatic organisms in streams that can be suitable
for aquatic life under baseflow conditions.
Episodes are generally accompanied by changes in at least two or more of the following chemical
parameters: ANC, pH, base cations, S042 , N03 . Aln+, organic acid anions, and DOC (Sullivan, 2000a).
The U.S. EPA Episodic Response Project (ERP) confirmed the chemical and biological effects of episodic
pH depressions in lakes and streams in parts of the U.S. (Wigington et al., 1993). The ERP illustrated that
episodic processes are mostly natural, that S042 and especially N03 attributable to atmospheric
deposition play important roles in the episodic acidification of some surface waters, and that the chemical
response that has the greatest effect on biota is increased Al; concentration. Similar findings had been
reported elsewhere, especially in Europe, but the ERP helped to clarify the extent, causes, and magnitude
of episodic acidification in portions of the U.S. (Sullivan, 2000a).
Water chemistry trends documented by long-term monitoring programs and reported here represent
recovery from chronic acidification. Most surface waters exhibit seasonally lower ANC and pH values
than would be captured by trend analysis that considers only chronic chemistry data. In many cases, sites
that are relatively low in ANC, but not chronically acidic, undergo short-term episodic acidification to
negative ANC values during spring snowmelt, or during intense rain events. Lawrence (2002) found that
16% of total stream reaches in the West Branch Neversink River, in the Catskill Mountains of New York,
were chronically acidic, whereas 66% of total stream reaches has a high likelihood of becoming acidic
during high flows.
Most research on episodic processes has been conducted on stream systems, which tend to be more
susceptible to such effects than lakes. Spatial variability can be considerable in lakes, and this complicates
efforts to quantify the magnitude of episodic effects (Gubala et al., 1991). Moreover, synoptic lake
surveys are typically conducted during the autumn "index period," during which time lakewater chemistry
exhibits low temporal variability. Although autumn is an ideal time for surveying lakewater chemistry in
terms of minimizing variability, lakewater samples collected during autumn provide little relevant data on
episodic processes, and in particular on the dynamics or importance of N as an agent of acidification.
Nitrate concentrations in lakewater are elevated during the autumn season only in lakes having
watersheds that exhibit fairly advanced symptoms of N saturation (Stoddard, 1994).
Mixing zones have received little attention despite the fact that they can be acutely toxic to aquatic
biota. Whether an area of acidic water that comes in contact with non-acidic water is a safe haven or a
toxic zone depends on many parameters, one of the most important of which is the amount and form of Al
species produced at the boundaries. For example, Al hydroxide (Al(OH3)) can precipitate out of solution
if pH is suddenly increased within a mixing zone. This form of Al is acutely toxic to fish.
The mechanisms that produce acidic episodes can include dilution of base cations and flushing of
N03~, S042 and/or organic acids from forest soils to drainage water (Kahl et al., 1992; Lawrence, 2002;
Wigington et al., 1996; Wigington, 1999). Acidic deposition can contribute to episodic acidification of
surface water both by supplying N which can produce pulses of N03 during high flow periods,
contributing hydrologically mobile S042 through dry deposition, and by lowering baseline pH and ANC,
so that episodes are sufficient to produce biologically harmful conditions (Stoddard et al., 2003).
Episodic acidification due to atmospheric deposition is most commonly associated with N
deposition, and effects tend to be most pronounced during snowmelt. However, snowmelt can flush into
surface waters N that was deposited from the atmosphere to the snowpack and also N that was
mineralized within the soil under the snowpack during winter. A substantial component of the N03 flux
B-47
-------
may have been derived from mineralization of organic N (Ley et al., 2004). Much of the N released from
the snowpack during the melting period is retained in underlying soils and only a component of that is
flushed to surface waters. Where soils are sparse, as in alpine regions of the western U.S., most snowpack
N is flushed to surface waters, and even though there is evidence through use of isotopic tracers that much
of the N was cycled microbially, snowpack N has been reported to caused temporary acidification of
alpine streams (Campbell et al., 2002; Williams and Tonnessen, 2000).
Episodic pH and ANC depressions during snowmelt are largely driven by base cation dilution and
N03 enrichment in most areas (Campbell et al., 1995; Stoddard, 1995; Wigington et al., 1991, 1993),
although Denning et al. (1991) found a significant decline of both pH and ANC associated with DOC
flushing from forest soils. Pulses of increased S042 during hydrological episodes are usually attributable
to S storage and release in soils (for example, in the southeastern U.S.) or wetlands. More commonly, lake
and streamwater concentrations of S042 decrease or remain stable during snowmelt. This is probably
because most stream flow during episodes is derived from water previously stored in watershed soils that
is then forced into streams and lakes by the piston effect.
In the Northeast, the most severe acidification of surface waters generally occurs during spring
snowmelt (Charles, 1991). Stoddard et al. (2003) found that on average, spring ANC values in New
England, the Adirondacks, and the Northern Appalachian Plateau were about 30 j^ieq/L lower than
summer values during the period 1990 to 2000 (Figure B-l 1). This implies that lakes and streams in these
regions would need to recover to chronic Gran ANC values above about 30 j^icq/L before they could be
expected to not experience acidic episodes (Stoddard et al., 2003). However, the estimate of 30 j^icq/L is
certain to be low because the comparison was made with non-episodic sampling in spring.
In the West, episodic acidification is an especially important issue for surface waters throughout
high-elevation areas. A number of factors pre-dispose western systems to potential episodic effects
(Peterson et al., 1998; Sullivan, 2000a), including:
¦ the abundance of dilute to ultradilute lakes which exhibit very low concentrations of base cations,
and therefore ANC, throughout the year;
¦ large snowpack accumulations at the high-elevation sites, thus causing substantial episodic
acidification via the natural process of base cation dilution; and
¦ short hydraulic retention times for many of the high-elevation drainage lakes, thus enabling
snowmelt to rapidly flush lake basins with highly dilute meltwater.
Based on measurements of microbial biomass, C02 flux through the snowpack, and soil N pools,
Williams et al. (1996a) concluded N cycling under the snowpack in Colorado during the winter and spring
was sufficient to supply the N03 measured in stream waters. Brooks et al. (1996) investigated soil N
dynamics throughout the snow-covered season on Niwot Ridge, CO. Sites with consistent snow cover had
a 3 to 8 cm layer of thawed soil under the snowpack for several months before snowmelt began. Nitrogen
mineralization in this thawed layer contributed to Nr pools that were significantly larger than the pool of
N stored in the snowpack. As snowmelt began, soil inorganic N pools decreased sharply, concurrent with
a large increase in microbial biomass N. As snowmelt continued, both microbial N and soil inorganic N
decreased, presumably due to increased demand by growing vegetation (Brooks et al., 1996).
B-48
-------
_ 150 -
O
z
<
# New England Lakes
O Adirondack Lakes
O Appalachian Streams
-
Mean Summer ANC (jJeq/L)
Source: Stoddard et al. (2003).
Figure B-11. Relationship between mean summer and spring ANC values at LTM sites in New England, the
Adirondacks, and the Northern Appalachian Plateau.
In the Sierra Nevada, the hydrology of alpine and subalpine ecosystems is dominated by snowfall
and snowmelt, with over 90% of the annual precipitation falling as snow. The relatively small loads of
acidic deposition can supply relatively high concentrations of S042 and NO;, to lakes and streams during
the early phase of snowmelt (Stoddard, 1995) through the process of preferential elution (Johannessen
and Henriksen, 1978).
Lakewater pH and ANC in the Sierra Nevada generally decrease with increasing runoff, reaching
minima near peak snowmelt discharge. Most other solutes exhibit temporal patterns that indicate dilution
or a pulse of increased concentration followed by either dilution or biological uptake. Williams and
Melack (1991) and Williams et al. (1995) documented ionic pulses (2 to 10 days in duration) in meltwater
concentrations in the Emerald Lake watershed twofold to twelvefold greater than the snowpack average.
Sulfate and NO;, concentrations in meltwater decreased to below the initial bulk concentrations after
about 30% of the snowpack had melted. The initial meltwater draining from the snowpack had
concentrations of N03 and NH4+ as high as 28 (ieq/L, compared to bulk snowpack concentrations
<5 (ieq/L (Williams et al., 1995). Stream water NO; concentrations peaked during the early snowmelt
period, with maximum streamwater concentrations of 18 j^icq/L. During summer, stream water NO;
concentrations were always near or below detection limit.
Stoddard (1995) reported results for two lakes in the Sierra Episodes Study, one of which (Treasure
Lake) typified the response of most high elevation lakes in the study and one whose response was most
extreme (High Lake). At Treasure Lake, ANC began to decline at the onset of snowmelt and reached a
minimum at peak runoff, corresponding with minimum base cation, N03~, and S042 concentrations. The
lakewater did not become acidic. High Lake watershed contained a deeper snowpack, and began melting
later in the season. ANC fell to 0 and below twice during the first 10 days of snowmelt. The ANC
minimum corresponded with maximum concentrations of base cations, NO; and Al. Nitrate
concentrations increased to values greater than 40 (ieq/L, exceeding concurrent increases in base cations
and causing the lake to become acidic for brief periods. Stoddard (1995) concluded that High Lake
appeared to be representative of the most extreme conditions of episodic acid-sensitivity in the Sierra
Nevada.
B-49
-------
o
2
<
E
Spring ANC (|Jeq/L)
Source: Sullivan et al. (2003).
Figure B-12. Minimum streamwater ANC sampled at each site during each year versus median spring ANC
for all samples collected at that site during that spring season. Data are provided for all
intensively studied streams within Shenandoah National Park during the period 1993-1999. A
1:1 line is provided for reference. The vertical distance from each sample point upwards to the
1:1 line indicates the ANC difference between the median spring value and the lowest sample
value for each site and year.
Data regarding episodic variability in streamwater ANC for six intensively studied sites within
Shenandoah National Park for the period 1993 to 1999 are presented in Figure B-12 (Sullivan et al, 2003).
The minimum measured ANC each year at each site (which generally is recorded during a large rain or
snowmelt episode) is plotted against the median spring ANC for that year at that site. Sites that exhibited
median spring ANC below about 20 j^icq/L (Paine Run, White Oak Run, Deep Run) generally had
minimum measured ANC about 10 j^icq/L lower than median spring ANC.
In contrast, at the high-ANC Piney River site (median spring ANC >150 (ieq/L), the minimum
measured ANC was generally more than about 40 j^ieq/L lower than the respective median spring ANC. At
sites having intermediate ANC values, with median spring ANC in the range of about 30 to 90 (ieq/L, the
minimum ANC measured each year was generally about 20 to 30 j^ieq/L lower than the respective median
spring ANC. Thus, there is a rather clear pattern of larger episodic ANC depressions in streams having
higher median ANC and smaller episodic ANC depressions in streams having lower median ANC. The
two sites that had median spring ANC between about 0 and 10 j^ieq/L consistently showed minimum
measured values below 0. Streams having low chronic ANC can be expected to experience relatively
small episodic ANC depressions. However, those depressions can result in minimum ANC values that are
associated with toxicity to aquatic biota.
A recent study by Deviney et al. (2006) Deviney used hourly ANC predictions over short time
periods to compute recurrence intervals of annual water-year minimum ANC values for periods of 6, 24,
72, and 168 h. They extrapolated the results to the rest of the Shenandoah National Park catchments using
catchment geology and topography. On the basis of the models, they conclude that large number of
Shenandoah National Park streams have 6- to 168-h periods of low ANC values, which may stress
resident fish populations (Deviney et al., 2006). Specifically, on the basis of a 4-year recurrence interval,
approximately 23% of the land area (44% of the catchments) can be expected to have conditions that are
indeterminate (ANC 20 to 50), episodically acidic (ANC 0 to 20) or chronically acidic (ANC less than 0)
for 72 continuous hours. Many catchments are predicted to have successive years of low-ANC values
potentially sufficient to extirpate some species (Deviney et al., 2006). The authors of the study reported
~
Paine Run
A
White Oak Run
O
Deep Run
~
Staunton River
•
NF Dry Run
¦
Piney River
150-
100-
B-50
-------
that smaller catchments are more vulnerable to episodic acidification than large catchments underlain by
the same bedrock. Catchments with similar topography and size are more vulnerable if underlain by less
basaltic and carbonate bedrock.
There are several different mechanisms of episodic acidification in operation in the streams in
Shenandoah National Park, depending at least in part on the bedrock geology of the stream. The most
acidic conditions in Shenandoah National Park streams occur during high-flow periods, in conjunction
with storm or snowmelt runoff. The general relationship between flow level and ANC is evident in Figure
B-13, which plots ANC measurements against flow for three intensively studied streams representing the
major bedrock types in the park. The response of all three streams is similar in that most of the lower
ANC values occur in the upper range of flow levels.
Consistent with observations by Eshleman (1988), the minimum ANC values that occur in response
to high flow are related to baseflow ANC values. Paine Run (siliciclastic bedrock) had a mean weekly
ANC value of about 6 j^icq/L and often had high-flow ANC values that were less than 0 (ieq/L. Staunton
River (granitic bedrock) had a mean weekly ANC value of about 82 j^icq/L and had only a few high-flow
ANC values less than 50 |_ieq/L. Piney River (basaltic bedrock) had a mean weekly ANC value of
217 (ieq/L and no values as low as 50 j^ieq/L.
Eshleman and Hyer (2000) estimated the contribution of each major ion to observed episodic ANC
depressions in Paine Run, Staunton River, and Piney River during a 3-year period. During the study, 33
discrete storm events were sampled and water chemistry values were compared between antecedent
baseflow and the point of minimum measured ANC (near peak discharge). The relative contribution of
each ion to the ANC depressions was estimated using the method of Molot et al. (1989), which
normalized the change in ion concentration by the overall change in ANC during the episode. At the low-
ANC (~0) Paine Run site on siliciclastic bedrock, increases in N03 and S042 , and to a lesser extent
organic acid anions, were the primary causes of episodic acidification. Base cations tended to compensate
for most of the increases in acid anion concentration. ANC declined by 3 to 21 j^ieq/L (median 7 j^ieq/L)
during the episodes studied.
B-51
-------
Pine Run {siliddastic bedrock class)
O
z
<
0.001 0.010 0.100 1.000 10 00
Staurton River (granitic bedrock class)
150 r 1
^ 125
g" 100
O 75
5 5„
25
0,010 0,100 1.000
Piney River (basaltic bedrock class)
400
350
300
1"
250
a.
200
o
z
150
<
100
50
0.001 0.010 0.100 1.000
Runoff (mm/hr)
Source: Sullivan et al. (2003).
Figure B-13 Relationship between ANC and runoff for streamwater samples collected at intensively
studied sites in Shenandoah National Park. The data represent samples collected during the
1992-1997 period.
At the intermediate-ANC (-60 to 120 j^icq/L) Staunton River site on granitic bedrock, increases in
S042 and organic acid anions, and to a lesser extent N03 , were the primary causes of episodic
acidification. Base cation increases compensated these changes to a large degree, and ANC declined by 2
to 68 (ieq/L during the episodes (median decrease in ANC was 21 j^icq/L).
B-52
-------
At the high-ANC (-150 to 200 j^icq/L) Piney River site on basaltic (69%) and granitic (31%)
bedrock, base cation concentrations declined during episodes (in contrast with the other two sites where
base cation concentrations increased). Sulfate and NO;, usually increased. The change in ANC during the
episodes studied ranged from 9 to 163 j^ieq/L (median 57 j^icq/L: Eshleman and Hyer, 2000).
Previous studies have shown that mobilization of dissolved A1 during episodic acidification is a
primary cause of fish mortality in streams that have low ANC under baseflow conditions (Wigington
et al., 1993). Streams with higher ANC during baseflow are less likely to become sufficiently acidic
during episodes to bring much Al into solution.
Figure B-14 provides an example of changes in ANC, pH, and total monomeric Al that occurred in
Paine Run, Staunton River, and Piney River during a high-flow episode in January 1995. Under baseflow
conditions, ANC at the Paine Run site was above 0 (ieq/L, pH was above 5.5, and Al concentration was
less than about 1 (.iM. Discharge levels increased dramatically during the episode, resulting in depression
of ANC to less than 0 (ieq/L, pH values less than 5.5, and an increase in Al concentration to near 3 (.iM.
above the threshold for adverse effects on some species of aquatic biota.
Pairie Run
(siliciclastic bedrock class)
20
15
10
5
0
-5
7.0
6.5
6.0
5.5
Staunton River
(granitic bedrock class)
Piney River
(basaltic bedrock class)
ANC ucu/L)
ANC (Meq/L)
ANC ueq/L
II I II I M II I M II M I I I II I I II M II
Total Monomeric Aluminum
(ug'L)
Total Monomeric Aluminum
(ug/L)
Total Monomeric Aluminum
(mq/l)
Runoff
(mm/day)
Runoff
(mm/day)
¦ i T111111 Tii iii'iimii
Runoff
ft
_ (mm/day)
Time
Source: Sullivan et al. (2003).
Figure B-14. Decrease in ANC and pH and increase in dissolved aluminum in response to a sharp increase
in streamflow in three watersheds within Shenandoah National Park during a hydrological
episode in 1995. The watersheds were selected to be representative of the three geologic
sensitivity classes within the park. Data are shown for the month of January 1995.
The same episode also resulted in substantial declines in ANC in the granitic (Staunton River) and
basaltic (Piney River) watersheds. However, ANC values at these two sites were relatively high before the
episode (about 75 and 175 (ieq/L, respectively) and did not decline to below about 50 j^ieq/L during the
episode at either site, and pH values remained above 6.0 and 6.5, respectively.
In general, pre-episode ANC is a good predictor of minimum episodic ANC and also a reasonable
predictor of episodic AANC. Higher values of pre-episode ANC lead to larger AANC values, but
minimum ANC values of such streams are generally not especially low. Lowest minimum ANC values are
reached in streams that have low pre-episode ANC, but the AANC values for such streams are generally
small.
Webb et al. (1994) developed an approach to calibration of an episodic acidification model for
VTSSS long-term monitoring streams in western Virginia that was based on the regression method
B-53
-------
described by Eshleman (1988). Median, spring quarter ANC concentrations for the period 1988 to 1993
were used to represent chronic ANC, from which episodic ANC was predicted. Regression results were
very similar for the four lowest ANC watershed classes, and they were therefore combined to yield a
single regression model to predict the minimum measured ANC from the chronic ANC. Extreme ANC
values were about 20% lower than chronic values, based on the regression equation:
ANCmin = 0.79 ANC chromc ~ 5.88 (r2 = 0.97; se of slope = 0.02, p = 0.001)
Because the model was based on estimation of the minimum ANC measured in the quarterly
sampling program, it is probable that the true minimum ANC values were actually somewhat lower than
20% below the measured chronic ANC. Nevertheless, regression approaches for estimation of the
minimum episodic ANC of surface waters, such as was employed by Webb et al. (1994) for western
Virginia, provide a basis for predicting future episodic acidification. It must be recognized, however, that
future episodic behavior might vary from current behavior if chronic conditions change dramatically.
The relative importance of the major processes that contribute to episodic acidification varies
among the streams within Shenandoah National Park, in part as a function of bedrock geology and
baseflow streamwater ANC. Sulfur-driven acidification was an important contributor to episodic loss of
ANC at all three study sites, probably because S adsorption by soils occurs to a lesser extent during high-
flow periods. This is due, at least in part, to diminished contact between drainage water and potentially
adsorbing soils surfaces. Dilution of base cation concentrations was most important at the high-ANC site.
The documented importance of N03 to episodic acidification was a relatively recent development,
attributed to the effects of gypsy moth (Lymantria dispar) infestation in many watersheds within
Shenandoah National Park (Webb et al., 1995). Consumption of foliage by the moth larvae converted
foliar N, which is normally tied up in long-term N cycling processes, into more labile N forms on the
forest floor.
Thus, episodic acidification of streams in Shenandoah National Park can be attributed to a number
of causes, including dilution of base cations and increased concentrations of sulfuric, nitric, and organic
acids (Eshleman et al., 1995; Hyer et al., 1995). For streams having low pre-episodic ANC, episodic
decreases in pH and ANC and increases in toxic Al concentrations can have adverse effects on fish
populations. Not all of the causes of episodic acidification are related to acidic deposition. Base-cation
dilution and increase in organic acid anions during high-flow conditions are natural processes. The
contribution of nitric acid, indicated by increased N03 concentrations, has evidently been (at least for
streams in the park) related to forest defoliation by the gypsy moth (Eshleman et al., 1998; Webb et al.,
1995). However, significant contributions of sulfuric acid, indicated by increased S042 concentrations
during episodes in some streams, is an effect of atmospheric deposition and the dynamics of S adsorption
on soils (Eshleman and Hyer, 2000).
B.4. Effects on Biota
Soil and surface water acidification involve changes in a number of chemical parameters, each of
which has the potential to influence the health and vigor of biological communities and the species that
comprise them. In most cases where biological effects of acidification have been documented, the most
important chemical parameters involved in those effects have been pH, Al;, and Ca2+. Less commonly, one
or more base cations other than Ca2+ (e.g., Mg2+, K+) or C are also involved. This is true for both aquatic
and terrestrial effects.
A number of authors have examined the complex interactions between pH, Al, and Ca2+ that must
be considered when attempting to determine the effects of acidification on both aquatic and terrestrial
biota (e.g., Mount et al., 1988; Wood et al., 1990). Calcium concentration significantly affects the
distribution of species and their ability to survive in acidified environments. Aluminum, leached by acid
B-54
-------
precipitation from soils in the watershed, complicates the response considerably because some forms of
A1 are highly toxic to both aquatic and terrestrial species. Aluminum and hydrogen ions interact both
synergistically and antagonistically depending on conditions (Havas, 1985; Rosseland and Staurnes,
1994). In the presence of naturally occurring organic acids, A1 toxicity can be reduced or eliminated. A
number of authors have examined the complex interactions between pH, Al, and Ca2+ that must be
considered when attempting to determine the effects of acidification on both aquatic and terrestrial biota
(e.g., Ingersoll et al., 1990a; Mount et al., 1988; Wood et al., 1990).
B.4.1. Types of Effects of Acidification on Biota
Ecological effects occur at four levels of biological organization: the individual; the population,
comprised of many individuals; the biological community, composed of many species, (Billings, 1978);
and the ecosystem. Several metrics have been developed to describe the effects of acidification at each of
these levels of organization. For the individual, effects are assessed in terms of sublethal effects on
condition. At the population level, effects are measured by changes in the population of a certain species.
At the community level, species richness and community structure can be used to evaluate effects, and at
the ecosystem level, changes in nutrient cycling and ecosystem processes are assessed. Most of these
indices have been applied primarily to aquatic ecosystems. Each is discussed below.
Baker et al. (1990b) conducted a rigorous review of the effects of acidification on aquatic biota for
the 1990 NAPAP State of Science/Technology reports. They evaluated hundreds of laboratory, in situ
bioassay, field surveys, whole-system field experiments, and smaller mesocosm studies on the effects of
acidification on aquatic biota. Their 381-page report is the most exhaustive source summarizing the
aquatic biological effects of acidification from acidic deposition. The summaries provided here in
Sections B.4 and B.6 rely heavily on this source.
In Shenandoah National Park, a statistically robust relationship between acid-base status of streams
and fish species richness was documented. The 3 year Fish in Sensitive Habitats (FISH) study of stream
acidification in Shenandoah National Park demonstrated negative effects on fish from both chronic and
episodic acidification (Bulger et al., 1999). Biological differences in low- versus high-ANC streams
included species richness, population density, condition factor, age, size, and field bioassay survival. Of
particular note was that both episodic and chronic mortality occurred in young brook trout exposed in a
low-ANC stream, but not in a high-ANC stream (MacAvoy and Bulger, 1995), and that blacknose dace
(Rhinichthys atratulus) in low-ANC streams were in poor condition relative to blacknose dace in higher-
ANC streams (Dennis et al., 1995; Dennis and Bulger, 1995).
B.4.1.1. Individual Condition Factor
Relatively little is known about changes in the condition of fish or other aquatic biota resulting
from acidification. It is expected that sublethal effects will occur in acid-sensitive species well before the
species is eliminated from a particular lake, stream, or terrestrial habitat. For that reason, loss of an acid-
sensitive species is not necessarily an ideal indicator of acid stress. Clearly, stress begins to occur before
species elimination. Sublethal effects are more difficult to quantify, but are nevertheless important.
Condition factor is one measure of sublethal effect that has been used to quantify effects of
acidification on fish. Condition factor is an index to describe the relationship between fish weight and
length. Expressed as fish weight/length3, multiplied by a scaling constant, this index reflects potential
depletion of stored energy reserves (Dennis et al., 1995; Everhart and Youngs, 1981; Goede and Barton,
1990). Condition factor is interpreted as depletion of energy resources such as stored liver glycogen and
body fat (Goede and Barton, 1990). Fish with higher condition factor are more robust than fish having
low condition factor.
B-55
-------
y = 1.194X +1.519, r2 = .777
10.5
o
o
n
LL
9.5
c
o
T3
C
o
u
8,5
7.5
5.2
5.4
5.6
5.8
6.2
6.4
6.6
6.8
7
7.2
6
mean pH
Source: Bulger et al. (1999).
Figure B-15. Length-adjusted condition factor (K), a measure of body size in blacknose dace (Rhinichthys
atratulus) compared with mean stream pH among 11 populations (n = 442) in Shenandoah
National Park. Values of pH are means based on quarterly measurements, 1991-94; K was
measured in 1994. The regression analysis showed a highly significant relationship (p <0.001)
between mean stream pH and body size, such that fish from acidified streams were less
robust than fish from circumneutral streams.
Field studies have shown lower condition factor in fish found in more acidic streams (Dennis et al.,
1995). Condition factor has been developed and applied mainly for blacknose dace. This species is widely
distributed in Appalachian Mountain streams and is moderately tolerant of low pH and ANC, relative to
other fish species in the region. However, the concept is probably applicable to other species as well.
Condition factor may be a useful metric for many species in aquatic ecosystems that are only marginally
affected by acidification.
Bulger et al. (1999) observed a positive relationship between condition factor and pH in streams in
Shenandoah National Park. Dennis and Bulger (1995) found a reduction in the condition factor for
blacknose dace in waters near pH 6.0. The four populations shown in Figure B-15 with the lowest
condition factor have mean habitat pH values within or below the range of critical pH values at which
Baker and Christensen (1991) estimated that negative population effects for blacknose dace are likely for
the species. The mean length-adjusted condition factor of fish from the study stream with the lowest ANC
was about 20% lower than that of the fish in best condition. Comparisons with the work of Schofield and
Driscoll (1987) and Kretser et al. (1989) suggest that pH in the low-pH Shenandoah National Park
streams is near or below the limit of occurrence for blacknose dace populations in the Adirondack region
of New York (Sullivan et al., 2003).
Chronic sublethal stress caused by pH below about 6.0 may have serious effects on a variety of
wild fish populations. There is an energy cost in maintaining physiological homeostasis; the calories used
to respond to stress are a part of the fish's total energy budget and are unavailable for other functions,
such as growth and reproduction (Schreck, 1981, 1982; Wedemeyer et al., 1990)
B-56
-------
Observed differences in condition factor may occur because maintenance of internal chemistry in
the more acidic streams would require energy that otherwise would be available for growth and weight
gain (Dennis and Bulger, 1999; Sullivan et al., 2003). The energy costs to fish for active iono-
osmoregulation can be substantial (Bulger, 1986; Farmer and Beamish, 1969). Because of the steep
gradient in Na+ and CI concentrations between fish blood and freshwater, there is constant diffusional
loss of these ions, that must be replaced by energy-requiring active transport. Low pH increases the rate
of passive loss of blood electrolytes (especially Na+ and CP), and Al elevates losses of Na+ and Cl~ above
the levels that occur due to acid stress alone (Wood, 1989).
It is also possible that the loss of sensitive individuals or early life stages within species may reduce
competition for food among the survivors, resulting in better growth rates, survival, or condition.
Similarly, competitive release (increase in growth or abundance subsequent to removal of a competitor)
may result from the loss of a sensitive species, with positive effects on the density, growth, or survival of
competitor population(s) of other species (Baker et al., 1990b). However, in some cases where
acidification continued, transient positive effects on size of surviving fish were shortly followed by
extirpation (Bulger et al., 1993).
Acid stress is at least partly responsible for the lower condition of blacknose dace populations in
Shenandoah National Park, though reduced access to food or lower food quality (Baker et al., 1990b),
either resulting from the nature of soft water streams or exacerbated by acidification, cannot be ruled out.
Primary productivity is low in headwater streams and lower still in soft water headwaters, which are more
likely to be acidified. Production of invertebrates is likely to be low in such streams as well (Wallace
et al., 1992). Thus, lower food availability cannot be discounted as a potential contributor to lowered
condition in Shenandoah National Park blacknose dace populations in low-pH streams. Nevertheless,
reduced growth rates have been attributed to acid stress in a number of other fish species, including
Atlantic salmon (Salmo salar), Chinook salmon (Oncorhynchus tshawytscha), lake trout (Salvelinus
namaycush), rainbow trout (Oncorhynchus mykiss), brook trout, brown trout (Salmo trutta), and Arctic
char (Salvelinus alpines).
B.4.1.2. Species Composition
Species composition refers to the mix of species that are represented in a particular ecosystem.
Acidification alters species composition in aquatic ecosystems. There are a number of species common to
many oligotrophic waters that are sensitive to acidic deposition and that cannot survive, compete, or
reproduce in acidic waters. In response to small to moderate changes in acidity, acid-sensitive species are
often replaced by other more acid-tolerant species, resulting in changes in community composition, but
little or no change in total community abundance or biomass. The effects of acidification are continuous,
with more species being affected at higher degrees of acidification. Therefore, the degree of alteration of
surface water biological community composition increases as surface waters become more acidic. There
is a consistent pattern of lower community diversity with increased acidification.
B.4.1.3. Taxonomic Richness
Taxonomic richness is a metric that is commonly used to quantify the effects of an environmental
stress such as acidification or eutrophication. The richness metric can be applied at various taxonomic
levels. For example, the number of fish species can be used as an index of acidification (Bulger et al.,
1999). Similarly, acidification effects on aquatic insects can be evaluated on the basis of the number of
families or genera of mayflies (order Ephemeroptera) (Sullivan et al., 2003).
Acidification results in the loss of acid-sensitive species, with more species lost with higher
degrees of acidification. A direct outcome of population loss caused by acidification is a decline in species
B-57
-------
richness (the total number of species in a stream or lake). This is a highly predictable outcome of regional
acidification, although the pattern and rate of species loss varies from region to region.
Decreases in ANC and pH and increases in Al; concentration contribute to declines in species
richness and abundance of zooplankton, macroinvertebrates, and fish (Keller and Gunn, 1995; Schindler
et al., 1985). Species richness is positively correlated with pH and ANC (Kretser et al., 1989; Rago and
Wiener, 1986) because of the elimination of acid-sensitive species (Schindler et al., 1985). Knowledge of
the spatial distribution of pH and other water quality variables is necessary to explain the presence or
absence of species within heterogeneous environments. Organisms that are mobile and can sense the pH
of their environment can move to areas (called refugia) that have more favorable water chemistry.
Although some species are favored by increased acidity, species diversity generally decreases as surface
water acidity increases.
Decreases in species richness have been observed for all major trophic groups of aquatic organisms
(Baker et al., 1990b). Baker et al. (1990b) discussed 10 selected studies that documented this
phenomenon, with sample sizes ranging from 12 to nearly 3,000 lakes and streams analyzed per study.
Lake and stream size can be an important complicating factor in interpreting species richness data.
Larger lakes and streams in larger watersheds would generally be expected to contain more species than
smaller lakes or streams in smaller watersheds, irrespective of acid-base chemistry. Nevertheless, when
adjusted for lake size, lakes with pH less than approximately 6.0 contain significantly fewer species than
lakes with pH above 6.0 (Figure B-17) (Frenette et al., 1986; Matuszek and Beggs, 1988; Rago and
Wiener, 1986; Schofield and Driscoll, 1987).
Studies in the Adirondack Mountains demonstrated the effect of acidification on species richness;
of the 53 fish species recorded in Adirondack lakes by the ALSC, about half (26 species) were absent
from lakes with pH below 6.0. Those 26 species included important recreational species, such as Atlantic
salmon, tiger trout (Salmo trutta X Salvelinus fontinalis), redbreast sunfish (Lepomis auritus), bluegill
(Lepomis macrochirus), tiger musky (/','S Ox masquinongy X lucius), walleye (Sander vitreus), alewife
(Alosapseudoharengus), and kokanee (Oncorhynchus nerka) (Kretser et al., 1989), plus ecologically
important minnows that serve as forage for sport fish. Fully 346 of 1,469 lakes surveyed by the ALSC
supported no fish at all at the time of the survey. These lakes were significantly lower in pH, dissolved
Ca2+, and ANC, and had higher concentrations of Al; than lakes hosting one or more species of fish
(Gallagher and Baker, 1990). Among lakes with fish, there was an unambiguous relationship between the
number of fish species and lake pH, ranging from about one species per lake for lakes having pH less than
4.5 to about six species per lake for lakes having pH >6.5 (Kretser et al., 1989; Driscoll et al., 2001).
Figure B-17 shows the mean number of fish species for pH classes from 4.0 to 8.0 in lakes in the
Adirondacks. It is important to note, however, that there are many possible causes of fish absence in
addition to acidification. These include lack of suitable habitat (especially for spawning), winter kill,
blocked access, etc.
B-58
-------
Source: Matuszek and Beggs (1988).
Figure B-16. Mean residual number of species per lake for lakes in Ontario, by pH interval. The residual
number of species for a lake is the deviation of the observed number from the number
predicted by lake area.
Sullivan et al. (2006b) developed a relationship between fish species richness and ANC class for
Adirondack lakes. Fish species richness observations, as a function of ANC (j^icq/L) class, were fit to a
logistic relationship by a non-linear regression analysis. Under chronically acidic conditions (summer
index or annual average ANC <0 (ieq/L), Adirondack lakes are generally Ashless. There was a marked
increase in mean species richness with increases in ANC up to values of approximately 100 (ieq/L. The
asymptote for the fish species equation was 5.7 species. This analysis suggests that there could be loss of
fish species with decreases in ANC below approximately 100 j^icq/L. It does not account, however, for the
possibility that lakes having higher ANC are often larger, and therefore support more fish species because
of increased habitat diversity and complexity.
As an element of the FISH project (Bulger et al., 1999), numbers of fish species were compared
among 13 Shenandoah National Park streams spanning a range of pH and ANC conditions. There was a
highly significant (p <0.0001) relationship between stream acid-base status (during the 7-year period of
record) and fish species richness among the 13 streams. The streams with the lowest ANC hosted the
fewest species (see Figure B-18).
B-59
-------
m
m
"5
&
to
¦C
m
5
E
3
z
T r i i r
-200 -100 0 100 200 300 400 500
ANC(peq/L)
10-
9 ¦
8 ¦
7 .
6 ¦
5 ¦
4 ¦
3 .
2 •
1 .
y = ,024x + 2.076, r2 = .763
D
O
~
~ US
'\
DL
"a ~
-50 0 50 100 150 200 250 300
mean ANC(peq/L)
Figure B-17. Number of fish species per lake or stream versus acidity statues, expressed either as pH or
ANC. (A) Adirondack lakes (Sullivan et al., 2006b); (B) streams in Shenandoah National Park
(Bulger et al., 1999). The data for the Adirondacks are presented as mean and range of species
richness within 10 |jeq/L ANC categories, based on data collected by the Adirondack Lakes
Survey Corporation.
Median stream ANC values and watershed areas are shown in Table B-ll for the 14 streams used
by Bulger et al. (1999) to develop the relationship between ANC and fish species richness shown in
Figure B-18. Despite the overall similarities, these study streams vary in watershed area by a factor of 10.
The streams that have larger watershed areas generally have more fish species than the streams having
smaller watershed areas. All of the "rivers" have watersheds larger than 10 km2 and ANC higher than
75 (ieq/L. In contrast, the majority (but not all) of the "runs" have watershed area smaller than 10 km2 and
ANC less than 20 (ieq/L. All of the streams that have watershed areas smaller than 10 km2 have three or
fewer known species of fish present. All of the streams having larger watersheds (>10 km2) have three or
more known fish species; seven of nine have five or more species; and the average number of fish species
is six. There is no clear distinction between river and run, but it is clear that as small streams in
Shenandoah National Park combine and flow into larger streams and eventually to rivers, two things
happen: acid-sensitivity generally declines, and habitat generally becomes suitable for additional fish
species (Sullivan et al., 2003).
B-60
-------
9
8
7
6
5
4
3
2
0
-25
0
25 50 75 100 125 150 175 200 225 250 275 300
Average ANC (peci/L)
Source: Redrawn from Bulger et ai. (1999).
Figure B-18. Number of fish species among 13 streams in Shenandoah National Park. Values of ANC are
means based on quarterly measurements, 1987-94. The regression analysis shoed a highly
significant relationship (p <0.0001) between mean stream ANC and number of fish species.
Streams having ANC consistently <75 peq/L had three or fewer species.
South of Shenandoah National Park the effects of surface water acidification on fish species
richness have been studied in some detail in the St. Marys River in Virginia. Fish species richness was
closely associated with surface water acid-base chemistry. Bugas et al. (1999) conducted electrofishing in
the St. Marys River in 1976, and every 2 years from 1986 through 1998. Systemic stream acidification
occurred during the study period. Sampling occurred at six sites between the downstream end of the St.
Marys Wilderness and the headwaters over a distance of about 8 km. The number of fish species in the St.
Marys River within the wilderness declined from 12 in 1976 to 4 in 1998. Three of the four species
present in 1998 (brook trout, blacknose dace, fantail darter [Etheostoma flcibellare]) are tolerant of low
pH and are typically the only fish species present in streams having similar levels of acidity in
Shenandoah National Park, which is also located in Virginia (Bulger et al. 1999). Bugas et al. (1999)
reported that successful brook trout reproduction in the St. Marys River occurred only 1 year out of 4
during the period 1995 through 1998. Eight of the fish species recorded in one or more early years have
not been observed in more recent years. Several, including blacknose dace, rainbow trout, and torrent
sucker (Thoburnia rhothoeca), showed a pattern of being progressively restricted over time to lower river
reaches, which generally have higher ANC. The number of fish species decreased with decreasing
minimum ANC, from nine species at ANC of about 160 (.ieq/L to one to three species at ANC near 0. The
best fit regression line suggested, on average, a loss of one species for every 21 |.ieq/L decline in annual
minimum recorded ANC value.
Dynamic water chemistry model projections have been combined with biological dose-response
relationships to estimate declines in fish species richness with acidification. A relationship derived from
the data in Figure B-18 was used by Sullivan et al. (2003) with stream ANC values predicted by the
MAGIC model to provide estimates of the expected number of fish species in each of the modeled
streams for the past, present, and future chemical conditions simulated for each stream. The coupled
geochemical and biological model predictions were evaluated by comparing the predicted species
richness in each of the 13 streams with the observed number of species that occur in each stream. The
agreement between predicted and observed species numbers was good, with a root mean squared error
(RMSE) in predicted number of species across the 13 streams of 1.2 species. Hie average error was 0.3
B-61
-------
species, indicating that the coupled models were unbiased in their predictions. Model reconstructions of
past species richness in the streams suggested that historical loss of species had been greatest in the
streams located on the most sensitive geological class (siliciclastic). The average number of species lost
from streams on the three bedrock types examined were estimated as: 1.6 species on siliciclastic bedrock;
0.4 species on granitic bedrock; and 0.4 species on basaltic bedrock. In the case of the siliciclastic
streams, the projected past changes were much larger than the average error and RMSE of the coupled
models, suggesting that the projections were reasonably robust.
It appears that fish species richness is controlled by multiple factors, of which both acidification
and watershed area can be important. Watershed area might be important in this context because smaller
watersheds may contain smaller streams having less diversity of habitat, more pronounced effects on fish
from high-flow periods, or lower food availability. Such issues interact with other stresses, including
acidification, to determine overall habitat suitability.
For Shenandoah National Park, Bulger et al. (1999) concluded that the most important cause of the
observed decline in species richness with decreasing ANC was acid stress associated with acidification.
However, an additional causal factor may have been the decrease in the number of available aquatic
niches when moving from downstream locations (which are seldom low in pH and ANC) to upstream
locations (which are often low in pH and ANC in this region; (Sullivan et al., 2003). The relative
importance of this latter factor, compared with the importance of acid stress, in determining this
relationship is unknown.
In the Adirondack region, Driscoll et al. (2001b) concluded that high-elevation lakes are more
likely to be Ashless than larger lakes at low elevation (Gallagher and Baker, 1990) because they have poor
access for fish immigration, poor fish spawning substrate, or low pH, or they may be susceptible to
periodic winter kills. Nevertheless, small, high-elevation Adirondack lakes with fish also had significantly
higher pH compared with Ashless lakes; acidity is likely to play an important role in the absences of fish
from such lakes (Driscoll et al., 2001b).
B.4.1.4. Community Structure
Ecosystem response to pollutant deposition is a direct function of the ecosystem's ability to
ameliorate resulting changes in individual species (Strickland et al., 1993). To determine ecosystem
response and the possible effects on community structure, species responses must be scaled in both time
and space and be propagated from the individual to the more complex levels of community interaction
within an ecosystem.
Individuals within a population vary in their ability to withstand a stress. The response of each
individual is based on its genetic constitution (genotype), its stage of growth at time of exposure to the
stress, and the microhabitat in which it lives (Levin, 1998). The range within which individuals in the
population can exist and function determines the ability of the population to survive when exposed to a
chronic stress. Those individuals that are able to cope with the stress survive and reproduce. The same
kinds of pressures act on populations of different species. Competition among species results in
community change overtime and eventually produces ecosystems composed of populations of species
that have the capability to tolerate the stress (Guderian et al., 1985; Rapport and Whitford, 1999;
U.S. EPA, 2004).
Work conducted on the biological effects of acidification has largely been focused on the response
of fish, especially salmonids (trout and salmon). This focus tends to be driven by the value people place
on fish and fishing, rather than any ecological consideration. Other vertebrate, invertebrate, plant, and
algal communities are also sensitive to acidification. In general, higher order trophic groups are more
susceptible to acidification. Thus, in terms of changes in community structure in response to aquatic
acidification, the general progression of sensitivity is as follows: fish >invertebrates (benthic and
zooplankton) >algae >microbes (Baker et al., 1990b). Population-level fish response to acidification is
primarily through recruitment failure, a result of increased mortality of early life stages or indirect effects
B-62
-------
through the food chain (loss of prey species). Al;, pH, and Ca2+ have been identified as the variables most
likely to have the greatest influence on fish community structure.
B.4.1.5. Indices of Ecological Effects
The most widely used index of acidification effect is the Acid Stress Index (ASI) developed by
Baker et al. (1990b). This index uses fish bioassay survival data fitted to a maximum likelihood logistic
regression model as a function of exposure to pH, Al, and Ca2+ to predict the probability of fish survival
expressed as a percent mortality. This approach can aid in determination of effects on species composition
by predicting the probability of occurrence of species of varying acid sensitivity. Separate ASI models
were developed for tolerant, intermediate, and sensitive fish species. Approximate ASI reference levels
were established for various species based on logistic regression of fish presence as a function of the
sensitive, intermediate, and tolerant ASI values for brown bullhead (Ameiurus nebulosus), brook trout,
lake trout, and common shiner (Luxilus cornutus). They are presented in Table B-12.
The ASI was deemed a useful index of stress by Baker et al. (1990b), even though the relationships
between ASIs and fish population status could not be quantified precisely because of confounding factors.
Such factors included the abundance and types of food species, competitors and predators present,
variations in habitat quality, and density-dependent effects on fecundity.
B.4.2. Timing of Effects
B.4.2.1. Life Stage Differences in Sensitivity
Episodic and chronic changes in the chemistry of surface waters can have different effects on
aquatic organisms and populations depending on species and the life history stages present. More is
known about the sensitivity to acidification of the life stages of fish than is known for other aquatic
organisms. In general, early life stages are more sensitive to acidic conditions than the young-of-the-year,
yearlings, and adults (Baker and Schofield, 1985; Baker et al., 1990b; Johnson et al., 1987). Also, small
fish, especially swim-up fry, are probably less mobile and less able to avoid exposure to adverse chemical
conditions than the relatively larger adults (Baker et al., 1996).
There are a number of issues of acidification timing that are important to determination of the
extent and magnitude of effects. One important issue concerns the timing of acidity exposure relative to
life stage. For example, adult fish are generally more tolerant of acidity than early life stages such as eggs,
fry and juveniles. There could be substantial differences in effect based on small differences in age or
timing of exposure to acidity. No definite pattern was observed by Baker et al. (1990b) across all studies
or species. This may reflect either differences in the test conditions or actual differences among species.
The presence of early life stages of brook trout, which are most sensitive to adverse effects from
acidification (Bulger et al., 2000), varies with season. For example, the most acid-sensitive stages of
brook trout development are present in Virginia streams throughout the cold season in general, and the
winter in particular (Figure B-19).
B-63
-------
Acid-Sensitive Life Stages of the Brook Trout
sacfry in mat
^
spawning hatching swim-up fry
~ ~
I L__J ¦ 1 1
JUL AUG SEP OCT NOV DEC JAN FEB MAR APR MAY JUN
Source: Sullivan et al. (2003).
Figure B-19. Life stages of brook trout.
The processes of oogenesis and fertilization in fish and aquatic invertebrates are especially
sensitive to low pH (Havas et al., 1995; Muniz, 1991). In fish, this sensitivity is most likely due to
adverse effects on the female spawner. For instance, Beamish et al. (1976) reported that reduced serum
and plasma Ca2+ in female fish in acidified Canadian lakes lead to a higher probability for failure in
producing viable eggs. A depletion of Ca2+ from bone and increased numbers of females with unshed eggs
have also been linked to sensitivity at this life stage (cf. Muniz, 1991; Rosseland, 1986).
After fertilization, the embryo seems to be susceptible to acidic waters throughout the whole period
of development. The periods shortly after fertilization and before hatching seem to be most critical
(Rosseland, 1986). The susceptibility of the embryo can be the result of direct exposure to elevated H
concentrations and also to the toxic effects of Al; at intermediate pH-values. Low pH in the surrounding
water also results in pH-depression inside the egg, leading to either a prolongation of the hatching or to a
reduced hatching success (Rosseland, 1986). Eggs lying in gravel on stream and lake beds are to some
extent protected from exposure to rapid changes in pH (Gunn and Keller, 1984b; Lacroix, 1985).
Nevertheless they can experience high mortality during periods of acid runoff, such as snowmelt (Gunn
and Keller, 1984a).
In fish, emergent alevins show susceptibility to the adverse effects of Al; and H that increases with
age (Baker and Schofield, 1982; Wood and McDonald, 1982). Rosseland (1986) indicated that this
increasing sensitivity results from changes that take place in the respiratory system. Shortly after hatch,
alevins still respire through their skin but gradually gills become the primary organ of gas and ion
exchange. Gills are the locus for interference of H+ and Al; with iono-regulatory exchange. Woodward
et al. (1989) exposed cutthroat trout (Oncorhynchus clarki) from the Snake River in Wyoming to pH
depressions from pH 4.5 to 6.5 in the laboratory. Fertilized egg, eyed embryo, alevin, and swim-up larval
stages were exposed to low pH for a period of seven days. Each life stage was monitored for mortality,
growth, and development for 40 days after hatching. Reductions in pH from 6.5 to 6.0 in low-Ca2+ water
(70 (ieq/L) did not affect survival, but reduced growth of swim-up larvae. The eggs, alevin, and swim-up
larval stages showed significantly higher mortality at pH 4.5 than at pH 6.5. Mortality was also higher at
pH 5.0 than at pH 6.5, but only statistically higher for eggs.
Woodward (1991) exposed greenback cutthroat trout (Oncorhynchus clarki stomias) in the
laboratory to 7-day pH depressions. Low-Ca2+ (65 j^ieq/L) water at pH 6.5 was experimentally reduced to
pH values of 6.0, 5.5, 5.0, and 4.5. Four life stages were exposed: freshly fertilized egg, eyed embryo,
alevin, and swim-up larva. Alevin survival was reduced at pH 5.0, whereas survival of eggs, embryos, and
swim-up larvae was reduced at pH 4.5. Swim-up larvae showed feeding inhibitions at pH 4.5. The authors
concluded that the threshold for effects of acidity on greenback cutthroat trout in the absence of Al was
pH 5.0 (Woodward, 1991).
B-64
-------
Yellowstone cutthroat trout (O. c. bouveri) were exposed to 7-day pH depressions by Farag et al.
(1993). Of the four life stages studied, eggs were most sensitive to low pH. Eggs exposed for seven days
to pH 5.0 test water showed a statistically significant reduction in survival compared with eggs exposed
for seven days to pH 6.5 water. Survival of alevin and swim-up larvae were significantly reduced from
near 100% at pH 6.5 to near 0% at pH 4.5. Intermediate pH values (6.0, 5.5) in all cases showed reduced
survival compared with the control (6.5) but not by statistically significant amounts. Eyed embryos were
not sensitive to any of the exposures.
According to Bulger et al. (1999), adult brook trout in Shenandoah National Park streams are more
tolerant of acidity than are adult blacknose dace. For both species, the early life stages are more sensitive
than the adults, and brook trout young are actually more sensitive than blacknose dace adults (Bulger
et al., 1999). Blacknose dace spawn during summer and the eggs and very young fry are therefore
somewhat protected from the most acidic episodes, which typically occur during cold-season, high-flow
conditions.
B.4.2.2. Biological Effects of Episodes
Episodic decreases in pH and ANC can produce chemical conditions in lakes, and especially in
streams, that are as harmful to biota as chronic acidification (Baker et al., 1996). Adverse effects on biota
occur particularly when changes involve pH, Alj. or Ca2+ (Baker et al., 1990b). Aquatic biota vary greatly
in their sensitivity to episodic decreases in pH and increases in Al; in waters having low Ca2+
concentration. However, Baker et al. (1990b) concluded that episodes are most likely to affect biota if the
episode occurs in waters with pre-episode pH above 5.5 and minimum pH during the episode of less than
5.0.
Results from the ERP demonstrated that episodic acidification can have long-term adverse effects
on fish populations. Streams with suitable chemistry during low flow, but low pH and high Al; levels
during high flow, had substantially lower numbers and biomass of brook trout than in non-acidic streams
(Wigington et al., 1996). Streams having acidic episodes showed significant mortality of fish.
Some brook trout avoided exposure to stressful chemical conditions during episodes by moving
downstream or into areas with higher pH and lower Alj. This movement of brook trout only partially
mitigated the adverse effects of episodic acidification, however, and was not sufficient to sustain fish
biomass or species composition at levels that would be expected in the absence of acidic episodes. Just as
spatially heterogeneous environments or refugia enable some species to survive in otherwise unfavorable
conditions, temporal heterogeneity often has the opposite effect. These findings suggest that stream
assessments based solely on chemical measurements during low-flow conditions will not accurately
predict the status of fish populations and communities in small mountain streams unless some adjustment
is made for episodic processes (Baker et al., 1996a, 1990b; Sullivan, 2000a; Wigington et al., 1996).
In Shenandoah National Park, MacAvoy and Bulger (1995) used multiple bioassays over 3 years in
one of the low-ANC streams as part of the FISH project to determine the effect of stream baseflow and
acid episode stream chemistry on the survival of brook trout eggs and fry. Simultaneous bioassays took
place in mid- and higher-ANC reference streams. Acid episodes (with associated low pH and elevated Al;
concentrations, and high streamwater discharge) induced rapid mortality in the low-ANC stream, while
the test fish in the higher-ANC stream survived (Bulger et al., 1999).
In the West, it has also been shown that native trout are sensitive to short-term increases in acidity.
For example, Woodward et al. (1989) exposed native western cutthroat trout to pH depressions (pH 4.5 to
6.5) in the laboratory. Reductions in pH from 6.5 to 6.0 in low-Ca2+ water (70 j^icq/L) did not affect
survival, but did reduce growth of swim-up larvae. Eggs, alevins, and swim-up larvae showed
significantly higher mortality at pH 4.5 as compared to pH 6.5. Mortality was also somewhat higher at pH
5.0, but only statistically higher for eggs. Some species of aquatic biota in western aquatic ecosystems
have been shown to be somewhat more sensitive to pH and ANC change than are cutthroat trout (Baker
et al., 1990b).
B-65
-------
Multiple logistic regression models were used by Van Sickle et al. (1996) to relate fish bioassay
mortality rates to summary statistics of time-varying stream chemistry over the 20-day bioassay periods.
Higher mortality of all three test fish species (brook trout, dace, sculpin [Cottus spp.]) during the in situ
bioassays was clearly associated with increased Al;. In addition, individual bioassays conducted during
chronically or episodically acidified conditions had higher median mortality than did those during non
acidic conditions, but no mortality differences were detected between chronically acidic and episodically
acidic conditions. Time-weighted median Al; was the best single predictor of 20-day mortality for both
brook trout and sculpin, whereas the number of days with Al; >200 (ig/L provided the best prediction of
blacknose dace mortality.
In the Northeast, Baker et al. (1996) studied the effects of episodic acidification on fish in 13 small
streams in the Adirondack and Catskill Mountains of New York and the Northern Appalachian Plateau in
Pennsylvania. They conducted in situ bioassays with brook trout and blacknose dace, mottled sculpin
{Cottus bairdi) or slimy sculpin (Cottus cognatus) depending on the region, to measure direct toxicity.
Movements of brook trout individuals in relation to stream chemistry were tracked using radiotelemetry.
Electrofishing surveys assessed fish community status and the abundance and biomass of brook trout in
each stream. Streams with suitable conditions during low flow, but moderate-to-severe episodic
acidification during high flow, had higher fish mortality in bioassays, higher net downstream movement
of brook trout during events, and lower brook abundance and biomass compared to nonacidic streams.
These streams lacked the more acid-sensitive fish species (blacknose dace and sculpin). Movement of
trout into refugia (areas with higher pH and lower Al) during episodes partially mitigated the adverse
effects of episodes.
Chemical measurements by ERP during high flow correlated with fish community status. In
general, reduced trout abundance occurred in ERP streams with median high flow pH <5.0 and Al; >100
to 200 (ig/L. Acid -sensitive fish species were absent from streams with median high flow pH <5.2 and
Alj >100 (ig/L. More recently, Baldigo et al. (2007) found that mortality of brook trout young of the year
occurred at concentrations as low as 54 j^ig/L .Al; was the single best predictor of fish mortality in ERP
bioassays (Van Sickle et al., 1996) and has been identified as an important toxic factor in other bioassays
and field studies (Ingersoll et al., 1990b; Mount et al., 1988; Rosseland et al., 1990). The relationships
between pH and Al; or ANC and Al; vary among streams (Wigington et al., 1996), and therefore
predictions of potential effects on fish based solely on pH or ANC may be misleading. High Al;
concentrations during episodes are probably the dominant cause of adverse effects on fish during episodic
acidity events.
Biological Effects of Chronic Acidification
Changes in surface water acid-base chemistry, including pH, ANC, Alj. and Ca2+, can affect in-
stream and in-lake biota. Adverse biological effects may be seen at pH less than about 6.0 to 6.5 and Al;
greater than about 30 to 50 (ig/L (1 to 2 It tends to increase with decreasing pH, and reaches
potentially toxic concentrations (>~2 |_iM) in surface drainage waters having pH less than about 5.5.
Effects vary substantially by organism, life stage, and the concentration of DOC. Inorganic Al in solution
is also toxic to plants.
Calcium can ameliorate the toxic effects of acidity and Al on biota. Most organisms can tolerate
lower pH and higher Al; at higher Ca2+ concentrations, but in natural environments, elevated
concentrations of Al; are only found in Ca2+-depleted systems. This effect is most important at low Ca2+
levels. Overall biological effects noted with decreasing pH are described in Table B-13 (Baker et al.,
1990b). The organisms most likely to respond to such changes in water chemistry include fish, aquatic
insects, zooplankton, and diatoms. In some cases, amphibians are also important sensitive biological
receptors. Most available data are for fish response.
In most stream or lake survey areas, direct quantification of biological responses to surface water
acidification is not possible, given the scarcity or absence of biological long-term monitoring and dose-
B-66
-------
response data. Few biological long-term monitoring studies have been conducted. Much of the available
in situ dose-response data have been generated from studies of streams in Virginia and Pennsylvania and
lakes in New York. Data with which to evaluate acidification relationships have been scarce in most other
regions.
Lakes
Fish status assessments for the eastern and upper midwestern U.S. were conducted by Baker et al.
(1990b), by region, using a variety of assessment methods. For the northeast region, two water chemistry
models were linked to fish response models: the Integrated Lake-Watershed Acidification Study (ILWAS)
model and MAGIC. For the Adirondack subregion, three process models were used: ILWAS, MAGIC,
and Regional MAGIC. For other areas in the eastern U.S. and for the Upper Midwest, analysis of fish
status was limited to application of the sensitive, intermediate, and tolerant toxicity models.
Assessment results reported by Baker et al. (1990b) for the Adirondack region are presented in
Table B-14 showing results based on the ASI. Table B-15 shows the estimated percentage of Adirondack
lakes with acid-base chemistry unsuitable for fish population survival according to various assessment
models based on responses for brook trout, lake trout, and common shiner. Assessment results for the
Northeast region are presented in Tables B-16 and B-17.
In acid-sensitive lakes in the western U.S., the focus is often mainly on native cutthroat trout. It is
important to note, however, that many high-elevation western lakes and streams were historically Ashless.
The top predators in such aquatic ecosystems were often amphibians or crustaceans. Thus, even though
cutthroat trout might be considered native to the region, they are not necessarily native to a particular lake
or stream.
Streams
In streams, the major organisms of concern with respect to water acidification are fish, amphibians,
benthic macroinvertebrates, and periphyton (attached algae). All of these groups have shown adverse
effects in response to acidification (see Annex B-6). Most available data are for fish and aquatic insects,
mainly in the southeastern U.S. Streams affected by acidic deposition tend to occur at high elevation.
They are often high-gradient and flow through base-poor geology.
Baker et al. (1990b) presented assessment results for the mid-Appalachian region as the
distribution (percent) of NSS lower node, upper node, and total streams classified in various ASI values
(Baker et al., 1990b). Most of the streams were classified in the lowest ASI category (Table B-18).
Assessment results for the interior Southeast region were similar (Table B-19).
Some fish response research has also been conducted for streams in the Catskill Mountains. Baker
and Christensen (1991) estimated that the fish species found in the Neversink River Basin in the Catskill
Mountains are typically lost when pH decreases to the range of 4.7 to 5.2 (brook trout), 5.5 to 5.9 (slimy
sculpin), 4.7 to 5.7 (brown trout), 5.6 to 6.2 (blacknose dace), and 4.9 to 5.3 (Atlantic salmon).
The Shenandoah National Park FISH Project evaluated the effects of streamwater acidification on
fish populations and communities in streams in Shenandoah National Park. Fish species richness,
population density, condition factor, age distribution, size, and bioassay survival were all lower in streams
having low-ANC compared to intermediate-ANC and high-ANC streams (Bulger et al., 1995; Dennis
et al., 1995; Dennis and Bulger, 1995; MacAvoy and Bulger, 1995).
Bulger et al. (2000) developed model-based projections using the MAGIC model to evaluate the
potential effect of reductions in S deposition of 40% and 70% from 1991 levels using data from VTSSS
and SWAS. Projections were based on four brook trout stream categories: Suitable , ANC >50 (ieq/L;
Indeterminate, ANC 20 to 50 |_icq/L: Marginal, ANC 0 to 20 j^ieq/L: and Unsuitable, ANC <0 j^ieq/L.
Three scenarios of future acidic deposition were modeled: constant deposition at 1991 levels, 40%
reduction from 1991 deposition levels, and 70% reduction from 1991 deposition levels. Based on
observed 1991 ANC values, approximately 30% of all trout streams in Virginia were marginal or
unsuitable for brook trout because they were either episodically (24%) or chronically (6%) acidic. In
B-67
-------
addition, another 20% of the streams were classified as indeterminate, and brook trout in these streams
may or may not have been affected. Based on the model simulations, 82% of these streams would not
have been acidic before the onset of acidic deposition and would likely have been suitable for brook trout.
The model projections suggested that neither the 40% nor the 70% reductions in acidic deposition
would be expected to increase the number of streams that were suitable for brook trout above the ambient
50%. In fact, the results suggested that a 70% reduction in deposition would be needed in the long-term
just to maintain the number of streams that were considered suitable for brook trout. Because of the length
of time required to restore buffering capacity in watershed soils, most of the marginal or unsuitable
streams were expected to remain marginal or unsuitable for the foreseeable future.
To develop projections of probable past and future responses of aquatic biota to changing S
deposition in Shenandoah National Park, the MAGIC model was coupled by Sullivan et al. (2003) with
several empirical models that linked biological response to past and future model projections of water
quality. Unlike MAGIC, which is a geochemical, process-based model, the biological effects estimates
were based on observed empirical relationships rooted in correlation and expressed as linear relationships.
Correlation does not necessarily imply cause, but an observed pattern of co-variation between variables
does provide a context for analysis of a possible relationship. In this case, the projections did not require
extrapolation and are, therefore, statistically robust. To the extent that the observed empirical relationships
used in the coupled models do in fact reflect the effects of acid stress on aquatic biota, the projections
were also biologically robust.
The geochemical and biological response models also differ in that MAGIC is a dynamic model
and explicitly predicts the time course of changing water quality, whereas the empirical relationships used
for estimating biological response were static. These relationships reflected a point in time (when the
observations were made) and provided no information concerning the dynamics of biological response.
That is, the empirical models predicted a new biological status for a new water chemistry, but gave no
indication of the time required to achieve the biological status once the water quality change had
occurred.
There are thus two considerations that must be kept in mind when interpreting the biological
responses predicted using a combination of process-based and empirical modeling approaches: the
causality of the relationship between water quality and response, and the dynamics of biological response.
With respect to the issue of causality, acidification is a disturbance and disturbance usually lowers species
richness. In turn, loss of species usually lowers ecosystem stability. Biodiversity loss is a predictable and
proven consequence of acidification, and there are abundant examples of this in North America and
Europe (Bulger et al., 2000). With respect to the timing of biological response, it can be variable and
difficult to predict.
B.4.2.3. Timing of Recovery from Acidification
Lakes and streams show spatial and temporal variability in response to a host of biotic and abiotic
factors. Against this background of variability, it is difficult to detect changes in biological communities
in response to changes in an individual environmental stressor without long-term biological data
(Schindler, 1990; Lancaster et al., 1996). Long-term data sets are rare, and there are few well-documented
instances of temporal changes in biological communities in response to changes in water chemistry.
Regardless, it is known that surface water acidification affects virtually all trophic levels (e.g., Flower and
Battarbee, 1983; Lancaster et al., 1996; 0kland and 0kland, 1986; Ormerod and Tyler, 1991; Rundle and
Hildrew, 1990; St. Louis et al., 1990; Siminon et al., 1993; Sullivan, 2000a).
Biological recovery can occur only if chemical recovery is sufficient to allow survival and
reproduction of acid-sensitive plants and animals. The time required for biological recovery is uncertain.
For terrestrial ecosystems, it may be decades after soil chemistry is restored because of the long life of
many plant species and the complex interactions of soil, roots, microbes, and soil biota. For aquatic
systems, research suggests that stream macroinvertebrate populations may recover relatively rapidly
B-68
-------
(within approximately 3 years), whereas lake populations of zooplankton recover more slowly (Gunn and
Mills, 1998).
The timing of fish recovery is highly uncertain, and probably will depend heavily on dispersal
opportunities. Stocking could accelerate fish population recovery (Driscoll et al., 2001b). Fish populations
have recovered in acidified lakes when the pH and ANC have been raised through liming or reduction of
acidic deposition (Hultberg and Andersson, 1982; Beggs and Gunn, 1986; Dillon et al., 1986; Keller and
Pitblado, 1986; Raddum et al., 1986; Gunn et al., 1988; Kelso and Jeffries, 1988).
Studies in Canada have improved understanding of the feasibility and complexity of biological
recovery in response to chemical recovery from acidification. Biological recovery of previously acidified
lakes is expected to be a slower process than chemical recovery. Sometimes there are other environmental
stresses in addition to acidity, such as metal contamination (Gundersen and Rasmussen, 1995; Havas
et al., 1995; Jackson and Harvey, 1995; McNicol et al., 1995; Yan et al., 1996b). Barriers can be imposed
by water drainage patterns between lakes that hinder re-colonization by some fish species (Jackson and
Harvey, 1995). Predation by non-acid-sensitive fish species can affect the recovery of zooplankton and
macroinvertebrate communities (McNicol et al., 1995). Finally, tributary-spawned fish can be preyed
upon when they move downstream into lakes inhabited by predatory fish and birds (Schofield and
Keleher, 1996).
Changes in surface water chemistry as a direct response to changes in S and N deposition are
difficult to predict. Both chemical and biological effects of changing deposition can lag as the ecosystem
comes into equilibrium with the modified deposition inputs. Soils or wetlands may continue to release S
at a high rate for many years subsequent to a decrease in S deposition. As a result, surface water S042
concentrations may decrease in the future as a consequence of deposition changes that have already
occurred. If soil base cations have become depleted, base cation concentrations in some surface waters
could decrease in the future irrespective of any further changes in S042 concentrations. This would be
expected to contribute to additional acidification.
Studies in the U.S., Canada, and Europe have illustrated the feasibility and complexity of biological
recovery in response to decreased surface water acidity. There is currently no theoretical basis on which to
predict the paths of biological recovery. At some scale, each stream or river is unique. The null hypothesis
is that recovery will proceed in the same fashion as acidification, only backwards. Thus, for example, the
last species lost (the most acid-tolerant) would be the first to return. However, time lags are expected to
differ widely among species and among water bodies. Biological recovery of previously acidified lakes or
streams can lag behind chemical recovery because of such factors as (a) limits on dispersal and
recolonization; (b) barriers imposed by water drainage patterns (Jackson and Harvey, 1995); (c) the
influence of predation (McNicol et al., 1995); and (d) other environmental stresses (Gunn et al., 1995;
Havas et al., 1995; Jackson and Harvey, 1995; McNicol et al., 1995; Yan et al., 1996a, 1996b).
Limitations on dispersal and recolonization can hamper biological recovery from chronic and
episodic acidification. If fish move into refugia areas during low pH and then return, behavioral
avoidance would reduce the overall effect of acidification on fish populations. However, if fish move out
of the stream system in response to sublethal episodes, as suggested by Baker et al. (1996), and do not
return or return in smaller numbers, then the population level effects of episodic acidification would be
greater than predicted based on mortality tests alone.
Stream macroinvertebrate communities are often dominated by immature life stages of flying
insects, such as mayflies, dragonflies, and stoneflies. Such species have rather rapid colonization times,
such that a functional stream macroinvertebrate community may return in only a few years in response to
improved chemistry. However, fish community recovery is expected to be quite variable, depending on
sources of colonists. In streams, fish could be introduced as soon as the water quality becomes suitable or
the macroinvertebrate community becomes established. In streams that had simple fish communities in
the past, a fish community might become rapidly established. It might take decades for complex
communities without species introductions.
The Sudbury region of Ontario, Canada has been important for studying the chemical and
biological effects of S deposition. Mining and smelting of copper-nickel ore began in the 1880s. By the
B-69
-------
1950s and 1960s, SO2 emissions from the mining and smelting operations peaked at over 5,000 tons/day
and extensive acidification of nearby surface waters was documented (Beamish and Harvey, 1972).
Emissions of SO2 then decreased during the 1970s to less than one-third of the peak values. This region
has been the focus of extensive chemical and biological effects work since the 1980s (Keller, 1992).
Sulfur emission reductions resulted in improved water quality in many lakes (Keller and Pitblado, 1986;
Keller et al., 1986), and some fisheries recovery was also documented (Gunn and Keller, 1990; Keller and
Yan, 1991). Griffiths and Keller (1992) found changes in the occurrence and abundance of benthic
invertebrates that were consistent with a direct effect of reduced lakewater acidity. A more recent
assessment of recovery of ecosystems in Canada provided further evidence of biological recovery, but
also showed that the spatial extent of recovery was limited to lakes that had been severely acidified by the
Sudbury smelter (Jeffries et al., 2003).
Whitepine Lake, located 90 km north of Sudbury, had low pH (5.4) and ANC (1 j^ieq/L) in 1980
and its fish populations displayed symptoms of acid stress. Acid-tolerant yellow perch (Perca flavescens)
were abundant, but the more acid-sensitive species lake trout and white sucker (Catostomus commersoni)
were rare and not reproducing. Fish populations were studied by Gunn and Keller (1990) from 1978
through 1987, and zooplankton were sampled at least monthly during the open-water periods of 1980
through 1988. During the period between 1980 and 1988, pH increased to 5.9 and ANC increased to
11 (ieq/L. Young lake trout first reappeared in 1982 and became increasingly abundant throughout the
study. The number of benthic invertebrate taxa increased from 39 in 1982/83 to 72 in 1988, and the
relative abundance of many of the invertebrates found in 1982 changed along with the changes in water
chemistry (Gunn and Keller, 1990). Research at Sudbury clearly documented that chemical recovery of
lakes was possible upon reduced emissions and deposition of S, and also that biological recovery,
involving multiple trophic levels, would soon follow.
Baker et al. (1990b) used field-based models to test the potential for biological recovery. The
models were calibrated from the observed among-lake or among-stream associations between fish status
and the chemical and physical characteristics measured in the surface water. The models were generally
calibrated using chemistry data collected in conjunction with surveys of fish status. It was assumed that
the systems surveyed were at steady state and that the observed status of the fish community was
determined by the observed chemical and physical conditions in the lake or stream. For each species
considered, the current presence or absence of the species was analyzed as a function of the water quality
variables associated with acidification (e.g., pH, Al, Ca2+, ANC, and DOC) using maximum likelihood
logistic regression (Reckhow et al., 1987). Models developed from data from the ELS and the ALSC were
calibrated against data from Ontario lakes.
The results from the various models were compared to their prediction of the change in the number
of Adirondack lakes with unsuitable acid-base chemistry, given a 50% decrease or a 30% increase in S
deposition relative to the existing conditions. All the models provided similar results (Figure B-20) with
the exception of those that relied on the pCa/pH term to predict fish status. Those models seemed to
overestimate the effect of Ca2+, and thus underestimate predicted fish response to changes in acidic
deposition.
B-70
-------
Brook Trout Models
Bayesian
LAF Framework
(Case 1)
Field-Based: pH
pCa/pH
pCa/ph, AI/DOC
Lake Trout Models
-20 10 0 10 20
Intermediate
ASI > BO
Field-Based: pH
pCa/pH
A!
-20 -10 0 10 20
Change in Percent Unsuitable
-20 -10 0 10 20
Change in Percent Unsuitable
Common Shiner Models
-30 -20 -10 0 10 20 30
| -50% scenario
+30% scenario
Sensitive
ASI > 80
Field-Based: ph
pCa/pH
-30 -20 -10 0 10 20 30
Change in Percent Unsuitable
Source: Baker et al. (1990b).
Figure B-20. Example model application. Projected changes in the percentage of Adirondack lakes
(Direct/Delayed Response project target population) with acid-base chemistry unsuitable for
the survival of fish populations in the year 2034, versus current simulated conditions, based
on projected changes in water chemistry from the Model of Acidification of Groundwater in
Catchments (MAGIC) and using alternative models offish response, given a 50% decrease in
deposition or a 30% increase in deposition, (a) brook trout, (b) lake trout, (c) common shiner.
An important consideration for measuring the success of S and N emissions controls is the
development of appropriate expectations for the magnitude of potential chemical recovery. Most lakes
inferred to have been measurably acidified by atmospheric deposition were already marginally acidic,
typically with pH less than about 6, before anthropogenic atmospheric pollution began before 1900.
Therefore, full recovery of currently acidic lakes would not be expected to yield neutral pH. Nevertheless,
increases in ANC may allow recovery of fish populations even if pH remains relatively low (Stoddard
et al., 2003).
B.4.3. Effects by Ecosystem Type
B.4.3.1. Terrestrial Ecosystems
Due to a strong dependency on atmospheric deposition and exposure to gaseous compounds as the
major sources of nutrients, lichens are affected by changes in these conditions. Vulnerability of lichens to
increased N input is generally greater than that of vascular plants (Fremstad et al., 2005). Even in the
Pacific Northwest, which receives uniformly low levels of N deposition, changes from acid-sensitive and
N-sensitive to pollution-tolerant and nitrophilic lichen taxa are occurring in some areas (Fenn, 2003). In
eastern North America and central Europe, areas experiencing relatively high levels of acidic deposition
have experienced noticeable reductions in cyanolichen abundance on both coniferous and deciduous trees
(Richardson and Cameron, 2004). Effects on lichen species biodiversity are also likely (McCune, 1988;
van Haluwyn and van Herk, 2002).
B-71
-------
Fenn et al. (2007) speculated that large, pollution-sensitive macrolichens, including epiphytic
cyanolichens, will be replaced by N-tolerant species in areas where development expands in western
Oregon and Washington into N-limited Coast Range forests. Currently, in the Pacific Northwest,
nitrophilic lichen species are common in and around Seattle, Portland, Spokane, the Tri-cities, Salem,
Oregon's agricultural lands in the northeast and southwest, and the Willamette Valley (Fenn et al., 2007).
The USDA Forest Service website contains information about lichen species pollution tolerance,
diversity, and preferred habitat in relation to exposure to N (http://www.nacse.org/lichenair).
In London, epiphyte diversity, including a majority of the lichen taxa, declined in areas where N02
surpassed 40 (ig/m3 and NOx surpassed 70 (ig/m3. Lichens remaining in areas affected by these levels of
exposure contained almost exclusively families Candelariaceae, Physciaceae and Teloschistaceae
(Davies et al., 2007).
Progressive decline in ectomycorrhizal fungal (EMF) community structure and species richness
was observed at five Alaskan coniferous forest sites (white spruce [Picea glauca\ dominant) along an
N deposition gradient (1 to 20 kg N/ha/yr) downwind from a large industrial complex on the Kenai
Peninsula. The effects were attributed to both acidification and fertilization processes (Lilleskov et al.,
2002). EMF communities are important in tree nutrition and C balance, and EMF trees tend to be
dominant in N-limited forest ecosystems. A shift in EMF community structure could result in changes in
tree species.
Westman et al. (1985) summarized the literature of negative effects of SO2 on native plants,
including decreased pollen germination and tube elongation in both angiosperms and gymnosperms. It is
often difficult to separate the effects of SO2 exposure on plants from the effects of S deposition. This is
because areas that experience high SO2 exposure generally also receive high S deposition. Kozlowski
(1985) summarized relative susceptibility of different trees, lichens, and bryophytes to SO2.
Available information is not sufficient to draw conclusions regarding the increased likelihood of
future effects on the condition of hardwood forests in the Southern Appalachian Mountain region
(Sullivan et al., 2002). Certainly, such effects are less likely for hardwood forests than for spruce-fir
forests. Red oak seedlings grown in a greenhouse in deciduous forest soils exhibited no response to
acidified soil (pH 4.0 from 9:1 H2S04:HN03) or to high or low S042 inputs (12.8 to 24.8 mg/L). The lack
of response suggested that red oak seedlings are not sensitive indicators of acidification effects from S
deposition (McClenahen, 1987).
Grasslands and Alpine Tundra
Alpine communities are considered very sensitive to changes in N deposition, but documented
effects in the scientific literature have been attributed to nutrient enrichment, rather than acidification
(Seastedt et al., 2004; Bowman et al., 2006). Lower-elevation grasslands, especially those in semi-arid
environments, would be expected to be even less sensitive to acidification because of low water leaching
potential and the common presence of base-rich Mollisol and Aridisol soils. However, some effects of
acidification may be manifested in mesic grasslands.
In a review of S02 effects on grasses in the United Kingdom, Bell (1985) suggested that damage
can occur at levels as low as 150 (ig/m3. However, he asserted that any ubiquitous critical load value must
be modified to include variations due to environmental conditions and combined effects with other
pollutants. He also suggested that many grass species exhibit a tolerance to S02, resulting from more
intraspecific competition in agricultural grasslands. Westman et al. (1985) also provided evidence of the
evolution of a tolerant grass species, Bromus rubens, in southern California coastal sage scrub, influenced
by an average of 3.7 (imol/m3 of S02 over 25 years.
Studies of S02 effects on timothy grass (Phleum pratense) showed diminished leaf production and
increased leaf senescence in seedlings exposed to 0.120 ppm S02 for 35 days (Mansfield and Jones,
1985). In another experiment, Mansfield and Jones (1985) reported that exposure to 0.120 ppm S02 in
seedlings over 40 days resulted in a 62% reduction in the dry weight of roots and 51% reduction in the
B-72
-------
dry weight of shoots, as well as a significant decline in leaf-area ratio (LAR) and specific leaf area (SLA)
by the end of the experiment. They suggested that decreased growth and shifts in LAR and SLA could
lead to decreased hardiness and increased susceptibility to water stress.
In a 5-year exposure of native mixed prairie grassland in Montana, Lauenroth and Milchunas
(1985) exposed grasses to a control (-20 (.ig/nr1) and three elevated levels of S02 (-60, 106, 184 (.ig/m').
Year-to-year S accumulation did not appear to occur over the 5-year course of the treatment, though
progressive increases in root and rhizome S concentrations were observed seasonally. No significant
negative effects on either above-ground net primary productivity or below-ground biomass dynamics in
grasses were observed, except a decrease in biomass for Bromus japonicus. However, lichen cover
declined after 1-year of exposure at the low treatment level. Though no biomass or cover effects were
observed at the community level, there were minor population changes. These results are consistent with
the nature of semi-arid grasslands that typically adjust well to perturbations (Lauenroth and Milchunas,
1985).
Arid Lands
At the time of the previous AQCD, it was believed that arid and semi-arid ecosystems were not as
susceptible to soil acidification and high N03 leaching as are forested ecosystems. This is because of a
scarcity of water for N03 leaching, except on an episodic basis, and because arid soils tend to be more
alkaline than soils in more humid environments. No new research has altered that conclusion. Arid lands
in the U.S. generally receive low levels of S deposition. However, N deposition can be quite high,
especially in southern California in the vicinity of the Los Angeles Basin. Little work has been done on
the effects of acidification on arid land ecosystems. As reviewed by Fenn et al. (2003), acidification
effects have not been demonstrated at the Central Arizona-Phoenix LTER site, despite the almost 30 kg
N/ha/yr of deposition received. Nevertheless, N deposition has the potential to increase plant growth and
denitrification and alter community composition in arid environments (Egerton-Warburton and Allen,
2000; Allen et al., 2005). Such changes could alter key ecosystem processes and, as such, merit
consideration. There has been little research to examine these issues and, therefore, the state of knowledge
is similar to what it was in 1993.
B.4.3.2. Aquatic Ecosystem
Chronic Effects
Sulfate
The study of Clow and Mast (1999) is unique in that S042 trends were evaluated with both raw
data and data adjusted for existing trends in flow. From 1967 to 1983, Clow and Mast (1999) showed a
decreasing trend in S042 concentrations in a Catskill river, no trend in three rivers in Maine,
Pennsylvania, and Virginia, and an increasing trend in a river in Ohio. The Maine river did show a
decreasing trend before flow adjustment of the data. From 1984 to 1996, however, Clow and Mast (1999)
found decreasing trends in S042 concentrations in all five rivers, both with and without flow adjustment.
The rivers in Pennsylvania, Ohio, and Virginia were south of the maximum southern extent of glaciation,
and therefore were more likely to be subject to the effects of S042 adsorption in soils of their watersheds.
In such streams, decreasing S-adsorption on soils would be expected to counteract the effects of
decreasing S deposition in terms of effects on stream S042 concentration.
Surface waters in other unglaciated regions exhibited decreasing trends in S042 by the 1980s.
Concentrations of S042 in 130 northeastern lakes in 1984 were compared to those in the same lakes in
2001 (Warby et al., 2005). Median concentrations in each subregion were lower in 2001 than 1984, and in
the region as a whole, the overall median decrease was 1.53 (ieq/L/yr. A decrease in S042 concentrations
B-73
-------
that averaged 2.16 (ieq/L/yr was also observed in 47 of 48 Adirondack lakes from 1992 to 2004, and a
similar decrease of 2.09 (ieq/L/yr was observed in a subset of these lakes from 1982 to 2004 (Driscoll
et al., 2007a).
The pattern of increasing concentrations of S042 in surface waters before the year of peak S
emissions in 1973, followed by widespread decreasing trends in S042 concentrations after the peak (with
the only exception being the Blue Ridge Mountain region in Virginia), provides convincing evidence of
the link between S emissions and S042 concentrations in surface waters. A similar link has been shown in
Europe (Stoddard et al., 1999). On this basis, continued decreases in S emissions would be expected to
result in further decreases in S042 concentrations in surface waters, although the rate of response is
uncertain due to an incomplete knowledge of S retention mechanisms in terrestrial systems. Also, in a
detailed analysis of flow effects on S042 trends, Murdoch and Shanley (Murdoch and Shanley, 2006)
found S042 that higher concentrations of S042 occurred at corresponding high, medium and low flows in
2000 to 2002 than in 1997 to 1999 in two of the rivers studied by Clow and Mast (1999), and at high and
medium flows in a third river. Continued monitoring of surface waters will be needed to verify a future
link between emissions and S042 concentrations in surface waters.
Nitrate
Driscoll et al. (1985) found that N03 concentrations in 20 lakes in the early 1980s in the
Adirondack region of New York averaged 12% of S042 concentrations, whereas Lovett et al. (2000)
found that baseflow N03 concentrations in 1994-97 were an average of 37% of S042 concentrations in
39 streams in the Catskill region of New York. Average concentrations of N03 in most southeastern
streams also tend to be considerably less than S042 concentrations (Webb et al., 2004).
High-frequency sampling in the study of Murdoch and Stoddard (1993) demonstrated the
importance of N03 during high-flow conditions in Catskill streams in which concentrations periodically
equaled or exceeded S042 concentrations. This study also reported increasing trends in N03
concentrations during the period of 1970 to 1990 in all 16 Catskill streams for which data were available.
A similar increase in N03 concentrations was reported for Adirondack lakes in the 1980s (Stoddard et al.,
1999). These increasing trends in N03 concentrations were attributed to N saturation in response to
atmospheric deposition (Aber et al., 1998).
The relationship between N deposition and surface water N03 concentrations up through the
1980s suggested that continued N deposition would further the accumulation of N in terrestrial
ecosystems and drive continued increases in surface water N03 concentrations. However, more recent
information on N03 concentrations have been less consistent with the concept of N saturation. Goodale
et al. (2003) resampled New Hampshire streams in 1996-97 that had been previously sampled in
1973-74 and found substantially lower N03 concentrations in the more recent sampling, despite two
decades of relatively stable levels of deposition to otherwise undisturbed forests. The lower N03
concentrations could not be accounted for by differences in flow or forest succession, but interannual
climate variation was proposed as a possible cause. The long-term record of dissolved inorganic N (which
is largely N03 ) concentrations at the HBEF showed a similar pattern; high concentrations in the late
1960s and 1970s, followed by decreases to minimum values in the mid-1990s (Aber et al., 2002). These
authors attributed this pattern to a combination of environmental factors, but did not identify a single most
important control variable. A reversal from increasing trends in N03 concentrations in the 1980s to
decreasing trends in the 1990s was also observed in Adirondack lakes (Driscoll et al., 2003a). A small
decrease in atmospheric deposition of N also occurred in this region through the 1990s, but was not
considered sufficient by these authors to explain the decreasing trend in lakewater N03 concentrations.
Rather, they proposed that increased concentrations of atmospheric C02 may have resulted in a
fertilization effect that increased N assimilation (Driscoll et al., 2007a).
In general, trends in surface water N03 concentrations during the 1990s were much smaller than
trends in S042 , with the only ecologically significant changes occurring in the two regions with the
highest ambient N03 concentrations. Lakes in the Adirondacks and streams in the Northern Appalachian
B-74
-------
Plateau both exhibited small but significant downward trends in N03 in the 1990s (Table B-6). Both of
these regions are central to the debate over whether N saturation is a legitimate threat to the health of
forests and surface waters (Aber et al., 1998; Stoddard, 1994). While declining N03 concentrations in
these regions is a positive development for these ecosystems, it is not known if these trends will continue,
especially because they do not appear to reflect changes in N emissions or deposition. The presence of
strong upward trends in N03 in these same regions in the 1980s (Murdoch and Stoddard, 1992; Stoddard,
1994) suggests that trends measured on the scale of a single decade may reflect variability in long-term
patterns of changing N03 leakage from forested watersheds. Such patterns are controlled by factors that
may take many years of additional research to determine. While great uncertainty exists and the time
scales ofN saturation may be longer than previously considered (e.g., centuries rather than decades), the
long-term retention of N deposition in forested regions is unlikely to continue indefinitely (Aber et al.,
2003).
In New England and the Upper Midwest, where ambient N03 concentrations are much lower than
in the Adirondacks and Northern Appalachian Plateau, N03 concentrations in surface waters were
unchanged during the 1990s. The Ridge/Blue Ridge province registered a small, but significant, decrease
in N03 during the 1990s, but interpretation of trends for N03 in this region is complicated by an
outbreak of gypsy moths that also occurred during this period. Forest defoliation by gypsy moths was the
most likely cause of a pulse in N03 export from many streams in this region in the mid-1990s (Eshleman
et al., 1998).
Some evidence of climate effects on long-term trends in N03 concentrations in surface waters was
provided by studies of Mitchell et al. (1996) and Murdoch et al. (1998). A synchronous pattern in N03
concentrations was observed from 1983 to 1993 in four small watersheds in New York, New Hampshire,
and Maine, which included anomalously high concentrations during the snowmelt period of 1990. The
region-wide spike in N03 concentrations followed an unusually cold December that may have disrupted
soil N cycling processes (Mitchell et al., 1996). Murdoch et al. (1998) also found that mean annual air
temperatures were strongly related to average annual N03 concentrations in most years in a Catskill
watershed with elevated N03 concentrations in stream water. Those relationships were explained by
microbial control of N release in watersheds that were considered to be N-saturated.
Efforts to explain the decreasing trends in N03 concentrations under conditions of reasonably
stable atmospheric N deposition have focused on terrestrial N cycling and N-saturation theory. However,
processes within lakes may have also played a role in the trends in Adirondack lakes. In a study of 30 of
the 48 lakes studied by Driscoll et al. (2003a; 2007a), Momen et al. (2006) found that concentrations of
N03 were inversely correlated with concentrations of Chi a in 11 lakes, and that Chi a was increasing in
concentration in 9 lakes. The increase in pH observed in most of these lakes may have stimulated
productivity so that N assimilation by plankton increased (Momen et al., 2006).
Thus, there is little or no apparent relationship between recent trends in N deposition and trends of
N03 concentrations in surface waters, in sharp contrast to S deposition and S042 concentrations. Rather
than disprove the concept of N-saturation; however, these studies more likely reflect the complexities of
N utilization within terrestrial and aquatic ecosystems. These complexities create considerable uncertainty
with regard to how future trends in N03 concentrations in surface waters will respond to changing levels
of deposition.
B-75
-------
Salmon Pond I New Cng?smJj
1 V«2 1H,«4 t'1,W 1/liW Ip'IW H1.W I'1*4 1'1« !'"« I'tTO I-1.1J2
Darts Lake (AdBomdacloO
. *. *• \
'
'1 \v\.-/7
l-'ltWZ U1JM law l.'KM HliW in#7 WHM VliM liiw VUOO f1»2
VtmtecooK Lake |Upp*r Midwtrt)
Figure B-21. Time series data for SO42", NO3", base cations [Ca2+ + Mg2+], Gran ANC, pH, and DOC in
example Long Term Monitoring Lakes and streams that have relatively low ANC. The example
surface waters include Salmon Pond, Maine (New England region); Darts Lake, NY
(Adirondack region); East Branch Neversink River (Appalachian Plateau region); and
Vandercook Lake, Wisconsin (Upper Midwest region). Significant trends are indicated by trend
lines. Shaded box indicates time period of analyses reported by Stoddard et al. (2003).
Base Cations
The earliest trends of base cation concentrations in acid-sensitive surface waters of the U.S. were
presented by Stoddard (1991) for 12 streams in the Catskill region. In 5 of 12 streams, concentrations of
(Ca" + Mg " ) increased from 1915-22 to 1945, but decreased from 1945-46 to 1990. In the remaining
seven streams, concentrations increased during both periods, but at a lower rate in the more recent period
in five of the seven streams. In streams that showed an increase in concentrations during both periods, the
average rate of increase from 1915 to 1922 was 2.8 (.icq/L. whereas the average rate of increase from
1945 to 1990 was 1.2 j.uxj/1.. Data on S042 trends were not available for the early period, but the trends
in (Ca2 + Mg" ) concentrations were consistent with the expected pattern of high rates of cation leaching
during the early stages of acidification from S deposition.
Clow and Mast (1999) observed trends in (Ca + Mg2+) concentrations that were generally
consistent with S042 trends in five eastern rivers from 1968 to 1983. Decreasing trends in concentrations
of (Ca2 , + Mg2") and SO_f concentrations were observed in a Maine river, and increasing trends in (Ca2"
B-76
-------
+ Mg2+) and S042 concentrations were observed in an Ohio river. None of the three other rivers showed a
decrease in concentrations of (Ca2+ + Mg2+), and only one showed a decreasing trend in S042
concentrations. For the period 1984 to 1996, the trend in S042 concentrations was negative in the Ohio
River and the concentrations of (Ca2+ + Mg2+) showed no trend. Also, a negative trend in (Ca2+ + Mg2+)
concentrations in a Virginia river was coupled with a negative trend in S042 concentrations. Relations for
the other three rivers were similar to the earlier period of 1984 to 1996.
The study of Likens et al. (1996) evaluated trends in base cations in relation to trends in (S042 +
N03 ) in the long-term record for the HBEF. This record showed an approximately linear, increasing
relationship between concentrations of base cations and (S042 + N03 ) from 1964 to 1969, then a
reversal in 1970 to a decreasing trend up to 1994. The slope of the phase with increasing anion
concentrations was steeper than the slope for the phase with decreasing anion concentrations. This
indicates lower base cation leaching per equivalent of mobile anion, and therefore suggests depletion of
base cations stored in soil. The study of Lawrence et al. (1999b) showed decreased concentrations of base
cations at a rate that exceeded decreases in (S042 + N03 ) in Catskill streams from 1984 to 1997. In
streams within western Virginia and in Shenandoah National Park, concentrations of base cations did not
exhibit significant trends from 1988 to 2001, perhaps due to the influence of S adsorption on streamwater
S042 concentrations.
Regional declines in base cation concentrations were measured in the LTM Program from 1990 to
2000 in New England lakes, Adirondack lakes, Appalachian streams, and upper Midwest lakes (Stoddard
et al., 2003). These results were consistent with decreased Ca2+ concentrations measured by Warby et al.
(2005) in 130 acid-sensitive lakes in the Northeast between 1984 and 2001. The rate of decrease
identified by Warby et al. (2005) for base cations (1.73 (ieq/L) was somewhat less than the rate of
decrease in S042 concentrations (1.53 (ieq/L). Driscoll et al. (2007), also documented decreasing trends
in base cation concentrations in 16 Adirondack Lakes from 1982 to 2004, and similar rates of decrease in
48 lakes (including the 16) from 1992 to 2004.
In summary, decreases in base cation concentrations over the past two to three decades are
ubiquitous and closely tied to trends in S042 concentrations in acid-sensitive regions of the U.S. Reports
of increases in concentrations of base cations in acid-sensitive regions were not found in the literature. In
most regions, rates of decrease for base cations have been similar to those for S042 and N03 . with the
exception of streams in Shenandoah National Park. Decreasing trends of base cation concentrations do
not necessarily indicate further acidification or recovery of surface waters, but do indicate lower leaching
rates in soils, a prerequisite for recovery of soil base saturation. However, decreased concentrations of
base cations, particularly Ca2+, would also be expected to lower productivity in oligotrophic surface
waters.
Acid Cations
Measurements of pH (sometimes expressed as H ) have been routinely collected in surface waters
in the U.S. where effects of acidic deposition have been monitored, but a long-standing reliance on
titrated ANC rather than pH as the primary chemical measurement has limited the amount of pH data
published. The longest continuous record of pH in surface waters dates back to 1963 at the HBEF
(Driscoll et al., 2001b). This record shows an overall increasing trend from 1963 to 1994, although most
of the increase occurred after 1980. In Adirondack lakes, 12 of 16 monitored from 1982 to 2004 showed
an increase in pH, but the rates of change among lakes were highly variable, and one lake showed a
decrease in pH (Driscoll et al., 2007a). When expressed as H concentration, the average increase for the
12 lakes was 0.18 (ieq/L/yr. In this same region, pH also increased in 31 of 48 lakes (including the 16
lakes monitored from 1982) from 1992 to 2004. Two lakes showed increases in pH over the 12 years.
Comparison of pH measurements of 130 lakes in 1984 and in 2001, in the northeastern U.S.,
showed an overall average increase in pH of 0.002 units (Warby et al., 2005). However, in this
assessment, lakes in the Adirondack region did not show a significant increase, nor did lakes in central
New England, or Maine. The Catskill/Poconos region of New York and Pennsylvania showed an average
B-77
-------
increase of 0.008 pH units per year, and southern New England showed an average increase of 0.002 pH
units per year. Through continuous monitoring from 1990 to 2000, Stoddard et al. (2003) found a decrease
in H4" (0.19 (ieq/L/yr) similar to that observed in the same Adirondack lakes by Driscoll et al. (200a7)
from 1992 to 2004 (0.18 (ieq/L), and an increase in Appalachian streams (0.08 (ieq/L/yr) and Midwest
lakes (0.01 (ieq/L/yr). No trends were found in New England lakes or Blue Ridge streams in Virginia in
this study. Stream monitoring in the Adirondack region from 1991 to 2001 showed an increase in H in
one stream, no trend in a second stream, and also an increase in a third stream near the outlet of a lake
(Lawrence et al., 2004). In summary, decreasing trends in pH in surface waters are common through the
1990s up to 2004, but many exceptions occur, and overall, the rates of change have been small.
The discovery that Al; was toxic to aquatic life resulted in a considerable amount of data on Al
concentrations in surface waters in the 1980s, but most of this sampling was done either once or for a
limited period of time (Johnson et al., 1981; Driscoll and Newton, 1985; Driscoll et al., 1987; Lawrence
et al., 1987; Cronan et al., 1990). Monitoring of Al; concentrations was begun in 16 Adirondack lakes in
1982 and expanded to 48 lakes in 1990. From 1982 to 2004, 5 of the original 16 Adirondack monitoring
lakes showed decreasing trends in Al; concentrations at rates that ranged from 0.02 (iM/yr to 0.18 (iM/yr
(Driscoll et al., 2007a). From 1992 to 2004, 24 of the 48 lakes showed decreasing trends in Al;
concentrations (Driscoll et al., 2007a). The analysis of Stoddard et al. (2003) also observed an average
decrease in Al j concentrations from 1990 to 2000 in the same group of Adirondack lakes reported on by
Driscoll et al. (2007a), but observed no trend for this period in New England lakes, Appalachian streams,
or Midwest lakes.
Monthly stream chemistry monitoring at the HBEF showed decreases in Al; concentrations at four
locations along the reference stream for the experimental forest from 1982 to 2000, but no trends at two
other locations along this stream (Palmer et al., 2004). These data also showed a surprising decrease in pH
at two of the locations where Al; decreased, and no pH trend at the other two locations where Al;
decreased (Palmer et al., 2004). Comparison of total Al concentrations in 130 lakes in the northeastern
U.S. in 1984 with those measured in 2001 showed lower average concentrations in 2001 in the
Adirondack region, the Catskill/Pocono region, central New England, southern New England, and Maine
(Warby et al., 2005). Because these measurements are of total Al, they are not directly comparable to Al;.
Most recently, Lawrence et al. (in review) found that 49 of 195 streams (25%) during August base flow in
the western Adirondack region had Al j concentrations above 2.0 (.iM. the level above which toxic effects
on biota have been shown (Driscoll et al., 2001b; Baldigo et al., 2007).
Acid Neutralizing Capacity
In response to reduced levels of acidic deposition required by the CAA and other emissions control
legislation, Stoddard et al. (2003) found trends during the 1990s toward increasing Gran ANC (Figure B-
21) in all of the glaciated regions of the eastern U.S. (i.e., New England, Adirondacks, Northern
Appalachian Plateau) and Upper Midwest, and decreasing Gran ANC in the Ridge/Blue Ridge province.
Changes were relatively modest compared with observed reductions in S042 concentrations. Only the
regional increases in the Adirondacks, Northern Appalachian Plateau, and Upper Midwest were
statistically significant (Figure B-21). Median increases of about +1 (ieq/L/yr in the Northern Appalachian
Plateau, Adirondacks and Upper Midwest represent significant movement towards ecological recovery
from acidification (Stoddard et al., 2003).
It has been hypothesized that decreases in acidic deposition will yield the most chemical recovery
in lakes and streams that have experienced the most severe acidification. Using data from all of the sites
in regions where decreases in surface water S042 and N03 have occurred, Stoddard et al. (2003) found
that acidic lakes and streams exhibited a highly significant median increase in Gran ANC of
+1.3 (ieq/L/yr during the 1990s. Low-ANC sites (0 to 25 j^icq/L) showed a smaller significant median
ANC increase of +0.8 (ieq/L/yr. Moderate ANC sites, those with mean ANC values greater than 25 (ieq/L,
showed no significant change in Gran ANC (Figure B-21).
B-78
-------
All of the glaciated regions in the eastern U.S. showed declines in base cation (Ca2+ + Mg2+)
concentrations during the 1990s, with the average changes in the range of-1.5 to -3.4 (ieq/L/yr. All of
the regional trends were highly significant (Figure B-21). Across the eastern U.S., surface water S042 has
decreased at a rate of about -2.5 (ieq/L/yr (the mean of regional median slopes), and N03 at a rate of
-0.5 (ieq/L/yr, in surface waters on glaciated terrain during the 1990s. These rates of change set an upper
limit to our expectation of ANC recovery of +3 (ieq/L/yr (i.e., the sum of S042 and N03 trend
magnitudes). The Gran ANC increase reported by Stoddard et al. (2003) was actually about one-third of
this maximum, +1 (ieq/L/yr. The difference between the observed Gran ANC trend and the maximum
trend estimated from rates of acid anion change can largely be explained by the average regional median
decline in (Ca2+ + Mg2+ concentrations, which was about -2.0 (ieq/L/yr (Stoddard et al., 2003).
Episodic Effects
Episodic acidification can result naturally from the mobilization of organic acids and from dilution
of base cation concentrations, but decreases in pH and ANC associated with increases in S042 and N03
are largely attributable to acidic deposition (Wigington et al., 1996). Episodic acidification is most
common in the early spring and late fall as a result of snowmelt and rainstorms, and is least common in
summer, when high flows tend to be infrequent. Seasonal variations in stream flow also result in seasonal
patterns of surface water chemistry at base flow. Lakes and streams at base flow tend to be more acidic in
early spring than at other times of the year and low flows in late summer tend to be least acidic (Lawrence
et al., 2007).
The transient nature of high flows makes episodic acidification difficult to measure. Therefore,
assessments have generally estimated the number of lakes and streams prone to episodic acidification by
combining episode information from a few sites with base flow values of ANC determined in large
surveys (Eshleman et al., 1995; Bulger et al., 2000; Driscoll et al., 2001b). Inclusion of episodically
acidified water bodies in regional assessments substantially increases estimates of the extent of surface
water acidification. For example, baseflow samples collected from 1991 to 1994 through the U.S. EPA
TIME Program indicated that 10% of the 1,812 lakes (>1 ha surface area) in the Adirondack region of
New York could be considered chronically acidic on the basis of ANC values less than 0 (ieq/L, but that
an additional 31% of these lakes had baseflow ANC values less than 50 |_icq/L and were, therefore,
estimated to be susceptible to episodic acidification (Driscoll et al., 2001b). Lawrence (2002) also
estimated the extent of episodically acidified stream reaches in a Catskill, NY watershed (area = 85 km2)
through the use of an index site at the base of the watershed that became episodically acidified at high
flows. Upstream sites with a lower base flow ANC than the index site at the same date and time were
found to have a high likelihood of becoming episodically acidified. Base flow sampling of 122 upstream
sites indicated that approximately 16% of the total upstream reaches were chronically acidified (ANC
<10 (ieq/L), but that 66% of the stream reaches became episodically acidified.
Stoddard et al. (2003) compared seasonal data from New England lakes, Adirondack lakes and
Northern Appalachian streams, collected monthly to quarterly, to evaluate the difference between the
chemistry of surface waters in the summer and in the spring. Results indicated that spring values of ANC
were an average of 30 j^icq/L lower than summer ANC. This study referred to samples collected in spring
as "episodic samples," although sampling was done independent of flow. Therefore, the 30 j^ieq/L
difference should be considered a seasonal effect rather than an episodic effect.
The most thorough characterization of episodic variations in stream chemistry was conducted
through the ERP, in which 13 low-order streams (watershed areas less than 24 km2) in the Adirondack and
Catskill regions of New York, and the Appalachian Plateau in Pennsylvania were monitored from 1988 to
1990 (Wigington et al., 1996). Acid episodes with chemical concentrations within the 90th percentile
involved decreases in ANC of up to 200 (ieq/L, decreases in pH of up to one unit, and increases in
concentrations of Al; of up to 15 (.iM (Wigington et al., 1996). Results also showed that acid episodes
reduced the size of fish populations and eliminated acid-sensitive species if median high-flow pH was less
than 5.2 and Al; concentrations exceeded 3.7 |_iM. despite the relatively short duration of episodes (Baker
B-79
-------
et al., 1996). Baker et al. (1996) concluded that effect on biota from episodic acidification were likely to
be similar to those from chronic acidification. Elimination of an annual age class can result from an
episode that occurred in the presence of a sensitive life stage. Largely on the basis of this study, the EPA
concluded that reversal of effects from episodic acidification could be used as a key ecological endpoint
for an acid deposition standard for protection of the environment (U.S. EPA, 1995a).
Despite the significance of the findings of ERP, little assessment or monitoring of episodes was
done in the 1990s. One exception was the work of Hyer et al. (1995) in three watersheds of differing
geology in Shenandoah National Park. Results suggested that episodic acidification was occurring
throughout the park on all bedrock types, although acidification was not sufficient to cause elevated Al
concentrations. Lawrence (2002) also documented severe episodic acidification in August 1998 in a
tributary of an ERP stream, where Al; concentrations increased from 1.6 to 7.3 (.iM in 6.5 h.
In the first large-scale study designed to sample streams during high-flow conditions, Lawrence
et al. (2007) found that 124 out of 188 (66%) western Adirondack streams were prone to acidification to
the level at which Al; becomes mobilized. Only streams accessible with less than a 60-min hike were
sampled in this study. The March, 2004 survey was chosen to represent episodic conditions, and a survey
conducted August 16-18, 2004 was chosen to represent base flow conditions. Based on this comparison,
35% of the streams were chronically acidified, 30% of the streams were episodically acidified, and 34%
were not acidified. Survey results were also used to estimate that 718 km of stream reaches were prone to
acidification, although 3085 km of stream reaches within the study region could not be assessed because
of inaccessibility.
There have been no studies in the U.S. that determine if either the severity or frequency of episodic
acidification has lessened. In a study of two streams in Nova Scotia (Laudon et al, 2002), trends in ANC
in four phases of storm hydrographs from 1983 to 1998, were not detected other than in the peak-flow
phase of one stream (an increase of 0.87 (ieq/L). In Sweden, the anthropogenic contribution to episodic
decreases in ANC were estimated to range from 40 to 80% in five streams from 1990 to 1999 (Laudon et
al, 2002).
B.5. Effects on Watersheds and Landscapes
B.5.1. Interactions among Terrestrial, Transitional, and Aquatic
Ecosystems
Acidification has pronounced effects on nutrient cycling in terrestrial, transitional, and aquatic
ecosystems. Of particular importance in this regard is the role of N deposition in influencing N cycling.
This topic is discussed in detail in Section 5. Also important are the influences of acidification on the
availability of Ca2+ and other nutrient base cations (Mg2+, K+).
In general, decomposition, nutrient cycling, productivity, and other system-level processes in
surface waters are not as sensitive as species composition and richness to relatively small amounts of
acidification. Such effects only seem to occur at high levels of acidification (e.g., pH <5). This is because
acid-sensitive species are often replaced by more acid-tolerant species that perform the same function
until acidification becomes severe. For example, whereas changes in microbial composition and
abundance have been observed with acidification, they appear to have minimal effect on overall microbial
respiration and nutrient cycling. At extreme levels of acidity, however, these system-level functions may
also decrease. Thus, system-level functions are not generally good indicators of light to moderate levels
of acidification.
B-80
-------
Integrating the effects of atmospheric deposition across spatial scales is difficult. The response of a
single plant, or small group of plants, cannot be easily scaled up to examine effects on plant communities,
ecosystems, watersheds, or geographic regions. Integration typically requires a combination of
approaches, including ecosystem modeling, experimental manipulation studies, surveys across pollution
deposition gradients, and long-term monitoring studies. Similarly, aquatic effects at the population level
can be readily quantified, but extrapolation to the community or aquatic ecosystem level is problematic.
Linking research results across scales will be an important component of future research. Measurements
that correlate with ecosystem processes, such as foliar N concentration, leaf area index, or spectral
reflectance, can in some cases be remotely sensed. They offer great promise for future assessment of
terrestrial effects across spatial scales.
Effects of atmospheric deposition of acidifying substances on soil, vegetation, and surface water
are manifested in specific processes, affecting energy, water and nutrient flow, intra- and inter-species
competitive interactions, and ecosystem primary production. Therefore, effects on sensitive species (only
some of which have been documented) have the potential to cascade throughout the ecosystem and
become manifested at a variety of scales. Such ecosystem- to landscape-scale effects from atmospheric
deposition of acidifying substances are known to occur, but the results of these interacting processes have
not been conclusively demonstrated.
It is also evident that acidification from natural and human-caused disturbances, including climatic
stressors (temperature, moisture availability, wind), insect infestation, disease, fire, and timber harvest,
can affect the severity of effect of atmospheric deposition of SOx and NOY. Although it is clear that such
interactions can occur, there are no studies that have clearly documented that acidic deposition at levels
that commonly occur across broad landscapes in the U.S. has conclusively altered ecosystem structure or
function. Similarly, although it is widely believed that acidic atmospheric deposition can make plants
more susceptible to the adverse effects of other natural and human-caused stressors, such effects have not
been conclusively demonstrated in more than a few cases. The data demonstrating and quantifying the
extent to which SOx and NOY deposition are altering natural terrestrial ecosystems via acidification
pathways are sparse. In particular, effects of soil and soil water acidification on soil ecosystem processes
and nutrient cycling are poorly known. Even less is known about effects on soil microorganisms and food
webs, or how such effects interact with the above-ground vegetation community.
B.5.2. Interactions with Land Use and Disturbance
The prevailing scientific consensus during the 1980s held that most lakes in eastern North America
that had pH less than about 5.5 to 6.0 had been acidified by acidic deposition. Reports that acidic lakes
and streams were rare or absent in similar areas not receiving acidic deposition were used as evidence of
acidification by acidic deposition in many regions (e.g., Baker et al., 1991b; Murdoch and Stoddard,
1992; Neary and Dillon, 1988; Sullivan et al., 1988). An alternative hypothesis had been advanced by
Rosenquist (1978), Krug (1989, 1991), and Krug and Frink (1983) that land use changes could explain
recent lake acidification in southern Norway and the northeastern U.S. According to this hypothesis,
natural soil processes and changes in vegetation can generate more acidity than is received from
atmospheric deposition. For example, an increase in acidic humus formation in response to decreased
upland agriculture was suggested as being responsible for regional acidification in southern Norway,
rather than acidic deposition (Rosenqvist, 1978). Subsequent acidic deposition effects research in some
cases seemed to be designed to refute this hypothesis rather than to explore the relationships between land
use and acid-base chemistry (cf. Havas et al., 1984; Birks et al., 1990). Research intended to discriminate
between acidic deposition and land use as the major cause of acidification generally concluded that acidic
deposition was the principal cause of regional acidification in certain areas of North America and Europe.
Perhaps more appropriate research questions might have focused on quantifying the relative importance
of land use activities or landscape change in exacerbating or ameliorating acidic deposition effects. The
B-81
-------
importance of acidic deposition as an agent of acidification does not preclude the importance of land use
and landscape changes which, in some cases, may actually be more important than acidic deposition
(Sullivan et al., 1996b).
There has not been a regional evaluation of land use changes in areas of the U.S. susceptible to
surface water acidification from acidic deposition. It has therefore not been possible to quantify the extent
or magnitude of land use effects on acidification. It is clear, however, that such changes can have
important effects on acid-base status (Sullivan, 2000a), especially as influenced by N deposition (Goodale
and Aber, 2001).
Changes in human land use activity, and associated changes in vegetative structure, influence
ecosystem response to external stressors such as acidic deposition, exposure to O3, natural disturbance
factors such as wind and fire, and climatic changes. Some activities contribute to the acidification of soil
and surface waters; other activities decrease acidity (Sullivan et al., 1996b) (Table B-21).
Forest management practices, especially those that have occurred over many generations of trees,
can have important effects on soil erosion, nutrient supplies, and organic material. Such effects can
influence the availability of base cations for acid neutralization and/or aspects of N cycling.
Forests are efficient at scavenging S and N from the atmosphere. Differences in forest canopy,
particularly between deciduous and coniferous trees, can cause large differences in dry deposition, and
therefore total deposition of S and N. In regions that receive high levels of acidic deposition, the presence
of forest vegetation, especially coniferous trees, enhances total deposition of acid-forming precursors
(Rustad et al., 1994). In addition to the enhanced deposition caused by the presence of large trees, there
are also differences in nutrient uptake. In particular, younger trees take up larger quantities of N and other
nutrients than do trees in older forests. Therefore, changes in the occurrence and age or species
composition of the forest can influence the rates of atmospheric deposition to the site as well as the fate of
atmospherically deposited substances.
Landscape processes and watershed disturbance can influence soil and water acidification in many
ways. Land use practices and vegetation patterns have been changing in various parts of the U.S. for
decades to centuries. These changes in human activity can influence the response of forested ecosystems
to external stressors, including atmospheric deposition of S or N, natural disturbance factors such as wind
and fire, and climatic changes. Some processes contribute to the acidification of soil and surface waters or
reduce the base saturation of the soils thereby increasing their sensitivity to acidic deposition. Other
processes cause decreased acidity (Sullivan et al., 1996b; Sullivan, 2000a).
Watershed disturbance from logging, blowdown, and fire disrupts the normal flow of water and can
cause increased contact between runoff water and soil surfaces, leading to increased base cation
concentration and ANC in drainage water. Recovery from disturbance can cause a decrease in drainage
water ANC as the system returns to pre-disturbance conditions. In particular, soil loss through erosion can
reduce the base cation pool size, thereby limiting the capacity of soils to neutralize atmospheric acidity. In
addition, forest harvesting has an important effect on forest N-demand, thereby reducing the likelihood of
future N-saturation in response to high N deposition. Forest management practices, especially those that
have occurred over many generations, have had important effects on soil chemistry, nutrient supplies, and
organic material.
Watershed disturbances, including road building, agriculture, mining, urbanization, logging,
blowdown, and fire can alter various aspects of ecosystems biogeochemistry. Such disturbances can
influence the water budget, base cation mobilization, routing of drainage water, nutrient input, and S and
N cycling in ways that affect the acid-base chemistry and nutrient dynamics of soils and drainage waters
(Sullivan et al., 1996b). The effects of such disturbances can greatly modify the response of a given
watershed to atmospheric inputs of S and N.
B-82
-------
B.5.2.1. Timber Harvest
Removal of the forest affects drainage water quality in several ways. Deposition of S and N are
reduced; leaching of N03 increases and, in some cases, causes a pulse of surface water acidification.
Base cations tied up in wood are lost when wood is transported off-site. Regrowth of the forest may
further affect drainage water quality through vegetative uptake of N and base cations. Trees accumulate
base cations to a greater degree than anions. To balance the charge discrepancy, roots release an
equivalent amount of protons and acidify the soil. Base cation accumulation by trees is age-dependent.
Young forests grow faster and are therefore more acidifying than older forests (Nilsson et al., 1982;
Nilsson, 1993). They also retain greater amounts of N.
Most forests in the northeastern U.S. are recovering from extensive human disturbance that
occurred over a period of about 200 years. Landscapes were mainly forested during pre-colonial times,
logged or cleared for agriculture in the mid to late 19th century, and are now largely early to mid-seral
stage regenerating forests (Niering, 1998).
In some areas that experience relatively high levels of acidic deposition, there is growing concern
about sustainable timber productivity (Adams, 1999). Harvest-induced leaching losses have been
estimated to range from 6 to 60 kg/ha/yr of N, 28 to 48 kg/ha/yr of Ca2+, and 7 to 16 kg/ha/yr of Mg2+
(Federer et al., 1989). Timber harvesting also increases leaching losses from the site because of the
reduction in transpirational water loss. The increased water flux after tree removal increases the
opportunity to leach base cations from the soil.
If base cations sequestered in tree wood are removed from the site by tree harvesting, the result is a
decrease in the available base cation pool on the site. Physical disturbances to forest soils during logging
operations and increased soil temperature that results from exposure of the forest floor to sunlight may
also cause a short-term increase in the rates of N mineralization and nitrification (Joslin et al., 1992). The
resulting increase in N03 production and leaching further depletes base cations from the soil pool.
Johnson et al. (1991a) measured short-term (3 years) effects of logging at HBEF in New
Hampshire on soil acid-base chemistry. Base saturation of the mineral soil Bh horizon decreased from 14
to 11% and pH decreased by 0.24 pH units.
Likens et al. (2002) reported results of a 34-year study of the biogeochemistry of forest ecosystems
at HBEF. Part of the study evaluated the effects of tree removal on S cycling and related biogeochemical
processes. Vegetation removal resulted in increased decomposition of organic matter and nitrification.
These changes, in turn, lowered soil water pH, enhanced S042 adsorption on mineral soil, and therefore
decreased the flux of S042 in stream water. With subsequent vegetation regrowth, the adsorbed S042
was released from the soil to drainage water, and streamwater S042 concentrations increased.
Baldigo et al. (2005) compared the effects of clear-cut and timber-stand improvement (TSI)
harvests on water chemistry and mortality of caged brook trout in three Catskill Mountain streams.
Harvests removed 73% of tree basal area from a clearcut subbasin, 5% basal area from a TSI subbasin,
and 14% basal area at a site below the confluence of both streams (the combined effect of the two harvest
methods). Water quality and trout mortality were affected only in the clearcut stream. Acidity and
concentrations of NO;, and Al; increased sharply during high flows after the first growing season (1997).
Acid-Alj episodes were severe during this period and decreased steadily in magnitude and duration
thereafter. All trout at the clearcut site died within 7 days during spring 1998, and 85% died during spring
1999. Only background mortality was observed in other years at this site and at the other three sites
during all tests. The effects of tree harvests on fish communities are of concern because they might
interact with effects of acidic deposition and produce more substantial effects on biota than either stress
factor on its own.
B-83
-------
B.5.2.2. Insect Infestation
Effects of insect-caused defoliation on the N cycle can be pronounced. The foliar N consumed by
insects is deposited on the forest floor as insect feces (frass), greenfall, and insect biomass. Some of this
deposited N is subsequently taken up by tree roots and soil microbes, with little effect on the nutritional
condition of the trees or the site. Where a sizable component of this N is leached in drainage water, the
nutritional consequences can be more significant. There are also various feedback mechanisms. For
example, low N supply can slow the population growth of defoliating insects (Mason, 1992) and enhance
the tree's chemical defenses against insects (Hunter and Schultz, 1995). The amount of N leaching loss is
generally small, relative to atmospheric deposition inputs and relative to the amount of N transferred to
the forest floor with the defoliation (Lovett and Ruesink, 1995; Lovett et al., 2002). Nevertheless, it can
be high enough to contribute to base cation depletion of soils and effects on downstream receiving waters.
The extent of N 03 leaching may be partly related to the extent of defoliation and tree mortality that
occurs and also the amount of precipitation that occurs immediately after the defoliation (Lovett et al.,
2002).
Forest insect infestation can have profound effects on the acid-base and nutrient chemistry of soils
and drainage waters. Effects of a gypsy moth infestation in Shenandoah National Park provide a good
example. Between the mid-1980s and the early 1990s, the southward expanding range of the European
gypsy moth traversed Shenandoah National Park and affected all of the University of Virginia's SWAS
study watersheds (Webb, 1999). Some areas of the park were heavily defoliated 2 to 3 years in a row. The
White Oak Run watershed, for example, was more than 90% defoliated in both 1991 and 1992. This
insect infestation of forest ecosystems in Shenandoah National Park resulted in substantial effects on
streamwater chemistry. The most notable effects of the defoliation on park streams were dramatic
increases in the concentration and export of N and base cations in streamwater. Following defoliation,
N03 export increased to previously unobserved levels and remained high for over 6 years before
returning to predefoliation levels. The very low levels of pre-disturbance N03 export in park streams
were consistent with expectations for N-limited, regenerating forests (Aber et al., 1989; Stoddard, 1994).
Release of N03 to surface waters following defoliation was likewise consistent with previous
observations of increased N export due to forest disturbance (Likens et al., 1970; Swank, 1988). The exact
mechanisms have not been determined, but it is evident that the repeated consumption and processing of
foliage by the gypsy moth larva disrupted the ordinarily tight cycling of N in Shenandoah National Park
forests.
Although N is thought to play an important role in the chronic acidification of surface waters in
some areas (Sullivan et al., 1997), the elevated concentrations of N03 in Shenandoah National Park
streams following defoliation did not appear to contribute to baseflow acidification in White Oak Run.
This was due to a concurrent increase in concentrations of base cations in streamwater (Webb et al.,
1995). Both N03 and base cation concentrations increased during high-runoff conditions, although the
increase in base cations did not fully compensate for the episodic increase in N03 . Episodic acidification
following defoliation thus became more frequent and more extreme in terms of observed minimum ANC
(Webb et al., 1995).
The full effect of the gypsy moth on aquatic resources in Shenandoah National Park is not well
understood. One consequence may be a reduction in the supply of available soil base cations and
associated effects on streamwater ANC. Repeated periods of defoliation would probably increase the
effect of episodic acidification on sensitive aquatic fauna and may determine the conditions under which
some species are lost. Ultimately such effects may depend upon both the severity of future gypsy moth or
other insect outbreaks and possibly on the amount of atmospheric N deposition. Gypsy moth populations
typically display a pattern of periodic outbreaks and collapse (Cambell, 1981). It remains to be seen what
the long-term pattern will be (Sullivan et al., 2003).
Webb et al. (1994) compared pre- and post-defoliation streamwater chemistry for 23 VTSSS
watersheds. Nitrate concentrations, measured quarterly, increased in most of the streams in response to
defoliation, typically by 10 to 20 j^icq/L or more. The increased streamwater N03 concentration was
B-84
-------
probably derived from the N content of the foliage that had been consumed by the gypsy moth larvae and
converted to feces on the forest floor. Sulfate concentrations and ANC also decreased in streamwater.
Although the mechanism for decreased S042 was not totally clear, Webb et al. (1994) hypothesized that
increased nitrification in response to the increased soil N pool may have caused soil acidification, which
would be expected to have increased soil S adsorption (Johnson and Cole, 1980). Decreased S deposition
during the comparison period may also have contributed to the S042 response.
Eshleman et al. (1998) reported N03 outputs from five small (<15 km2) forested watersheds in
Virginia and Maryland from 1988 to 1995. The study watersheds varied in geology, vegetation, and acid
sensitivity, with baseflow ANC typically in the range of 0 to 10 j^ieq/L in Paine Run to the range of 150 to
350 (ieq/L in Piney River. Oak species (Quercus spp.), which are a preferred food source of gypsy moth
larvae, occupied about 60% to 100% of the study watersheds. Nitrate concentrations increased in at least
three of the watersheds in association with intense defoliation by the gypsy moth larva during the late
1980s to early 1990s, to peak annual average N03 concentrations of about 30 to 55 j^icq/L. Most of the
increased N03 leaching occurred during storm flow conditions.
A number of other studies have been conducted that examined the effects of gypsy moth, or other
forest insect pests, on watershed biogeochemistry. Defoliation of poplars (Populus sp.) by gypsy moth
larvae in southwestern Michigan did not result in appreciable N03 leaching (Russell et al., 2004).
Other pest species can have similar effects. For example, spruce-fir forests throughout the southern
Appalachian Mountains have been subjected to significant disturbance, especially from the balsam wooly
adelgid, a European pest which has infested Fraser fir since about the 1960s. Severe fir mortality has
occurred in many areas. This disturbance factor has the potential to interact with acidic deposition and
other ecosystem stresses, and contribute to multiple-stress tree mortality and to changes in
biogeochemical cycling.
Defoliation by the elm spanworm (Ennomos subsignarius Hiibner) larvae in old-growth hemlock-
hardwood forests on the Allegheny High Plateau of northwestern Pennsylvania increased streamwater
N03 concentrations from pre-defoliation levels of about 29 j^icq/L to peak values the summer after
defoliation of about 100 |_icq/L (Lewis and Likens, 2007).
B.5.3. Wind or Ice Storm Damage
Forest blowdown might affect surface water acid-base chemistry by changing the pathway
followed by drainage water through watershed soils (Dobson et al., 1990). Pipes formed in the soil by
decaying tree roots can alter hydrologic flow so that less water enters the soil matrix, where neutralization
processes buffer the acidity of rainwater and snowmelt. Pipes tend to occur most commonly in near-
surface soil horizons where most tree rooting occurs. Contact between drainage water and mineral soil is
reduced when runoff is routed through them. If enhanced pipeflow, resulting from sudden extensive tree
mortality, affects a large portion of a watershed, runoff water may have less opportunity for acid
neutralization than would be the case in the absence of such pipeflow.
Severe canopy damage occurred in 1998 in response to an ice storm at HBEF and surrounding
areas in the White Mountains. Houlton et al. (2003) reported effects of this disturbance on N cycling and
leaching losses. Subsequent to the ice storm, drainage water N03 concentrations increased sevenfold to
tenfold. Peak streamwater N03 concentrations during spring months reached or exceeded 50 j^ieq/L at
many sites. There were no significant differences, however, in N mineralization, nitrification, or
denitrification rates between damaged and undamaged areas. Houlton et al. (2003) interpreted these
results as an indication that increased N03 leaching was probably due to decreased root uptake rather
than accelerated N cycling by soil microbes. The amount of N03 leaching loss was estimated to be more
than half of the entire year's worth of atmospheric N deposition.
B-85
-------
B.5.3.1. Fire
Fire can increase concentrations of N03 and S042 in soils and drainage water (cf. Chorover et al.,
1994; Riggan et al., 1994). Fenn and Poth (1998) hypothesized that successful fire suppression efforts
may have contributed to the development of N-saturation in fire-adapted ecosystems in southern
California by allowing N to accumulate in soil and in the forest floor, and by maintaining dense mature
stands with reduced N demand.
The effects of fire on N03 leaching in chaparral stands in the San Gabriel Mountains, CAthat
received high atmospheric N deposition were investigated by Riggan et al. (1994). Study watersheds were
burned with fires of different intensity and, after rainfall, N03 and NH4+ were measured in watershed
streams. Nitrogen release was up to 40 times greater in burned watersheds than in unburned watersheds,
and the amount and concentration of N release were found to be related to fire intensity.
Chorover et al. (1994) evaluated the effects of fire on soil and stream water chemistry in Sequoia
National Park, CA. Burning increased concentrations of N03 and S042 in soil water and stream water.
Sulfate concentrations increased 100 fold. Nitrate concentrations also increased and remained higher in
soils and stream water for about 3 years. These results suggest that successful fire suppression may have
contributed to the development of N saturation in fire-adapted ecosystems in southern California by
allowing N to accumulate in the soil and forest floor, and by maintaining dense mature stands with
reduced N demand (Fenn and Poth, 1998).
B.5.3.2. Multiple Stress Response
Acidification-related effects of S and N deposition do not occur in isolation; they interact with
disturbances of various types, both natural and human-caused. They also influence a range of
biogeochemical processes that may be difficult to predict. Overall, the interactions between disturbance
and ecosystem acidification as a consequence of acidic deposition are not well understood.
It is believed that high rates of N deposition cause increased susceptibility of forests to other
stressors, including reducing the resistance of some tree species to frost, insect damage, or drought. The
effects of acidic deposition can interact with a variety of stressors, both natural and human-caused. The
end result might include adverse effects that would not occur solely in response to acidic deposition, or in
response to any one of the other stressors.
Watershed disturbance might also effect Hg cycling and its relationship to S deposition. For
example, Garcia and Carignan (2000), in a study of 20 watersheds in Quebec, Canada, found that the
average Hg concentration in 560-mm northern pike (A'SOx lucius) was significantly higher in lakes
whose watersheds had recently (1995) been logged (3.4 jj.g/g), as compared with reference lake
watersheds (1.9 jj.g/g), that had remained undisturbed for at least 40 years. Fish tissue Hg concentration
also increased with increasing DOC and lakewater S042 concentration, and with decreasing pH.
B.6. Ecological indicators of acidification
B.6.1. Biological Indicators
Surface water acidification from acidic deposition causes effects on organisms at all trophic levels.
Early studies focused on the loss of fish populations, especially salmonids. Later studies also reported that
many species of phytoplankton, zooplankton, insect larvae, crayfish, snails, and freshwater mussels are
sensitive and are often reduced or absent from acidified lakes and streams (Havas, 1986; Baker et al.,
B-86
-------
1990b). Similarly, many species of microrhizal fungi and lichens have been reported to be particularly
sensitive to acidic deposition in terrestrial ecosystems.
Effects of acidification on aquatic biota have been demonstrated in laboratory and field bioassays
(e.g., Baker et al., 1996), whole-ecosystem acidification experiments (e.g., Schindler et al., 1985), and
field surveys (e.g., Baker and Schofield, 1982; Gallagher and Baker, 1990). Many of the species that
commonly occur in acid-sensitive surface waters susceptible to acidic deposition cannot reproduce or
survive if the water is acidic. Some sensitive species of fish, invertebrates, and algae cannot survive at
moderate levels of acidity. For example, some zooplankton predators, sensitive mayfly species, and
sensitive fish species are affected at pH values below the range of 5.6 to 6.0 (Baker and Christensen,
1991). Such pH values generally equate to ANC below about 25 to 50 j^ieq/L.
There are few published examples of long-term monitoring data for biological assemblages in acid-
sensitive surface waters, and none in the U.S. Therefore, conclusions about the effect of acidic deposition
on the distribution of sensitive species are based on other kinds of data (Stoddard et al., 2003). For
example, the number of fish species increases with increasing pH and ANC when evaluated for multiple
water bodies across the landscape. This result has been shown for streams in Virginia, lakes in the
Adirondacks, and both high-elevation and seepage lakes in Maine (Figure B-17).
Given the available data, it is clear that acidification from acidic deposition limits the distribution
of acid-sensitive fish, benthic invertebrate, phytoplankton, and zooplankton species, but a lack of
adequate data makes it difficult to quantify the magnitude of change in biota from historical condition or
in response to recent (past two to three decades) decreases in acidic deposition in individual lakes or
streams. Studies in Canada and Europe have illustrated the feasibility and complexity of biological
recovery in response to decreased acidity.
Threshold pH levels for adverse biological effects have been summarized for a variety of aquatic
organisms (Haines and Baker, 1986; Baker et al., 1990b). The effects of low pH are specific to the
organism, and perhaps region, under consideration and depend also upon the concentrations of other
chemical constituents in the water, notably Al; and Ca2+. In general, populations of salmonid fish are not
found at pH levels less than 5.0, and smallmouth bass (Micropterus dolomieu) populations are usually not
found at pH values less than 5.2 to 5.5 (Haines and Baker, 1986). A number of synoptic surveys indicate
loss of species diversity and absence of many other fish species in the pH range of 5.0 to 5.5 (Haines and
Baker, 1986). Levels of pH less than 6.0 to 6.5 have been associated with adverse effects on populations
of dace, minnows, and shiners (family Cyprinidae), and bioassays suggest that given sufficient Al
concentrations, pH less than 6.5 can lead to increased egg and larval mortality in blueback herring (Alosa
aestivalis) and striped bass (Morone saxatilis) (Hall, 1987; Klauda et al., 1987).
Mycorrhizal fungi have been suggested as possible biological indicators of atmospheric deposition
effects by Lokke et al. (1996) because they are intimately associated with tree roots, depend on plant
assimilates, and play essential roles in plant nutrient uptake. Thus, mycorrhizal fungi can influence the
ability of their host plants to tolerate different anthropogenically generated stresses. Mycorrhizae and
associated fine roots have short lifespans and their turnover appears to be controlled by environmental
factors. Changes in mycorrhizal species composition, or the loss of dominant mycorrhizal species in areas
where diversity is already low, may cause increased susceptibility of plants to stress (Lokke et al., 1996).
Mycorrhizal fungi are dependent for their nutrition on the supply of assimilates from the host plant.
Stresses that shift the allocation of C reserves to the production of new leaves at the expense of supporting
tissues will be reflected rapidly in decreased fine root and mycorrhizal biomass (Winner and Atkinson,
1986). Decreased C allocation to roots could also affect soil carbon and rhizosphere organisms. For
example, earthworms are believed to decrease in abundance, and in species number, in acidified soils
(Lokke et al., 1996). Soil dwelling animals, including earthworms, are important for decomposition, soil
aeration, and nutrient redistribution in the soil. They contribute to decomposition and nutrient availability,
mainly by increasing the accessibility of dead plant material to microorganisms.
B-87
-------
B.6.1.1. Phytoplankton
Phytoplankton are the small microscopic plants or plant-like organisms that live suspended in the
water column of lakes and large rivers. Acidification results in decreased species richness and diversity of
phytoplankton communities. There is also a shift in the composition of dominant taxa, but species
composition shifts cannot be accurately predicted (though it is clear that community restructuring occurs
with acidification). This effect is most prevalent in the pH 5 to 6 range (Baker et al., 1990b). Acidification
has also been found to cause decreases in food web complexity (indicated by the number of trophic links
or species) in the Adirondack Mountains (Havens and Carlson, 1998). Both Al toxicity and P limitation
may also be responsible for shifts in phytoplankton community composition. Neither grazing pressure nor
changes in water clarity associated with acidification seem to have a major effect on phytoplankton
community structure. There is no consistent pattern of acidification effects on phytoplankton biomass.
Various lakes have shown increases, decreases, or no change in phytoplankton biomass with acidification
(Baker et al., 1990b). Leavitt et al. (1999) suggested that the complex interactions between pH, DOC, and
light explain the high variability in the algal biomass-acidification relationship. In most lakes,
acidification has a negligible effect on primary productivity.
Diatoms, which comprise an important component of the phytoplankton, are excellent indicators of
environmental change in aquatic ecosystems, including acidity, nutrient status, salinity, and climatic
features (Stoermer and Smol, 1999; Sullivan and Charles, 1994). There are thousands of different species,
many of which have rather narrow ecological tolerance ranges. Diatoms have been widely used as
indicators of past lake acidification. Inference based on diatom fossil remains preserved in lake sediments
is an excellent approach for quantifying historical chemical change (Charles and Norton, 1986).
Paleolimnological reconstructions of past lakewater chemistry are based on transfer functions
derived from relationships between current lakewater chemistry and diatom (or, in some cases,
chrysophyte) algal remains in surface sediments. Predictive relationships are developed using regional
lake datasets, and are then applied to diatom assemblage data collected from horizontal slices of lake
sediment cores to infer past lakewater conditions (Charles et al., 1990a; Husar and Sullivan, 1991).
Periphyton are the small microscopic plants (or plant-like organisms) that live on submerged
substrates in aquatic systems (e.g., stream or lake bottoms). As seen in phytoplankton communities,
acidification results in decreased species richness, community alteration, and emergence of new dominant
species in periphyton communities. Many diatom and blue-green bacterial periphyton species cannot
tolerate acidic conditions. On the other hand, green algae, particularly the filamentous Zygnemataceae,
increase in relative abundance at lower pH (Baker et al., 1990b). Unlike for phytoplankton, there is
evidence that the biomass of attached periphyton increases at lower pH.
Studies of phytoplankton recovery from acidification indicate that there is an increase in
phytoplankton species richness and diversity as pH increases. In the Experimental Lakes area of Ontario,
previously acidified lakes have been experimentally de-acidified. In Lake 223, there was little increase in
phytoplankton diversity as pH changed from 5.0 to 5.8 but a strong recovery of diversity at pH above 6
(Findlay and Kasian, 1996). In Lake 302S, profound change began at pH 5.5; phytoplankton assemblages
at pH below 5.5 resembled acidified lakes. Cyanobacteria were among the first to recover at pH 5.5 to 5.8
(Findlay et al., 1999). In the Killarney Park area of Ontario, Findlay (2003) reported that lakes that were
previously low in pH (5.0 to 5.5) and are now above pH 6 have shifted towards phytoplankton
assemblages typical of circumneutral environments.
B.6.1.2. Zooplankton
Field survey and experimental lake studies both indicate that acidification reduces zooplankton
species richness. Effects of acidification on community biomass and abundance, however, were not
definitive. Some studies indicated a lower biomass under low pH conditions, whereas other studies
showed no consistent pattern in the biomass-pH relationship. Limited data indicated that acidification
B-88
-------
does not alter zooplankton community grazing rates. Zooplankton species that have been shown to be
sensitive to low pH include Diaptomus sicilis, Epischura lacustris, Tropocyclops parsinus mexicanus,
Daphnia galeata mendotae, Daphnia rosea, Diaphanosoma birgei, Leptodora kindtii, Asplanchna
priodonta, and Conochilus unicornis. In North America, species reported to have increased dominance in
acidic lakes (acid-tolerants) include Keratella taurocephala, Bosmina longirostris, and Diaptomus
minutus. Possible mechanisms for zooplankton sensitivity to low pH include ion regulation failure,
reduced 02 uptake, inability to reproduce, and A1 toxicity. Indirect effects of acidification on zooplankton
communities are also possible due to pH-induced shifts in higher trophic level zooplankton predators.
This mechanism is probably of less importance than the direct effects of low pH. It is also probable that
under acidic conditions, zooplankton communities are less able to ameliorate nutrient additions or control
algal densities (Baker et al., 1990b).
Reported pH thresholds for zooplankton community alteration range from 5 to 6. Holt and Yan
(2003) reported a threshold of community change at pH 6 for lakes in southern Ontario. Locke and
Sprules (1994) reported that acidification below pH 5 in the 1970s overcame the resistance stability of the
zooplankton community in Ontario Precambrian Shield lakes. The subset of study lakes that showed pH
recovery from acidification 20 years later in 1990 also showed recovery in the stability of the zooplankton
community. Holt and Yan (2003) also noted recovery in zooplankton community composition (based on
similarity to neutral lakes) in the subset of Killarney Park (Ontario) lakes in which the pH increased to
over 6 during the 1971 to 2000 study period. They did not, however, note any time trend of increasing
species richness between recovering lakes and non-recovering lakes.
Recovery in experimentally acidified Lake 223 back to pH 6.1 was studied by Malley and Chang
(Malley and Chang, 1995). They reported that the zooplankton community was still in a state of flux.
Species diversity that had been reduced during the acidification phase had partially returned to
preacidification levels. Rotifers had recovered less than crustaceans. One decade after cessation of the
experimental acidification of Little Rock Lake in Wisconsin, recovery of the zooplankton community was
complete (Frost et al., 2006). Recovery did not follow the same trajectory as the initial acidification,
however, indicating a substantial hysteresis in zooplankton community recovery. About 40% of the
zooplankton species in the lake exhibited a lag of 1 to 6 years to recover to levels noted in the neutral
reference basin.
In situ enclosure studies were conducted for 35 days at Emerald Lake in the Sierra Nevada by
Barmuta et al. (1990). The lake sediments were included within the experimental enclosures. This allowed
the investigators to document the response of zoobenthos as well as zooplankton. Treatments included a
control (pH 6.3) and acid addition to reach pH levels of 5.8, 5.4, 5.3, 5.0, and 4.7. Results indicated that
zooplankton were sensitive to acidification but zoobenthos were unaffected by the experimental
treatment. Daphnia rosea and Diaptomus signicauda decreased in abundance below the range of pH 5.5
to 5.8 and were eliminated below about pH 5.0. Bosnia longirostris and Keratella taurocephala generally
became more abundant with decreasing pH. Barmuta et al. (1990) concluded that even slight acidification
of high-elevation lakes in the Sierra Nevada might alter the structure of the zooplankton community.
Sullivan et al. (2006a) found that zooplankton taxonomic richness varied with ANC in Adirondack
lakes (Table B-22). Taxonomic richness expressed as number of species of crustaceans, rotifers, and total
zooplankton, increased with increasing ANC. In general, lakewater ANC explained nearly half of the
variation in total zooplankton and crustacean taxonomic richness, but less for rotifer richness. These
results (Table B-22) provided the basis for estimating changes in zooplankton richness in response to past
or future changes in lakewater ANC. Several zooplankton species found in lakes in the Sierra Nevada are
also known to be sensitive to acidity status (Gerritsen et al., 1998).
B.6.1.3. Benthic Invertebrates
Within stream systems, macroinvertebrate communities are among the most sensitive life forms to
disturbances, including those associated with atmospheric deposition (Cairns and Pratt, 1993). In
B-89
-------
addition, they are relatively easy to sample in the field (Plafkin et al., 1989; Resh et al., 1995; Karr and
Chu, 1999).
Acidification results in the loss of acid-sensitive benthic invertebrates and decreases in pH of one
unit or more typically result in species loss. Invertebrate taxa that are most sensitive to acidification are
mayflies, amphipods, snails, and clams. At low levels of acidification (pH 5.5 to 6.0), acid-sensitive
species are replaced by more acid-tolerant species, yielding little or no change in total community species
richness, diversity, density, or biomass. If pH decreases are larger, more species will be lost without
replacement, resulting in decreased richness and diversity. Many sites also note decreases in invertebrate
biomass and productivity (more so in streams than lakes). High levels of acidification (pH <5) were found
to virtually eliminate all mayflies, crustaceans and mollusks from French streams (Guerold et al., 2000).
Examples of sensitive benthic invertebrate species include Baetis rhodani, Gammarus lacustris, Hyalella
azteca, Asellus aquaticus, Orconectes rusticus, and O. propinquous. Stoneflies are generally more acid-
tolerant than mayflies and caddisflies.
Possible mechanisms for acidification effects on invertebrates include direct toxicity of H and Al,
disruption of ion regulation, and reproductive failure. Indirect effects due to acidification-induced changes
to invertebrate predator populations are also possible (Baker et al., 1990b). Acidic episodes in streams can
cause increased downstream drift of acid-sensitive species, particularly Baetis (Kratz et al., 1994; Smock
and Gazzera, 1996).
It has been well documented that low streamwater pH can be associated with reductions in
invertebrate species richness or diversity (Townsend et al., 1983; Raddum and Fjellheim, 1984; Burton
et al., 1985; Kimmel et al., 1985; Hall and Ide, 1987; Peterson and Van Eechhaute, 1992; Rosemond et al.,
1992; Sullivan et al., 2003), and sometimes density (Hall et al., 1980; Townsend et al., 1983; Burton et al.,
1985; Kimmel et al., 1985). Effects on invertebrate density are not universal; a number of studies have
found no density effects (Harriman and Morrison, 1982; Simpson et al., 1985; Ormerod and Tyler, 1987;
Winterbourn and Collier, 1987). However, a decrease in species richness with decreasing pH has been
found in almost all such studies (Rosemond et al., 1992), and this finding has been especially pronounced
in streams for order Ephemeroptera (mayflies).
The Ephemeroptera-Plecoptera-Tricoptera (EPT) Index is a common measure of stream
macroinvertebrate community integrity. The EPT metric is the total number of families present in those
three insect orders (mayflies, stoneflies, and caddisflies, respectively). The total number of families is
generally lower at acidified sites because species within those families tend to exhibit varying acid
sensitivity (USDA, 1996). Mayflies tend to be most sensitive of the three, and stoneflies tend to be least
sensitive (Peterson and Van Eechhaute, 1992).
There has been some recovery in benthic invertebrate communities in surface waters exhibiting
chemical recovery from acidification. In Scotland, Soulsby et al. (1995) reported an increase in acid-
sensitive mayflies in some streams that showed recent ANC increases. However, no increases in
invertebrates were observed in the most acidic streams despite observed increases in ANC. They
suggested that further acidic deposition reductions and sufficient time for reversal of soil acidification
may be required before biotic recovery can occur. Tipping et al. (2002) noted increases of invertebrate
richness and diversity in the English Lake District in their study streams that had pH increases of 0.3 to
0.5 units since about 1970.
Responses of aquatic macroinvertebrates to acidification were evaluated by Kratz et al. (1994) in
12 streamside channels in Sequoia National Park, CA. Replicated treatments included a control (pH 6.5 to
6.7) and experimental exposure at pH levels of 5.1 to 5.2 and 4.4 to 4.6. Invertebrate drift was monitored
continuously and benthic densities were determined before and after acidification. Single 8-h acid pulses
increased the drift of sensitive taxa, and benthic densities were reduced. Baetis showed reduced density
post-treatment to less than 25% of control densities in both pH reduction treatments (5.2, 4.6) and two
different experimental exposures. Densities of Paraleptophlebia appeared to be reduced by the
acidification, but most treatment effects were not statistically significant. Kratz et al. (1994) suggested
that the effects of acid inputs on benthic species densities depended on microhabitat preferences. Baetis
nymphs are epibenthic and active. They are often found on the upper surfaces of rocks where they are
B-90
-------
directly exposed to acidified water. This may have been responsible for their greater response to
acidification.
B.6.1.4. Fish
By 1990 it was well established that pH in the range of 4.0-6.5 could cause significant adverse
biological effects on fish. Low pH was one of the most important factors resulting in adverse effects. The
toxicity of pH was, in most cases, the result of impaired body salt regulation. Decreased pH in the water
inhibited the active uptake of Na+ and CF and stimulated the passive loss of these ions (Baker et al.,
1990b).
The response to acidification was not uniform, however. Some species and life stages experienced
significant mortality in bioassays at relatively high pH (e.g., pH 6.0-6.5 for eggs and fry of striped bass
and fathead minnow) (Buckler et al., 1987; McCormick et al., 1989), whereas others were able to persist
at quite low pH without adverse effect (Mudminnow; [Umbra spp.] at pH 4.0 and Umbra pygmaea at pH
3.5) (Dederen, 1987). Many minnows and dace (Cyprinidae) are sensitive to acidity (threshold effects at
pH <5.5 to 6.0), but some common game species such as brook trout, largemouth bass, and small mouth
bass are relatively insensitive (threshold effects at pH <5.0 to 5.5). A summary of studies that
demonstrated the difference among species is shown in Table B-23. Table B-24 summarizes the results
from a variety of studies that determined the threshold values of pH for various taxa and kinds of effects.
The effect of acidification on aquatic organisms, especially fish, is due in large part to the toxic
effect of Al; that is released from watershed soils. A number of studies reviewed by Baker et al. (1990b)
reported threshold values of Al; for various species and effects. Those results are presented in Table B-25.
The effects of low pH and high Al j can be ameliorated to an extent in the presence of increased Ca2+
concentration. A summary of the effect of increasing Ca2+ concentration is presented in Table B-26.
Fish populations in acidified streams and lakes of Europe and North America have declined, and
some have become extinct as a result of atmospheric deposition of acids and the resulting changes in
water quality (Baker et al., 1990b). A variety of factors, including Al;, DOC, and Ca2+, along with the
timing and magnitude of episodic fluctuations in toxic acid and Al; concentrations, are related to the
degree to which surface water acidification influences fish survival in natural systems (Baker et al.,
1990b; Baldigo and Murdoch, 1997; Gagen et al., 1993; Siminon et al., 1993; Van Sickle et al., 1996).
Aluminum fractionation and Al; concentration are directly dependent upon pH levels (Driscoll et al.,
1985).
Fish communities of acid-sensitive streams and lakes may contain a variety of species, but are
often dominated by trout. Across the eastern U.S., brook trout is often selected as an indicator of
acidification effects on aquatic biota because it is native to many eastern streams and lakes and because
residents place great recreational and aesthetic value on this species. It must be emphasized, however, that
brook trout is a relatively acid-tolerant species. Many other fish species, including rainbow and brown
trout, as well as a variety of other fish species, are more acid-sensitive than brook trout. In many
Appalachian Mountain streams that have been acidified by acidic deposition, brook trout is the last
species to disappear; it is generally lost at pH near 5.0 (MacAvoy and Bulger, 1995), which usually
corresponds in these streams with ANC near 0 (Sullivan et al., 2003).
Although there are known differences in acid sensitivity among fish species, experimentally
determined acid sensitivities are available for only a minority of freshwater fish species. Baker and
Christensen (1991) reported critical pH values for 25 species of fish. They defined critical pH as the
threshold for significant adverse effects on fish populations. The reported range of pH values represents
the authors' estimate of the uncertainty of this threshold. The range of response within species depends on
differences in sensitivity among life stages, and on different exposure concentrations of Ca2+ and Al. To
cite a few examples, blacknose dace is regarded as very sensitive to acid stress, because population loss
due to acidification has been documented in this species at pH values as high as 6.1; in field bioassays,
B-91
-------
embryo mortality has been attributed to acid stress at pH values as high as 5.9. Embryo mortality has
occurred in common shiner at pH values as high as 6.0. Although the critical pH range for rainbow trout
is designated as 4.9-5.6, adult and juvenile mortality have occurred at pH values as high as 5.9. Brown
trout population loss has occurred over the pH range of 4.8-6.0, and brook trout fry mortality has
occurred over the range of 4.8-5.9 (Baker and Christensen, 1991). Relative sensitivities can be suggested
by regional surveys as well, although interpretation of such data is complicated by factors that correlate
with elevation. Such factors, including habitat complexity and refugia from high-flow conditions, often
vary with elevation in parallel with acid sensitivity. It is noteworthy, however that about half of the 53 fish
species found in Adirondack Mountain waters in New York never occur at pH values below 6.0 (Kretser
et al., 1989; Driscoll et al., 2001b); for those species whose acid tolerances are unknown, it is probable
that acid sensitivity is responsible for at least some of these absences. It is the difference in acid tolerance
among species that produces a gradual decline in species richness as acidification progresses, with the
most sensitive species lost first.
Effects on biota can be assessed as effects on a particular sensitive species or species perceived to
be important, or as effects on the richness or diversity of fish or other potentially sensitive life form. For
example, Bulger et al. (2000) developed ANC thresholds for brook trout in Virginia, which are presented
in (Table B-27). These values were based on annual average stream water chemistry, and therefore
represent chronic exposure conditions. The likelihood of additional episodic stress is incorporated into the
response categories in the manner in which they are interpreted. For example, the episodically acidic
response category, which has chronic ANC in the range of 0 to 20 (ieq/L, represents streams that are
expected to acidify to ANC near or below 0 during rainfall or snowmelt episodes. In such streams,
sublethal and/or lethal effects on brook trout are possible (Bulger et al., 2000; Sullivan et al., 2003).
Fish species richness, population density, condition factor, age distribution, size, and bioassay
survival have all been shown to be reduced in low-ANC streams as compared to intermediate-ANC and
high-ANC streams (Bulger et al., 1995; Dennis et al., 1995; Dennis and Bulger, 1995; MacAvoy and
Bulger, 1995). Fish species richness is a good indicator of acidification response. Lakes or streams having
pH below about 5.0 or ANC below about 0 generally do not support fish. Depending on the region, waters
having pH above about 6.5 and ANC above about 50 j^icq/L support large, but variable, numbers of
species. There is often a positive relationship between pH and number of fish species, at least for pH
values between about 5.0 and 6.5, or ANC values between about 0 and 50 to 100 j^icq/L (Bulger et al.,
1999; Sullivan et al., 2006a). Such observed relationships are complicated, however, by the tendency for
smaller lakes and streams, having smaller watersheds, to also support fewer fish species, irrespective of
acid-base chemistry. This pattern may be due to a decrease in the number of available niches as stream or
lake size decreases. Nevertheless, fish species richness is one of the most useful indicators of biological
effects of surface water acidification.
Acidification and the associated elevated concentrations of Al; in surface waters have adversely
affected fish populations and communities in parts of the Adirondack Mountains of northern New York
(Baker and Schofield, 1982; Johnson et al., 1987; Kretser et al., 1989; Schofield and Driscoll, 1987;
Siminon et al., 1993) and in acid-sensitive streams of the Catskill Mountains of southeastern New York
(Stoddard and Murdoch, 1991) and the Appalachian Mountains from Pennsylvania to Tennessee and
South Carolina (USDA, 1996; Bulger et al., 1999, 2000).
Adverse effects of low pH and high Al j concentration on fish include increased mortality, decreased
growth, decreased reproductive potential, and ionoregulatory impairment. A partial list of studies
demonstrating such effects is provided in Table B-28 from Baker et al. (1990b). It has been shown,
however, that there is marked variability among species and among life stages within species in the
specific levels of pH and Al; that produce measurable responses.
Surface-water acidification can affect fish populations by a number of mechanisms ranging from
increased morality and emigration to decreased food supplies (Baker et al., 1990b). The primary reason
for population decline and extinction, however, is usually the failure of a species to successfully recruit
young-of-the-year fish (Mills et al., 1987; Brezonik et al., 1993). The response of aquatic communities to
B-92
-------
acidification, therefore, should appear first as changes in age distribution and decreased health of
individual fish (growth and condition), then as decreased biomass and density in populations of acid-
intolerant fish species, and finally as elimination of sensitive species (Baker et al., 1990b).
The primary mechanism for the toxic effects of low pH and elevated Al on fish involves disruption
of normal ion regulation at the gill surface resulting in increased rates of ion loss and inhibition of ion
uptake (McWilliams and Potts, 1978; Leivestad, 1982; Wood and McDonald, 1987; Bergman et al.,
1988). Additional effects might include disruption of Ca2+ metabolism (Peterson and Martin-Robichaud,
1986; Gunn and Noakes, 1987; Reader et al., 1988), and decreased hatching success (Runn et al., 1977;
Peterson et al., 1980; Haya and Waiwood, 1981; Waiwood and Haya, 1983).
Prominent physiological disturbance for fish exposed to acid waters are iono- and osmoregulatory
failure, acid-base regulatory failure, and respiratory and circulatory failure. Most of these effects can be
directly attributed to effects on gill function or structure. The acute toxicity of low pH in acidic waters
results in the loss of Ca2+ from important binding sites in the gill epithelium, which reduces the ability of
the gill to control membrane permeability (McDonald, 1983; Havas, 1986; Exley and Phillips, 1988).
The energy costs to fish for active iono-osmoregulation can be substantial (Farmer and Beamish,
1969; Bulger, 1986). The concentrations of serum electrolytes (such as Na+ and CI ) are many times
higher (often 100-fold higher) in fish blood than in the fresh waters in which they live. The active uptake
of these ions occurs at the gills. Because of the steep gradient in Na+ and CI concentrations between the
blood and fresh water, there is constant diffusional loss of these ions, which must be replaced by energy-
requiring active transport. Low pH increases the rate of passive loss of blood electrolytes (especially Na+
and CF); and Al elevates losses ofNa+ and CI above the levels due to acid stress alone (Wood, 1989).
For example, dace in an acidified stream maintain whole-body Na+ at levels similar to dace in a high-
ANC stream (Dennis, 1995), despite probable higher gill losses of electrolytes due to acid/Al stress.
Therefore, the homeostatic mechanisms at the gill responsible for maintaining blood electrolyte levels
must work harder and use more energy to maintain these levels for dace in the acidified stream.
Whole lake experiments and artificial stream channel experiments have shown that acidification
can lead to loss of fish species. A summary of the work on Lake 223 in the Experimental Lakes Area in
Canada is provided in Table B-29. Work at Little Rock Lake in Wisconsin suggested that rock bass
suffered recruitment failure at pH 5.6 or below. Artificial channel studies showed poor survival and
reproductive success for fathead minnow at pH 5.9 to 6.0.
ANC criteria have been used for evaluation of potential acidification effects on fish communities.
The utility of these criteria lies in the association between ANC and the surface water constituents that
directly contribute to or ameliorate acidity-related stress, in particular pH, Ca2+, and Al. Bulger et al.
(2000) developed ANC thresholds for brook trout response to acidification in forested headwater
catchments in western Virginia (See Table B-27). Note that because brook trout are comparatively acid
tolerant, adverse effects on many other fish species should be expected at relatively higher ANC values.
Streams with chronic ANC greater than about 50 j^icq/L are generally considered suitable for brook
trout in southeastern U.S. streams because they have a large enough buffering capacity that persistent
acidification poses no threat to this species, and there is little likelihood of storm-induced acidic episodes
lethal to brook trout. In such streams, reproducing brook trout populations are expected if the habitat is
otherwise suitable (Bulger et al., 2000), although some streams may periodically experience episodic
chemistry that affects species more sensitive than brook trout. Streams having annual average ANC from
20 to 50 (ieq/L may or may not experience episodic acidification during storms that can be lethal to
juvenile brook trout, as well as other fish. Streams that are designated as episodically acidic (chronic ANC
from 0 to 20 j^ieq/L) are considered marginal for brook trout because acidic episodes are likely (Hyer
et al., 1995), although the frequency and magnitude of episodes vary. Streams that are chronically acidic
(chronic ANC less than 0 j^ieq/L) are not expected to support healthy brook trout populations (Bulger
et al., 2000).
Field surveys provided a regional context for fish response to acidification. Although there were
some variations, the results of field surveys generally confirmed the results of bioassays, field
B-93
-------
experiments, and other intensive field studies. The results of many field surveys were summarized in
Baker et al. (1990b) and are compiled in (Table B-30).
It is important to note, however, that the absence of fish from a given lake or stream in an area that
experiences surface water acidification does not necessarily imply that acidification is responsible for the
absence of fish. For example, results of fisheries research in the Adirondacks has indicated that many
Adirondack lakes always had marginal spawning habitat for brook trout (Schofield, 1993), and some of
the currently Ashless acidic lakes probably never supported fish.
Many of the data for the assessment of fish status in the Adirondack region of New York come
from the reports by Kretser et al. (1989) and Baker et al. (1990a). The status of fish and of the presence of
individual species were related to a variety of lake characteristics. Of the lakes without fish, 42% had high
organic acid content that may have caused the observed low pH, 13% were bog lakes of high acidity and
naturally poor fish habitat, 9% had pH >5.5 suggesting other factors were likely responsible for the lack
of fish, and 3% were small high-elevation lakes that were unlikely to have fish regardless of acid-base
chemistry. However, 34% of the lakes surveyed (112 lakes) that had no fish at the time of survey had low
pH that was most likely the result of acid deposition and no other obvious explanation for the lack of fish.
Multivariate regression of the presence/absence of brook trout in Adirondack waters produced a
ranking of factors that appeared to influence the presence of brook trout when biological factors were
excluded from the analysis (stocking, presence of associated species, presence of competitors). Among
contributing factors, including SiC>2, ANC, DOC, substrate, and distance to the nearest road, pH ranked
first as a predictor of brook trout presence (Christensen et al., 1990). The results of this analysis supported
the hypothesis that 1990 levels of pH and related variables restricted the distribution of some fish in
Adirondack waters.
Fish toxicity models have been developed as mathematical regression functions fit to observations
of fish mortality when exposed to constant levels of pH, Al;, and Ca2+ in laboratory toxicity tests. These
models had the advantage that they dealt directly with the interaction effects of pH, Al, and Ca2+, but they
did not account for the effects of variations in other aspects of surface water quality, and they could not be
directly interpreted in terms of population-level response.
The many bioassays conducted of pH effects were screened by Baker et al. (1990b) to provide data
most suitable for model development. Bioassays selected for inclusion were those that measured the
mortality of early life stages, those that incorporated different combinations of pH, Al, and Ca2+, and those
that used fish of varying sensitivity (Bergman et al., 1988).
Acidity and Al toxicity are not the only stress factors that influence the distribution of fish in acid-
sensitive streams. Other habitat characteristics, including water temperature and stream channel
morphology, can be important (Sullivan et al., 2003). In addition, it is probable that some trout
populations have been affected by competition with other introduced species (Larson and Moore, 1985).
B.6.1.5. Amphibians
Some species of amphibians are considered to be highly sensitive to changes in environmental
conditions and some species have probably been adversely affected by acidic deposition in some areas.
Furthermore, several species of amphibian have exhibited marked declines in abundance throughout the
western U.S. in recent decades and there has been much speculation concerning the cause(s) of these
declines in abundance.
Populations of many species of amphibians have declined or become eradicated throughout the
world in recent decades (Barinaga, 1990; Wake, 1991). The causes have not been evident and some of the
declines have occurred in remote pristine areas. For example, in the Sierra Nevada, at least two of five
species of aquatic-breeding amphibians, Rana muscosa (mountain yellow-legged frog) and Bufo canorus
(Yosemite toad) have been declining (Phillips, 1990). A number of hypotheses have been proposed for
amphibian decline, including acidic deposition. In the western U.S., however, acidic deposition has been
discounted as the primary cause of the decline of R. muscosa and B. canorus in the Sierra Nevada and of
B-94
-------
R. pipiens and B. boreas in the Rocky Mountains (Corn et al., 1989; Bradford et al., 1992). Grant et al.
(2005) reported little relationship between streamwater ANC and the adjacent salamander community in
Shenandoah National Park.
In some cases, population fragmentation as a consequence of fish predation may be a more likely
cause (Bradford et al., 1993). It is generally recognized that R. muscosa was eliminated by introduced fish
early in the 20th century in many lakes and streams in Sequoia and Kings Canyon National Parks. The
amphibians have been eliminated from nearly all waters inhabited by fish, presumably by predation on
tadpoles. Before 1870, virtually all of the high-elevation (>2500 m) lakes in the Sierra Nevada were
barren of fish, but have since been stocked with fish. Fish introductions may have contributed to recent
amphibian declines because amphibian populations are now more isolated from each other than formerly.
The role of atmospheric deposition as an additional stressor is not clear.
The acidification sensitivity of temporary ponds, where many amphibians live or reproduce, have
not been well studied. These ponds tend to fill directly from rain or snowmelt and thus can be more acidic
than surrounding lakes and streams. There is a correlation between pond acidity and amphibian
abundance.
There are both acid-sensitive and acid-tolerant amphibians. Examples of acid-sensitive amphibians
include the spotted salamander (Ambystoma maculatum) and Jefferson salamander (Ambystoma
jeffersonianum). Embryos of acid-sensitive species are killed by water with pH less than about 4.5. Acid-
tolerant embryos may survive at a pH of 3.7. Toxicity is not solely a matter of pH, but is also influenced
by Ca2+, Al;, and DOC concentrations. It is also dependent on the life stages present and water
temperature (Baker et al., 1990b). Large-scale amphibian extinctions in any geographic region due to
acidic deposition have not been detected.
Although acidic deposition may play a role in some areas, there is no evidence to suggest that it is a
primary factor. Other issues, including fish introductions, are probably more important as stressors on
amphibian populations across broad regional to national scales.
B.6.1.6. Fish-Eating Birds
Relative to other trophic groups, there are few studies assessing acidification effects on fish-eating
birds. Limited data suggest that fish-eating birds are adversely affected by acidification. Acidification
effects on birds may be indirect, related to changes in the quantity and quality of food. Other potential
causal pathways include delayed egg laying, lighter/thinner egg shells, and reduced chick growth in acidic
waters (Tyler and Ormerod, 1992). There is also concern about increased metal and Hg concentrations in
fish-eating birds associated with bioaccumulation from contaminated fish in known areas of acidification
(Baker etal., 1990b).
Fish-eating birds can serve as biological indicators of lakes affected by acidic deposition (McNicol,
2002). Lack of prey resources, decreased food quality, and elevated lake water methylmercury (MeHg)
concentrations that could be associated with acidification may negatively effect foraging, breeding, and/or
reproduction for the common loon (Gavia immer), common merganser (Mergus merganser), belted
kingfisher (Ceryle alcyori), osprey (Pandion haliatus), American black duck (Anas rubripes), ring-necked
duck (Aythya collaris), eastern kingbird (Tyrannus tyrannus), and tree swallow (Tachycineta bicolor)
(Table B-31) (Longcore and Gill, 1993). Breeding distribution for the common goldeneye (Bucephala
clangula), an insectivorous bird, may be positively effected by acidic deposition (Longcore and Gill,
1993). Reduced prey diversity and quantity have been observed to create feeding problems for nesting
pairs of loons on low-pH lakes in the Adirondacks (Parker, 1988).
Since the mid 1980s, a statistically significant increase in fish-eating birds has been observed in the
Sudbury region of Ontario, Canada, which has corresponded with a decreasing abundance of common
goldeneye (McNicol, 2002). This interaction has been attributed to an increase in prey for pisciverous
birds and a decrease in available prey for insectivorous birds as a result of stricter S emissions controls in
the U.S. and Canada (McNicol, 2002). Logistic regression modeling with measured pH and species
B-95
-------
occurrence data for acid-sensitive lakes in the Algoma region of Ontario showed that the occurrence of
fish, common loons, and common mergansers is positively related to lake water pH (McNicol, 2002).
Predictions of common loon and merganser recovery for this area were made using the Waterfowl
Acidification Response Modeling System (WARMS) under varying S emissions control scenarios
targeted for 2010 (McNicol, 2002). The modeled emissions scenarios include:
¦ SI: sulfate emissions equal to those in the early 1980s (base case)
¦ S2: sulfate emissions equal to that in 1994 (full Canadian emissions reductions based on the 1991
Canada/U.S. Air Quality Agreement)
¦ S3: expected sulfate emissions in 2010 (full implementation of U.S. emissions reductions based
on the 1991 agreement)
¦ S4: a hypothetical 50% reduction in expected 2010 sulfate emissions
¦ S5: a hypothetical 75% reduction in expected 2010 sulfate emissions
The number of lakes projected to be suitable for supporting breeding pairs and broods increased
with lake pH and stricter emissions controls (Table B-32) (McNicol, 2002).
Marginal improvements to fish-eating bird habitat were predicted to occur by 2010 (S3), with more
significant improvements expected under hypothetical S emissions reductions of 50% and 75% (S4 and
S5) for lakes with pH below 6.5 (McNicol, 2002). Fundamental to the predicted improvement of these
fish-eating bird populations is the expected increase in food availability with lake pH recovery.
Elevated MeHg accumulation in fish-eating birds in Wisconsin and the northeastern U.S. has been
linked to lake acidification (Meyer et al., 1995; (Hrabik and Watras, 2002)Evers et al., 2007). This form
of Hg is toxic, bioavailable, and accumulates in top predators to levels of concern for both human health
and the environment (Table B-33) (Evers et al., 2007).
Acidic deposition might contribute to Hg toxicity in fish-eating birds because S042 addition to
wetland environments could stimulate the production of MeHg, thereby increasing lake water
concentrations of MeHg (Jeremiason et al., 2006). Kramar et al. (2005) determined that the extent of
wetland located in close proximity (less thanl50 m) to loon territory was positively correlated with Hg
concentrations in loon blood. Wetland MeHg production is discussed in greater detail in Section 6.3.
Accumulation of MeHg in fish-eating birds can result in damage to nervous, excretory, and
reproductive systems (Wolfe et al., 1998). Table B-34 lists several studies indicating effects related to
mercury bioaccumulation in avian eggs and tissues. Reproduction is considered one of the most sensitive
endpoints to chronic low-level MeHg exposure for fish-eating birds (Wolfe et al., 1998). Reduced clutch
size, increased number of eggs laid outside the nest, eggshell thinning, and increased embryo mortality
have all been documented (Wolfe et al., 1998).
Table B-1. N-saturated forests in North America, including estimated N inputs and outputs.
Location
Forest Type
Elevation (m)
N Input
(kg/ha/yr)
N Output
(kg/ha/yr)
Reference
Adirondack Mts.,
northeastern New York
Northern hardwoods or
hardwood/conifer mix
396-661
9.3a
Stage 1 N
lossb
Driscoll and Van
Dreason (1993)
Catskill Mts.,
southeastern New York
Mainly hardwood;
some eastern hemlock
335-675
10.2a
Stage 1 and 2
N lossb
Stoddard (1994)
Turkey Lakes Watershed,
Ontario, Canada
Sugar maple and
yellow birch
350-400
7.0-7.7
(as throughfall)
17.9-23.6
Foster et al. (1989);
Johnson and
Lindberg (1992)
B-96
-------
Whitetop Mt.,
southwestern, Virginia
Red spruce
1,650
32c
47c
Joslin and Wolfe
(1992); Joslin etal.
(1992)
Fernow, West Virginia
Mixed hardwood
735-870
15-20
6.1
Gilliam etal.;
Peterjohn et al.
(1996)
Great Smoky Mts.
National Park, Tennessee
American beech
1,600
3.1 d
2.9
Johnson and
Lindberg (1992)
Great Smoky Mts.
National Park, Becking
Site, North Carolina
Red spruce
1,800
10.3d
19.2
Johnson et al.
(1991c)
Great Smoky Mts.
National Park, Tower Site,
North Carolina
Red spruce
1,740
26.6
20.3
Johnson et al.
(1991c)
Front Range, Colorado
Alpine tundra,
subalpine conifer
3,000-4,000
7.5-8.0
7.5
Williams et al.
(1996a)
San Dimas, San Gabriel
Mts., southern California
Chaparral and
grasslands
580-1,080
23.3e
0.04-19.4
Riggan et al. (1985)
Camp Paivika, San
Bernardino Mts., southern
California
Mixed conifer
1,600
30
7-26
Fenn etal. (1996)
Klamath Mts., northern
California
Western coniferous
NA
Mainly
geologicg
NAg
Dahlgren (1994)
Thompson Forest,
Cascade Mts.,
Washington
Red alder
220
4.7 plus >100
as N2 fixation
38.9
Johnson and
Lindberg (1992)
8 Estimated total N deposition from wet deposition data is from Driscoll et al. (1991) for the Adirondacks, and from Stoddard and Murdoch (1991) for the Catskills. Total deposition was
estimated based on the wet deposition: total N deposition ratio (0.56) at Huntington Forest in the Adirondacks (Johnson, 1992). N deposition can be higher in some areas, especially at
high-elevation sites such as Whiteface Mountain (15.9 kg N/ha/yr); (Johnson, 1992). bStage 1 and 2 of N loss according to the watershed conceptual model of Stoddard (1994). N
discharge (kg N/ha/yr) data are not available, only stream water NO3" concentration trend data were collected.c Values appear high compared to other sites, especially N leaching
losses. Joslin and Wolfe (1992) concede that "there is considerable uncertainty associated with the estimates of atmospheric deposition and leaching fluxes." However, elevated NO3"
concentrations in soil solution and lack of a growth response to N fertilization ratio (Joslin and Wolfe, 1994) support the hypothesis that the forest at Whitetop Mountain is N -saturated.d
Estimated total N deposition from throughfall data. Total deposition was estimated based on the throughfall/total N deposition ratio from the nearby Smokes Tower site (Johnson, 1992).
e Annual throughfall deposition to the chaparral ecosystem.f N output is from unpublished stream water data (Fenn and Poth, 1999). The low value represents a year of average
precipitation, and the high value is for 1995, when precipitation was nearly double the long-term average. N output includes N export in stream water and to groundwater, a Annual input
and output data are not known, although N deposition in this forest is probably typical for much of the rural western U.S. (2-3 kg N/ha/yr; Young et al., 1988). Excess N is from
weathering of ammonium in mica schist bedrock. The ammonium was rapidly nitrified, leading to high NO3" concentrations in soil solution (Dahlgren, 1994).
Table B-2 Summary of measured ANC, pH, and Al concentrations compared with reference values in the
six high-interest areas.
Area
n*
N*
Percent of Population with
ANC > 0
pH > 5.5
Ah >100 ng/L
ADIRONDACKS
Southwest lakes
52
450
38
51
36
Other lakes
84
707
0
3
0
B-97
-------
NEWENGLAND
Seaboard Lowlands lakes
94
848
8
11
0
Highland lakes
354
3,574
2
5
2
MID-ATLANTIC HIGHLANDS
Forested lakes
91
433
10
9
1
Other lakes
52
791
0
0
0
Forested streams
78
11,631
12
17
8
Other streams
69
10,172
0
2
0
ATLANTIC COASTAL PLAIN
Northeast lakes
22
187
11
15
7
Pine Barrens streams
12
675
56
92
56
Other streams
31
7,452
10
24
15
FLORIDA
Northern Highland lakes
32
522
63
53
10
Northern Highland streams
18
669
28
55
0
EASTERN UPPER MIDWEST
Low silica lakes
155
1,254
16
19
1
High silica lakes
125
1,673
3
4
2
* n = sample size, N = estimated number of lakes or upstream reach ends in population.
Source: Baker etal. (1990b).
Table B-3 Sources of data and sample sizes for datasets analyzed by Stoddard et al. (2003), along with
estimates of the condition of surface waters in each region in the 1980s.
Source of Data and Region
No. of Sites1
Size of Population2
Percent Acidic in 1980s3
STATISTICAL SURVEYS
New England Lakes4
30
4,327 lakes
5%
Adirondack Lakes4
43
1,290 lakes
14%
Appalachian Plateau Streams
31
72,000 stream miles
6%
SENSITIVE SURFACE WATERS
New England Lakes
24
N.A.
5%
Adirondack Lakes
48
N.A.
14%
Northern Appalachian Streams
9
N.A.
6%
Upper Midwest Lakes
38
N.A.
3%
B-98
-------
Ridge/Blue Ridge Streams
69
N.A.
5%
1 Number of monitoring sites with monitoring data available (1990-2000). 2 Total number of lakes, or stream length, for which statistical survey results can be inferred. Site selection for
LTM (sensitive surface waters) is not statistically based, and results cannot be expanded to population level.3 Estimates of extent of acidification, based on National Surface Water
Survey results (Linthurstet al., 1986; Kaufmann et al., 1988).4 Estimates are for lakes with surface areas >4 ha; estimates based on populations including smaller lakes are likely to be
higher, due to the increased incidence of acidification in small lakes.
Statistical survey data are from the EMAP and TIME projects. Sensitive surface water data are from the LTM project, as well as other contributed studies. Source: Stoddard et al. (2003)
Table B-4 Estimates of change in number and proportion of acidic surface waters in acid-sensitive
regions of the North and East, based on applying current rates of change in Gran ANC to past
estimates of population characteristics from probability surveys.
Results of Regional Survey Results of Monitoring during 1990s
Region
Population
Size
Number
Acidic1
%
Acidic2
Time
Period of
Estimate
Rate of
ANC
change3
Estimated
Number
Acidic in
2000
%
Acidic
in 2000
% Change in
Number of
Acidic
Systems
New England
6,834 lakes
386 lakes
5.6%
1991-94
-K).3
374 lakes
5.5%
-2%
Adirondacks.
1830 lakes
238 lakes
13.0%
1991-94
+0.8
149 lakes
8.1%
-38%
No.
Appalachians
42,426 km
5014 km
11.8%
1993-94
+0.7
3600 km
8.5%
-28%
Ridge/Blue
Ridge
32,687 km
1634 km
5.0%
1987
-0.0
1634 km
5.0%
0%
Upper
Midwest
8,574 lakes
251 lakes
2.9%
1984
+1.0
80 lakes
0.9%
-68%
Source: Stoddard et al. (2003)
Table B-5 Regional trend results for long-term monitoring sites for the period 1990 through 2000.
Region
S042" NOs-
(|jeq/L/yr) (|jeq/L/yr)
Base Cations
[Ca2+ + Mg2+]
(Meq/L/yr)
Gran ANC
(Meq/L/yr)
Hydrogen
(Meq/L/yr)
DOC
(mg/L/yr)
Aluminum
(Mg/L/yr)
New England Lakes
-1.77**
+A.01ns
-1.48**
+0.11ns
-0.01ns
+0.03*
+0.09ns
Adirondack Lakes
-2.26**
-0.47**
-2.29**
+1.03**
-0.19**
+0.06**
Appalachian Streams
-2.27*
-1.37**
-3.40**
+0.79*
-0.08*
+0.03ns
+0.56ns
Upper Midwest Lakes
-3.36**
+fl.02ns
-1.42**
+1.07**
-0.01*
+0.06**
-0.06ns
Ridge/Blue Ridge Streams
+0.29**
-0.07**
-0.01ns
-0.07ns
+0.01 ns
NA
NA
ns regional trend not significant (p >0.05). * p <0.05. ** p <0.01. NA insufficient data. Note: Values are median slopes for the group of sites in each region.Source: Stoddard et al. (2003)
B-99
-------
Table B-6 Slopes of trends in Gran ANC in acidic, low, and moderate ANC lakes and streams, 1990-2000.
ANC Class
Number of Sites
Change in Gran ANC (jjeq/L/yr)
Acidic (ANC <0 peq/L)
26
+1.29**
Low ANC (0 0.05). ** p <0.01. Note: Analysis includes all sites in New England, Adirondacks, Appalachian Plateau, and Upper Midwest; Ridge and Blue Ridge
Source: Stoddard et al. (2003)
i sites excluded.
Table B-7 Changes in key chemical characteristics during periods of record in Maine aquatic systems.
Years
Change in (all in Heq/L)
Sulfate Nitrate
Base Cations
Calculated ANCa
ANC
DOC"
Acadia NP lakes (22)
17
-10 0
-17
-7
0
-2
LTM lakes @ Tunk Mtn (6) - spring
17
-9 0
-10
-1
1
1
LTM lakes @ Tunk Mtn (6) - fall
17
-7 0
-9
-2
-2
1
LTM lakes since 1990 - fall only
8
-9 0
-10
-1
-1
0
High elevation lakes (90)
12
-16 1
-23
-8
-2
4
Seepage lakes (120)
12
-9 1
-1
7
7
4
East Bear Brook at BBWM
11
CO
1
CM
CM
1
-44
-6
-4
1
RLTM lakes (16)
7
-6 1
-17
-12
-4
2
0 Calculated ANC = [change in base cations] minus [change in (sulfate + nitrate)].b DOC (peq/L) =
DOC in mg/l * 4 (e.g., Kahl et al., 1999). Source: Kahl et al. (1999)
Table B-8 Projected changes (|jeq/L) in median values of streamwater chemistry at the regional
modeling sites from 1995 to 2040 in each of the three emissions control strategies, stratified
into two segments of the SAMI region (northeast and southwest) and by physiographic
province.
Physiographic Province
Number of Sites
A Sulfate
A Nitrate A SBC
AANC
A2 STRATEGY1
VIRGINIA AND WEST VIRGINIA
Blue Ridge
16
1.8
0.03 -2.2
-4.0
Valley and Ridge
41
-0.45
0.02 -6.8
-6.6
Appalachian Plateau
34
-31.2
-3.5 -33.8
-4.4
B-100
-------
NORTH CAROLINA, TENNESSEE, SOUTH CAROLINA, GEORGIA, AND ALABAMA
Blue Ridge
33
8.8
0.15
1.0
-8.0
Appalachian Plateau
6
15.3
0.15
-1.2
-15.9
VIRGINIA AND WEST VIRGINIA
Blue Ridge
16
-2.7
-0.04
-3.0
-1.0
Valley and Ridge
41
-5.6
-0.37
-8.2
-4.7
Appalachian Plateau
34
-36.3
-4.9
-38.4
-1.4
NORTH CAROLINA, TENNESSEE, SOUTH CAROLINA, GEORGIA, AND ALABAMA
Blue Ridge
33
5.6
-0.48
-0.60
-5.4
Appalachian Plateau
6
11.6
-0.23
-1.9
-13.3
B3 STRATEGY1
VIRGINIA AND WEST VIRGINIA
Blue Ridge
16
-7.4
-0.09
-5.1
2.9
Valley and Ridge
41
-13.8
-0.36
-10.3
-0.83
Appalachian Plateau
34
-40.4
-5.8
-39.3
2.6
NORTH CAROLINA, TENNESSEE, SOUTH CAROLINA, GEORGIA, AND ALABAMA
Blue Ridge
33
3.2
-1.0
-2.3
-3.2
Appalachian Plateau
6
7.2
-0.53
-3.1
-10.4
1 Emissions control strategies were based on existing regulations (A2), moderate additional controls (B1), and more aggressive additional controls (B3)
Source: Sullivan et al. (2004)
Table B-9. Population estimates of water chemistry percentiles for selected lake populations in the
western U.S.3
Population n N
PH
ANC
(Meq/L)
SBC
(Meq/L)
S042"
(Meq/L)
no3-
(Meq/L)
DOC
(mg/L)
P1
P5
P1
P5
P1
P5
P95
P99
P95
P99
P50
P99
Sierra Nevada 114 2,119
5.84
6.31
15
16
21
26
90
386
8
10
0.8
2.7
Cascades 146 1,473
5.95
6.25
11
18
20
31
60
97
3
6
1.3
2.6
Idaho Batholith 88 937
6.34
6.42
21
33
30
45
30
43
3
4
1.2
2.4
NW Wyoming 38 648
6.56
6.56
38
38
64
66
41
2,909
13
32
1.0
4.8
Colorado 121 1,173
Rockies
6.02
6.65
25
42
58
80
915
2,212
10
13
1.3
5.7
aData from Landers et al. (1987). Excluding Fern Lake (4D3-017) which is naturally acidic. Note: The 1st and 5th percentiles (P1, P5) are presented forpH, ANC (peq/L), and SBC
(peq/L) and the 95th and 99th (P95, P99) percentiles are shown for SO42" (peq/L) and NO3" (Meq/L). The median (P50) and 90th percentiles are shown for DOC (mg/L).
B-101
-------
Table B-10. Population estimates of the percentage of lakes in selected subregions of the West with ANC
and N03~ within defined ranges.
ANC (Heq/L)
NOs" (kieq/L)
<0
<25
<50
>5
>10
Sierra Nevada
0
8.7
39.3
10.6
1.5
Cascades
0
10.2
22.4
1.5
0.0
Idaho Batholith
0
2.0
23.6
4.6
3.9
NY Wyominga
0
2.3
12.8
22.8
8.9
Colorado Rockies
0
0.9
5.5
9.8
1.8
a Excluding Fern Lake (4D3-017) which is a naturally acidic lake. Source: Landers et al. (1987)
Table B-11. Median streamwater ANC and watershed area of streams in Shenandoah National Park that
have water chemistry and fish species richness data.
Site ID
Watershed Area (km2)
Median ANC (Heq/L)
# Fish Species
SMALLER WATERSHEDS (<10 KM2)
North Fork Dry Run
2.3
48.7
2
Deep Run
3.6
0.3
N.D.a
White Oak Run
4.9
16.2
3
Two Mile Run
5.4
10.0
2
Meadow Run
00
CO
-3.1
1
LARGER WATERSHEDS (>10 KM2)
Brokenback Run
10.1
74.4
3
Staunton River
10.6
76.8
5
Piney River
12.4
191.9
7
Paine Run
12.7
3.7
3
Hazel River
13.2
86.8
6
White Oak Canyon
14.0
119.3
7
N. Fork Thornton River
18.9
249.1
9
Jeremy's Run
22.0
158.5
6
Rose River
23.6
133.6
8
0 No data were available regarding the number of fish species in Deep Run. Source: Sullivan (2003)
B-102
-------
Table B-12. Reference levels for the Acidic Stress Index based on logistic regression of fish presence as a
function of the sensitive intermediate and tolerant ASI values for brown bullhead, brook trout,
lake trout, and common shiner.
Reference Acid Stress Index
— Fish Response
Lakes
Streams
TolerantASI >30
Intermediate ASI >30
Loss of all fish species
TolerantASI >10
Sensitive ASI >30
Loss of brook trout
Intermediate ASI >80
Loss of other sport fish, such as smallmouth bass and lake trout
Sensitive ASI >80
Sensitive ASI >10
Loss of acid-sensitive species, such as minnows.
Source: Baker etal. (1990b).
Table B-13. General summary of biological changes anticipated with surface water acidification,
expressed as a decrease in surface water pH.
^ General Biological Effects
Decrease °
6.5 to 6.0 Small decrease in species richness of plankton and benthic invertebrate communities resulting from the loss of a few highly
acid-sensitive species, but no measurable change in total community abundance or production.
Some adverse effects (decreased reproductive success) may occur for highly acid-sensitive fish species (e.g., fathead
minnow, striped bass).
6.0 to 5.5 Loss of sensitive species of minnows and dace, such as fathead minnow and blacknose dace; in some waters, decreased
reproductive success of lake trout and walleye, which are important sport fish species in some areas.
Visual accumulation of filamentous green algae in the near-shore zone of many lakes and in some streams.
Distinct decrease in species richness and change in species composition of plankton and benthic invertebrate communities,
although little if any change in total community abundance or production.
Loss of some common invertebrate species from zooplankton and benthic communities, including many species of snails,
clams, mayflies, and amphipods, and some crayfish.
5.5 to 5.0 Loss of several important sport fish species, including lake trout, walleye, rainbow trout, and smallmouth bass, as well as
additional nongame species such as creek chub.
Further increase in the extent and abundance of filamentous green algae in lake near-shore areas and streams.
Continued shift in species composition and decline in species richness of plankton, periphyton, and benthic invertebrate
communities; decreases in total abundance and biomass of benthic invertebrates and zooplankton may occur in some waters.
Loss of several additional invertebrate species common in surface waters, including all snails, most species of clams, and
many species of mayflies, stoneflies, and other benthic invertebrates.
Inhibition of nitrification.
5.0 to 4.5 Loss of most fish species, including most important sport fish species such as brook trout and Atlantic salmon. A few fish
species are able to survive and reproduce in water below pH 4.5 (e.g., central mudminnow, yellow perch, and in some waters,
largemouth bass).
B-103
-------
Measurable decline in the whole-system rates of decomposition of some forms of organic matter, potentially resulting in
decreased rates of nutrient cycling.
Substantial decrease in number of species of plankton and benthic invertebrates and further decline in species richness of
plankton and periphyton communities; measurable decrease in total community biomass of plankton and benthic invertebrates
of most waters.
Loss of additional species of plankton and benthic invertebrate species, including all clams and many insects and crustaceans.
Reproductive failure of some acid-sensitive species of amphibians, such as spotted salamanders, Jefferson salamanders, and
the leopard frog.
Source: Baker etal. (1990b).
Table B-14. Estimated percentage of Adirondack lakes with and Acidic Stress Index exceeding the
reference levels for effects on fish populations, based on diatom-inferred historical (pre-
industrial) chemistry and present-day measured and inferred acid-base chemistry.
DDRP Target Population
ASI Reference Level
Diatom Inferred
Measured
ELS/NSWS Target Population Measured
Historical
Current
Net Change
TolerantASI >30
0.0
3.6
+3.6
1.8
2.2
TolerantASI >10
0.0
9.1
+9.1
10.9
6.5
Intermediate ASI >80
7.3
21.8
+14.5
21.8
15.2
Sensitive AS I >80
28.5
41.2
+12.7
32.7
20.0
Source: Baker etal. (1990b).
Table B-15. Estimated percentage of Adirondack lakes with acid-base chemistry unsuitable for fish
population survival, based on diatom-inferred historical (pre-industrial) chemistry and
present-day measured and inferred acid-base chemistry.
DDRP Target Populationb
ELS/NWS
ALSC
Fish Species
Diatom-Inferred0
Target Population11
Historical
Current
Net Change
Measured
Measured
Measured
BROOK TROUT
Bayesian
2.7
13.0
+10.3
14.2
10.1
21.8
LAF framework
-
-
-
-
15.8
24.6
PH
2.3
11.3
+9.0
12.8
9.3
22.2
pCa/pH
16.0
13.3
-2.6
14.5
10.3
23.0
pCa/pH, AI/DOc
16.6
15.6
-1.0
19.2
13.9
23.5
LAKE TROUT
B-104
-------
Fish Species
Model3
DDRP Target Populationb
ELS/NWS
ALSC
Diatom-Inferred0
Target Population11
Historical
Current Net Change
Measured
Measured
Measured
PH
6.4
18.1 +11.2
21.4
14.4
30.9
pCa/pH
31.7
25.1 -6.6
29.0
18.9
-
Inorg.AI 23.9 38.6 +14.7 26.2 17.1
COMMON SHINER
pH 19.2 29.6 +10.5 33.5 21.3 42.3
pCa/pH 45.8 37.7 -8.1 40.2 29.1
8 All models, except the brook trout Bayesian model (Section 3.5) and LAF framework (Section 3.4), are field-based acidification response models as defined in Section 3.3.3.
bELS/NSWS target population in Subregion 1A, defined in Section 3.1 (N = 1,290 lakes); a subset of these lakes was considered for the DDRP and sediment diatom analyses, for
example, excluding lakes with ANC >400 peq/L and with site depths <1.5 m (N = 675 lakes). See Sullivan (1990) for further details.c Estimates of acid-base chemistry inferred from
sediment diatom analysis; methods and water chemistry described in Sullivan (1990). - Analysis not conducted. Source: Baker et al. (1990b).
Table B-16. Estimated percentage of the lakes in the Northeast and Upper Midwest, ELS/NSWS target
population with an Acidic Stress Index exceeding the reference levels for fish populations
defined in Table C-12.
ASI Reference Level
Subregion 1A
Northeast Region
Upper Midwest Region
TolerantASI >30
2.2
1.0
0.5
TolerantASI >10
6.5
2.4
1.0
Intermediate ASI >80
15.2
5.7
2.0
Sensitive AS I >80
20.0
8.6
3.1
Source: Baker etal. (1990b).
Table B-17. Estimated percentage of lakes in the Northeast, ELS/NSWS target populations with acid-base
chemistry unsuitable for fish population survival.
Fish Species/Model
Subregion 1A
Entire Northeast
BROOK TROUT
Bayesian
10.1
3.7
LAF Framework
15.8
8.9
PH
9.3
3.5
pCa/pH
10.3
4.4
pCa/pH, AI/DOC
13.9
7.0
LAKE TROUT
B-105
-------
Fish Species/Model
Subregion 1A
Entire Northeast
PH
14.4
5.8
pCa/pH
18.9
8.9
Inorganic Al
17.1
6.3
COMMON SHINER
PH
21.3
9.5
pCa/pH
29.1
19.7
Source: Baker etal. (1990b)
Table B-18. Distribution of acidic stress index values among the NSS-1 Target populations for the mid-
Appalachian region.
Number (%)
Total Length (%)
Lower Node
Upper Node
SENSITIVE ASI
>10
84.6
66.7
76.1
10-30
10.1
18.9
14.0
30-50
1.4
1.9
1.9
50-80
1.8
2.6
1.6
>80
2.0
9.8
6.4
INTERMEDIATE ASI
>10
97.8
89.3
88.9
10-30
0.2
2.5
0.7
30-50
0.6
1.3
0.4
50-80
0.1
1.4
0.4
>80
1.3
5.4
1.9
TOLERANT ASI
>10
99.4
97.1
98.1
10-30
0.6
1.4
0.9
30-50
0.0
0.6
0.3
50-80
0.0
0.9
0.7
>80
0.0
0.0
0.0
Source: Baker etal. (1990b).
B-106
-------
Table B-19. Distribution of acidic stress index values among the NSS-1 target populations for the interior
Southeast region.
Number (%)
Total Length (%)
Lower Node
Upper Node
SENSITIVE ASI
>10
79.4
70.1
75.9
10-30
18.8
21.1
18.8
30-50
0.0
1.7
2.0
50-80
1.7
5.2
2.5
>80
0.0
1.7
0.7
INTERMEDIATE ASI
>10
100.0
98.3
99.3
10-30
0.0
0.0
0.0
30-50
0.0
0.0
0.0
50-80
0.0
0.0
0.0
>80
0.0
1.7
0.7
TOLERANT ASI
>10
100.0
100.0
100.0
10-30
0.0
0.0
0.0
30-50
0.0
0.0
0.0
50-80
0.0
0.0
0.0
>80
0.0
0.0
0.0
Source: Baker etal. (1990b).
B-107
-------
Table B-20. Comparison of solution and tissue chemistries at threshold treatment levels where significant
impacts on tree growth or nutrient content were first observed. In many cases, adverse
impacts were observed at the lowest Al treatment level. Hence, the actual threshold Ca/AI ratio
may be higher than reported. Results are from a variety of studies reported in the literature.
Study
Solution
Al
(|jmol/L)
Solution Foliar Foliar
Ca/AI Ca Al
(M) (mg/km) (mg/km)
Foliar
Ca/AI
(M)
Root Ca Root Al Root
(mg/km) (mg/kg) Ca/AI(M)
Type of
Study or
Experiment3
Al
Response Analysis
Variableb Used in
Ratio0
NORWAY SPRUCE
Godbold
100
1.3
H
N
AN
etal. (1988)
Matzner
100+
0.3 to 1.8
F
N
Alt
etal. (1989)
Stienen and
1500
0.66 1470 32
31
770 1890 0.28
H
N,B
Alt
Bauch
(1988)
Schroder
2000
1
H
N
AN
etal. (1988)
RED SPRUCE
Thornton
250
1 1100 65
11.4
650 6000 0.07
H
B
AN
etal. (1987)
Hutchinson
185
2.2
S
B
Alt
etal. (1986)
Joslin and
200
nd -3000
-2000
S
B
AN
Wolfe (1988)
Schier
1850
1.35
12.9
0.43
H
B,N
AN
(1985)
Ohno et al.
250
0.8
14
S
N
Ala in soil
(1988)
paste
Joslin and
0.45
F
B
Alt
Wolfe (1992)
WHITE SPRUCE
Nosko et al.
50
0.2
H
B
Alt
(1988)
RED OAK
Joslin and
300
4.05
11.9
0.06
S
B
Alt in
Wolfe (1989)
SrCI2
DeWald
115
4.48 3630 75
32.4
3630 6415 0.38
S
B
Alt
etal. (1990)
B-108
-------
Study
Solution
Al
(|jmol/L)
Solution Foliar Foliar
Ca/AI Ca Al
(M) (mg/km) (mg/km)
Foliar
Ca/AI
(M)
Root Ca Root Al Root
(mg/km) (mg/kg) Ca/AI(M)
Type of
Study or
Experiment3
Al
Response Analysis
Variableb Used in
Ratio0
McCormick
and Steiner
(1978)
7405
0.54
H
B Alt
HONEYLOCUST
Thornton 50 1.1
et al.
(1986b; 198
6c)
Sucoff et al.
(1990)
100
1.4
0.35
S
B
AN
Wolfe and
Joslin (1989)
100
4.3
0.71
S
B
AN
SUGAR MAPLE
Thornton
et al.
(1986a)
100-600
0.42 to
2.5
-2000
-190
9.9
-1500
-2700
0.4
H
B,N
AN
LOBLOLLYPINE
Cronan et al.
(1989)
Thornton
(unpubl.)
500-3000
0.5
900
260
2.3
3700
7770
0.32
H
N
AN
AMERICAN BEECH
Cronan et al.
(1989)
500-3000
0.5
2670
69
26.1
1140
7930
0.1
H
N
AN
EUROPEAN BEECH
Asp and
Berggren
(1990)
300
0.35
3.8
0.2
H
N
AN
Cronan et al.
(1989)
500
0.5
H
N
AN
PEACH
Edwards
and Horton
(1977)
222
10.8
0.008
H
N
AN
SCOTCH PINE
llvesniemi
(1992)
185
nd
400
300
0.9
2300
S
N,B
Alt
McCormick
and Steiner
(1978)
2960
1.35
H
B
Alt
0.21 to
0.32
All
B-109
-------
Study
Solution
Al
(|jmol/L)
Solution Foliar Foliar
Ca/AI Ca Al
(M) (mg/km) (mg/km)
Foliar
Ca/AI
(M)
Root Ca Root Al Root
(mg/km) (mg/kg) Ca/AI(M)
Type of
Study or
Experiment3
Al
Response Analysis
Variableb Used in
Ratio0
VIRGINIA PINE
McCormick
and Steiner
(1978)
2960
1.35
H
B Alt
PITCH PINE
McCormick
and Steiner
(1978)
2960
1.35
H
B Alt
Cumming
and
Weinstein
(1990)
50
20
S
N,B Alt
BIRCHES: GRAY, PAPER AND YELLOW
McCormick
and Steiner
(1978)
4444
0.9
H
B Alt
EUROPEAN BIRCH
Goransson
and
Eldhuset
(1987)
1000
0.02
0.17
H
N Ali
RADIATA PINE
Truman et al.
(1986)
17
10.5 2120 800
1.8
1320 1850 0.48
H
N Alt
DOUGLAS-FIR
Keltjens and
Van Loenen
(1989)
370
0.54 2300 300
5.2
H
B Alt
LARCH
Keltjens and
Van Loenen
(1989)
555
0.36 1800 250
4.9
H
B Alt
a Types of study include hydroponic (H), soil or sand culture (S), or existing forest (F). b Response variables include biomass (B), or nutrient content (N).
c Aluminum measurements include Ali (Ali), monomeric Al (Ala), and total Al (Alt). Since most of the lab studies were conducted under conditions of low pH and minimal DOC,
measurements of total Al, Ali, and labile Al are very comparable. Source: Cronan and Grigal (1995)
B-110
-------
Table B-21. Overview of selected major processes by which landscape change can alter drainage
water acid-base chemistry
Landscape Change
Effect on Acid-Base Chemistry
Logging, blowdown
Dilution
Lower deposition, less acidity
Pulse of NO3" acidity initially
Less base cation neutralization, more acidity
Less water contact with mineral soils, less neutralization of acidic deposition inputs
Road building and construction
More base cation neutralization, less acidity initially
Depletion of base cation reserves in soils, more acidity long-term
Drainage of wetlands
Re-oxidation of stored S, pulses of acidity with increased discharge
Drought
Reduced groundwater inputs to seepage lakes with consequent increased acidity
Increased relative baseflow to drainage waters with consequent decreased acidity
Lake shore development
Decreased acidity
Insect damage
Pulse of NO3" acidity initially
Source: Sullivan (2000)
Table B-22. Observed relationships between zooplankton species richness (R) and lakewater ANC.
Taxonomic Group
Equation r2
P
Total Zooplankton
R= 15.65+ 0.089ANC 0.46
0.001
Crustaceans
R = 6.35 + 0.028ANC 0.47
0.001
Rotifers
R = 9.04 + 0.053ANC 0.30
0.001
Source: Sullivan et al. (2006a). Reprinted with permission.
B-111
-------
Table B-23. Threshold response of increased mortality of fish to low pH listed from least sensitive to most
sensitive.
Study
Species
Increased Mortality
Threshold, pH
Study Conditions
Johnson et al. (1987)
Blacknose dace, creek chub
5.9-6.0
In situ bioassay with early life stages in
Adirondack surface waters
Brook trout
4.8-5.1
Holtze and Hutchinson
(1989)
Common shiner
Lake whitefish, white sucker,
walleye
Smallmouth bass
5.4-6.0
5.1-5.2
Laboratory exposure of early life stages to pH
andAI.
Johansson et al. (1977) Atlantic salmon
Brown trout
Brook Trout
5.0
4.5-5.0
4.5
Laboratory tests with eggs exposed to low pH,
noAI.
Swenson et al. (1989) Black crappie
Rock bass
Yellow perch, largemouth
bass
5.5
5.0
4.5
Laboratory tests with early life stages exposed
topH andAI.
Mills etal. (1987)
Fathead minnow
5.9
Slimy sculpin
5.6-5.9
Lake Trout
5.6
Pearl dace
5.1
White sucker
5.0-5.1
Whole-lake treatment (fish population
recruitment failure)
Source: Baker etal. (1990b).
B-112
-------
Table B-24. Threshold values of pH for various taxa and effects.
Reference
Type of Study
Taxa
PH
Observed Effects
Buckler etal. (1987)
Lab bioassay
Striped bass
6.5
>50% larval mortality
McCormick et al. (1989)
Lab bioassay
Fathead minnow
6.0
Significant decrease in embryo survival
Mills etal. (1987)
Whole-lake
experiment
Fathead minnow
5.9
Population recruitment failure
Klaudaetal. (1987)
Lab bioassay
Blueback herring
5.7
>50% mortality of larvae
Holtze and Hutchinson (1989)
Lab bioassay
Common shiner
5.4
>50% embryo mortality
Baker and Schofield (1980)
Lab bioassay
White sucker
5.2
Substantial reduction in embryo survival
Kane and Rabeni (1987)
Lab bioassay
Smallmouth bass
5.1
>50% mortality of larvae after 30-day exposure
Leino et al. (1987)
Whole-Lake
experiment
Adult fathead minnow
5.2-
5.8
Increased numbers of chloride (ionoregulatory)
cells on the gills
Lacroix (1985a)
Field survey
Atlantic salmon parr
(age 1+)
4.9-
5.3
Significantly lower blood CI levels; high K levels
McDonald and Milligan (1988)
Lab bioassay
Adult brook trout
5.2
Reduced Na transport activity
McWilliams and Potts (1978)
Lab bioassay
Adult brown trout
5.0
Net Na loss; major shift in the gill tansepithelial
potential
Tietgeet al. (1988)
Lab bioassay
Adult brook trout
4.9
Increased volume density of lamellar chloride
cells on gills
Booth etal. (1988)
Lab bioassay
Adult brook trout
4.8
Net loss of Na and CI
Audet and Wood (1988)
Lab bioassay
Adult rainbow trout
4.8
Decreased plasma Na and CI levels
Peterson and Martin-
Robichaud (1986)
Lab experiment
Atlantic salmon larvae
4.5
Reduced accumulation of Na, K, and Ca
Powell and McKeown (1986)
Lab bioassay
Coho salmon parr and
smolts
4.4
Net decrease in plasma Na
Source: Baker etal. (1990b).
B-113
-------
Table B-25. Threshold values of Al for various species and effects (form of Al not specified for most
studies).
Reference
Type of Study
Taxa
PH
Al
(M9/L)
Observed Effect (at similar pH without
added Al)
Sadler and Lynam
(1988)
LB
Brown trout
5.2
30
Significant reduction in fish growth
Turnpenny et al.
(1987)
Field survey
Broen trout
—
40
Fish absent or rare in streams in Wales and
England
Holtze and
Hutchinson (1989)
LB
Walleye
4.9
50
>50% mortality of embryos to 4-d post-hatch
Skogheim and
Roseland (1986)
Field mesocosm
experiment
Atlantic salmon
5.1
75
>50%mortality of smolts
Klauda and Palmer
(1987)
LB
Blueback herring
5.5-
5.6
100
>50% larval mortality
Rosseland and
Skogheim (1984)
LB
Atlantic salmon
4.9-
5.0
130
Significant increase in mortality of presmolts
Baker and Schofield
(1982)
LB
White sucker
5.2
200
>50% larval mortality
Fjellheim et al. (1985)
LB
Eel
5.1
230
Significant increase in elver mortality
Brown (1983)
LB
Brown trout
4.5-
5.4
250
>50% fry mortality
Schofield and Trojnar
(1980)
Field study
Brook trout
4.9
286
No survival of trout stocked into lakes with higher
total Al (even after accounting for pH effects).
Ormerod et al. (1987)
Whole-stream
experiment
Atlantic salmon and
brown trout
5.0
350
>50% mortality of young-of-the-year.
Source: Baker et al. (1990b)
B-114
-------
Table B-26.
The effects of increasing Ca2+ to ameliorate low pH and high Al.
Reference
Type of
Study
Taxa
PH
Al Range
(M9/L)
Ca2+ Range
(Meq/L)
Observed Response to Increasing Ca2+
Brown (1982,
1983)
LB
Brown trout
4.5-
5.1
—
12-400
Increased embryo survival and hatch
Wright and
Snekvik (1978)
Field
survey
Brown trout
4.5-
7.5
—
20-200
Trouot status significantly correlated with log
Ca2+and pH.
Brown (1983)
LB
Brown trout
4.5-
5.4
0-500
12-100
Increased fry survival in low pH or high Al
waters
McDonald (1983)
LB
Rainbow trout
4.3
—
69-223
Decreased adult mortality and net loss of Na
and CI
Edwards et al.
(1987)
LB
Brown trout
4.2
—
100-5600
Lower loss of plasma Na and CI ions
Freda and
McDonald (1988)
LB
Common shiner,
rainbow trout
4.0
—
70-1000
Significant decrease in Na loss
Mount et al.
(1988)
LB
Brook trout
5.0-
6.3
0-500
25-400
Increased survival and growth of adults;
increased progeny survival with adult exposure
to low and high Al
Sadler and
Lynam (1988)
LB
Brown trout
5.2
0-80
8-800
Increased yearling survival and growth in
waters with elevated Al.
Source: Baker etal. (1990b).
Table B-27. Brook trout acidification response categories developed by Bulger et al. (2000) for streams in
Virginia.
Response Category
Chronic ANC Range (Meq/L)
Expected Response
Suitable
>50
Reproducing brook trout expected if other habitat features are also suitable
Indeterminate
20 to 50
Brook trout response expected to be variable
Episodically acidic
0 to 20
Sub-lethal and/or lethal effects on brook trout are possible
Chronically acidic
<0
Lethal effects on brook trout probable
B-115
-------
Table B-28. Partial listing of bioassays demonstrating decreased fish survival in waters with low pH and
(or) elevated aluminum.
Reference
Species
Life Stage
Lab/Field
Johansson and Kihlstrom (1975)
Northern pike
Fry
Lab
Johansson and Milbrink (1976)
Roach
European perch
Egg
Lab, field
Johansson etal. (1977)
Brown trout; Brook trout
Egg and fry
Lab
Trojnar (1977)
White sucker
Egg and fry
Lab
Peterson et al. (1980)
Atlantic salmon
Egg
Lab
Schofield and Trojnar (1980)
Brook trout
Fry
Lab
Baker and Schofield (1982)
Brook trout; White sucker
Egg and fry
Lab
Brown (1983)
Brown trout
Fry
Lab
Hulsman et al. (1983)
Walleye
Rainbow trout
Egg
Fry
Field
Sharpe et al. (1983)
Brook trout; Brown trout
Rainbow trout; Mottled sculpin
Fry and adult
Field
Jagoe et al. (1984)
Arctic char
Egg and fry
Lab
Lacroix (1985b)
Atlantic salmon
Egg and fry
Field
Ingersoll (1986)
Brook trout
Egg and fry
Lab
Buckler etal. (1987)
Striped bass
Fry
Lab
Johnson et al. (1987)
Brook trout; Lake trout
Creek chub; Blacknose dace
Egg, fry, and young-of-year
Field
Klauda and Palmer (1987)
Blueback herring
Egg and fry
Lab
Lacroix and Townsend (1987)
Atlantic salmon
Juvenile
Lab
Wales and Liimatainen (1987)
Walleye
Egg
Field
Palmer et al. (1988)
Bluegill; Fathead minnow
Juvenile
Lab
Gunn (1989)
Lake trout
Egg and fry
Lab, field
Holtze and Hutchinson (1989)
Common shiner; Lake whitefish
White sucker; Walleye
Smallmouth bass; largemouth bass
Egg and fry
Lab
Hutchinson etal. (1989)
Lake trout
Brook trout
Egg and fry
Lab
B-116
-------
Table B-29.
Mills et al., 1987. Shows effect of various pH on fish forage fish and lake trout.
1977
1978
1979
1980
1981
1982
1983
DlOld
pH 6.13
pH 5.93
pH 5.64
pH 5.59
pH 5.02
pH 5.09
pH 5.13
Forage
Fathead
Fathead
Increase in
White sucker
Recruitment
Recruitment
fish
minnow
minnow near
abundance of
recruitment
failure for all
failure for all
experience
extinction;
pearl dace,
failure; no effect
species
species
recruitment
slimy sculpin
Suckers very
on adult growth
failure
decline in
abundant.
and survival
abundance
Lake
Increase
Increased
Lake trout
Recruitment
Lake trout
Lake trout
trout
in
abundance of
recruitment failure;
failure; no effect
condition poor;
condition very
condition
young-of-the-
condition similar to
on adult growth
recruitment
poor; recruitment
i.e., "fatter"
year
preacidifi cation
and survival
failure; reduced
failure; reduced
adult survival
adult survival
Source: (Bakeret al., 1990b).
Table B-30. Range of minimum pH of fish species occurrence in 11 lake surveys.
Family and Species
High Minimum pH
Low Minimum pH
CYPRINIDAE
Bluntnose minnow
6.6
5.6
Fathead minnow
6.3
5.1
Blacknose dace
6.8
5.6
Pearl dace
5.9
4.7
Northern redbelly dace
5.9
4.7
Common shiner
6.2
4.9
Golden shiner
5.5
4.5
Creek chub
5.9
4.6
SALMONIDAE
Brook trout
5.6
4.6
Lake trout
5.2
4.9
Brown trout
5.0
4.6
Atlantic salmon
6.3
5.3
CENTRARCHIDAE
Smallmouth bass
7.0
4.9
Largemouth bass
5.0
4.6
B-117
-------
Family and Species
High Minimum pH
Low Minimum pH
Pumpkinseed
6.6
4.6
Bluegill
4.5
4.5
Rock bass
6.2
4.6
Black crappie
5.6
5.6
PERCIDAE
Yellow perch
5.8
4.4
Walleye
6.9
5.2
Johnny darter
6.2
4.9
Iowa darter
6.2
4.6
ESOCIDAE
Northern pike
5.9
4.0
CATASTOMIDAE
White sucker
5.5
4.6
ICTALURIDAE
Brown bullhead
5.6
4.5
UMBRIDAE
Central mudminnow
4.5
4.2
GASTEROSTEIDAE
Brook stickleback
5.4
4.6
Source: Baker etal. (1990b).
B-118
-------
Table B-31. Studies3 that either did (yes) or did not (no) yield evidence that acidic deposition affected
certain species of birds
Species
Diet/ Foraging
Breeding Distribution
Reproductive Measures
Yes No
Yes No
Yes
No
Reference3
Common loon
X
X X
X
X
1-3,19,20
Common merganser
X
X
19
Belted kingfisher
X
4
Osprey
X
X
X
5,6
Black duck
X
X
xb
7-9
Common goldeneye
xb
8
Ring-necked duck
X
X
10,11
Eurasian dipper
X
X
X
12-14
Eastern kingbird
X
X
15
Tree swallow
X
X
X
16-18
a1= Alvo et al. (1988); 2 = Parker (1988); 3 = Wayland and McNicol (1990); 4 = Goriup (1989); 5 = Eriksson (1983); 6 = Eriksson (1986); 7 = Hunter et al. (1986); 8 = DesGranges and
Darveau (1985); 9 = Rattneret al. (1987); 10,11 = McAuley and Longcore (1988a,b) 12,13 = Ormerod et al. (1985, 1986); 14 = Ormerod and Tyler (1987); 15 = Glooschenko et al.
(1986); 16,17 = Blancher and McNicol (1988,1991); 18 = St. Louis et al. (1990); 19 = Blancher and McNicol (1991); 20 = Blair (1990).. bThe effect was beneficial
Source: Longcore and Gill (1993)
Table B-32. Predicted habitat suitability for lakes in the Algona Model Dataset
Number of Lakes with Suitable Habitat Under Each Emission Scenario
Group Model „ ¦5jj[re.n*if Current pH <6 Current pH 6-6.5 Current pH >6.5
Lanes
S1
S2
S3
S4
S5
S1
S2
S3
S4
S5
S1
S2
S3
S4
S5
Fish
526
338
29
34
40
77
100
97
100
107
124
133
196
196
197
197
197
Common loon
pairs
433
100
12
14
14
16
17
22
22
23
24
24
66
66
66
67
67
Common loon
broods
433
36
2
2
2
2
2
6
6
7
7
7
28
28
28
28
28
Common
merganser pairs
433
52
6
12
27
68
86
14
17
20
33
44
31
31
32
34
35
Common
merganser broods
433
31
6
10
18
69
89
14
15
16
21
29
12
11
10
10
10
Results are expressed as the number of lakes with suitable habitat for fish, common loons (pairs and broods) and common mergansers (pairs and broods) under each emission
scenario (S1, S2, S3, S4, S5) according to current pH classes (<6, 6-6.5, >6.5). Habitat suitability is calculated by probability of presence at time t from WARMS output (#of suitable
lakes at time t/total number of lakes), for fish [n = 526], and for loons and mergansers [n = 433],
B-119
-------
Table B-33. Summary statistics of biological data layers for mercury (Hg) concentrations in fish and
wildlife (|jg/g) in the northeastern U.S. and southeastern Canada.
Hg Concentrations
Category/Species
Sample
Size
Data layer
Designation
Mean ±
Standard
Deviation
Range
Hg Level
of Concern
(Tissue Type)
Percentage of
Samples with
Concentration >
Level of Concern
HUMAN HEALTH
Yellow percha
4089
Primary
0.39 ± 0.49
<0.05-5.24
0.30 (fillet)
50
Largemouth bassb
934
Secondary
0.54 ± 0.35
<0.05-2.66
0.30 (fillet)
75
ECOLOGICAL HEALTH
Brook trout
319
Secondary
0.31 ±0.28
<0.05-2.07
0.16 (whole fish)
75
Yellow perchc
(841 )d
Secondary
0.23 ± 0.35
<0.05-3.18
0.16 (whole fish)
48
Common loone
1546
Primary
1.74 ±1.20
0.11-14.20
3.0 (blood)
11
Bald eagle
217
Secondary
0.52 ± 0.20
0.08-1.27
1.0 (blood)
6
Mink
126
Secondary
19.50 ± 12.1
2.80-68.50
30.0 (fur)
11
River otter
80
Secondary
20.20 ±9.30
1.14-37.80
30.0 (fur)
15
Note: All data are in wet weight except for fur, which is on a fresh-weight basis
aFillet Hg in yellow perch is based on individuals with a standardized length of 20 cm.
bFillet Hg in largemouth bass is based on individuals with a standardized length of 36 cm.
cWhole-fish Hg in yellow perch is based on individuals with a standardized length of 13 cm. Whole-fish Hg for yellow perch was converted to fillet Hg.
dThe sample population of 841 yellow perch examined for whole-fish Hg is included with the 4089 fillets (i.e., the total number of all biotic data layers does not double-count yellow
perch).
eEgg Hg for the common loon was converted to the adult blood equivalent
Table B-34. Mercury concentrations in avian eggs and tissues and related effects.
Tissue
Concen.
(ppm)
Wet (w)
or Dry
(d)
Endpoint
Species
Reference
Liver
1.06
w
No effect
Common tern
Gochfeld (1980)
Liver
22.2
w
Abnormal feather loss in juveniles
Common tern
Gochfeld (1980)
Liver
5
w
Conservative threshold for major toxic
effects
Water birds
Zillioux etal. (1993)
Liver
7.2
w
Increased disease and emaciation
Common tern
Spalding and Forrester
(1991)
Liver
9.08
w
Nesting success
Common tern
Finley and Stendall (1978)
Liver
20.7
w
Hatching success
Common tern
Finley and Stendall (1978)
B-120
-------
Tissue
Concen.
(ppm)
Wet (w)
or Dry
(d)
Endpoint
Species
Reference
Liver
30
W
Neurologic effects
Osprey
Heinz (1974)
Liver
35
w
Death
Common loon
Wiemeyeret al. (1987)
Liver
54.5
w
LD33a
European starling
Finley et al. (1979)
Liver
97.7
w
Death
Gannet
Liver
103.6
w
LD33
European starling
Finley et al. (1979)
Liver
126.5
w
LD33
Red-winged
blackbird
Finley etal. (1979)
Liver
306 total/
20.4 MeHg
d
No adverse effects observed
Black-footed
albatross
Gochfeld (1980)
Brain
4-6
w
Failure to hatch
Black duck
Hoffman and Moore (1979)
Brain
20
w
25% mortality
Zebra finch
Scheuhammer (1988)
Egg
1-5/0.2-1.0
d
Reduced productivity in one half of the
population
Merlin
Newton and Hass (1988)
Egg
0.5-1.5
w
Decreased hatchability
Pheasant
Heinz (1979)
Egg
0.86
w
Aberrant nesting behavior
Common loon
Heinz (1979)
Egg
1.0
w
Successful reproduction
Common tern
Finley and Stendall (1978)
Egg
1.0-3.6
w
Residue threshold for significant toxic
effects
Variety of water
birds
Zilliouxetal. (1993)
Egg
2-16
w
No decreased hatchability
Herring gull
Finley and Stendall (1978)
Egg
3.65
w
27% hatching, 10-12% fledging
Common tern
Finley and Stendall (1978)
Kidney
37.4 total/
6.2 MeHg
d
No adverse effect observed
Black-footed
albatross
Kim etal. (1996)
Kidney
40.4
w
LD33
Grackle
Finley etal. (1979)
Kidney
74.3
w
LD33
Red-winged
blackbird
Finley etal. (1979)
Kidney
86.4
w
LD33
European starling
Finley etal. (1979)
aLD33 = lethal dose, 33%. Source: Wolfe et al. (1998)
B-121
-------
Annex C. Nutrient Enrichment Effects
from Nitrogen
C.1. Effects on Biogeochemical Pathways and Cycles
C.1.1. Nitrogen Cycling in Terrestrial Ecosystems
C.1.1.1. Nitrogen Deposition Effects on DON Leaching
Some N fertilization experiments suggest that increasing N deposition drives an increase in
production of dissolved organic nitrogen (DON) in soil (McDowell et al., 2004; Seely and Lajtha, 1997),
but there is little evidence that elevated N deposition increases the export and loss of DON from terrestrial
ecosystems. Essentially all of the increase in N export across gradients ofN deposition occurs as an
increase in N03 rather than DON export. The latter is typically less than 2 kg N/ha/yr from most
northeastern-forested watersheds (Campbell et al., 2000b; Aber et al., 2003; Goodale et al., 2000; Lovett
et al., 2000).
C.1.1.2. Interactions Between Snow Melt and Nitrate Leaching
Changes in other environmental parameters can also be important. Measurement of nutrient
concentrations in Emerald Lake (Sierra Nevadas) over a period of 19 years suggested that N03
concentration declined between 1983 and 1995. This was likely caused by changes in the snow regime
induced by a drought during the period 1987 to 1992 (Sickman et al., 2003). Years that had shallow and
early melting snowpacks generally had lower snowmelt N03 concentration. In addition, declines in N03
concentration during the growing season even in the wet years of 1993 through 2000 were likely the
result of increased P loading to Emerald Lake and the consequent release of phytoplankton from P
limitation (Sickman et al., 2003).
C.1.1.3. Denitrification: NO and N2O Flux
Davidson et al. (2000) described N gas loss from terrestrial ecosystems using a conceptual model
called "hole-in-the-pipe." In this model, production of NO, N20, and N2 gas are functions of the general
rate of N cycling processes through soil (i.e., the N flux "flowing through the pipe"), combined with
information on soil water content, a key determinant of the ratio of NO:N20 (relative "hole size" for NO
and N20 gas "leakage"). The model formulation has been supported by a range of field measurements in
temperate and especially tropical ecosystems (Davidson et al., 2000), and suggests that processes that
increase the rate of N cycling through soils should also increase the rate of N gas loss from these systems.
Production of NO and N20 tend to be lower in temperate than in tropical ecosystems, largely because of
colder temperatures and slower rates of N cycling in temperate systems, and the frequency of P rather
than N limitation in tropical systems. However, increased availability of N through fertilization can
increase the rate of NO and N20 gas loss from temperate forests.
C-1
-------
Early studies of N gas emission in response to N fertilization experiments at the Harvard Forest,
MA, found small increases in N20 production in response to the highest N treatment (150 kg N/ha/yr) to a
red pine (Pinus resinosa) stand, but N20 losses accounted for <0.4% of N additions (Magill et al., 2000).
However, later studies found that NO emission rates can be more than an order of magnitude greater than
N20 emissions, with NO emissions amounting to 3-4% to 8% (4 to 5 kg N/ha/yr) of the N additions to the
fertilized pine stands (Venterea et al., 2003, 2004). Emissions of NO and N20 increased with fertilization
rate (0, 50, and 150 kg N/ha/yr) in both the red pine and a nearby red oak (Quercus rubra)lxQ& maple
(Acer rubrum) stand (Venterea et al., 2003, 2004). A study of the response of Scots pine (Pinus sylvestris)
stands across a gradient of N deposition in Germany found a threefold to fourfold increase in the rate of
NO and N20 production as N deposition increased from 15 to 22 kg N/ha/yr. In these forests, both gases
were produced in roughly equal amounts, although as N deposition increased, the rate of NO production
increased more steeply than did the rate of N20 production (Butterbach-Bahl et al., 2002).
At Hoglwald, a German site receiving 20 to 30 kg N/ha/yr in throughfall, Butterbach-Bahl et al.
(2002) reported higher emissions of NO than N20 in both a spruce and a beech stand, with N oxide
emissions totaling 4.5 to 6.8 kg N/ha/yr. Intensive laboratory studies suggested additional emissions of N2
gas amounting to 7.2 and 12.4 kg N/ha/yr in the spruce and beech stands, respectively. This is the only
known forest site for which a complete (NO, N02, N20, and N2) N gas budget has been estimated, and in
total, these measurements suggest that soil emissions may balance 46% to 78% of the N received in
throughfall at this site. This result suggests somewhat higher rates of N gas loss than might be inferred
from a series of 15N tracer studies in conifer stands across Europe, which spanned a range of rates of N
input from atmospheric and experimental sources of 3 to 91 kg N/ha/yr. Across all sites, total recovery of
added 15N in soil, vegetation, and lysimeter leachate after 9 to 21 months amounted to 65% to 105% of
added 15N (Tietema et al., 1998), providing a broad constraint on N gas emissions of no more than 35% of
added 15N. The lowest rates of 15N recovery (65% to 67%) occurred at Speuld and Ysselsteyn, two sites in
The Netherlands with the highest rates of chronic throughfall N input (35 to 53 kg N/ha/yr). Although
much more work is needed on complete N gas budgets, several lines of evidence suggest that trace gas
emissions of N may constitute an increasing pathway ofN loss with increasing rates of N deposition.
C.1.1.4. Climate and N2O Interactions
Rainfall events are an important feature controlling N20 produced via denitrification. Rainfall
increases soil moisture. This inhibits 02 diffusion creating anoxic conditions, which increases rates of
denitrification. A study of a spruce forest sites under ambient and elevated N deposition (20 and 30 kg
N/ha/yr, respectively) indicated through most of the study period N20 emission was equivalent between
the sites (Mohn et al., 2000). However, after rainfall events the maximum rate of N20 emission was much
higher for the +N plots, especially when rainfall caused low soil redox potential (an indicator of anoxic
conditions). Another study of mixed spruce, pine and birch forest (100 years old) under well- and poorly
drained soil moisture conditions indicated poorly drained soils produced 1/3 more N20 (118 g N20 N/
ha/yr) than well drained soils. In this study, N deposition was increased in the poorly drained soil from
ambient (12 kg N/ha/yr) to elevated (42 kg N/ha/yr), N20 emissions increased by a factor of more than 2
(254 kg N/ha/yr) (Klemedtsson et al., 1997).
In addition to soil moisture, temperature also influences denitrification. The PnET-N-DNDC model
is designed to simulate and predict soil C and N biogeochemistry in temperate forest ecosystems and to
simulate the emissions of N20 and NO from forest soils. The model couples the PnET model
(Photosynthesis-Evapotranspiartion-Model), the Denitrification-Decomposition (DNDC) model, and an N
module that are further described in Li et al. (1992; 1996; 2000a), Li (2000), and Stange et al. (2000). The
PnET-N-DNDC model is designed to simulate and predict soil C and N biogeochemistry in temperate
forest ecosystems and to simulate the emissions of N20 and NO from forest soils. Denitrification is
described in the model as a series of sequential reductions driven by microorganisms using N oxides as
electron acceptors under anaerobic conditions. As intermediates of the processes, NO and N20 are tightly
C-2
-------
controlled by the kinetics of each step in the sequential reactions. The capacity of this model to simulate
N trace gas emissions from forest soils was tested by comparing model results with results form field
measurements at 19 different field sites across Europe and 1 site in the U.S. (Kesik et al., 2005). Possible
feedbacks of temperature and precipitationchange on forest soil NO and N20 emissions in Europe were
investigated using PnET-N-DNDC (Kesik et al., 2006). The model results indicated decreasing
precipitation and increasing temperature in areas with light texture soils (below 15%) resulted in
decreased soil moisture values; in turn, N20 production by denitrification decreases. Under these same
environmental conditions, NO production by nitrification increases. Most laboratory studies show
increasing temperature increases N20 production, however if water filled pore space (WFPS) increases to
70-80%, then N2 rather than N20 is the main product of denitrification and N20 emissions go down. This
illustrates how N20 emissions increase with increasing soil moisture until soil moisture become more
conducive to N2 emission.
C.1.2. Nitrogen Cycling in Transitional Ecosystems
C.1.2.1. Denitrification: Measurement Techniques
There are a variety of methods for measuring denitrification rates in wetland, freshwater and
marine sediments, including measurements of N03 loss, N2 production, N20 accumulation in response to
acetylene inhibition of N20 reduction, isotopic methods, and N2:argon (Ar) measurement by membrane
inlet mass spectrometry (MIMS) (Smith, 2006). Most direct measurements of denitrification have
measured only rates of production of N20, or used the "acetylene block" technique of inhibiting
transformation of N20 to N2 and monitoring the accumulation of N20 as a surrogate of the sum of N20
and N2. The acetylene block method is highly problematic, however, as it also inhibits rates of
nitrification, and so denitrification rates are strongly underestimated where nitrification and denitrification
processes are coupled closely in space or time (Groffman et al., 2006).
Techniques can be based on laboratory incubation of sediment cores or in situ studies. Each
method has advantages and disadvantages and many studies have been conducted to compare results
among the various methods (e.g., Bernot et al., 2003; Seitzinger, 1988, 1993, 2002b; Smith et al., 2006a;
Groffman et al., 2006). Kana et al. (1998) described the MIMS method to measure small changes in
dissolved N2 caused by denitrification in sediments. This technique allows measurement of N2 flux in
unperturbed sediment cores with high temporal resolution (Kana et al., 1998). This is especially useful
during summer conditions when N03 concentration in the water is typically low, but the high
temperature can support high rates of denitrification. Under such conditions, it is likely that a coupled
sequence of nitrification and denitrification accounts for substantial N loss from estuarine sediments
(Kemp et al., 1990). Constraints regarding field and analytical methods have seriously limited
understanding of the magnitude and controls on denitrification (Groffman et al., 2006).
C. 1.2.2. Nitrogen Deposition Effects on Methane
Increased N loading to transitional ecosystems can affect both methane (CH4)-producing and de-
oxidizing microbial activity. The difference between the CH4 production and oxidation determines the
magnitude of CH4 emission from soils. There is evidence to support that ammonium compounds reduce
CH4 oxidation (Gulledge et al., 1997; King and Schnell, 1994; Steudler et al., 1989), but ammonium
compounds have also been observed to increase methanotropic bacterial activity (Bodelier et al., 2000). In
general CH4 emissions from saturated soils have been observed to increase with N addition (Granberg
et al., 2001; Saarnio et al., 2003; Zhang et al., 2007). A hypothesis for explaining this effect is that
increases in vegetative cover caused by N addition increase C availability through root exudates, which in
C-3
-------
turn stimulates methanogenic bacteria and CH4 emissions (Granberg et al., 2001; Saarnio et al., 2003;
Zhang et al., 2007).
Saarnio et al. (2003) observed moderate increases in CH4 emissions from boreal wetland soils with
N fertilization rates of 30 kg N/ha/yr as ammonium nitrate (NH4NO3). Comparable N application rates
and effects on CH4 emissions were also observed by Granberg et al. (2001) in a similar ecosystem type.
Zhang et al. (2007) observed elevated CH4 emissions from freshwater wetland soils with experimental N
additions of 240 kg N/ha/yr. They postulated that additional N increased abundance of Deyeucia
angustifolia which increased CH4 emissions by supplying methanogenic bacteria with additional substrate
in the form of root exudates. Other studies have shown that N addition had little or no effect on CH4
emissions across a variety of ecosystem types (Ambus and Robertson, 2006; Saarnio et al., 2000; Silvola
et al., 2003). Note that the N enrichment rates employed in all of the above reported studies related to N
effects on soil CH4 emissions were greater (30 to 240 kg N/ha/yr) than atmospheric N inputs in most areas
of the U.S. that are heavily effected by elevated atmospheric N deposition. See ISA Section 3.4 for a
discussion of methane flux from terrestrial, transition and aquatic ecosystems.
C.1.3. Nitrogen Cycling in Estuarine Ecosystems
C.1.3.1. Denitrification and Anammox in Estuarine Ecosystems
Denitrification is a major factor governing the loss of N from estuarine ecosystems. Denitrification
by microbes found in estuarine and marine sediments releases much of the added N inputs back into the
atmosphere (Arrigo, 2005; Vitousek et al., 1997). Collection of quantitative data on this process has been
hampered, however, by the complexity of environmental controls on the denitrification process and
difficulties in measuring denitrification rates (Kana et al., 1998). Major environmental controls include
temperature and the availability of N03~, 02, and organic materials (Rysgaard et al., 1994; Seitzinger,
1988).
Marine microbial ecology is highly complex and poorly understood. Relatively new knowledge
about anammox bacteria has completely altered scientific understanding of N cycling in the oceans.
Although it was previously believed that denitrification was responsible for virtually all of the transfer of
Nr in the ocean to the atmosphere as N2 gas, it now appears that anaerobic ammonium oxidation
(anammox) may account for up to 50% of the N2 production in the oceans (Dalsgaard et al., 2005; Devol,
2003; Kuypers et al., 2005; Ward, 2003). This reaction uses N02 as the primary electron acceptor and is
catalyzed by planctomycete bacteria of the genera Brocadia, Kuenenia, and Scalindua. That NH/ could
be oxidized under anoxic conditions was theorized several decades ago based on calculations of the ratios
among N, P, and C in marine ecosystems. Nevertheless, the process was not experimentally documented
until the 1990s (van de Graaf et al., 1995). More recently, anammox has been detected in a variety of
freshwater, estuarine, and marine waters and sediments (Dalsgaard et al., 2005; Devol, 2003; Engstrom
et al., 2005; Jetten et al., 2003, 2005; Kuypers et al., 2005; Op Den Camp et al., 2006; Pilcher, 2005;
Rysgaard et al., 2004; Ward, 2003)
C.1.3.2. Nitrogen Budgets
The greatest uncertainty in the development of detailed N budgets for coastal ecosystems is
quantifying how much of the N deposited on the watershed is transferred through the terrestrial watershed
to the estuary. The difficulty stems from multiple agricultural and mobile and stationary fuel combustion
emission sources in an estuary watershed; quantifying dry deposition to the estuary surface and to the
watershed; measuring gaseous losses of NH3 and NOx compounds to the atmosphere; and complex N
flow pathways through the watershed (NRC, 2000). Published estimates of the contribution of
C-4
-------
atmospheric deposition to estuary N load exhibit wide variability. Such estimates for a specific estuary
may differ. Some examples are summarized in Table C-l. Despite the variability, it appears that
atmospheric sources of N loading to estuaries in the U.S. can be quantitatively important. The major
sources of N to estuaries and near-coastal marine waters in the U.S., in addition to atmospheric
deposition, include wastewater effluent derived mainly from food imports and consumption, fertilizer
application, livestock feed imports, and N-fixing crops (Boyer et al., 2002; Driscoll et al., 2003).
Table C-1. Estimated percent of total N load to Delaware Bay and Hudson River/Raritan Bay contributed
by atmospheric deposition.
Reference
Percent of N Load Contributed by Atmospheric Deposition
Delaware Bay Hudson River/Raritan Bay
Paerl (1985)
44
—
Hinga et al. (1991)
—
33
Scudlark and Church (1993)
15
—
Paerl (1995)
—
68
Jaworski et al. (1997)
44
68
Alexander et al. (2001)
22
26
Castro et al. (2001)
—
10
Stacey et al. (2001)
Land-based
16
10
Sparrow model
25
27
Castro and Driscoll (2002)
20
17
Castro et al. (2003)
23
18
Boyer et al. (2002) estimated that atmospheric deposition averaged 31% of total N inputs over the
combined area of the 16 northeastern river basins. Contributions from atmospheric deposition ranged
from 60% of N inputs for the basins in northern Maine to 15 to 20% for the Schuylkill and Potomac River
Basins, the latter of which had large agricultural N inputs. Across all basins, estimated riverine export of
N amounted to 25% of total N inputs, and ranged from 11% to 40%. This result is consistent with a
similar analysis by Howarth et al. (1996), who found that basins draining to the North Atlantic exported
approximately 25% of anthropogenic N inputs on average.
Turner et al. (2001) found a strong correlation between population density (persons/ km2) and the
total N loading from watershed to estuary (r2 = 0.78) for coastal watersheds in the U.S. This finding is
likely due to the prevalence of automobiles in heavily populated areas, along with their associated N
emissions and deposition, plus the myriad non-atmospheric sources of N from human activities,
particularly sewage releases. They also determined that direct atmospheric deposition becomes
increasingly more important as a contributor to the total N loading to an estuary as the water surface area
increases relative to total watershed area (terrestrial plus water surfaces). Turner et al. (2001) found that,
on average, direct atmospheric deposition of N accounted for more than an estimated 25% of the estuarine
C-5
-------
N load when the estuary surface occupied 20% or more of the overall watershed area. Few of the estuaries
in the eastern U.S. comprise such a large percentage of their watershed (Castro et al., 2001).
The estimates of the effect of direct atmospheric deposition to estuary surfaces are hampered by
uncertainties in dry deposition rates. Many published studies have assumed that dry N deposition is equal
to measured wet deposition (Fisher and Oppenheimer, 1991; Hinga et al., 1991; Scudlark and Church,
1993), and this is probably biased high (Baker, 1991a). However, other studies have assumed dry N
deposition rates for estuarine and near coastal areas are equal to 40% of wet (Jaworski et al., 1997) or
67% of wet (Meyers et al., 2000b). Of particular importance, the rate of dry deposition to open water
surfaces is much lower than the rate of dry deposition to vegetated terrestrial surfaces. Paerl et al. (2001a)
estimated that dry deposition to open estuarine surfaces is three to five times lower than to vegetated
surfaces. This difference is seldom considered in N-budgeting studies, and can have a substantial effect
on estimates of direct atmospheric loading to estuary surfaces, which is especially important for estuaries
having low watershed area to estuary surface area ratio.
A number of empirical approaches have been developed to quantify N fluxes to the coastal zone
which rely on estimates of N sources within the watershed and characteristics of the landscape. Alexander
et al. (2002) compared several of these empirical methods, the most accurate and least biased of which
was that of Howarth et al. (1996). A modified version of the Howarth et al. (1996) methodology was
published by Boyer et al. (2006). More mechanistic approaches include those of Bouwman et al. (2005),
Van Drecht et al. (2003), and Green et al. (2004).
C.1.4. Timing of Chemical Change
C.1.4.1. Interannual Change: Nitrate Leaching
Interannual changes in N cycling can be reflected in changes in streamwater chemistry. N03
leaching from terrestrial ecosystems throughout the 1980s was observed in many of the original lakes in
EPA's Adirondack Long Term Monitoring (ALTM) program (Driscoll and Van Dreason, 1993), which
was followed by a decline during the 1990s. As a consequence of this subsequent decline, Driscoll (2003)
reported an overall significant (p <0.1) decrease in N03 concentration for the period 1982 to 2000 for 8
of the 16 original ALTM monitoring sites. Only the one mounded seepage lake in the study (Little Echo
Pond) had a small, but statistically significant, increase in N03 concentration (0.01 (ieq/L/yr, p <0.06). It
is not clear why many Adirondack watershed soils leached N03 to a lesser extent during the 1990s than
they did during the 1980s (Driscoll et al., 2003). Decreasing stream N03 concentrations during the 1990s
was also observed in the Catskill Mountains (Stoddard et al., 2003) and in New Hampshire (Goodale
et al., 2003). There was not a substantial change in N emissions or deposition in the Northeast region over
that period. Climatic factors, insect defoliation, increases in atmospheric C02, and interactions with
increasing availability of DOC have been proposed as possible contributing factors to regional decreases
in N03 leaching (Aber et al., 2002; Driscoll et al., 2003; Goodale et al., 2003; 2005; Mitchell et al.,
1996), but the driver of this decadal scale pattern remains uncertain.
C.1.4.2. Episodic Change
Nutrient enrichment effects of N deposition are controlled to a large degree by biological and
hydrological processes that operate on episodic (hours to days), seasonal, and interannual time scales.
Nitrogen uptake and transformation reactions and processes vary greatly with season and with climatic
factors. In particular, N export from terrestrial and transitional ecosystems to aquatic ecosystems is
governed by seasonal fluctuations in temperature and biological uptake, and episodic fluctuations in water
movement associated with rainstorms and snowmelt. The role of N in driving biotic change in stream
C-6
-------
ecosystems due to episodic pulses of N03 associated with spring snowmelt are discussed in detail in ISA
Section 3.2.
C.1.4.3. Reversibility of Impacts
Although there are relatively few studies of the reversibility of the biogeochemical effects of
elevated N deposition, the few to date suggest the possibility of recovery. Experimental studies in Europe
suggest that some ecosystem processes and characteristics are likely to recover rapidly following a
reduction in N deposition. In a study in northern Sweden, high levels of fertilization (90 kg N/ha/yr) over
20 years induced substantial soil acidification, including loss of over half of the base cations in the
mineral soil, a decrease in pH, and an increase in soluble A1 (Hogberg et al., 2006). However, 10 years
after this treatment was stopped, the pH of the mineral soil had increased, and extractable N03 was no
higher than in the control plot. Stem volume growth did not substantially increase relative to the
acidification period. "Clean roof' experiments that prevent N deposition inputs at sites receiving >40 kg
N/ha/yr ambient atmospheric N deposition in The Netherlands increased wood and root production soon
after the roof was installed (Boxman et al., 1998), and N03 exports below the rooting zone were reduced
dramatically within 2 years (Bredemeier et al., 1998).
A study of alpine lake sediment cores in Rocky Mountain National Park, CO (Wolfe et al., 2003)
suggested the possible reversibility of N enrichment effects on lake biota. Although increased dominance
of mesotrophic diatom species was correlated with increased N inputs during the 20th century, it did not
appear that any of the oligotrophic species had been totally lost from study lakes. Thus, reduced future N
loading may allow renewed dominance by oligotrophic diatom species.
It is not necessarily true, however, that nutrient enrichment effects of N deposition will, in all
cases, be easily reversible. For example, it has been suggested that vegetation conversion in the coastal
sage scrub community in California has altered hydrologic function to an extent that may be difficult to
reverse. The depth of rainwater percolation into soil has been reduced as a result of invasion of non-native
annual grasses. This hydrologic change inhibits the growth of deep-rooting native shrubs (Wood et al.,
2006).
C-7
-------
C.1.5. Tables Supporting Cross Ecosystem Evaluation of N2O, CH4 and
CO2 Flux
Table C-2 summarizes key information from the experiments included in the meta-analysis
presented in ISA Section 3.3.4.
Table C-2 The study site, experimental condition, ecosystem type, N form, amount of N addition and
citations is presented for all studies used in NEE, EC, CH4 uptake, CH4 emission and N20
emission meta analyses.
Site
Experimental
Condition
Ecosystem
N Form
N
Addition
(kg
ha/yr)
Reference
NET ECOSYSTEM CARBON EXCHANGE (NEE)
Swiss FACE (Lolium)
field
grassland
NH4NO3
320
Aeschlimann etal. 2005
Swiss FACE (Trifolium)
field
grassland
NH4NO3
320
Aeschlimann etal. 2005
Ottawa, Canada
field
wetland
NH4NO3
32
Basiliko et al. 2006
Ottawa, Canada
field
wetland
NH4NO3
64
Basiliko et al. 2006
Ottawa, Canada
field
wetland
NH4NO3
16
Bubrie et al. 2007
Ottawa, Canada
field
wetland
NH4NO3
32
Bubrie et al. 2007
Ottawa, Canada
field
wetland
NH4NO3
64
Bubrie et al. 2007
Laqueuille, France
field
grassland
175
Soussana et al. 2007
Oensingen, Switzerland
field
grassland
slurry
200
Soussana et al. 2007
Toolik Lake, AK(lnlet)
field
tundra
NH4NO3
100
Shaver etal. 1998
Toolik Lake, AK (Outlet)
field
tundra
NH4NO3
100
Shaver etal. 1998
Orange county, CA
field,w/o water addition
grassland
NO3-
100
Harpole et al., 2007
Orange county, CA
field, with water addition
grassland
NO3-
100
Harpole et al., 2007
Switzerland
field
grassland
40
Diemer 1997
Finland
field
wetland
NH4NO3
30
Saarnio et al. 2003
Abisko, Sweden
field
tundra
NH4NO3
100
Christensen etal. 1997
ecosystem carbon content (ec)
WA
field
coniferous
Canary et al. 2000
Placerville CA
field
coniferous
NH4+
100
Johnson et al. 2006
Placerville CA
field
coniferous
nh4+
200
Johnson et al. 2006
C-8
-------
Clayey, AL
field
coniferous
DAP
45
Leggettet al. 2006
Sandy, MS
field
coniferous
DAP
45
Leggettet al. 2006
Harvard forest, MA
field
deciduous
NH4NO3
150
Magill et al. 2004
Harvard forest, MA
field
coniferous
NH4NO3
150
Magill et al. 2004
Harvard forest, MA
field, w/o S addition
deciduous
NH4NO3
50
Magill et al. 2004
Harvard forest, MA
field, with S additon
deciduous
NH4NO3
50
Magill et al. 2004
Harvard forest, MA
field, w/o S addition
coniferous
NH4NO3
50
Magill et al. 2004
Harvard forest, MA
field, with S addition
coniferous
NH4NO3
50
Magill et al. 2004
BBWM, ME
field
deciduous
NH4+
25.2
Parker etal. 2001; Elvir
etal. 2006
Ml (site A)
field
deciduous
NO3-
30
Pregitzer et al. 2008;
Burton et al. 2000
Ml (site B)
field
deciduous
NO3-
30
Pregitzer et al. 2008;
Burton et al. 2000
Ml (site C)
field
deciduous
NO3-
30
Pregitzer et al. 2008;
Burton et al. 2000
Ml (site D)
field
deciduous
NOs-
30
Pregitzer et al. 2008;
Burton et al. 2000
FL
field
coniferous
Shan et al. 2001
CH4 EMISSION
Netherland (eutrophic site)
Incubation
wetland
NH4+
Aerts and Caluwe 1999
Netherland (mesotrophic site)
incubation
wetland
nh4+
Aerts and Caluwe 1999
Polish (mesotrophic site)
incubation
wetland
nh4+
Aerts and Caluwe 1999
Netherland
incubation
wetland
nh4+
Aerts and Caluwe 1999
Swiss FACE
field, ambient CO2
grassland
NH4N03
84
Baggs and Blum 2004
Swiss FACE
field, elevated CO2
grassland
NH4N03
84
Baggs and Blum 2004
Minnesota
Incubation
wetland
nh4+
20
Keller etal., 2005
Minnesota
incubation
wetland
nh4+
60
Keller etal., 2005
Niwot Ridge, CO
field
grassland
urea
250
Neef etal. 1994
Finland
incubation
wetland
NH4NO3
30
Nykanen et al. 2002
Finland
incubation
wetland
NH4NO3
100
Nykanen et al. 2002
Salmisuo, Finland
field, ambient CO2
wetland
NH4NO3
30
Sarrnioand Silvola 1999
Salmisuo, Finland
field, elevated CO2
wetland
NH4NO3
30
Sarrnioand Silvola 1999
Sanjiang mire, China
field
wetland
NH4NO3
240
Zhang et al. 2007
C-9
-------
ch4 uptake
Quebec, Canada
incubation
coniferous
NH4+
Adamsen and King 1993
Quebec, Canada
incubation
coniferous
N03-
Adamsen and King 1993
Michigan
field
coniferous
NH4NO3
10
Ambus and Robertson
2006
Michigan
field
coniferous
NH4NO3
30
Ambus and Robertson
2006
Michigan
field
deciduous
NH4NO3
10
Ambus and Robertson
2006
Michigan
field
deciduous
NH4NO3
30
Ambus and Robertson
2006
Michigan
field
grassland
NH4NO3
10
Ambus and Robertson
2006
Soiling, Germany
field, with clean rain roof
coniferous
NH4NO3
18.5
Borken et al. 2002
Soiling, Germany
field, with clean rain roof
coniferous
NH4NO3
24.2
Borken et al. 2002
Soiling, Germany
field, w/o clean rain roof
coniferous
NH4NO3
18.5
Borken et al. 2002
Soiling, Germany
field, w/o clean rain roof
coniferous
NH4NO3
24.2
Borken et al. 2002
Bousson, PA
field
deciduous
NH4NO3
100
Bowen et al. 2000
Perridge Forest, UK
incubation
deciduous
NO3-
LowN
Bradford et al. 2001
Perridge Forest, UK
incubation
deciduous
NO3-
High N
Bradford et al. 2001
MT Ascutney, VT
field
coniferous
NH4+
31.4
Castro etal. 1992
Bousson, PA
field
deciduous
NH4NO3
100
Chan et al. 2005
Abisko, Sweden
field
tundra
NH4NO3
100
Christensen etal. 1997
Finland
field
wetland
nh4+
100
Crilletal. 1994
Finland
field
wetland
no3-
100
Crilletal. 1994
Finland
field
wetland
urea
100
Crilletal. 1994
Villingen, Germany
field
coniferous
NH4+
150
Gulledge and Schimel
2000
LTER Alaska
field
deciduous
NH4NO3
66.7
Gulledge and Schimel
2000
LTER Alaska
field
coniferous
NH4NO3
12.3
Gulledge and Schimel
2000
LTER Alaska
field
deciduous
NH4NO3
171.4
Gulledge and Schimel
2000
LTER Alaska
field
coniferous
NH4NO3
142.9
Gulledge and Schimel
2000
C-10
-------
Swiss FACE
field, ambient CO2
grassland
NH4NO3
560
Ineson etal. 1998
Swiss FACE
field, elevated CO2
grassland
NH4NO3
560
Ineson etal. 1998
Scotland
incubation
heathland
NH4+
40
Macdonald et al. 1997
Scotland
incubation
heathland
NH4NO3
40
Macdonald et al. 1997
Scotland
incubation
heathland
NO3-
40
Macdonald et al. 1997
Finland
field
coniferous
NH4NO3
200
Maljane et al. 2006
Swale, CO
field
grassland
urea
450
Mosieretal., 1991
Midslope, CO
field
grassland
urea
450
Mosieretal., 1991
Passture, CO
field
grassland
urea
450
Mosieretal., 1991
Niwot Ridge, CO
field
grassland
urea
250
Neef etal. 1994
Villingen, Germany
field
coniferous
NH4+
150
Papen et. al 2001
Finland
field
wetland
NH4NO3
30
Saarnio et al. 2003
Gjovelandsnesset, Norway
incubation
coniferous
NH4NO3
30
Sitaula et al. 1995
Gjovelandsnesset, Norway
incubation
coniferous
NH4NO3
90
Sitaula et al. 1995
Harvard forest, MA
field
coniferous
NH4NO3
37
Steudler et al. 1989
Harvard forest, MA
field
deciduous
NH4NO3
37
Steudler et al. 1989
Harvard forest, MA
field
coniferous
NH4NO3
120
Steudler etal. 1989
Harvard forest, MA
field
deciduous
NH4NO3
120
Steudler etal. 1989
Belgium (Bulrush site)
incubation
wetland
NH4+
VanderNatetal. 1997
Belgium (Reed site)
incubation
wetland
nh4+
VanderNatetal. 1997
Costa Rica (Loam site)
field
tropical
forest
nh4+
65
Weitzetal. 1999
Costa Rica (Clay site)
field
tropical
forest
nh4+
65
Weitzetal. 1999
N20 EMISSION
Michigan biological Station
Incubation, ambient CO2
deciduous
NH4NO3
LowN
Ambus and Robertson
1999
Michigan biological Station
Incubation, ambient CO2
deciduous
NH4NO3
High N
Ambus and Robertson
1999
Michigan biological Station
Incubation, elevated CO2
deciduous
NH4NO3
LowN
Ambus and Robertson
1999
Michigan biological Station
Incubation, elevated CO2
deciduous
NH4NO3
High N
Ambus and Robertson
1999
Michigan
field
coniferous
NH4NO3
10
Ambus and Robertson
2006
C-11
-------
Michigan
field
deciduous
NH4NO3
10
Ambus and Robertson
2006
Michigan
field
grassland
NH4NO3
10
Ambus and Robertson
2006
Michigan
field
coniferous
NH4NO3
30
Ambus and Robertson
2006
Michigan
field
deciduous
NH4NO3
30
Ambus and Robertson
2006
Hyytiala
incubation
coniferous
NH4+
Ambus et al. 2006
Speuiderbos
incubation
coniferous
nh4+
Ambus et al. 2006
San Rossore
incubation
coniferous
nh4+
Ambus et al. 2006
Giencorse
incubation
coniferous
nh4+
Ambus et al. 2006
Nyirjes
incubation
coniferous
nh4+
Ambus et al. 2006
Achenkirch
incubation
coniferous
nh4+
Ambus et al. 2006
Hoglwald
incubation
coniferous
nh4+
Ambus et al. 2006
Sor0
incubation
deciduous
nh4+
Ambus et al. 2006
Bosco negri
incubation
deciduous
nh4+
Ambus et al. 2006
Schottenwald
incubation
deciduous
nh4+
Ambus et al. 2006
Hyytiala
incubation
coniferous
NO3-
Ambus et al. 2006
Speuiderbos
incubation
coniferous
NO3-
Ambus et al. 2006
San Rossore
incubation
coniferous
NO3-
Ambus et al. 2006
Giencorse
incubation
coniferous
NOs-
Ambus et al. 2006
Nyirjes
incubation
coniferous
NOs-
Ambus et al. 2006
Achenkirch
incubation
coniferous
NO3-
Ambus et al. 2006
Hoglwald
incubation
coniferous
NO3-
Ambus et al. 2006
Soro
incubation
deciduous
NOs-
Ambus et al. 2006
Bosco negri
incubation
deciduous
NOs-
Ambus et al. 2006
Schottenwald
incubation
deciduous
NO3-
Ambus et al. 2006
Swiss FACE
field, ambient CO2
grassland
NH4NO3
84
Baggs and Blum 2004
Swiss FACE
field, elevated CO2
grassland
NH4NO3
84
Baggs and Blum 2004
Swiss FACE (Lolium)
field, ambient CO2
grassland
NH4NO3
420
Baggs et al., 2003
Swiss FACE (Lolium/Trifolium)
field, ambient CO2
grassland
NH4NO3
420
Baggs et al., 2003
Swiss FACE (Trifolium)
field, ambient CO2
grassland
NH4NO3
420
Baggs et al., 2003
Swiss FACE (Lolium)
field, elevated CO2
grassland
NH4NO3
420
Baggs et al., 2003
C-12
-------
Swiss FACE (Lolium/ Trifolium) field, elevated C02 grassland NH4NO3 420 Baggs etal., 2003
Swiss FACE (Trifolium) field, elevatedC02 grassland NH4NO3 420 Baggs etal., 2003
Soiling, Germany
field
coniferous
NH4NO3
18.5
Borken et al. 2002
Soiling, Germany
field
coniferous
NH4NO3
24.2
Borken et al. 2002
Soiling, Germany
field
coniferous
NH4NO3
18.5
Borken et al. 2002
Soiling, Germany
field
coniferous
NH4NO3
24.2
Borken et al. 2002
Harvard forest
field
deciduous
NH4NO3
37
bowden etal. 1991
Harvard forest
field
coniferous
NH4NO3
37
bowden etal. 1991
Harvard forest
field
deciduous
NH4NO3
120
bowden etal. 1991
Harvard forest
field
coniferous
NH4NO3
120
bowden etal. 1991
Soiling, Germany
field
deciduous
NH4+
140
Brumme and Beese
1992
Gullane
incubation
deciduous
nh4+
Castaldi and Smith 1998
Gullane
incubation
deciduous
NO3-
Castaldi and Smith 1998
Ascutney, VI
field
coniferous
nh4+
31.4
Castro etal. 1993
Allt, UK (site M3)
field
heathland
NH4N03
20
Curitis et al., 2006
Allt, UK (site M4)
field
heathland
NH4N03
20
Curitis et al., 2006
Afon Gwy, UK (site G1)
field
grassland
NH4N03
20
Curitis et al., 2006
Afon Gwy, UK (site G2)
field
grassland
NH4N03
20
Curitis et al., 2006
Afon Gwy, UK (site G3)
field
grassland
NH4N03
20
Curitis et al., 2006
River Etherow, UK (site E1)
field
heathland
NH4N03
20
Curitis et al., 2006
River Etherow, UK (site E2)
field
heathland
NH4N03
20
Curitis et al., 2006
St. James Parish, LO
field
wetland
no3-
100
Delauneetal. 1998
St. James Parish, LO
field
wetland
nh4+
100
Delauneetal. 1998
Saratoga
field
grassland
NH4N03
168
Delgadoetal. 1996
Puerto Rico
field
tropical
forest
urea
300
Erickson et al. 2001
Jasper Ridge, CA
incubation, ambient CO2
grassland
urea
200
Hungateetal. 1997
Jasper Ridge, CA
incubation, elevated CO2
grassland
urea
200
Hungateetal. 1997
Swiss FACE
field, ambient CO2
grassland
NH4NO3
560
Ineson etal. 1998
Swiss FACE
field, elevated CO2
grassland
NH4NO3
560
Ineson etal. 1998
Manaus, Brazil (site 1)
field
tropical
forest
NO3-
Keller etal. 1988
C-13
-------
Manaus, Brazil (site 2)
field
tropical
forest
no3-
Keller etal. 1988
Manaus, Brazil (site 3)
field
tropical
forest
no3-
Keller etal. 1988
Manaus, Brazil (site 4)
field
tropical
forest
no3-
Keller etal. 1988
Manaus, Brazil (site 1)
field
tropical
forest
nh4+
Keller etal. 1988
Manaus, Brazil (site 2)
field
tropical
forest
nh4+
Keller etal. 1988
Manaus, Brazil (site 3)
field
tropical
forest
nh4+
Keller etal. 1988
Manaus, Brazil (site 4)
field
tropical
forest
nh4+
Keller etal. 1988
gardsjon watershed
field, well drained
coniferous
nh4no3
35
Klemedtsson et. al 1997
gardsjon watershed
field, wet
coniferous
nh4no3
35
Klemedtsson et. al 1997
Avoyelles, LO
field
wetland
nh4+
100
Lindau etal. 1994
Avoyelles, LO
field
wetland
nh4+
300
Lindau etal. 1994
Avoyelles, LO
field
wetland
no3-
100
Lindau etal. 1994
Avoyelles, LO
field
wetland
NOr
300
Lindau etal. 1994
Harvard forest
field
coniferous
NH4N03
113
Magill etal. 1997
Finland
field
coniferous
200
Maljane et al. 2006
Swale, CO
field
grassland
urea
450
Mosieretal., 1991
Midslope, CO
field
grassland
urea
450
Mosieretal., 1991
Passture, CO
field
grassland
urea
450
Mosieretal., 1991
Puerto Rico
field, low tide
wetland
NH4+
15.4
Munoz-Hincapie et al.
2002
Puerto Rico
field, low tide
wetland
nh4+
130.2
Munoz-Hincapie et al.
2002
Puerto Rico
field, low tide
wetland
nh4+
266
Munoz-Hincapie et al.
2002
Puerto Rico
field, low tide
wetland
NOr
15.4
Munoz-Hincapie et al.
2002
Puerto Rico
field, low tide
wetland
no3-
130.2
Munoz-Hincapie et al.
2002
Puerto Rico
field, low tide
wetland
no3-
266
Munoz-Hincapie et al.
2002
Puerto Rico
field, high tide
wetland
nh4+
15.4
Munoz-Hincapie et al.
2002
C-14
-------
Puerto Rico
field, high tide
wetland
NH4+
266
Munoz -Hincapie et al.
2002
Puerto Rico
field, high tide
wetland
NO3-
15.4
Munoz-Hincapie et al.
2002
Niwot Ridge, CO (wet meadow)
field
grassland
urea
250
Neef etal. 1994
Niwot Ridge, CO (dry meadow)
field
grassland
urea
250
Neef etal. 1994
Villingen
field
coniferous
NH4+
150
Papen et. al (2001)
Duke FACE
incubation, ambient CO2
coniferous
NO3-
Phillips et al. 2001
Duke FACE
incubation, elevated CO2
coniferous
NOs-
Phillips et al. 2001
Finland
field
wetland
NOs-
100
Reginaetal. 1998
Finland
field
wetland
NH4+
100
Reginaetal. 1998
Finland
field
wetland
urea
100
Reginaetal. 1998
Mojave
incubation
desert
NO3-
Schaeffer et al. 2003
Gjovelandsnesset, Sweden
field, pH=3
coniferous
NH4NO3
30
Sitaula et al. 1995
Gjovelandsnesset, Sweden
field, pH=3
coniferous
NH4NO3
90
Sitaula et al. 1995
Gjovelandsnesset, Sweden
field, pH=4
coniferous
NH4NO3
30
Sitaula etal. 1995
Gjovelandsnesset, Sweden
field, pH=4
coniferous
NH4NO3
90
Sitaula etal. 1995
Gjovelandsnesset, Sweden
field, pH=5
coniferous
NH4NO3
30
Sitaula etal. 1995
Gjovelandsnesset, Sweden
field, pH=5
coniferous
NH4NO3
90
Sitaula etal. 1995
Scotland
field
coniferous
NH4NO3
48
Skibaetal. 1998
Deepsyke forest
field
coniferous
NH4NO3
48
Skibaetal. 1999
Deepsyke forest
field
coniferous
NH4NO3
96
Skibaetal. 1999
Costa Rica (Loam site)
field
tropical
forest
65
Weitzetal. 1999
Costa Rica (Clay site)
field
tropical
forest
65
Weitzetal. 1999
Sanhuabg Mire, China
field
wetland
NH4NO3
240
Zhang et al. 2007
C.2. Terrestrial Ecosystems
The following sections are organized by ecosystem type and combine information that is
supplemental to Section 3.3 of the ISA.
C-15
-------
C.2.1. General C Cyling
C cycling is a complex process that includes C capture from the atmosphere by autotrophic biota,
the primary producers of the ecosystem, and respiration (autotrophic + heterotrophic). In general,
atmospheric nutrient (e.g., N) deposition on an ecosystem that is deficient in that nutrient will often cause
an increase in growth, at least initially, especially of the primary producers. If that same nutrient is
deposited on an ecosystem that has an adequate supply of that nutrient, there may be no appreciable
nutrient enrichment effect, at least up to a point. Nutrient input that is greatly in excess of biological
demand will often cause toxicity, reduced growth, or problems other than those associated with nutrient
enrichment (i.e., N-saturation, acidification, base cation depletion) (Figure C-l).
Deficiency
Sufficiency
Toxic
15
c
ra
CL
>
Nutrient Supply
Source: U.S. EPA (1993a)
Figure C-1. Schematic representation of the response of vegetation to nutrient addition.
C.2.2. Forest Growth Interactions with Herbivores
Light availability, nutrient balance, and C:N:P stoichiometry are closely relate, and affect the
composition of autotrophic species that will occupy a particular habitat. The resulting stoichiometric
balance of C:N:P in the autotrophic community can have additional feedbacks on nutrient cycling by
herbivores, detritivores, and decomposers (Sterner and Elser, 2002). Such effects also extend to
herbivores, and likely other members of the food web. Forkner and Hunter (2000) altered plant growth of
oak (Onerous primis and O. rubra) saplings through fertilizer (N, P, K) addition and then censused the
densities of insect herbivore guilds and predaceous anthropods on experimental and control trees. In
general, leaf chewers, phloem feeders, and leaf miners were more common on fertilized, as compared
with non-fertilized, trees. Predaceous arthropods were also more abundant on fertilized trees and their
densities were correlated with herbivore densities.
C.2.3. Southern California Coniferous Forest
Wet N deposition is generally low throughout the region, in the range of 1 to 3 kg N/ha/yr.
However, dry deposition is highly variable, but ranges up to about 30 kg N/ha/yr or more (Bytnerowicz
and Fenn, 1996; Fenn, 1997; Takemoto et al., 2001). Available data (e.g., Minnich et al., 1995) suggest
progression toward less needle retention, higher shoot:root biomass ratios, increasing depth of litter, and
high NO;, in soil solution in response to high N deposition. These changes may eventually lead to
C-16
-------
replacement of pine species with nitrophilous and 03-tolerant species such as fir and cedar (Takemoto
et al., 2001).
Streamwater N03 concentrations in montane watersheds that are downwind of the greater Los
Angeles area are the highest in North America. Some streams in the San Gabriel and the San Bernardino
Mountains have been documented to have levels of N03 in stream water with peaks as high as 370 (ieq/L
(Fenn and Poth 1999), reflecting very high N deposition and N-saturation of the terrestrial ecosystem. In
contrast, N leaching is low in most watersheds in the Sierra Nevada, and N03 concentrations in streams
are usually below 1 (ieq/L. Nevertheless, some of the higher elevation watersheds in the Sierra Nevada
export appreciable N03 from the terrestrial environment, particularly during the early phases of
snowmelt. Fenn et al. (2002) reported springtime peaks of N03~concentration in lakewater up to 38 j^ieq/L
at high elevation and for watersheds dominated by talus. At lower elevation areas, however, most of the
inorganic N deposition loading is retained within the watersheds and concentrations of N03 in stream
and lake waters are low (Fenn et al., 2003a). Surface water N03 concentrations in these areas provide an
index reflecting the general levels of N deposition. For example, where surface water N03
concentrations are high, N deposition to the terrestrial watershed is also high.
C.2.4. Boreal Forests
The boreal forest represents the largest terrestrial biome on Earth, and as such can have a large
influence on global cycling of N and other nutrients. Plant growth in the boreal forest is limited mainly by
N availability, in part because of slow mineralization of organic materials in the harsh climate (Vitousek
and Howarth, 1991). Conceptual models of N cycling in the boreal forest have typically assumed that
mineralization of organic N is required for plant uptake of N (Nasholm et al., 1998). However, it has been
demonstrated in laboratory studies (Chapin et al., 1993) and field studies (Nasholm et al., 1998) that some
boreal plants are capable of directly taking up amino acids from the soil, and therefore bypassing the need
for prior mineralization. The interactions among soil abiotic processes, mycorrhizal associations,
microbes, and plants are complex and poorly understood. Nevertheless, these interactions are important to
global N cycling and to boreal plant species composition because organic N concentrations are typically
high in the soil of boreal forests. It appears that atmospheric N deposition and climate warming have the
potential to alter boreal forest plant communities by shifting nutritional processes from organic to
inorganic N uptake (Nasholm et al., 1998).
C.2.5. Alpine
The western U.S. contains extensive land areas that receive low levels of atmospheric N deposition,
interspersed with hot spots of relatively higher N deposition downwind of large metropolitan centers and
agricultural areas (Fenn, 2003). Alpine plant communities occur in some of the areas that receive
moderately elevated atmospheric N deposition such as those located in the Sierra Nevada in southern
California, the Front Range in Colorado, and the Cascade Mountains in Washington (Figure C-2).
C-17
-------
Existing Vegetation Typ«
billing V«g«blton Typ«
Source: Vegetative distribution data were taken from the national map LAND FIRE (September 2006) (http://qisdata.usqs. net/web site/I a ndfire/V
Figure C-2. Distribution of alpine vegetation in three western regions that are in close proximity to urban
and agricultural sources of atmospheric N emissions : a) the Denver-Fort Collins region of
Colorado, b) the Seattle-Tacoma region of Washington, and c) the Fresno-Los Angeles area of
California (the blue line on the map is the California/Nevada border). Alpine vegetation in
these areas is sensitive to nutrient enrichment effects from atomospheric N deposition.
C-18
-------
Alpine plant species are typically adapted to low nutrient availability and their soil-forming
processes are poorly developed, therefore they are often sensitive to effects from N enrichment (Bowman
et al., 2006) including changes in species composition (Bowman et al., 1995; Seastedt and Vaccaro,
2001). Other reasons alpine tundra are sensitive to N enrichment include factors such as low rates of
primary production, short growing season, low temperature, and a wide variation in moisture availability
(Bowman and Fisk, 2001).
Nitrogen cycling in alpine environments is strongly tied to variations in moisture regime (Bowman
et al., 1993; Bowman, 1994; Fisk et al., 1998). Blowing snow is transported across alpine landscapes by
wind and tends to accumulate in certain depression areas. These areas receive much higher levels of
moisture and winter season N deposition than other more wind-swept portions of the alpine environment
(Bowman, 1992). Fenn et al. (2003) suggested that as much as 10 kg N/ha/yr may leach through the snow
during the initial phases of snowmelt in some of the alpine areas in Colorado that accumulate substantial
snowpack. It is these moist meadow areas that may be most affected by N deposition and are also the
areas most likely to show changes in plant species composition and impacts on N cycling (Bowman and
Steltzer, 1998).
Nitrogen deposition to the alpine tundra of Niwot Ridge in the Colorado Front Range altered N
cycling and provided the potential for replacement of some native plant species by more competitive,
faster-growing native species (Baron, 2000; Bowman and Steltzer, 1998; Bowman, 2000). Many plants
that grow in alpine tundra, as is true of plants growing in other low resource environments (e.g., infertile
soil, desert), tend to have some similar characteristics, including slow growth rate, low photosynthetic
rate, low capacity for nutrient uptake, and low soil microbial activity (Bowman and Steltzer, 1998;
Bowman, 2000). Such plants generally continue to grow slowly when provided with an optimal supply
and balance of resources (Pearcy et al., 1987; Chapin, 1991). In addition, plants adapted to cold, moist
environments grow more leaves than roots as the relative availability of N increases. These patterns of
vegetative development and their response to added N affect plant capacity to respond to variation in
available resources and to environmental stresses such as frost, high winds, and drought. Vegetation in the
southern Rocky Mountains responds to increased N supply by increasing plant productivity for some
species, but this increase in productivity is also accompanied by changes in species composition and
abundance (Bowman et al., 1993). Many of the dominant plant species do not respond to additional N
supply with increased production. Rather, many subdominant species, primarily grasses and some forbs,
increase in abundance when the N supply is increased (Fenn et al., 2003c).
In alpine ecosystems, changes in plant species composition due to N deposition can result in
increased leaching of N03 from the soils because the plant species favored by higher N supply are often
associated with greater rates of N mineralization and nitrification than the pre-existing species (Bowman
et al., 1993, 2006; Steltzer and Bowman, 1998; Suding et al., 2006).
Total organic N pools in the soils of dry alpine meadows are large compared to pools of NH/ and
N03 (Fisk and Schmidt, 1996). However, positive response to inorganic N fertilization has been
demonstrated, and thus some plant species appear to be restricted in their ability to take up organic N
from the soil and are growth-limited by the availability of inorganic N (Bowman et al., 1993, 1995;
Theodose and Bowman, 1997). Miller and Bowman (2002) analyzed patterns of foliar 15N, N03
reductase activity, and mycorrhizal infection compared with N uptake quantified by stable isotope tracer
additions in the greenhouse. 13C enrichment subsequent to 13C, 15N-glycine addition indicated that all of
the 11 genera studied were able to take-up labeled glycine to some extent. Glycine uptake ranged from
about 35% to more than 100 % of NH/ uptake. Only Festuca (fescue grass) showed glycine uptake
exceeding both NH/ and N03 uptake (Miller and Bowman, 2002).
C-19
-------
C.2.6. Arctic Tundra
Soluble N in tundra soil solution is dominated by organic N, including free amino acids, rather than
NH4+ or N03 (Kielland, 1995). Tundra plants appear to exhibit a range of interspecific differences that
allow coexistence under conditions that reflect a single limiting element. Species differ in rooting depth,
phenology, and uptake preferences for organic and inorganic forms of N (Shaver and Billings, 1975;
Chapin et al., 1993; Kielland, 1994; McKane et al., 2002). McKane et al. (2002) demonstrated, based on
15N field experiments, that arctic tundra plant species were differentiated in timing, depth, and chemical
form of N utilization. Furthermore, the species that exhibited greatest productivity were those that
efficiently used the most abundant N forms.
Ericoid mycorrhizae provide host plants with the capacity to take up N in the form of amino acids
(Stribley and Read, 1980; Bajwa and Read, 1985). This is important in arctic plant communities that
occur on acidic organic soils because amino acids are typically readily available in such soils, and N
availability generally limits primary productivity.
Future climate warming could have important effects on N cycling in arctic tundra ecosystems. In
the past, organic materials have accumulated in tundra soils, largely because decomposition has been
slower than plant growth. Climate warming may increase the decomposition of soil organic matter,
thereby increasing the availability of stored N (Weintraub and Schimel, 2005). The distributions of woody
plant species are also increasing in response to warming, with likely feedbacks on C and N cycling. For
example, the dominant shrub species in the arctic tundra in Alaska, Betula nana, is expanding its
distribution in tussock vegetation communities (Weintraub and Schimel, 2005).
Poor soil aeration is caused by permafrost, resulting in poor water drainage and the development of
anaerobic conditions. Vegetation composition and primary productivity vary in response to differences in
soil moisture and aeration (Everett and Brown, 1982; Gebauer et al., 1995). Reduced soil 02 can limit
nutrient availability. For example, under anaerobic conditions, N mineralization and nitrification rates
decrease while denitrification increases (Ponnamperuma, 1972; Gebauer et al., 1995).
C.2.7. Arid Land
From 1989 to 2004 in the Chihuahuan desert, Baez et al. (2007) observed a 43% increase in
ambient N deposition, from 1.71 to 2.45 kg N/ha/yr, resulting in an additional 5.88 kg N/ha/yr deposition
over that time period. They suggest that these deposition trends may result in significant plant community
changes, as indicated by fertilization studies of blue gramma (Bouteloua gracilis) and black gramma (B.
eriopoda). In a field addition with additions of 20 kg N/ha/yr in one season, blue gramma was favored
over black gramma, the current dominant species (Baez et al., 2007).
C.2.8. Lichens
There are several potential uses of lichens for air pollution and deposition monitoring. These
include measurement of tissue lichen concentrations of specific pollutants (i.e., lichens as passive
monitors), determination of changes in species composition or the presence/absence of sensitive species,
and identification of areas having relatively high levels of air pollution, where monitoring instrumentation
could be installed to more quantitatively measure pollution levels. Assessment of long-term change in the
epiphytic lichen community can be especially valuable to provide an early indication of either improving
or deteriorating air quality and atmospheric deposition. Such monitoring was incorporated in 1994 into
the USFS Forest Inventory and Analysis (FIA) Program (See Annex A).
C-20
-------
Lichen communities in the Pacific Northwest show signs of air pollution damage under current air
pollution levels. Symptons include decreases in the occurrences of sensitive taxa and replacement by
pollution-tolerant and nitrophilous taxa (Fenn et al., 2003a; Geiser and Neitlich, 2007). Indicators of clean
sites and polluted sites (Table C-3) were used by Geiser and Neitlich (2007) to create six lichen zones of
air quality within the region, from worst (all sensitive species absent) to best (all sensitive species
present). Air pollution was associated with effects on community composition of lichens, rather than
species richness. The most widely observed effects included paucity of sensitive, endemic species, and
enhancement of nitrophilous and non-native species (Geiser and Neitlich, 2007). The strongest
relationship was with wet NH44" deposition, consistent with findings in California (Jovan and McCune,
2005) and Europe (van Dobben et al., 2001). The zone of worst air quality was associated with absence of
sensitive lichens, enhancement of nitrophyllous lichens, mean wet NH/ deposition >0.06 mg N/L, lichen
tissue N and S concentrations >0.6% and 0.07 %, and S02 levels harmful to sensitive lichens.
Jovan and McCune (2005) constructed a model based on non-metric multidimensional scaling
ordination to analyze lichen species distribution from 98 FIA plots in the greater Central Valley of
California. The model used epiphytic macrolichen community data to reflect air quality and climate in
forested areas. Some species respond negatively to NOx and SOx deposition (Gauslaa, 1995; McCune,
1988; van Haluwyn and van Herk, 2002). Other species respond positively to NHY deposition (de Bakker,
1989; van Dobben and de Bakker, 1996; van Herk, 1999, 2001; Jovan and McCune, 2005).
Similarly, Jovan and McCune (2006) developed a model of NH3 exposure to epiphytic
macrolichens in the Sierra Nevada region. They found that lichens provide a relatively inexpensive tool
for estimating fine-scale distributions of NH3 exposure to terrestrial ecosystems. Because NH3 has a high
deposition velocity (Asman and van Jaarsveld, 1992), dry deposition of reduced N exhibits high spatial
variability. Monitoring of species composition of epiphytic lichen communities can therefore help
quantify spatially variable eutrophication risk to forest health in the Sierra Nevada region (Jovan and
McCune, 2006).
Table C-3. Principal Air Quality Indicator Lichen Species in Oregon and Washington
Group
Sub-Group
Indicator Species
Clean Air
Regional distribution
Bryoria capillaris, Lobaria oregana,
Sphaerophorus globosus, Usnea filipendula,
Usnea scabrata
Sub-regional distribution
Ahtiana pallidula, Alectoria sarmentosa, Bryoria
fuscescens, Hypogymnia enteromorpha,
Nephroma bellum, Nodobryoria oregana
Polluted Air
Regional nitrophytes
Candelaria concolor, Physcia adscendens,
Xanthoria polycarpa
* Includes only species with highest indicator value, used by Geiser and Neitlich (2007) to define air quality
zones.
Source: Geiser and Neitlich (2007).
C-21
-------
C.3. Transitional Ecosystems
The sensitivity of wetlands is particularly important given that they contain a disproportionately
high number of rare plant species (Figure C-3) (Moore et al., 1989). EPA reported that, of the 130 plant
species from the conterminous U.S. that were listed as threatened or endangered in 1987, 14% occurred
principally in wetlands (U.S. EPA, 1993b). Bedford and Godwin (2003) indicated that a
disproportionately high number of rare plant species occur in fens relative to their percent land cover
(Table C-4) (Bedford and Godwin, 2003). For example, fens comprise only 0.01% of northeastern Iowa
but contain 12% of the region's rare plant species and 17% of the listed endangered, threatened, and
species of concern (Table C-4).
S
s
g
(A a „
© H
a
©
Cl
i
CM
100
200 300 400 500
Standing crop (gl.25 m1)
600
TOO
Source: Moore et al. (1989).
Figure C-3. Number of nationally rare species versus standing crop in each of 401 quadrants from
wetlands in Ontario, Quebec, and Nova Scotia.
Table C-4. Contribution of fens to support of plant species diversity in selected states.
Number of
Percent of
Non-
Number of
Uncommon &
Rare Species
Found in Fens
Percent of State
Uncommon &
Rare Species
Found in Fens
Vascular
State Vascular
Vascular
Percent of
Species
Found in Fens
(# Native)
Flora Found in
Fens(%
Native)
Species
Found in
Fens
State Area
in Fens
Colorado
-500
-14
20
3.3
0.08-0.15
Idaho
327
20
35
12.0
Iowa
320
18
134
12.0
0.01
(307)
(17.2)
Montana
174
60
40
0.0015
C-22
-------
New
Hampshire
340
17.2
91
52
13.7
0.078
Calcareous
fens
0.0058
Open fens
0.0726
New Jersey
96
13.5
0.00733
Calcareous
fens
245
60
0.0073
Acidic
seeps
169
13
36
0.00003
New York
440
13.8
77
55
7.0
0.07
(397)
(19.0)
North
Carolina
77
11.0
0.0023
Source: Bedford and Godwin (2003).
C.4. Aquatic Ecosystems
Aquatic systems can be subdivided into major types based on hydrology. At the broadest level,
freshwater aquatic ecosystems can be classified as riverine, lacustrine, and palustrine systems. Riverine
systems can be identified at varying scales, including valley segment, river reach, and channel unit.
Lacustrine systems include deepwater habitats associated with lakes and reservoirs. Palustrine systems
include small, shallow, or intermittent water bodies, including ponds. Each type of aquatic ecosystem is
potentially sensitive to nutrient enrichment effects from N deposition. Nevertheless, available data
documenting such effects are limited.
The dose-response data for aquatic organisms such as those cited here are generally expressed in
concentration units, as mg/L or |_imol/L of N, for example. Such exposure concentration data cannot be
directly related to ecosystem exposure, which is generally expressed in such units as kg/ha. This is
because a given N deposition exposure can result in widely varying concentrations of N compounds
(especially N03 ) in water. For convenience, a concentration of 1 mg/L of N (as, for example, in the case
of N03~-N or NH/-N) is equal to 71.4 (imol/L or 71.4 (ieq/L ofN03~ or NH/.
C.4.1. History of Evaluating Nitrogen Enrichment in Freshwater
Aquatic Ecosystems
The role of N deposition in freshwater eutrophication and acidification processes has been
considered secondary to P and S, and only within the past 20 years have there been studies questioning
the established science and showing N-limitation in some fresh waters, N excess in some terrestrial
systems, and N-caused acidification in poorly buffered fresh waters. A number of things have conspired to
prevent extensive evaluation of the effects of atmospheric N deposition on aquatic organisms via nutrient-
C-23
-------
enrichment pathways. These include assumptions, or prevailing paradigms, that have channeled scientific
thought in one direction and away from others.
First were the assumptions for many years that atmospheric deposition was caused primarily by
sulfur (S) emissions and that effects on aquatic ecosystems were primarily caused by acidification
processes. Only after S emissions began to decrease substantially in response to the CAAA did the role of
NOx, and still later, NH3, emissions become recognized as potential agents of environmental change. And
even then, that role was assumed to be restricted mainly to acidification from N03 . a strong acid anion,
not eutrophication (Reuss and Johnson, 1985). Second, because N is the nutrient most limiting to primary
production in most ecosystems, it was assumed until fairly recently that N was tightly cycled in terrestrial
systems, and that excess N03 leaching rarely occurred in natural environments (Vitousek and Howarth,
1991). Finally, the attention of aquatic biologists has been strongly focused on the role of P in
eutrophication of freshwaters for the past 40 years, largely due to the demonstrated role of P in causing
large increases in algal productivity worldwide (Schindler et al., 1971; Schindler, 1974). P is an essential,
and often limiting, nutrient to aquatic organisms. A large number of highly influential studies in the 1960s
and 1970s exposed the role of wastewater, in particular phosphate detergents, in causing excessive algal
production and anoxia in Lake Mendota (Wisconsin), Lake Washington (Washington), Lake Erie, and
many other locations (Hasler, 1947; Vollenweider, 1968; Edmondson, 1969, 1991). Because of the
emphasis on P as a major cause of fresh water eutrophication, Downing and McCauley (1992) wrote as
recently as 1992: "opinions differ on the role of N as a limiting nutrient in lakes."
C.4.2. Interactions between Nitrogen and P loading
Results from surveys, paleolimnological reconstructions of past conditions, experimental results,
and meta-analyses of hundreds of studies all consistently show N-limitation to be common in fresh
waters, especially in remote areas, and there is a nearly universal eutrophication response to N-enrichment
in lakes and streams that are N-limited. Surveys of lake N concentrations and trophic status along
gradients of N deposition show increased inorganic N and increased productivity to be strongly related to
atmospheric N deposition. Where N-enrichment has occurred, P limitation, N+P colimitation, and a few
instances of Si depletion have been reported. Paleolimnological records show increases in productivity
and changes in algal assemblages in the recent past (since 1950) that are correlated with increased societal
use of synthetic N fertilizers and human. The paleolimnological evidence is strongest in regions with the
highest N deposition, and is weaker where N deposition is lower (Wolfe et al., 2001; 2003; 2006; Saros
et al., 2003). In additions to changes in productivity, algal community reorganization has been observed
in the paleolimnological record, experiments, and observations of N-enriched lakes, especially those
where enrichment has come from N deposition. A summary of additional studies addressing N-limitation
is given in Table C-5.
It is generally believed that the Laurentian Great Lakes are P-Limited (Schelske, 1991; Downing
and McCauley, 1992; Rose and Axler, 1998). Water quality in the open waters of these lakes has been
improving in recent years in response to controls on point sources of P (Nicholls et al., 2001). Work by
Levine et al. (1997), however, suggested a more complicated pattern of response to nutrient addition for
Lake Champlain. They added nutrients to in situ enclosures and measured indicators of P status, including
alkaline phosphatase activity and orthophosphate turnover time. Although P appeared to be the principal
limiting nutrient during summer, N addition also resulted in algal growth stimulation. During spring,
phytoplankton growth was not limited by P, N, or silica (Si), but perhaps by light or temperature (Levine
et al., 1997).
Data from 28 Sierra Nevada lakes sampled in 1985 and again in 1999 suggested that N03
concentrations decreased during that period and total P concentrations increased in more than 70% of the
lakes sampled. Sickman et al. (2003a) concluded that lakes throughout the Sierra Nevada appear to be
experiencing measurable eutrophication in response to atmospheric deposition of nutrients, but N
C-24
-------
deposition is only part of the process. Based on the evidence of increased P loading throughout the Sierra
Nevada, Sickman et al. (2003a) concluded that site-specific P sources were unlikely to be the cause of
observed trends. They proposed that atmospheric deposition and accelerated internal cycling of P in
response to changes in climatic factors were the most likely sources of increased P loading to the Sierra
Nevada Lakes, but it is not known why atmospheric deposition of P to these lakes has increased over
time. Possibilities include use of organo-phosphate pesticides and aeolian transport of soils and dust that
are high in P from the San Joaquin Valley to the Sierra Nevada Mountains (Bergametti et al., 1992;
Lesack and Melack, 1996; Sickman et al., 2003a).
Data from a survey of 44 lakes east and west of the Continental Divide in Colorado indicated that
lakes on the western side of the Continental Divide averaged 6.6 (ieq/L ofN03 . whereas lakes on the
eastern side of the Continental Divide averaged 10.5 j^ieq/L of N03 concentration. In the Colorado Front
Range, N03 concentrations in lakes above 15 |_icq/L have commonly been measured, suggesting some
degree of N-saturation (Baron, 1992). A meta-analysis of 42 regions in Europe and North America
suggested that a majority of lakes in the northern hemisphere were limited by N in their natural state
(Bergstrom and Jansson, 2006). While many of these lakes now receive sufficient N from deposition that
they are no longer N-limited, some lakes in remote regions still maintain their original oligotrophic or
ultra-oligotrophic status.
Table C-5. Summary of additional evidence for N limitation on productivity of freshwater ecosystems.
Region
Endpoint
Observation
Ecosystem
Type
Reference
Sweden
N
limitation
a consistent pattern of nutrient limitation showing N limitation for
deposition below approximately 2.5 kg N/ha/yr, co-limitation of N and P
for deposition between -2.5 and 5.0 kg N/ha/yr, and P limitation in areas
with N deposition greater than 5.0 kg N/ha/yr.
lakes
Bergstrom
etal. (2005)
Rocky
Mountains of
Colorado and
Wyoming
N
limitation
Review: the author concluded that the effects of atmospheric N deposition
were uncertain and that a widespread shift from N to P limitation had not
been clearly demonstrated.
Burns (2004)
Texas
N
limitation
some instances of seasonsal N-limitation, and other instances of year-
round N-limitation
rivers
Stanley et al.,
(1990)
Rocky
Mountains of
Colorado and
Wyoming
N
limitation
Review: Author stated that recent studies did suggest a change in diatom
species dominance in the 1950s, but widespread species changes across
lakes in the region and the role of N deposition in these changes needed
confirmation. Thus, the available data were not clear at that time, but
suggested that some changes had likely occurred in some aquatic
ecosystems.
lake
Burns (2004)
C.4.3. Aquatic Species Affected
The following Section contains studies in which the amount of N added was less than 10 mg N03
N/L, or 714 (.iM, and most studies tested the effects of 5 mg N03 -N/L or less. Many of these studies are
summarized in Table C.6. Overall, several major effects were reported on biota treated with N
enrichment: the effects on algae included growth stimulation, increased cell densities, decline or
C-25
-------
stimulation of individual taxa, and decline in diversity; the amount of N required to stimulate growth in
phytoplankton is extremely low: 3 (.iM or less; and animal responses included no response, decreased
reproductive capability, declines in growth rate and biomass, mortality, and in one case, increased fitness
because N03 was detrimental to a fungal parasite.
C.4.3.1. Phytoplankton and Plants
Two species of diatom, Asterionella formosa and Fragilaria crotonensis, now dominate the flora of
at least several alpine and montane Rocky Mountain lakes (Baron et al., 2000; Interlandi and Kilham,
1998; Saros et al., 2003, 2005; Wolfe et al., 2001, 2003). These species are opportunistic algae that have
been observed to respond rapidly to disturbance and slight nutrient enrichment in many parts of the world.
They were among the first diatoms to increase in abundance following watershed settlement and
agricultural development in European lake watersheds in the 12th and 13th centuries (Anderson et al.,
1995; Lotter, 1998), and North American settlements in the 18th and 19th centuries (Christie and Smol,
1993; Hall et al., 1999). In these studies, as well as in a Swedish lake influenced by acidic deposition,
these two diatom species expanded following initial disturbance, and were later replaced by other species
more tolerant of either acidification or eutrophication (Renberg et al., 1993; Hall et al., 1999). Moreover,
the growth of A. formosa has been stimulated with N amendments during in situ incubations, using
bioassays and mesocosms (6.4 to 1616 (iM N/L; McKnight et al., 1990) (76 (iM N/L; Lafrancois et al.,
2004) (18 (iM N/L; Saros et al., 2005).
It may seem obvious that additions of N stimulate cell growth, but not all species of diatoms or
other algae are equally responsive to N supply. A. formosa and F. crotonensis have extremely low
resource requirements for P, enabling them to outcompete other algae for resources and such differences
in resource requirements allow some species to gain a competitive edge over others upon nutrient
addition, and as a consequence, shifts in assemblages have been observed (Lafrancois et al., 2004; Saros
et al., 2005; Wolfe et al., 2001; 2003). This is in keeping with findings of Interlandi and Kilham (2001),
who demonstrated that maximum species diversity was maintained when N levels were extremely low
(<3 (.iM N) in lakes in the Yellowstone National Park (Wyoming, Montana) region. The implication is
that species diversity declines with increasing availability of N, and this finding complements the results
of terrestrial studies that also showed a negative relationship between species diversity and N availability
(Gilliam 2006; Suding et al., 2005; Stevens 2004).
P limitation and co-limitation of both N and P are reported for fresh waters in the literature,
particularly during summer (Downing and McCauley, 1992; Elser 1990; Morris and Lewis 1988;
Sickman et al., 2003b). Because diatoms in northern temperate freshwaters respond rapidly and favorably
to N enrichment and also have relatively high Si requirements, Si can be depleted, at least seasonally,
from waters that are relatively high in N and P.
Silica depletion due to nutrient enrichment has been reported for the Great Lakes (Conley et al.,
1993). Increased growth of silicate-utilizing diatoms as a result of N03 and phosphate (P043 )-induccd
eutrophication, and subsequent removal of fixed biogenic Si via sedimentation has brought about changes
in the ratios of nutrient elements Si, N, and P. In turn, such changes can cause shifts from diatoms to non-
siliceous phytoplankton in large rivers and coastal marine regions (Ittekot, 2003). Reduction in dissolved
Si in lakewater corresponded to phytoplankton blooms under ice and large numbers of diatoms during
spring in Loch Vale Watershed, Rocky Mountain National Park (Campbell et al., 1995). This is a
potential seasonal issue in water bodies underlain by aluminosilicate rocks because mineral weathering
can replenish the Si supply.
C-26
-------
C.4.3.2. Seasonal Nitrogen Input and Cyanobacteria
Some ecosystems are seasonally enhanced with N from atmospheric deposition, either from
snowmelt flushing of accumulated N in winter snow, or from flushing during dormancy of terrestrial
vegetation (Stoddard, 1994). While many eutrophic and hypereutrophic freshwater ecosystems have
seasonal or perennial cynaobacteria that fix atmospheric N, obviating the need for an external source of N
(Wetzel, 2001), only one oligotrophic lake with obligate N-fixing bacteria has been reported (Reuter
et al., 1985). N-fixation is energy expensive and sometimes limited by trace metal availability, so obligate
N-fixing cyanobacteria (formerly called blue-green algae) are rarely found in ultra-oligotrophic waters
(McKnight et al., 1990; Vitousek and Howarth, 1991). Because of this, oligotrophic and ultraoligotrophic
waters are extremely sensitive to even low inputs of N from atmospheric deposition. Anabaena circinalis,
an obligate N-fixing cyanobacterium, was suppressed with additions of 500 |_iM/L N (Higley et al., 2001),
and DIN levels >~200 (iM/L N completely inhibited N fixation in Castle Lake, CA (Reuter et al., 1985).
C.4.3.3. Nitrate Toxicity: Invertebrates
Toxic responses to N exposure by aquatic invertebrates have been identified in a number of studies.
Toxic response thresholds are typically much higher than the levels of N in surface waters that could be
attributable to N deposition in the U.S. Safe Concentrations (SC), or threshold values of N, were
determined by Camargo and Ward (1995) for several aquatic insects at different life stages. Early instars
are generally more sensitive to N in solution than later or adult stages. The SC for late instars of
Hydropsyche occidentalis, a caddis fly, was found to be 171 and concentrations greater than this
value induced mortality. The SC was 100 |_iM/L for early instars of the same species (Camargo and Ward,
1995). Another caddis fly, Cheumatopsyche pettiti, tolerated higher concentrations, with safe
concentrations of 171 and 250 (.iM N/L, respectively for early and late instars (Camargo and Ward, 1995).
Two species of amphipod did not survive after 120-h exposure to N03 concentrations of 200 (.iM N/L for
one species, and 314 |_iM N/L for the other (Camargo et al., 2005). No observable effect concentrations
above which Ceriodaphinia dubia exhibited reduced reproductive capability ranged broadly in laboratory
experiments, but some effects were seen at concentrations greater than 507 (.iM N/L (Scott and
Crunkilton, 2000). A decline in Daphnia spp. was observed in mesocosm nutrient enrichment
experiments where 75 (.iM N/L was added, but this was attributed to lower food quality of the algal
assemblage that replaced the original species as a result of fertilization (Lafrancois et al., 2004). Thus,
toxic responses seem to occur at N concentrations that are much higher than the concentrations required
to elicit a response in competitive interactions.
A whole-ecosystem experiment at the Bear Brook watershed, ME simulated the effects of N and S
deposition by means of experimental (NLL^SC^ addition over a period of 10 years. Researchers found
that elevated N inputs had minimal effect on stream detritus processing (Chadwick and Huryn, 2003).
They also found that N additions had no significant effect on stream macroinvertebrate secondary
production or varying production by functional feeding groups. They concluded that climate-related
variables such as flow duration and litter inputs controlled secondary production when N was not limiting
(Chadwick and Huryn, 2005).
Changes to aquatic food webs have not been as thoroughly explored as changes to algal
assemblages, but a few studies have shown declines in zooplankton biomass (Paul et al., 1995; Lafrancois
et al., 2004) in response to N-related shifts in phytoplankton biomass toward less palatable taxa with
higher C:P ratios (Elser et al., 2001). N enrichment of arctic streams not only increased periphyton
biomass and productivity, but also stimulated the entire ecosystem, increasing decomposition rates, fungal
biomass, and invertebrates (Benstead et al., 2005).
C-27
-------
C.4.4. Nitrate Toxicity: Amphibians and Fish
A summary of studies on the effects of nitrate on amphibians and fish is given by Table C-6. It
appears that very high N03 concentrations in surface water are required to elicit a toxic response in
amphibian populations. Concentrations that caused no observed effects and no observed adverse effects
ranged from 357 to 714 (.iM N/L for frogs, salamanders, and the American toad (Bufo americanus)
(Hecnar, 1995; Laposata and Dunson, 1998; Johansson et al., 2001; Romansic et al., 2006). In one
experiment, the red-legged frog {liana aurora) exhibited a decreased susceptibility to Saprolegnia mold
when exposed to elevated N03 concentrations (Romansic et al., 2006).
According to one review, adverse direct effects of N deposition on fish due to nutrient enrichment
are probably minimal (Burns, 2004). N concentrations alone are not high enough to influence fish
metabolism, and the extent of eutrophication is insufficient (due to induced P limitation in oligotrophic
waters) to cause 02 depletion.
Other research suggests that the eggs and fry of rainbow trout (Oncorhyncus mykiss\ including
steelhead), cutthroat trout (O. clarki), and chinook salmon ((). tshawytscha) are susceptible to elevated
concentrations ofN03 , with rainbow trout mortality occurring after 30 day incubations in concentrations
>79 (.iM N/L (Kincheloe et al., 1979). There were no observed effects reported below this concentration.
Chinook salmon and cutthroat trout eggs and fry responded to slightly higher concentrations; no observed
effects occurred below 164 (.iM N/L, but mortality occurred at higher concentrations (Kincheloe et al.,
1979). Lake whitefish (Coregonus clupeaformis) and lake trout (Salvelinus namaycush) embryos
displayed developmental delays at concentrations greater than 446 and 114 (.iM N/L, respectively
(McGurk et al., 2006). All of these toxic threshold concentrations are much higher than the concentrations
of N03 in surface water that would routinely be expected to occur solely in response to atmospheric N
deposition in the U.S. Nevertheless, such high concentrations of streamwater N03 have been measured in
the Great Smoky Mountains, NC (Cook et al., 1994) and in mixed conifer forests in southern California
(Fenn and Poth, 1999).
Table C-6. Summary of effects of N enrichment on aquatic biota in freshwater ecosystems.
Species
Common
Name
Life
Stage
N Concentration
(mg NOs-N/L)
Observed Effects
Reference
ALGAE: PHYTOPLANKTON
Asterionella formosa
diatoms
0.252 mg/L
stimulated growth
Saros et al.
(2005)
Asterionella formosa
diatoms
6.4 [jmol/L
stimulated growth
McKnight et al.
(1990)
Multiple species
diatoms
1.06 mg/L
(low ambient N-deposition):increase in
chlorophyll-a content and growth rate;
no cell density effect
Lafrancois
etal. (2004)
Asterionella formosa
diatoms
5.7 x 10-4 (>0.041
uM) (high light)
increased growth rate (measured at
half the maximum growth rate)
Michel et al.
(2006)
Fragilaria crotonensis
diatoms
0.252 mg/L
stimulated growth
Saros et al.
(2005)
Multiple species
diatoms
1.06 mg/L
(low ambient N deposition):increase in
chlorophyll-a content and growth rate;
no cell density effect
Lafrancois
etal. (2004)
C-28
-------
Fragilaria crotonensis
diatoms
3.5 x 10-4 (>0.028
uM) (high light)
increased growth rate (measured at
half the maximum growth rate)
Michel et al.
(2006)
Fragilaria crotonensis
diatoms
8.4 x 10-6 (>0.006
uM) (med light)
increased growth rate (measured at
half the maximum growth rate)
Michel et al.
(2006)
Staurosirella pinnata
diatoms
8.4 x 10-6 (>0.006
uM) (med light)
increased growth rate (measured at
half the maximum growth rate)
Michel et al.
(2006)
Tetracyclus glans
benthic diatoms
1.7 x 10-4 (>0.012
uM) (low light)
increased growth rate (measured at
half the maximum growth rate)
Michel et al.
(2006)
Multiple species
phytoplankton
assemblages
3.0 uM
N saturation value for maximum
diversity in WY low N lakes
Interlandi et al.
(1999)
not identified
phytoplankton
assemblages
0.5 mg/L
NO3"- stimulated growth seasonally,
while tributary periphyton communities
were P limited
Stanley et al.
(1990)
not identified
phytoplankton
assemblages
0.3 uM
NH4 additions more effective at
stimulating growth than NO3"
Levine and
Whalen (2001)
not identified
phytoplankton
assemblages
100 |jg/L
stimulated NO3" uptake
Axler and
Reuter (1996)
Multiple species
crysophytes
1.21 mg/L
(elevated ambient N deposition): no
response to NO3" additions; increased
chlorophyll-a and cell density when
NO3" combined with acid and P
Lafrancois
etal. (2004)
Multiple species
epilimnetic algae
0.012 mg/L
increased growth rate (measured at
half the maximum growth rate)
Priscu et al.
(1985)
Multiple species
hypolimnetic
algae
0.050 mg/L
increased growth rate (measured at
half the maximum growth rate)
Priscu et al.
(1985)
PERIPHYTON
attached benthic
algae
0.5 M NaNOs in 2%
agar
biomass increased in response to N
and N & P additions during period of
seasonal N-limitation (July-August)
Smith and Lee
(2006)
attached benthic
algae
2.5 M
NO3" stimulated stream algal growth
during seasonal N-limitation
Bushong and
Bachmann
(1989)
not identified
attached benthic
algae
0.5 M NaNOs in 3%
agar
NO3" alone stimulated stream algal
growth during seasonal N-limitation,
while N & P co-limited growth in other
times
Wold and
Hershey
(1999)
not identified
attached benthic
algae
0.5 M NaNOs in 3%
agar
NO3" alone stimulated stream algal
growth during seasonal N-limitation,
while N & P co-limited growth in other
times
Allen and
Hershey
(1996)
not identified
attached benthic
algae
0.036 mg/L
stimulated NO3" uptake
Axler and
Reuter (1996)
not identified
attached benthic
algae
0.5 M NaNOs in 2%
agar
no growth response
Higley et al.
(2001)
C-29
-------
not identified
epilithic
-700 |jg NOs-N
suppressed l\l2-fixation
Reuter et al.
(1985)
not identified
attached benthic
algae
0.16 mg/L
increased summer growth rate
(measured at half the maximum growth
rate)
Reuter et al.
(1985)
not identified
attached benthic
algae
0.32 mg/L
increased winter growth rate
(measured at half the maximum growth
rate)
Reuter et al.
(1986)
not identified
sublittoral
epilithic algae
0.259 mg/L
increased growth rate (measured at
half the maximum growth rate)
Reuter and
Axler (1992)
not identified
eulittoral epilithic
algae
0.126 mg/L
increased growth rate (measured at
half the maximum growth rate)
Reuter and
Axler (1992)
not identified
epipelic algae
0.713 mg/L
increased growth rate (measured at
half the maximum growth rate)
Reuter and
Axler (1992)
CYANOBACTERIA
Anabaena circinalis
N fixing
cyanobacteria
0.5 M NaNOs in 2%
agar
decreased abundance
Higley et al.
(2001)
Microcycstis sp.
non-N-fixing
cyanobacteria
0.28 mg/L
increased growth rate; increased
mycrocystin and anatoxin-a
concentrations
Gobler et al.
(2007)
INVERTEBRATES
Hydropsyche
occidentalis
caddis fly
early
instaar
1.4 (SC)
mortality
Camargo and
Ward (1995)
late
instaar
2.2 (SC)
mortality
Camargo and
Ward (1995)
Cheumatopsyche
pettiti
caddis fly
early
instaar
2.4 (SC)
mortality
Camargo and
Ward (1995)
late
instaar
3.5 (SC)
mortality
Camargo and
Ward (1995)
Echinogammarus
echinosetosus
amphipod
adult
2.8 (120 h LC0.01)
mortality
Camargo et al.
(2005)
Eulimnogammarus
toletanus
amphipod
adult
4.4 (120 h LC0.01)
mortality
Camargo et al.
(2005)
Ceriodaphnia dubia
water
flea/cladoceran
adult
7.1-56.5 (7d
NOEC)
decreased reproductive ability; fewer
neonates produced per female
Scott and
Crunkilton
(2000)
Daphnia pulex
Water flea
adult
1.06
(low ambient N-deposition): decreased
biomass in response to NO3"
Lafrancois
etal. (2004)
Daphnia schoedleri
Water flea
adult
1.06
(low ambient N-deposition): decreased
biomass in response to NO3"
Lafrancois
etal. (2004)
VERTEBRATES: AMPHIBIANS
Rana temporaria
common frog
larvae
5 (70-d NOEC)
delayed development, lower growth
rate and body mass at metamorphosis
Johansson
etal. (2001)
C-30
-------
Rana sylvatica
wood frog
fertilized
eggs
9 (NOAEL)
no effect of NO3" on survivorship
Laposata and
Dunson (1998)
Ambystoma
jeffersonianum
Jefferson's
salamander
fertilized
eggs
9 (NOAEL)
no effect of NO3" on survivorship
Laposata and
Dunson (1998)
Ambystoma
maculatum
spotted
salamander
fertilized
eggs
9 (NOAEL)
no effect of NO3" on survivorship
Laposata and
Dunson (1998)
Ambystoma gracile
northwestern
salamander
larvae
5-20
no effect of NO3" on survivorship
Romansic
etal. (2006)
Rana aurora
red-legged frog
larvae
5-20
no effect of NO3" on survivorship; NO3"
decreased susceptibility to Saprolegnia
mold
Romansic
etal. (2006)
Hyla regilla
Pacific tree frog
larvae
5-20
no effect of NO3" on survivorship
Romansic
etal. (2006)
Pseudacris triseriata
striped chorus
frog
tadpole
10 (100-d LOEC)
mortality
Hecnar (1995)
Rana pipiens
northern leopard
frog
tadpole
10 (100-d LOEC)
mortality
Hecnar (1995)
Bufo americanus
American toad
fertilized
eggs
9.0 (NOAEL)
no effect of NO3" on survivorship
Laposata and
Dunson (1998)
FISH
Oncorynchus mykiss
(anadromous)
steelhead
eggs
1.1 (30-d NOEC)
mortality occurred above this value
Kincheloe
etal. (1979)
Oncorynchus mykiss
(nonanadromous)
rainbow trout
eggs
1.1 (30-d NOEC)
mortality occurred above this value
Kincheloe
etal. (1979)
Oncorynchus mykiss
(nonanadromous)
rainbow trout
fry
1.1 (30-d NOEC)
mortality occurred above this value
Kincheloe
etal. (1979)
Oncorynchus
tshawytscha
chinook salmon
fry
2.3 (30-d NOEC)
mortality occurred above this value
Kincheloe
etal. (1979)
Salmo clarki
cutthroat trout
(Lahontan)
eggs
2.3 (30-d NOEC)
mortality occurred above this value
Kincheloe
etal. (1979)
Salmo clarki
cutthroat trout
(Lahontan)
fry
4.5 (30-d NOEC)
mortality occurred above this value
Kincheloe
etal. (1979)
Coregonus
clupeafbrmis
lake whitefish
embryo
6.25 (~120-d
NOEC)
hatching and developmental delays
McGurk et al.
(2006)
MISCELLANEOUS
Saprolegnia spp.
pathogenic water
mold
5-20 mg N03"/L
decreased ability to infect and kill the
larvae of the red-legged frog, Rana
aurora
Romansic
etal. (2006)
NOEC = No-observed-effect concentration; NOAEL = No-observed-adverse-effect level; SC = Safe concentration
C-31
-------
C.5. Estuary and Coastal Ecosystems
There are a variety of factors that govern the sensitivity of estuaries and near-coastal marine waters
to eutrophication from atmospheric N deposition. Of critical importance is the total N input from all
sources, including both atmospheric and non-atmospheric sources. Other key elements include the
dilution capacity of the watershed, which reflects the volume of water available to dilute added N, and
flushing rate, which reflects the time required for inflowing water to replace estuary volume, (Bricker
et al., 1999; NRC, 2000). Other potentially important factors can include the following (NRC, 2000):
¦ Physiography (geomorphology, dominant biological communities, biogeographic province);
¦ Type of primary production base (i.e., seagrasses, phytoplankton, coral, attached intertidal algae,
etc.);
¦ Stratification and extent to which phytoplankton occupy the nutrient-rich photic zone; and
¦ Allochthonous inputs of organic matter.
A number of factors control the N loading rates to estuaries and the potential effects of N
deposition on nutrient loading and algal blooms. Estuaries communicate with fresh water on the upstream
side and with the ocean on the downstream side. The flushing of fresh river water through the system and
the movement and mixing of salt water from the ocean are complicated and are always changing in
response to weather and tidal cycles. The surface area, volume, and depth of the estuary are also critical
factors governing the sensitivity of an estuary to N inputs. Decreases in grazer, filter-feeder, and higher
trophic level populations of fish and shellfish exacerbate problems associated with nutrient over-
enrichment (Jackson et al., 2001).
At the upstream end of an estuary, the water is primarily fresh much of the time. Discharge of N
from the land surface, only a part of which is of atmospheric origin (mainly as deposition to the land that
subsequently leached to the river water), dominates new N inputs. Further downstream within the estuary,
where fresh water is more thoroughly mixed with saltwater, much of the terrestrial N load is assimilated
by phytoplankton and benthic flora or removed by microbes in the process of denitrification (Paerl, 2002).
The importance of atmospheric N as a contributor to the total N load beyond this zone probably increases,
but there are no data to evaluate that.
The principal watershed features that control the amount of increased N flux to estuaries in the U.S.
include human population, agricultural production, and the size of the estuary relative to its drainage
basin (Caddy, 1993; Fisher et al., 2006; Peierls et al., 1991). Dense human populations generate large
volumes of nutrient-rich wastewater. Tertiary sewage treatment can reduce effluent N concentrations to
less than 35 (.iM. but these technologies have not been promoted as aggressively in the U.S. as elsewhere
(Conley et al., 2002; U.S. EPA, 2003). Agricultural production is heavily dependent on fertilizer
application to generate high yields from small areas. Fertilizer application has dramatically increased
N03 concentrations in ground water in many agricultural areas (Bohlke and Denver, 1995), which can
leach to surface waters. Large terrestrial drainage basins that drain into small estuaries tend to have high
nutrient flux if the land is heavily populated or used for agriculture.
In addition to estuaries, coastal marine ecosystems are highly susceptible to nutrient enrichment,
especially from N. Land clearing, agricultural land use, sewage treatment discharge, and atmospheric
deposition can all result in high loadings of N to the coastal zone. Excessive N inputs contribute to a
range of impacts, including enhanced algal blooms, decreased distribution of seagrasses, and decreased
dissolved oxygen (DO) concentration (Borum, 1996; Bricker et al., 1999; Nixon, 1995; Valiela et al.,
1992). Because of human population growth and the great popularity of coastal areas, there is substantial
potential for increased N loading to coastal ecosystems from both atmospheric and non-atmospheric
sources.
C-32
-------
C.5.1. Interacting Factors with Productivity
Results of empirical observations and short-term (3 weeks) marine mesocosm experiments suggest
that there can be wide variation in the response of autotroph biomass to nutrient addition (Cloern, 2001;
Olsen et al., 2006). Such variation may be attributable to the time scale of the observations, rate of water
exchange, grazing pressure, and other environmental factors (Olsen et al., 2006).
C.5.2. Hydrology Interactions with Phytoplankton Biomass
River discharge has a huge influence on the hydrology and nutrient cycling of estuaries. For
example, discharge from the large watershed of the Susquehanna River is important to the seasonal and
interannual variability in the hydrology of Chesapeake Bay (Fisher et al., 1988; Malone et al., 1988).
When discharge from the Susquehanna River is low, summer phytoplankton biomass in Chesapeake Bay
tends to be low compared to spring conditions, and the phytoplankton community is dominated by small
and flagellated forms (Marshall and Lacouture, 1986). Under higher river flows, summer phytoplankton
biomass in the bay is higher, and has an increased prevalence of diatoms (Paerl et al., 2006).
Hydrologic variation interacts with nutrient supply to control phytoplankton seasonal patterns in
Chesapeake Bay. High biomass during the spring diatom bloom leads to consequent sedimentation of
organic material out of the photic zone during the transition to summer (Malone et al., 1996; Harding
et al., 2002). Microbial decomposition of this material then fuels the pattern of summer anoxia in bottom
waters (Paerl et al., 2006). N loading to Chesapeake Bay and its tributaries during spring high runoff
periods contributes to periods of P limitation and co-limitation (Boynton et al., 1995). The ecosystem then
returns to N limitation during low flow summer months (Paerl, 2002).
C.6. Watersheds, Landscapes and Disturbance
C.6.1. Interactions among Terrestrial, Transitional, and Aquatic
Ecosystems
Streams, and to a lesser extent lakes, can serve as indicators of regional environmental change
(Seastedt et al., 2004), partly because they integrate conditions within their watersheds including
atmospheric, edaphic, geologic, and hydrologic conditions. Streams reflect the terrestrial environment
most closely during high flow when much of the stream water enters the channel from the upper soil
where most of the biological activity occurs. The terrestrial signal can be less clear in lakes because they
have the capacity to store water and modify water chemistry through internal processes to a greater degree
than streams (Lawrence et al., 2008).
Young and Sanzone (2002) provided a checklist of ecological attributes that should be considered
when evaluating the effects of an environmental stressor on the integrity of ecological systems
(Table C-7). The Essential Ecological Attributes (EEAs) listed in the table represent groups of related
ecological characteristics (Harwell et al., 1999), including landscape condition, biotic condition, chemical
and physical characteristics, ecological processes, hydrology and geomorphology, and natural disturbance
regimes. The first three ecological attributes listed in Table C-7 can be classified primarily as "patterns,"
whereas the last three are "processes" (Bormann and Likens, 1979). They can be affected by a variety of
environmental stressors (Figure C-4).
C-33
-------
Of concern in this annex are relationships between NOx atmospheric deposition, derived from
anthropogenic sources, and one or more of the EEAs. The ranges of likely changes in ecosystem patterns
and processes associated with changes in deposition are discussed in the subsections that follow.
The following discussion assesses and characterizes the overall condition or integrity of
ecosystems within the U.S. that are affected by the deposition of atmospheric N and its role as a nutrient.
The six EEAs - landscape condition, biotic condition, chemical/physical characteristics, ecological
processes, hydrology/geomorphology, and natural disturbance regimes (Table C-7) - provide a
hierarchical framework for assessing ecosystem status. Characteristics related to structure, composition,
or functioning of ecological systems may be determined by the use of endpoints or ecological indicators
of condition that are measureable and significant either ecologically or to society (Harwell et al., 1999).
Table C-7. Essential ecological attributes and reporting categories.
Landscape Condition
Ecological Processes
Extent of Ecological System/Habitat Types
Energy Flow
Landscape Composition
Primary Production
Landscape Pattern and Structure
Net Ecosystem Production
Biotic Condition
Growth Efficiency
Ecosystems and Communities
Material Flow
Community Extent
Organic Carbon Cycling
Community Composition
N and P Cycling
Trophic Structure
Other Nutrient Cycling
Community Dynamics
Hydrology and Geomorphology
Physical Structure
Surface and Groundwater flows
Species and Populations
Pattern of Surface flows
Population Size
Hydrodynamics
Genetic Diversity
Pattern of Groundwater flows
Population Structure
Salinity Patterns
Population Dynamics
Water Storage
Habitat Suitability
Dynamic Structural Characteristics
Organism Condition
Channel/Shoreline Morphology, Complexity
Physiological Status
Extent/Distribution of Connected Floodplain
Symptoms of Disease or Trauma
Aquatic Physical Habitat Complexity
Signs of Disease
Sediment and Material Transport
Chemical and Physical Characteristics
Sediment Supply/Movement
(Water, Air, Soil, and Sediment)
Particle Size Distribution Patterns
Nutrient Concentrations
Other Material Flux
C-34
-------
N
Natural Disturbance Regimes
P
Frequency
Other Nutrients
Intensity
Trace Inorganic and Organic Chemicals
Extent
Metals Duration
Other Trace Elements
Organic Compounds
Other Chemical Parameters
PH
Dissolved O2
Salinity
Organic Matter
Other
Physical Parameters
Source: Young and Sanzone (2002).
The relationships among the EEAs are complex because all are interrelated. Changes in one EEA
may affect, directly or indirectly, every other EEA. Ecological processes create and maintain
environmental patterns, and these patterns affect how the processes are expressed (Young and Sanzone,
2002). Changes in patterns or processes can result in changes in the status and functioning of an
ecosystem.
Changes in the biodiversity, composition, and structure of ecosystems relate directly to functional
integrity. Changes in biodiversity are of particular significance in altering ecosystem function. The energy
obtained by plants (producers) from sunlight during photosynthesis and the chemical nutrients taken up
by those plants from the soil and the atmosphere are transferred to other species (consumers) within the
ecosystem through food webs. The movement of chemical nutrients through an ecosystem is cyclic, as the
nutrients are used or stored and eventually returned to the soil by microorganisms and fungi
(decomposers). Energy is transferred among organisms through the food webs and eventually is
dissipated into the environment as heat. The flows of energy and cycling of nutrients provide the
interconnectedness among the elements of the ecosystem and transform the community from a random
collection of numerous species into an integrated whole.
Human existence and welfare on this planet depend on life-support services provided by the
interaction of the EEAs. Both ecosystem structure and function play essential roles in providing goods
and services (Table C-8; Daily, 1997). Ecosystem processes provide diverse benefits including absorption
and breakdown of pollutants, cycling of nutrients, binding of soil, degradation of organic waste,
maintenance of a balance of gases in the air, regulation of radiation balance and climate, and fixation of
solar energy (Daily, 1997; Westman, 1977; World Resources Institute, 2000). These ecological benefits, in
turn, provide economic benefits and values to society (Costanza et al., 1997; Pimentel et al., 1997). Goods
such as food crops, timber, livestock, fish, and drinking water have market value. The values of
ecosystem services such as flood-control, wildlife habitat, cycling of nutrients, and removal of air
pollutants are more difficult to measure (Goulder and Kennedy, 1997). See discussion in Annex F.
Biodiversity is an important consideration at all levels of biological organization, including species,
individuals, populations, and ecosystems. Human-induced changes in biotic diversity and alterations in
the structure and functioning of ecosystems are the two most dramatic ecological trends of the past
C-35
-------
century (U.S. EPA, 2004; Vitousek et al., 1997b). The deposition of nutrient N from the atmosphere has
the potential to alter ecosystem structure and function by altering nutrient cycling and changing
biodiversity. It is important to understand how ecosystems respond to stress to determine the extent to
which anthropogenic stresses, including N deposition, affect ecosystem services and products (Table C-8).
Particular concern has developed within the past decade regarding the consequences of decreasing
biological diversity (Ayensii et al., 1999; Chapin et al., 1998; Hooper and Vitousek, 1997; Tilman, 2000;
Wall, 1999). Human activities that decrease biodiversity also alter the complexity and stability of
ecosystems, and change ecological processes. In response, ecosystem staicture, composition and function
can be affected (Figure C-5) (Chapin et al., 1998; Daily and Ehrlich, 1999; Levlin, 1998; Peterson et al.,
1998b; Pimm, 1984; Tilman and Downing, 1994; Tilman, 1996; Wall, 1999).
Hydrolic Alteration
Habitat Conversion
Habitat Fragmentation
Climate Change
Invasive Non-native Species
Turbidity/Sedimentation
Pesticides
Disease/Pest Outbreaks
Nutrient Pulses
Metals
Dissolved Oxygen Depletion
Ozone (Tropospheric)
Nitrogen Oxides
Nitrates
Hydrolic Alteration
Habitat Conversion
Habitat Fragmentation
Climate Change
Over-Harvesting Vegitation
Large-Scale Invasive
Species Introduction
Large-Scale Disease/Pest
Outbreaks
Hydrolic Alteration
Habitat Conversion
Climate Change
Turbid ity/Sedimen tation
Pesticides
Nutrient Pulses
Metals
Dissolved Oxygen Depletion
Ozone (Tropospheric)
Nitrogen Oxides
Nitrates
Sulfates
Salinity
Acidic Deoosition
Landscape
Condition
Chemical/
Physical
Hydrology/
Geomorphology
Hydrolic Alteration
Habitat Conversion
Climate Change
Over-Harvesting Vegitation
Disease/Pest Outbreaks
Altered Fire Regime
Altered Flood Regime
Hydrolic Alteration
Habitat Fragmentation
Climate Change
Turbidity/Sedimentation
Hydrolic Alteration
Habitat Conversion
Climate Change
Pesticides
Disease/Pest Outbreaks
Nutrient Pulses
Dissolved Oxygen Depletion
Nitrogen Oxides
Nitrates
Sulfates
Source: Young and Sanzone (2002).
Figure C-4. Sample stressors and the essential ecological attributes they affect.
Table C-8. Primary goods and services provided by ecosystems.
Ecosystem
Goods
Services
Agroecosystems Food crops
Fiber crops
Crop genetic resources
Maintain limited watershed functions (infiltration, flow control, partial soil protection)
Provide habitat for birds, pollinators, and soil organisms important to agriculture
Sequester atmospheric carbon
Provide employment
C-36
-------
Coastal
Fish and shellfish
Moderate storm impacts (mangroves, barrier islands)
Ecosystems
Fishmeat (animal feed)
Provide wildlife (marine and terrestrial (habitat and breeding
Seaweeds (for food and industrial
areas/hatcheries/n urseries
use)
Maintain biodiversity
Salt
Dilute and treat wastes
Genetic resources
Provide harbors and transportations routes
Provide human and wildlife habitat
Provide employment
Contribute aesthetic beauty and provide recreations
Forest
Timber
Remove air pollutants, emit O2
Ecosystems
Fuelwood
Cycle nutrients
Drinking and irrigation water
Maintain array of watershed functions (infiltration, purification, flow control, soil
Fodder
stabilization)
Nontimber products (vines, bamboos,
Maintain biodiversity
leaves, etc.)
Sequester atmospheric carbon
Food (honey, mushrooms, fruit, and
Moderate weather extremes and impacts
other edible plants; game)
Generate soil
Genetic resources
Provide employment
Provide human and wildlife habitat
Contribute aesthetic beauty and provide recreation
Freshwater
Drinking and irrigation water
Buffer water flow (control timing and volume)
Fish
Dilute and carry away wastes
Hydroelectricity
Cycle nutrients
Genetic resources
Maintain biodiversity
Provide aquatic habitat
Provide transportation corridor
Provide employment
Contribute aesthetic beauty and provide recreation
Grassland
Livestock (food, game, hides, and
Maintain array of watershed functions (infiltration, purification, flow control, soil
Ecosystems
fiber)
stabilization)
Drinking and irrigation water
Cycle nutrients
Genetic resources
Remove air pollutants and emit O2
Maintain biodiversity
Generate soil
Sequester Atmospheric carbon
Provide human and wildlife habitat
Provide employment
Contribute aesthetic beauty and provide recreation
Source: World Resources Institute (2000).
C-37
-------
Water Use and
Nutrient Loss
Water Availability
Hydrologic
C02 and
Temperature
Changes /
Land
Transformation
Erosion, \
Changes \
in Water Flow \
and Temperature
Precipitation
and Temperature
Loss of Crop
Genetic Diversity
Loss and
Fragmentation
of Habitat /
Habitat Loss
Change in
Transpiration
and Albedo
Reduced Resilience
to Change
Habitat
Change
Climate Change
Freshwater
Supply and Demand
Biodiversity Loss
Food Supply
and Demand
Forest Product
Supply and Demand
Source: Modified from Ayensu et al. (1999).
Figure C-5. Linkages among various ecosystem goods and services (food, water, biodiversity, forest
products) and other driving forces (climate change).
C.6.2. Interactions with Land Use and Disturbance
Scientific understanding of N cycling in forested watersheds is complicated by ecosystem response
to climatic variation, human land use, and various kinds of landscape disturbance, including insect
infestation, wind storm, fire, and disease (Aber and Driscoll., 1997; Goodale et al., 2000; Mitchell et al.,
2006). N dynamics in watersheds of mixed land use (i.e., agriculture, urban, forest) are even more
complicated. It is clear that disturbances have major impacts on nutrient enrichment from N deposition,
and that these effects can be long-lasting. Nevertheless, the scientific community is only in the early
stages of learning how to quantify these interactions.
Changes in land use can affect nutrient heterogeneity in the mineral soil of forest stands. For
example, Fraterrigo et al. (2005) found that patterns of variance in soil C, N, and Ca2+ concentration
increased with the extent of intensive past land use in western North Carolina. Land use might alter the
local patchiness of soil nutrients by decoupling interactions among microclimate, topography, vegetation,
and soil biota. In particular, mechanical soil mixing and maintenance of agricultural monocultures can
homogenize soils in cultivated systems (Robertson et al., 1993; Paz-Gonzalez and Taboada, 2000). Such
effects may be important if the land use is changed to forest. Similarly, changes in species composition
can alter the spatial distribution of nutrients in litter inputs (Dijkstra and Smits, 2002; Fraterrigo et al.,
2005).
In the northeastern U.S., concentrations of N in streams of upland forested watersheds tend to be
considerably lower than in streams draining watersheds with other land uses. In a comparison of small
watersheds in eastern New York, concentrations of N were highest and most variable in a stream draining
a watershed where the predominant land use was row crop production. Total dissolved N concentration in
streams in sewered suburban and urban watersheds were somewhat lower and less variable than in
streams draining the agricultural watershed. Streams in urban and suburban watersheds may also
experience high episodic N loading caused by combined sewer overflows (Driscoll et al., 2003c).
C-38
-------
C.6.3. Timber Harvest and Fire
Timber harvest contributes to nutrient removal from the ecosystem via biomass export and
acceleration of leaching losses (Bormann et al., 1968; Mann et al., 1988). In particular, logging
contributes to loss of N and Ca2+ from the soil (Tritton et al., 1987; Latty, 2004). The extent of nutrient
loss is determined, at least in part, by the intensity of the logging and whether or not it is accompanied by
fire (Latty, 2004). The species composition of the regrowth vegetation also has important effects on
nutrient cycling. Fire is sometimes followed by establishment of N-fixing vegetation that provides
substantial sources of Nr (Johnson, 1995; Johnson et al., 2004). Tree species also vary dramatically in
their N cycling properties, especially in their influence on litter mass and quality (Finzi et al., 1998;
Ferrari, 1999; Ollinger, 2002). Thus, over time, the extent of effect of logging and fire on nutrient cycling
can increase, depending largely on shifts in tree species composition and the degree to which C and N
pools are altered in the mineral soil and the forest floor.
Dissolved N exports have been clearly shown to increase substantially after major watershed
disturbance, often reaching peak concentrations in streamwater within 1 to 3 years of disturbance, and
then returning to background concentrations after about 5 to 10 years (Likens et al., 1978; Bormann and
Likens, 1979; Eshleman et al., 2000). Such transient N03 leakage has been shown to occur subsequent to
both logging (Dahlgren and Driscoll, 1994; Martin et al., 1984; Yeakley et al., 2003) and insect
infestation (Eshleman et al., 1998, 2004).
The extent to which timber harvesting influences leaching of N03 and base cations from soils to
drainage waters depends on changes in primary productivity, nutrient uptake by plants and
microorganisms within the terrestrial ecosystem, and hydrological pathways for transferring nutrients to
drainage water (Hazlett et al., 2007). Because of the variety of responses and interactions of
environmental and forest litter and soil conditions, it is difficult to generalize about the influence of
harvesting on N cycling (Grenon et al., 2004; Hazlett et al., 2007).
It is known, however, that land use history constitutes a major influence on N leaching from
forested watersheds that receive moderate to high levels of atmospheric N deposition (Pardo et al., 1995;
Aber and Driscoll, 1997; Goodale et al., 2000; Lovett et al., 2000a). The severity of effect and length of
the recovery period probably vary according to the nature of the past disturbance. Extensive past logging
appears to have considerable and long-lasting effects on nutrient cycling (Goodale and Aber, 2001; Fisk
et al., 2002). Latty et al. (2004) compared soil nutrient pools and N cycling among three forest stands in
the Adirondack Mountains: old growth, selectively logged, and selectively logged and then burned. The
logging and fire had occurred about 100 years previously. Results suggested that even relatively light
logging, plus burning, may influence the extent of subsequent N limitation overtime scales of decades to
centuries (Latty, 2004). Models of forest ecosystem response to disturbance incorporate such long-lasting
effects of land use on C and N storage, cycling, and release (Aber et al., 1997).
Chen and Driscoll (2004) simulated the response of five forested watersheds in the Adirondack and
Catskill regions of New York to changes in atmospheric deposition and land disturbance. Simulation
results suggested that forest harvesting caused increased leaching of base cations and N03 from the
watersheds. These changes also affected model projections of future pH and acid neutralizing capacity
(ANC) of lake water. Model results suggested that lakewater pH and ANC were lower in response to
forest cutting as compared with undisturbed conditions.
Nitrification rates at old growth sites in the White Mountains of New Hampshire (63 ± 4.3 kg
N/ha/yr) were approximately double those at previously burned (34 ± 4.4 kg N/ha/yr) and previously
logged (29 ± 4.7 kg N/ha/yr) sites (Goodale, 2001). Fire and logging disturbances had occurred about 100
years previously on these study sites. Nitrification increased as forest floor C:N ratio decreased, resulting
in higher N03 concentrations in streamwater. These results suggest that forest disturbance can have long-
lasting effects on N cycling and the potential for N saturation.
Thus, disturbance can affect N cycling and the response of forest ecosystems to N deposition. In
addition, it also appears that vegetative changes stimulated by N deposition may affect the frequency and
C-39
-------
severity of disturbance. Excess Nr deposition is thought to be impacting essential ecological attributes
associated with terrestrial ecosystems and how they respond to disturbance. Effects of Nr deposition
influence habitat suitability, genetic diversity, community dynamics and composition, nutrient status,
energy and nutrient cycling, and frequency and intensity of natural fire disturbance regimes. For example,
several lines of evidence suggest that Nr deposition may be contributing to greater fuel loads and thus
altering the fire cycle in a variety of ecosystem types (Fenn et al., 2003c). Invasive grasses, which can be
favored by high N deposition, promote a rapid fire cycle in many locations (D'Antonio and Vitousek,
1992). The increased productivity of flammable understory grasses increases the spread of fire and has
been hypothesized as one mechanism for the recent conversion of coastal sage shrub (CSS) to grassland
in California (Minnich and Dezzani, 1998).
High grass biomass has also been associated with increased fire frequency in the Mojave Desert
(Brooks, 1999; Brooks and Esque, 2002; Brooks et al., 2004). This effect is most pronounced at higher
elevation, probably because the increased precipitation at higher elevation contributes to greater grass
productivity. Increased N supply at lower elevation in arid lands can only increase productivity to the
point at which moisture limitation prevents additional growth. Fire was relatively rare in the Mojave
Desert until the past two decades, but now fire occurs frequently in areas that have experienced invasion
of exotic grasses (Brooks, 1999).
C.6.4. Insect Infestation and Disease
Insect infestation and plant disease, via atmospheric N deposition, can alter the effects of nutrient
enrichment on forest ecosystems. Such disturbances alter the pool of Nr in the forest floor with short-term
impacts on N03 and base cation leaching. Positive influences of N deposition on root and seed biomass
of an annual plant, common ragweed, were suppressed by herbivory, which increased with higher
available plant shoot N (Throop, 2005).
Eshleman et al. (2004) applied a regional lithology-based unit N export response function model to
simulate N03 export to streams in the Chesapeake Bay watershed. The model considered the geographic
distribution of bedrock classes and the timing and extent of defoliation by gypsy moth (Lymantria dispar)
larvae. Modeling results suggested that the regional annual N03 -N export increased during the year
following peak insect defoliation by about 1500%, from an initial rate of 0.1 kg N/ha/yr to nearly 1.5 kg
N/ha/yr.
Between the mid-1980s and the early 1990s, the southward expanding range of the gypsy moth
traversed Shenandoah National Park, VA (Webb, 1999). Some areas of the park were heavily defoliated 2
to 3 years in a row. The White Oak Run watershed, for example, was more than 90% defoliated in both
1991 and 1992. The gypsy moth population in White Oak Run then collapsed due to pathogen outbreak.
This insect infestation of the forest ecosystem resulted in substantial effects on streamwater chemistry.
The most notable effects of the defoliation on park streams were dramatic increases in the concentration
and export of N and base cations in streamwater. Figure C-6 shows the increase in N03 export that
occurred in White Oak Run. Following defoliation, N03 export increased to previously unobserved
levels and remained high for over 6 years before returning to predefoliation levels. The very low baseline
levels of N03 export in park streams were consistent with expectations for N-limited, regenerating
forests (e.g., Aber et al., 1989; Stoddard, 1994). Release of N03 to surface waters following defoliation
was likewise consistent with previous observations of increased N export due to forest disturbance
(e.g., Likens et al., 1970; Swank, 1988). The exact mechanisms have not been determined, but it is
evident that the repeated consumption and processing of foliage by the gypsy moth larva disrupted the
ordinarily tight cycling of N in the forests within this park.
C-40
-------
Change in Annual Streamwater Nitrate Flux
Following Watershed Defoliation
6
IB
f 4
0)
•*->
CO
t_
+->
5 2
0
80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99
Year
Source: Sullivan et al. (2003).
Figure C-6. Effect of watershed defoliation by the gypsy moth caterpillar on NO3" flux in streamwater.
White Oak Run was heavily defoliated for three consecutive years. The watershed area
defoliated was 46.5% in 1990, 92.9% in 1991, and 90.4% in 1992. In 1993, the gypsy moth
population collapsed and there was no further defoliation.
Hie elevated concentrations of NO, following defoliation did not appear to contribute to baseflow
acidification in White Oak Run. This was due to a concurrent increase in concentrations of base cations in
streamwater (Webb et al., 1995). Both N()3 and base cation concentrations also increased during high-
runoff conditions, although the increase in base cations did not fully compensate for the episodic increase
in NO; . As a consequence, episodic acidification became more frequent and more extreme (Webb et al.,
1995).
Large trees in old growth forests may resorb less N from foliage than do younger trees on
previously logged or burned sites (Killingbeck, 1996). This process would be expected to contribute to an
alleviation of N limitation on plant processes in old-growth forests (Lattv et al.. 2004). This effect might
also extend to herbivores, which are often N-limited (Mattson, 1980). Latty et al. (2004) attributed the
high severity of beech bark disease in old growth forests to such a mechanism. Beech bark disease is
caused by an introduced scale insect {Crytococcus fctgisugct), which has high N requirements (Wargo,
1988; Houston, 1994). The N-rich foliage in the old growth forest may improve insect fitness,
contributing to a higher rate of infestation in the old growth stands (Latty et al,, 2004).
C.6.5. Urbanization
Perhaps the most noteworthy impact of urban land use on processes of nutrient enrichment from N
deposition concerns the transport of Nr to N-limited estuarine and near-coastal waters. In agricultural, and
especially in forested areas, it is generally expected that most atmospherically deposited N is taken up by
terrestrial vegetation. Relatively little of the deposited N is available for transport to downstream
White Oak Run
First Year of
Mapped Defoliation
C-41
-------
receiving waters. This is not the case for urban land use. Urbanization often involves substantial clearing
of vegetation and compaction of soil (Poff et al., 1997; Burges et al., 1998; Jones et al., 2000; Trombulak
and Frissell, 2000; Alberti et al., 2007). Due to the relatively large impervious surface area in the urban
landscape (buildings, roads, parking lots, etc.), a higher percentage of precipitation is routed directly to
surface waters, with less opportunity for vegetative uptake of deposited N (Arnold and Gibbons, 1996;
Montgomery and Buffington, 1998). Therefore, atmospheric N deposition contributes proportionately
more N03 to surface waters in urban settings than it does with other land uses. The reduction in riparian
and wetland coverage and functionality also diminishes the ability of the urban watershed to filter
contaminants from runoff, including atmospherically deposited N (Peterjohn and Correll, 1984). Because
many large urban areas are both located close to the coastline and expected to receive relatively high NOx
deposition, they can constitute sizeable sources of body contribution to estuarine and marine waters.
C.6.6. Agriculture
Agricultural ecosystems are not sensitive to levels of N deposition typically found in the U.S.
Rather, such ecosystems often act as net sources of NH3 emissions rather than as sinks (Griinhage et al.,
1992; Krupa, 2003). Atmospheric N deposition can contribute a quantitatively important component of
the Nr requirements of pastures and croplands. In such settings, atmospheric N deposition provides an
additional chronic source of N fertilizer. This may be viewed as a beneficial outcome. Nevertheless, some
of the N that is atmospherically deposited on agricultural land may eventually leach to drainage waters
and contribute to eutrophication, especially in estuarine and near-coastal marine environments. Industrial
livestock operations also contribute substantial amounts of NHY to the atmosphere, some of which is
deposited on coastal waters.
Agriculture also experiences indirect effects of NOx emissions through the formation of ground-
level 03. Such effects are not considered in this review.
C.6.7. Other Disturbances
In some ecosystems, chronic additions of atmospherically derived N may have had irreversible
consequences that involve interactions with invasive, non-native plants. For example, California has
many plant species that occur in shrub, forb, and grasslands that receive high N deposition. There are up
to 200 sensitive plant species in southern California coastal sage scrub (CSS) communities alone (Skinner
and Pavlik, 1994). About 25 plant species are thought to be extinct in California, most of them forbs that
occurred in sites that have experienced conversion to annual grassland (U.S. EPA, 2005a). As CSS
vegetation continues to convert to grassland dominated by invasive species, loss of additional rare plant
species may occur. Invasive plant species are often identified as a major threat to rare native plant species.
However, the occurrence of invasive species may combine with other stress factors, including N
deposition, to promote increased productivity of invasive species at the expense of native species.
As sensitive vegetation is lost, wildlife species that depend on these plants can also be adversely
affected. There are several threatened or endangered wildlife species listed by the U.S. Fish and Wildlife
Service, including the desert tortoise (Gopherus agassizii) and checkerspot butterfly that are native to
plant communities in California thought to be sensitive to atmospheric N input. A native to the San
Francisco Bay area, the bay checkerspot butterfly has declined in association with invasion of exotic
grasses that replaced the native forbs on which the butterfly depends. In particular, the larval stage of the
butterfly is dependent on Plantago erecta, which is increasingly being outcompeted by exotic grasses
(U.S. EPA, 2005a).
Decline in the population of the desert tortoise may be due to a number of co-occurring stresses,
including grazing, habitat destruction, drought, disease, and a declining food base. In the desert shrub
C-42
-------
inter-spaces, sites where native forbs once flourished, invasive grasses now dominate, reducing the
nutritional quality of foods available to the tortoise (Fenn et al., 2003a; Nagy et al., 1998). N deposition
contributes to the productivity and density of grasses at the expense of native forbs (Brooks, 2003).
C.6.8. Multiple Stress Response
Ecosystems are often subjected to multiple stressors, of which nutrient enrichment from
atmospheric deposition of N is only one. Additional stressors are also important, including 03 exposure,
climatic variation, natural and human disturbance, the occurrence of invasive non-native plants, native
and non-native insect pests, and disease. Atmospheric N deposition interacts with these other stressors to
affect ecosystem patterns and processes in ways that are only beginning to beunderstood.
For example, terrestrial ecosystems at many locations are subjected to high N deposition and high
exposure to 03. This is especially true in portions of southern California and the Appalachian Mountains.
Mixed conifer forests in the San Bernardino and San Gabriel mountains in southern California are
exposed to high levels of atmospheric 03 and receive atmospheric N deposition in the range of about 5 to
over 30 kg N/ha/yr (Takemoto et al., 2001). Spatial variability in N deposition is high due to the patchy
characteristics of these forests and associated canopy effects on dry deposition processes. The forest
ecosystems have reached N-saturation, as evidenced by high N03 concentrations in stream water.
However, evaluation of N effects on vegetation is complicated by the concurrent effects of 03, which has
damaged sensitive plant species, especially Ponderosa and Jeffrey pine. Bytnerowicz (Bytnerowicz, 2002)
summarized N/03 interactions and consequent effects.
Peak diurnal concentrations of atmospheric 03 and N02 co-occur at Tanbork Flat in the San
Bernardino Mountains (Bytnerowicz et al., 1987). They can have counteracting effects, with 03 reducing
growth and N deposition enhancing growth of pine trees (Grulke and Balduman, 1999).
Jeffrey and Ponderosa pine are the most sensitive western coniferous tree species to injury from 03
pollution (Miller etal., 1983; Duriscoe and Stolte, 1989). In some areas of the western Sierra Nevada
Mountains, 03 concentrations have been high enough to cause visible foliar injury to these species and
reduced needle retention (Bytnerowicz, 2002). Reduced radial growth has also been observed (Peterson
et al., 1987, 1991). In the San Bernardino Mountains, trees of these species that exhibit severe foliar
injury from 03 do not show growth reductions (Arbaugh et al., 1999), and this has been attributed to the
fertilizing effects of high N deposition (Bytnerowicz, 2002; Takemoto et al., 2001). This may be an
example of counteracting effects from 03 and N air pollution. It is also possible that N deposition in the
western Sierra Nevada Mountains may increase growth of pines, especially on nutritionally poor granitic
soils (Takemoto et al., 2001).
C-43
-------
Table C-9.
Ecological effects of N deposition described for study sites in the Western U.S.
Ecological or
Environmental
Impact
Location
Level of
Uncertainty
Possibility of
Broader Occurrence
(at other sites)
Reference
EFFECTS IN AQUATIC SYSTEMS
Elevated NO3" in runoff;
most severe in southern
California and in
chaparral catchments in
the southwestern Sierra
Nevada
Transverse ranges of
southern California; low-
elevation catchments in
the Sierra Nevada; high-
elevation catchments in
the Colorado Front Range
Well-documented
response
It is unclear how
widespread this
phenomenon is outside
the ecosystems listed,
because there is littler
information from low-
elevation systems in the
Sierra Nevada and
elsewhere.
Fenn and Poth (1999)
Fenn et al. (2003a)
Williams et al. (1996a)
N enrichment and shifts
in diatom communities
in alpine lakes
Colorado Front Range;
Lake Tahoe
(California/Nevada border)
Documented for two
lakes east of the
Continental Divide
and Lake Tahoe
These effects seem
likely in other N-
enriched lakes but have
not been investigated.
Baron et al. (2000)
Goldman (1988)
Wolfe et al. (2001)
Reduced lake water
clarity and increased
algal growth
Lake Tahoe
(California/Nevada
border); high-elevation
lakes throughout central
and southern Sierra
Nevada
Well-documented
response; N and P
deposition believed
to be important
factors
Lake Tahoe is an
unusual case because
of its renowned lake
clarity; extent of
occurrence elsewhere in
northern Sierra Nevada
is unknown.
Jassby et al. (1994)
Sickman et al. (2003a)
Increased NO3"
concentrations in high-
elevation lakes
Several regions, mainly
downwind of urban
centers
Fairly well
established from
lake surveys, but
more data needed
for improved
definition of
frequency and
severity
Evidence suggests that
urban plumes and
agricultural emissions
affect lake NO3" levels.
There is also evidence
of impacts on low-
elevation lakes.
Sickman et al. (2002)
EFFECTS IN TERRESTRIAL SYSTEMS
Enhanced growth of
invasive species
Costal sage scrub,
southern California; San
Francisco Bay area
N deposition,
fertilization studies,
and plant
community data
supportive, but
moderate
uncertainty remains
It is not known if this
effect occurs elsewhere,
but it is expected that
nitrophilous species will
be selected for if N
accumulates in soil.
Allen etal. (2008)
Weiss (1999)
Lichen community
changes
Parts of the Pacific
Northwest; many areas in
California; north and
central Colorado
Well-established
response; a highly
sensitive air
pollution indicator
Because of the
sensitivity of many
lichen species, it is likely
that this effect occurs
elsewhere.
Nash and Sigal (1999)
C-44
-------
Deleterious effects on
threatened and
endangered species
San Francisco Bay area;
southern California
Supportive
evidence, but high
degree of
uncertainty about
the precise role of N
deposition
There is a high
likelihood of effects in
some habitats where N
accumulates in soils
Weiss (1999), Brooks (2003)
Altered fire cycle
Coastal sage scrub in
southern California
Hypothesis based
on observations,
fertilization studies,
and N deposition
and N cycling data;
high level of
uncertainty
Because it has not been
studied elsewhere, it is
uncertain whether this
effect occurs in other
areas.
Allen etal. (2008)
Altered forest C cycling
and fuel accumulation
San Bernardino Mountains
Documented
response
It is uncertain whether
this effect occurs in
other areas.
Grulke and Balduman (1999)
Physiological
perturbation of
overstory species
San Bernardino Mountains
Documented
response
This effect has not been
widely studied but is
expected for sensitive
plant species exposed
to O3 and adapted to N
limitation but growing in
N-enriched soils.
Grulke etal. (1998)
Grulke and Balduman (1999)
Takemoto et al. (2001)
Forest expansion into
grasslands
Great Plains of western
Canada
Supportive
evidence found, but
high degree of
uncertainty as to the
role of N deposition
It is not known whether
this effect occurs in
other areas.
Kochy and Wilson (2001)
N emissions as a major
contributor to regional
haze problem
National forests and parks
throughout California, the
Pacific Northwest, and
some sites in the Interior
West
Well-established
effect; contribution
from Nous
pollutants has been
quantified
This is known to occur
in areas far removed
from emissions sources
because of long-range
transport.
Fenn et al., (2003a)
IMPROVE data (4 March 2003;
http://vista.circa.colostate.edu/
improve)
NOx emissions as
precursors for
phytotoxic levels of O3,
leading to O3 injury to
sensitive plant species
Southern California; Sierra
Nevada
Well-established
effect
Significant O3 injury to
vegetation has not been
reported from other
sites downwind of urban
centers but cannot be
ruled out as urban
regions expand.
Miller and McBride (1999)
Carroll etal. (2003)
Note: Summary includes the degree of uncertainty regarding the role of N deposition in each effect and the likelihood that these effects may occur elsewhere in the West.
C-45
-------
Annex D. Critical Loads
D.1. Background
Critical loads and critical levels are used to express how much deposition of an atmospheric
pollutant (a "load") or how large a concentration of an airborne pollutant (a "level") can be tolerated by
natural or artificial systems without significant harm or change occurring in those systems (see
Section D.2.1). The critical load and critical level approaches to quantifying the effects of pollutants
attempt to estimate the atmospheric deposition load or concentration that would be likely to cause
environmental harm. The expectation is that environmental harm can be avoided by keeping pollution
levels or loads below these critical values. This approach is commonly used to estimate loads or levels of
pollution required to protect lakes, streams, or forest soils from environmental harm. The basic principles
are, however, transferable to any sensitive receptor. Since the present evaluation deals primarily with the
effects of atmospheric deposition of S and N compounds, this chapter focuses on critical loads more than
critical levels.
Most critical load studies in North America have been undertaken in Canada. The critical load
approach has been used in Canada to design emission reduction programs (Jeffries and Lam, 1993;
RMCC, 1990). Modeling of critical loads for the 1997 Canadian Acid Rain Assessment (Jeffries, 1997)
was conducted for six regional clusters of lakes, four in eastern Canada, one in Alberta, but also the
Adirondack Mountains in New York. More recently, critical loads have been determined and mapped for
waters (Aherne et al., 2004; Dupont et al., 2005; Henriksen et al., 2002; Hindar et al., 2001; Watmough
et al., 2005) and forest soils (Arp et al., 1996; Moayeri et al., 2001; Ouimet et al., 2006; Watmough and
Dillon, 2003), for a number of regions in eastern Canada. There have also been a number of regional
critical loads studies (cf. Henriksen and Dillion, 2001; Ouimet et al., 2006) focused on acid-sensitive
lakes on the Canadian pre-Cambrian shield. Much of this work is summarized and presented, along with
steady state critical load maps for eastern Canada, in the 2004 Canadian Acid Deposition Science
Assessment (Jeffries et al., 2005).
At the regional, cross-border level, critical loads in northeastern North America have been
developed by a joint U.S.-Canadian cooperative. The Conference of New England Governors and Eastern
Canadian Premiers (NEG/ECP) has undertaken a program with the objective to "estimate sustainable
acidic deposition rates and exceedances for upland forests representative of the New England States and
the Eastern Canadian Provinces..." (NEG/ECP Forest Mapping Group, 2001). The Forest Mapping
Working Group within the NEG/ECP conducts regional assessments of the sensitivity of northeastern
North American forests to current and projected S and N emissions levels. The group is charged with
identifying specific forested areas most sensitive to continued S and N deposition and estimating
deposition rates required to maintain forest health and productivity at large spatial scales (Miller and
McDonald, 2006). The NEG/ECP has also provided estimates of critical loads for surface waters in
northeastern North America (Dupont et al., 2005).
Aside from the NEG/ECP studies, the use of critical loads to assess S and N deposition effects in
the U.S. has not been as geographically extensive as elsewhere in North America or Europe. Most critical
loads studies in the U.S. have focused on smaller sub-regional areas or individual sites. Critical loads
studies for forests in the U.S. have been centered in the northeast and have usually used a catchment-
based approach (Aber et al., 2003; Driscoll et al., 2003d; Pardo and Driscoll, 1993, 1996). Critical load
studies for surface waters have been more extensive along the eastern seaboard. Critical loads have been
estimated for lakes in the Northeast (Driscoll et al., 2001b; Pembrook, 2004) and for streams in the Mid-
Atlantic States and central Appalachians (Sverdrup et al., 1992; Sullivan et al., 2004). In the western U.S.
there have been a few studies of critical loads for acidification of surface waters (Sullivan et al., 2004).
D-1
-------
The primary concern in the West, however, has been the critical load of N deposition affecting both
terrestrial and aquatic resources through eutrophication and/or through N enrichment and its impact on
community structure (Baron et al., 1994, 2000; Baron, 2006; Bowman et al., 2006; Burns, 2004; Fenn,
2003a; Nydick et al., 2003, 2004b; Stevens et al., 2004; Williams and Tonnessen, 2000; Wolfe et al.,
2003).
The critical load approach has been used extensively in Europe for organizing information about
effects, and for specifying emissions reductions that would be required to protect ecosystems and other
sensitive receptors from the harmful effects of atmospheric S and N) deposition. During the 1970s, it was
recognized that transboundary air pollution in Europe had adverse ecological and economic
consequences. In response, the countries of the UN Economic Commission for Europe (UNECE)
developed the Convention on Long-range Transboundary Air Pollution (LRTAP), the first international
legally binding instrument to deal with problems of air pollution on a broad regional basis (see
http://www.unece.org/env/lrtap). Signed in 1979, it entered into force in 1983. The LRTAP Convention
requires that its Parties cooperate in research into the effects of S compounds and other major air
pollutants on the environment, including agriculture, forestry, natural vegetation, aquatic ecosystems, and
materials. To this end, the Executive Body for the Convention established a Working Group on Effects
(WGE) that is supported by a number of International Cooperative Programmes (ICPs). The ICP for
Mapping and Modeling generated maps of critical loads for all of Europe in 1995 (Posch et al., 1995).
Those maps are modified on a continuing basis (e.g., Posch et al., 2001). By comparing current or
expected future deposition to the critical loads maps, mapped estimates of exceedances have been
generated. An exceedance is the amount of S and N deposition that occurs at some specific time (past,
current, or future), above the critical load of deposition that would be required to protect against adverse
effects on the environment. The maps of estimated exceedances are used in negotiations to regulate
pollutant emissions in Europe (for example, the 1999 Gothenberg Protocol via the UNECE Convention
on LRTAP).
Outside of North America and Europe, there is an increasing use of critical loads for assessment
purposes, and to inform policy development. Examples include studies in Siberia (Bashkin et al., 1995),
Thailand (Milindalekha et al., 2001), and South Africa (Van Tienhoven et al., 1995). In China, several
studies have been carried out to study the sensitivity of surface waters to acidification and the critical
loads of acid deposition (Duan et al., 2000; Li et al., 2000; Ye et al., 2002; Hao et al., 2001), and to
calculate the critical loads of S and N acidity for soils at both the local and regional scales (Xie et al.,
1995; Zhao and Seip, 1991; Duan et al., 2000b, 2001).
D.1.1. The Critical Load Process
The process of estimating critical loads is not a purely scientific enterprise. Management or policy
input to the process is needed to insure that the appropriate science is included and the appropriate
questions are addressed. The critical load process integrates knowledge of a multitude of physical,
chemical and biological mechanisms affected by ambient air quality, and presents the current scientific
understanding in a format that is most useful for assessing current or future management practices and
policy decisions regarding air quality, or the resources affected. The critical loads process provides
decision-making insight based on both scientific evidence and policy priorities.
Science and policy are closely coupled in the critical loads process. In the development of critical
load estimates, it is important to identify those elements that are essentially scientific in nature as opposed
to those elements that are driven by management or policy priorities. The scientific elements include tasks
such as: relating ambient air quality to pollutant deposition, quantifying the relationships between
pollutant deposition and resource responses, identifying the resources at risk to adverse effects,
understanding the temporal and spatial responses of resources to pollutant deposition, and more. The
policy-dependent elements include tasks such as: identifying the environmental resources to be protected,
D-2
-------
establishing appropriate criteria for different land use areas (e.g., Class I areas, national parks, wildlife
refuges), defining significant harm to protected resources, and more. When all elements are integrated, it
is apparent that the critical load process provides a framework for alternate ways of examining and
understanding the cascade of effects from ambient air quality to resource effects, described in the
preceding four annexes. Changing scientific assumptions or understanding may result in different critical
load estimates for the same resources. Changing policy or management assumptions or priorities may also
result in different critical load estimates.
There is, therefore, no single "definitive" critical load for a natural resource. Critical load estimates
are explicitly linked to policy, but their reliability is conditioned on the soundness of the underlying
science. As elements of the critical load process change, the critical load estimates will change to reflect
both the current state-of-knowledge and policy priorities. Changes in scientific understanding may
include: new dose-response relationships, better resource maps and inventories, larger survey datasets,
continuing time series monitoring, improved numerical models, etc. Changes in the policy elements may
include: new definitions of harm, new mandates for resource protection, focus on new pollutants, or
inclusion of perceived new threats that may exacerbate the pollutant effects (e.g., climate change).
The critical load process is thus an iterative process — as science changes, the content is updated;
as policy needs change, the content is re-directed. Being iterative, the process allows incremental
improvement in understanding resource responses to ambient air quality. Individual elements of the
process can be replaced as needed to reflect new science or policy. Continuing to update the process may
reduce uncertainty and risk, as new data or techniques allow refinement of existing pieces. The piecewise
nature of the process provides adaptability as new policy concerns arise, such as new pollutants or
mandates. As the critical load process advances, a "library" of critical load estimates will result.
Examining and comparing these accumulated results, and their underlying scientific and policy bases,
may produce a "weight of evidence" consensus, even if any single estimate entails substantial uncertainty.
D.1.2. Organization of this Annex
This Annex is intended as a review of the current state of critical loads science. It is not the
intention to address questions of management or policy other than to point out where these activities
influence the critical loads process. The material in Section D.2 presents necessary definitions and
describes the conceptual framework for a critical load analysis. This framework identifies those elements
that are primarily scientific in nature and those elements that require policy input. The framework also
describes the steps that are taken in deriving a critical load estimate for a given resource. It is not an
objective of this Annex to provide details of all critical loads studies that have been implemented in the
U.S. or elsewhere. The conceptual framework, however, provides a generalized summary of the steps
most critical loads studies have followed. Section D.3 discusses the time frame of responses for
implementation of a critical load. Time frames of resource response are often ignored, or assumed
implicitly, in defining a critical load analysis. The time required to implement the policy and technology
to achieve a critical load can also affect the responses of the resources at risk. The time frames of
response are important for selecting the data and models used to estimate critical loads. Section D-4
summarizes the tools (models and modeling approaches) commonly used in calculating critical loads. The
Annex concludes in Section 5 with a summary of the current agreement on critical loads uses in the U.S.
that was a product of the Multi-Agency Workshop on Critical Loads held in 2006. The workshop
produced a series of recommendations for current and future activities related to critical loads analyses in
the U.S.
D-3
-------
D.2. Definitions and Conceptual Approach
D.2.1. Critical Load Definitions
Critical loads and critical levels are used to express how much deposition of an atmospheric
pollutant (a "load") or how large a concentration of an airborne pollutant (a "level") can be tolerated by
natural or artificial systems without significant harm or change occurring in those systems. The generally
accepted definition of a critical load or a critical level of atmospheric pollutants emerged from a pair of
international workshops held in the late 1980s. The workshop participants defined a critical load or a
critical level as "a quantitative estimate of an exposure to one or more pollutants below which significant
harmful effects on specified sensitive elements of the environment do not occur according to present
knowledge."
This evaluation deals primarily with the effects of atmospheric deposition of S and N compounds.
This Annex, therefore, will deal exclusively with the concept of critical loads of S and N compounds from
atmospheric deposition. Critical levels of pollutant concentration will not be addressed. As discussed in
previous annexes, the deposition of both S and N has acidifying effects on receptors (Annex B), and the
deposition of oxidized and/or reduced N compounds can produce eutrophication or nutrient-enrichment
effects in receptors (Annex C). The following material, therefore, will focus on critical loads of S and N
for acidification effects, and on critical loads of N for nutrient effects.
In addition to the generic definition of a critical load/level presented above, the participants in the
second international workshop (the Skokloster Workshop) developed a number of specific definitions
related to known atmospheric pollutants. Two of those definitions are relevant to this Annex.
Recognizing that both S and N compounds contribute to the acidity of deposition, the workshop
participants developed a definition for critical loads of S and N for acidification of an ecosystem: "the
highest deposition of acidifying compounds that will not cause chemical changes leading to long-term
harmful effects on ecosystem structure and function according to present knowledge." Recognizing that N
in both oxidized (NOY = NO +, N02 + N02 +, N03 ) and reduced (NHX = NH3 + NH/) forms in
deposition may influence the eutrophication and nutrient balances of ecosystems, the workshop
participants defined the critical load of N for nutrient effects in an ecosystem as "the highest deposition of
N as NHX and/or NOY below which harmful effects in ecosystem structure and function do not occur
according to present knowledge."
All three definitions can be applied to different receptors in a number of different environments
(e.g., terrestrial ecosystems, transitional ecosystems, aquatic ecosystems, groundwater, agricultural crops,
etc.). A sensitive element can constitute a part of, or the whole of, an ecosystem. Harmful effects can
occur to individual organisms, to populations, or to entire communities within an ecosystem. Harmful
effects can also be defined at the level of the ecosystem itself as changes in ecosystem processes,
structure, and/or function.
While the concepts expressed in these definitions of critical loads and levels are easily understood
and intuitively satisfying, the application of the critical load concept requires careful consideration and
definition of a number of terms and procedures. It is apparent that there can be many different critical
load values for a given atmospheric pollutant depending on the receptor or sensitive element(s) being
considered. There can also be multiple different atmospheric pollutants that can produce the same harmful
effects in a given receptor. Therefore, the critical load of a given pollutant can potentially be dependent on
the deposition and/or atmospheric concentration of other pollutant species. Finally, the same atmospheric
pollutant can produce a variety of different disturbances in a sensitive ecosystem that might occur at
different pollutant loads. For example, N deposition produces both nutrient and acidification effects and
the critical load of N for each type of disturbance may be different.
D-4
-------
Therefore, to derive a quantitative estimate of the critical load of an atmospheric pollutant, a
number of factors must be identified and defined Figure D-l. These include disturbance type, receptor,
sensitive elements, and definition of what constitutes significant harm. In addition, a numerical
relationship between pollutant deposition and the identified receptor response must be formulated,
generally based on either an empirical dose-response relationship or a steady state or dynamic numerical
model simulation. The next section outlines the steps (decisions) that must be taken to implement this
process.
D.2.2. Critical Load Analysis Procedures
The development of a quantitative critical load estimate requires a number of steps. In this
discussion, Figure D-l is used to illustrate the procedure. The figure is simplified to facilitate general
discussion and does not represent the full complexity of the choices that must be made, or the scientific
understanding underlying those choices.
Table D-1 An example of the matrix of information that must be considered in the definition and
calculation of critical loads. Note that multiple alternative biological indicators, critical
biological responses, chemical indicators, and critical chemical limits
1
V
\
There are eight general steps that must be taken to define the basic critical load question in any
analysis.
1. Identify the ecosystem disturbance of concern (acidification, eutrophication, etc.). Not all
disturbances will occur in all regions or at all sites, and the degree of disturbance may vary across
landscape areas within a given region or site.
1) Disturbance
Acidification
Eutrophication
2) Receptor
Forest
Lake
Grassland
Lake
3) Biological
indicator
Sugar
Maple
Norway
Spruce
Brook trout
Fish species
richness
Species
diversity
Primary
productivity
4) Critical
biological
response
Failure to
reproduce
Seedling
death
Presence
absence
Species
loss
Species
loss
Excess
productivity
5} Chemical
indicator
Soil % Base
Saturation
Soil Ca/AI
ratio
Lakewater
ANC
Lakewater
ANC
Soil C/N
ratio
Lakewater
no3
8) Critical
chemical
limit
10%
1,0
0 peq/L
50 peq/L
20
10 peq/L
7} Atmospheric
pollutant
S04, NGg,
mh4
so4. no3,
NH4
so4. no3.
nh4
so4, no3.
mb4
NO,, NH4
no3. nh4
8} Critical
pollutant load
???
???
???
???
???
???
D-5
-------
2. Identify the landscape receptors subjected to the disturbance (forests, surface waters, crops, etc.)
Receptor sensitivity may vary locally and/or regionally, and the hierarchy of receptors most
sensitive to a particular type of disturbance may vary as well.
3. Identify the biological indicators within each receptor that are affected by atmospheric deposition
(individual organism, species, population, or community characteristics). Indicators will vary
geographically and perhaps locally within a given receptor type
4. Establish the critical biological responses that define "significant harm" to the biological
indicators (presence/absence, loss of condition, reduced productivity, species shifts, etc.).
Significant harm may be defined differently for biological indicators that are already at risk from
other stressors, or for indicators that are perceived as "more valued."
5. Identify the chemical indicators or variables that produce or are otherwise associated with the
harmful responses of the biological indicators (stream water pH, A1 concentration, soil base
saturation, etc.). In some cases, the use of relatively easily measured chemical indicators (e.g.,
surface water pH or ANC) may be used as a surrogate for chemical indicators that are more
difficult to measure (e.g., A1 concentration).
6. Determine the critical chemical limits for the chemical indicators at which the harmful responses
to the biological indicators occur (e.g., pH <5, base saturation <5%, A1 concentrations >100 (ig/L,
etc.). Critical limits may be thresholds for indicator responses such as presence/absence, or may
take on a continuous range of values for continuous indicator responses such as productivity or
species richness. Critical limits may vary regionally or locally depending on factors such as
temperature, existence of refugia, or compensatory factors (e.g., high calcium concentration
mitigates the toxicity of A1 to fish and plant roots).
7. Identify the atmospheric pollutants that control (affect) the pertinent chemical indicators
(deposition of S042 , N03, NH4, HN03, etc.). Multiple pollutants can affect the same chemical
variable. The relative importance of each pollutant in producing a given chemical response can
vary spatially and temporally.
8. Determine the critical pollutant loads (often in units of kg/ha/yr deposition of S or N03~N, etc.) at
which the chemical indicators reach their critical limits. Critical pollutant loads usually include
both wet and dry forms of pollutant deposition. The critical pollutant load may vary regionally
within a receptor or locally within a site (as factors such as elevation or soil depth vary) and may
vary temporally at the same location (as accumulated deposition alters chemical responses).
The definition of the critical load problem for a region or individual site generally requires working
down the table from top to bottom (Table D-l) What is the disturbance? What receptors are affected?
What indicator organisms are, or were previously present and observable? What chemical indicators are
changing and can be measured? What atmospheric pollutant is driving the changes in the chemical
indicators?
The derivation of a quantitative estimate of a critical load generally requires working from the
bottom of the table back towards the top, as indicated by the arrows in Table D-l. What is the maximum
load of a pollutant that will cause a shift in the chemical indicator to its critical limit such that a critical
indicator response occurs, or does not occur? From this point of view, it can be seen that steps 8 and 6
require the development of dose-response functions for the components of the ecosystem being
considered (arrows in Table D-l). Step 8 describes the response of the chemical indicator as a function of
the pollutant load, and Step 6 describes the responses of the biological indicator as a function of the
chemical variable. As discussed in later sections, these response functions can be derived using empirical
(e.g., statistical) or process-based (e.g., mechanistic) models that are either time-invariant (static or steady
state) or time-variable (dynamic).
D-6
-------
Each step in the development of the critical load, as summarized in Table D-l can be classified as
either a predominantly scientific task or as a task benefiting from, or perhaps requiring, collaboration and
input from scientists, decision-makers, and other interested parties. For instance, tasks 1, 2, and 3 can be
viewed as predominantly scientific tasks that can be completed by asking questions of fact. At task 4,
however, questions of what defines "significant harm" entail subjective elements that cannot be
determined by scientific techniques alone. In anticipation of the ultimate use of the critical load definition
to set policy or establish management strategies, it is appropriate that political, socioeconomic, or perhaps
ethical considerations be brought to bear in defining "significant harm." To define "harm" is to imply a
corrective action, the cost of which will have to be borne by someone. Having reached agreement on task
4, however, tasks 5, 6, 7, and 8 are again predominantly scientific in nature, requiring determination of the
causal links, represented as response functions or models, leading from the loading of the pollutant to the
defined "significant harm."
This procedure will almost certainly result in calculation of multiple critical load values for a given
pollutant and analysis location. The multiple solutions derive from the nested sequence of disturbances,
receptors, and biological indicators that must be considered for a given pollutant. Multiple critical load
values may also arise from an inability to agree on a single definition of "significant harm" at step 4.
Finally, there is the inescapable heterogeneity of all natural environments. Consider soils for instance. The
high spatial variability of soils almost guarantees that for any reasonably sized soil-based "receptor" that
might be defined in a critical load analysis, there will be a continuum of critical load values for any
indicator chosen. The range of this continuum of values may be narrow enough to be ignored, but in any
critical load analysis there is nevertheless an a priori expectation of multiple values, or of a range of
values.
The existence of multiple estimates of critical loads for a given pollutant and receptor should
present no real problem. Examination of the range of critical loads derived may be deemed useful in
subsequent discussions of the analysis, and in the decision-making steps that may follow critical load
calculation. For instance, the lowest critical load of all those derived may be adopted as "the" critical
load, as is often done in Europe. This however, is a policy choice. The scientific task is the derivation of
the multiple values using best available information.
D.2.3. Target Load Definition
As seen in the previous section, it is expected that a potentially large number of critical load values
may be objectively determined for a given atmospheric pollutant and a given receptor. Like the definition
of "significant harm," the choice of which critical load value to use for management or decision-making
is subjective, and should be driven by socioeconomic, political, and ethical considerations. The target load
concept was developed to address these issues. Target loads are deposition loads of a given pollutant,
based on critical load estimates for the pollutant, which incorporate policy and/or management decisions
about the amount of pollutant deposition, and therefore the amount of resource damage that is deemed
acceptable. Target loads can be set at, above, or below the various critical loads. If the target load is set
above some of the estimated critical loads, one accepts the inevitability that some of the ecosystem
components, generally the most sensitive, will be adversely affected. If the target load is set below all of
the estimated critical loads, a safety margin has been established to account for uncertainty inherent in the
process.
Given the spatial heterogeneity of natural systems, target loads might also be used to provide some
measure of "cumulative resource protection." As discussed above, there typically exists a range of critical
loads for a particular "significant harm" in a particular receptor. Selecting a target load within the range
will provide protection for the fraction of the receptor with critical loads above the chosen target load,
whereas that fraction with critical loads below the chosen target load will be expected to suffer some harm
at that deposition level. In this way, it is possible to use the target load to define protection for some
D-7
-------
cumulative proportion of the receptor (i.e., a target load for protection of 95% of the resource from
"significant harm").
While most of the steps involved in estimating critical loads depend on sound, objective scientific
analysis, the selection of target loads is almost entirely a subjective judgment. The selection of target
loads must begin with reliable estimates of critical loads to set the constraints, and define the expected
consequences of the target load choices. Nevertheless, the final decisions of which indicators are the key
indicators, how much cumulative resource should be protected, how much sooner or later resource
protection will be implemented, cannot be answered scientifically. Political, socioeconomic, and ethical
considerations will form the basis of the final target load selections. Frequently, the legal mandates for
various public lands would have a determinant influence on the selection of target loads. For instance,
Federal Class I areas may be held to one standard of harm because of mandates to protect "natural
condition," whereas Federal mixed-use lands may be held to a different standard of harm, and cropland to
yet another standard of harm.
It is also important to note that scientific understanding, modeling approaches, and the data used to
estimate critical loads are continually improving. Furthermore, the political, economic, and social
environments surrounding selection of target loads are also constantly shifting. Therefore, the analysis
and estimation of critical and target loads must be an iterative process.
D.3. Time Frame of Response
The critical load definitions and procedures discussed in the previous section do not explicitly
consider the time frame of ecosystem response. When is "significant harm" expected? How long will it be
before existing harm is reversed? When should critical loads be implemented? How long should a critical
load be maintained? The use of critical and target loads in resource management always has some time
frame of expected response, and some context of management priorities. For instance, it may be that a
target load well below the critical load would hasten the recovery of a receptor with existing harm. Or, it
may be that a receptor that has not yet been damaged can sustain a target load above the critical load for
some finite period before incurring "significant harm." Such time frames can be very long (many decades
or centuries).
The time frame of response between implementation of a critical load and the corresponding
changes in biological or chemical indicators is a potentially important factor in establishing critical load
analysis procedures and in selecting the final target load. Analyses can be designed to provide estimates
of either "steady state critical loads" or "dynamic critical loads" depending on the perceived, or
mandated, importance of the time frame of response and the types of models (transfer functions) used.
Steady state critical loads analyses provide estimates of the long-term sustainable deposition of a
pollutant that will not cause "significant harm" to a receptor. This is the relevant information needed for
any receptor to provide protection from damage by the pollutant in perpetuity as the receptor comes into
equilibrium with the pollutant critical load (the implicit purpose of steady state analyses). However, no
information is given concerning the time to achieve the equilibrium or what may happen to the receptor
along the path to equilibrium. Estimated steady state critical loads for receptors that are currently
damaged provide no information concerning when the desired long-term sustainable protection will occur
and the existing "significant harm" will be mitigated. There exists the possibility that receptors with no
current damage could suffer "significant harm" while waiting for implementation of the critical load. The
possible occurrence, timing, and duration of such "interim periods of harm" are not the subject of steady
state analyses.
Dynamic critical loads analyses provide estimates of a specifically scheduled deposition load of a
pollutant that will not result in "significant harm" to a receptor at a specified time. This is the relevant
information needed for any receptor to provide protection from damage by the pollutant within a specified
time frame (the explicit purpose of dynamic analyses). However, care should be taken in interpreting the
D-8
-------
results of dynamic analyses to ensure that "significant harm" to the receptor does not occur after the
specific timetable has been completed. Many receptors can tolerate higher loads of a pollutant for a few
decades (a common length of specified schedules for dynamic analyses) than can be sustained over longer
periods. The use of dynamic critical load estimates in such cases may provide protection from harm
during a time frame of immediate interest, but ultimately fail to provide long-term protection, unless these
issues are considered.
D.3.1. Steady State Critical Loads
If the time frame of response is not important, for instance, if the target load is to provide long-term
sustainable protection and the immediacy of the responses is not relevant, the use of static or steady state
models (response functions) is justified in the critical load analysis procedure. Using steady state models
to estimate critical loads and compare the estimated critical load to current or future deposition, only two
cases can be distinguished: current or future deposition is below the critical load; or current or future
deposition exceeds the critical load. In the first case, no problem is apparent, and no target load is deemed
necessary, unless increases in deposition are anticipated. In the second case, there is by definition an
increased risk of "significant harm" to the receptor and selection of a target load for resource protection is
indicated.
The lack of explicit consideration of time in a steady state critical load analysis can lead to
assumptions that are frequently not warranted. The critical load derived in a steady state analysis is an
estimate of the long-term, constant deposition that a receptor can tolerate with no significant harm after it
has equilibrated with the critical load deposition. However, biological and geochemical processes that
affect a receptor may delay the attainment of equilibrium (steady state condition) for years, decades, or
even centuries. By definition, steady state critical loads do not provide any information on these time
scales. As a result, it is often assumed that reducing deposition to, or below the steady state critical load
value will immediately eliminate or mitigate "significant harm." That is, it is assumed that the chemical
indicator affected by the atmospheric pollutant immediately attains a non-critical value upon
implementation of the critical load, and that there is immediate biological recovery as well. As discussed
in the next section, these assumptions may not be valid.
D.3.2. Dynamic Critical Loads
The time frame of receptor response is important if the establishment of target loads is tied to
defined schedules of deposition change or receptor responses. The use of time-dependent or dynamic
model response functions will be necessary if the critical load analysis considers the response time frame.
In the cascade of events that occur from changed deposition of an atmospheric pollutant to development
of responses of key biological indicators, there are many processes in natural systems that are time and/or
resource dependent and therefore can introduce delays in the response pattern. In the decision-making
process leading to the adoption of target loads, there are likewise considerations of when deposition
changes can be initiated and completed and when biological indicator responses are desired. With
dynamic models, either empirical or process-based, a wide range of estimated critical loads can be
derived for comparison with current or future deposition depending on the temporal constraints imposed
on the critical load analysis. Temporal constraints that can be imposed on a given critical load analysis are
determined by: the receptor responses—the characteristic time scales and inherent lags of the receptor
being analyzed (a function of hydrobiogeochemical processes in the receptor); and the deposition
schedules—the years designated for beginning and completing the changes in deposition and for
evaluating the indicator responses (a function of political, socioeconomic, and management constraints).
D-9
-------
D.3.3. Receptor Responses
The general conceptual model of the linkages among pollutant deposition and the responses of
chemical and biological indicators can be characterized as a series of delays. In the causal chain from
deposition of pollutant to damage to key biological indicators there are two major links that can give rise
to delays. First, hydrological and biogeochemical processes in catchments can delay the responses of
chemical indicators. Second, biological processes and population dynamics can further delay the response
of biological indicators. The pattern of chemical and biological indicator responses can be represented
conceptually (Figure D-l) (adapted from Jenkins et al., 2003; Posch et al., 2003). Five stages in the
conceptual pattern can be distinguished (Figure D-l):
¦ Stage 1: Pollutant deposition is below the critical load for either the critical chemical limit or the
critical biological response, and there is no "significant harm" to the receptor. As long as
deposition stays below the critical load, this is the 'ideal' situation.
¦ Stage 2: Pollutant deposition rises above the critical load, but chemical and biological indicators
still do not violate their respective criteria because there is a delay. No damage is likely to occur
at this stage, despite the exceedance of the critical load. The time between the first exceedance of
the CL and first violation of the biological criterion (first occurrence of "significant harm") is
called the Damage Delay Time (DDT = t3—tl).
¦ Stage 3: Pollutant deposition is above the critical load and both the chemical and biological
criteria are violated. Measures to reduce emissions are taken to avoid further harm to the receptor
and pollutant deposition begins to decrease.
¦ Stage 4: Pollutant deposition has been reduced to a level below the critical load, but the chemical
and biological criteria are still violated, and thus "recovery" has not yet occurred. The time
between the first non-exceedance of the critical load and the subsequent non-violation of both
criteria can be called the Recovery Delay Time (RDT = t6-t4).
¦ Stage 5: This stage is similar to Stage 1. Pollutant deposition has been reduced to a level below
the critical load and neither the chemical nor biological criteria are violated. Only at this stage can
the receptor be considered to have recovered to an undamaged level.
Stages 2 and 4 can be further subdivided into two sub-stages each: chemical damage and recovery
delay times (DDTc = t2—tl and RDTc = t5-t4; dark grey in Figure D-l and (additional) biological damage
and recovery delay times (DDTb = t3-t2 and RDTb = t6-t5; light grey). Given opportunities for
"confounding effects" (i.e., mechanisms not related to acidic deposition but affecting biological
indicators, such as forest pest infestation or climate change) occurring during the "delay periods," it is
clear that unambiguous short-term patterns of recovery of biological indicators are unlikely to be
observed, even in the presence of rather large declines in pollutant deposition. This has important
implications for recovery expectations.
D.3.4. Deposition Schedules
Dynamic critical loads, by definition, must explicitly account for the receptor time scales and lags
described above. Therefore, the process of estimating dynamic critical load values for a given pollutant
must be based on a planned or assumed deposition schedule for changing the pollutant deposition and for
assessing the receptor responses. Three different time periods are specified, as illustrated in Figure D-2.
[The nomenclature used here for the three years specified in the deposition schedule conforms to that used
in European "dynamic target loads analyses" (Posch et al., 2003)].
D-10
-------
Stage 1 Stage! Stages Stage 4 stages
c
Critical Load
Critical Limit
Critical Response
DDT RDT
Figure D-1. Conceptual patterns of pollutant deposition effects on a chemical indicator and a
corresponding biological indicator during increasing and decreasing deposition. Critical
limits and responses for the chemical and biological indicators are indicated as horizontal
lines, along with the critical load of deposition that produces these levels. The delays between
the exceedance of the critical load (t1), the violation of the critical chemical limit (t2), and the
crossing of the critical biological response (t3) are indicated in grey shades, highlighting the
DDT. Similar delays in chemical and biological recovery during deposition reductions (t4, t5,
and t6) define the RDT of the system.
D-11
-------
Historical Deposition Pattern
Future Deposition Patterns
Protocol Year
Time
Target Year
Implementation Year'
Figure D-2. Pollutant deposition patterns for defining the temporal parameters of dynamic critical loads
analyses. The deposition schedule requires that three years be specified: (1) the year in
which changes in pollutant deposition are begun, called the protocol year; (2) the year in
which changes in pollutant deposition are completed, called the implementation year; and (3)
the year in which the chemical or biological response indicator is evaluated, called the target
year.
The first time period is the protocol year when deposition changes moving toward the critical load
are begun. It will be the case that voluntary or mandated changes in deposition will require a number of
years to get underway once a critical load has been calculated or target load has been selected. These
delays in moving toward the critical load will affect the dynamic responses of the chemical and biological
indicators and, therefore, must be included in the dynamic modeling of receptor response. Before the
protocol year, it must be assumed that pollutant deposition will be continuing along the pattern of recent
or historical deposition change or along the pattern dictated by future deposition scenarios already
planned and assumed to take effect.
The second time period is the implementation year when deposition changes are complete and
pollutant deposition has reached the desired critical or target load. It is likely that a number of years will
elapse between the time changes in deposition toward the critical load are initiated, and the time when
they are completed. During this transition period, pollutant deposition continues at a rate higher, or lower,
than the critical load. The effects of these years of deposition inputs above or below the critical load value
will affect the dynamic responses of the chemical and biological indicators, and must also be included in
the dynamic modeling of receptor response. It is assumed in dynamic critical loads analyses that the
pollutant deposition to the receptor remains constant at the critical load for all years after the
implementation year.
The final time period is the target year when the biological indicators are evaluated. Recognizing
that there are inherent lags in receptor responses following changes in pollutant deposition, it a number of
years will frequently be allowed to elapse after the implementation year before the receptor responses are
assessed. It must also be recognized that receptor responses will continue to change over time. Thus,
selection of the target year will affect attainment of the critical limit.
The deposition schedule for a dynamic critical load analysis can be driven by a number of
considerations, and can be organized from protocol year to target year or vice versa. The selection of the
protocol and implementation years is often a matter of political will and economic possibility. Large-scale
pollution abatement programs take time to negotiate. Costs or engineering difficulties may delay the start
of the abatement program and affect the length of time it takes to complete the program once it is begun.
Once these constraints have been established, it is then possible to select a reasonable target year for
D-12
-------
evaluation of the receptor responses. Alternately, resource management mandates might require that
"significant harm" to receptor indicators be mitigated or eliminated by a certain time. This establishes the
target year for the dynamic analysis and the protocol and implementation years must be selected to allow
time for any lags in the receptor responses to occur.
Clearly, there is tension between the two approaches when developing a deposition schedule. It is
possible, for instance, to defer the protocol and implementation years so far into the future that extensive
"significant harm" occurs to the receptor indicators in the intervening years. If that damage is especially
severe, the critical load, when the target year is finally reached, may not be achievable. Similarly, if a
receptor is currently suffering harm, it is possible to choose a target year for receptor response too close to
the present day to allow time for the receptor to recover, even if the pollutant deposition was reduced to
zero immediately.
Both of these hypothetical scenarios raise important points about dynamic critical load estimates.
Because time is explicitly incorporated, there are certain dynamic critical loads questions that have no
answer. Commonly called "you can't get there from here" problems, these deposition schedules choose
protocol, implementation, or target years that are inconsistent with the time scales of receptor response.
For example, setting a target year 5 years in the future for achievement of no "significant harm" in a
receptor that is currently badly damaged, and has a history of high pollutant loading, may be asking the
impossible. Critical load estimates derived in this case would require having set the pollutant deposition
to zero some years in the past. In other words, the state of no significant harm cannot be reached within
the specified five years regardless of how deposition is changed within that 5 year period ("you can't get
there from here"). This problem is moot for steady state critical loads. Steady state critical loads analyses
will always provide some sensible estimate (zero or finite) of long-term sustainable pollutant deposition
for every receptor because time is not a factor. Dynamic critical loads analyses, on the other hand, may
frequently provide non-quantitative results, but these results nonetheless convey useful information
concerning the current status of the receptor and point out the necessity to continue the analysis with
modified assumptions or expectations to develop realistic and achievable target load values.
D.3.5. Long-Term Implications
The explicit inclusion of time in critical loads estimation provides useful information for managers
or policymakers when deciding when and how much to alter pollutant emissions and deposition, but the
dynamic approach leads to implicit assumptions that must be recognized. Focus on the near-term aspects
of receptor responses (the years included in the deposition schedule) can be misleading. The implicit
assumption is that having attained the desired biological or chemical response in the target year, nothing
more will happen, or at least that further changes in the receptor, if they do occur, will not produce
"significant harm." Available dynamic model critical load estimates suggest that this is not always true,
and the long-term implications of dynamic critical load estimates should be examined carefully.
The dynamic critical loads procedure assumes that pollutant deposition remains constant at the
critical load from the implementation year until the target year, and the assessment of receptor response.
Model simulations can be continued, assuming deposition at the constant critical load value, for a number
of years after the specified target year to ensure that lags in the receptor response will not result in
"significant harm" appearing at some later date, even though it was not present in the target year. Some
receptors have chemical or biological lags that are many decades long, or longer. Critical loads analyses
based on deposition schedules that cover only 20 to 30 years can produce the unwanted result of
estimating a dynamic critical load that avoids "significant harm" to the receptor in the target year, only to
have the receptor suffer damage some years later.
It is important to determine which receptor responses and which deposition schedules might lead to
such an unwanted result. There are some general guidelines concerning this potential problem. For any
receptor for which there is currently no "significant harm," the estimated critical load provided by the
D-13
-------
dynamic approach will be one that brings the biological or chemical indicators to the threshold of harm
without crossing it (the definition of the critical load). However, this dynamic critical load has then put
the biological or chemical indicator on a trajectory away from its currently good status and toward the
threshold of harm. In most cases, the trajectory toward harm will continue past the target year and
"significant harm" will occur in these receptors some time after the target year.
On the other hand, in any receptor for which there currently is "significant harm," the estimated
critical load provided by the dynamic approach will be one that brings the biological and chemical
indicators to the threshold of harm and crosses it, just to return to a state of no harm. This dynamic critical
load puts the biological or chemical indicator on a trajectory away from harm and towards good status. In
most cases, it is likely that the upward trajectory will continue past the target year and significant further
recovery will occur in these receptors after the target year.
These are, however, merely generalizations. Depending on the geochemical and biological process
affecting a receptor, there exist possibilities that upward trajectories could become downward trajectories
sometime after the target year and vice versa. The most straightforward procedure is to run the model(s)
used in the dynamic critical loads analyses for a sufficiently long time after the target year so that any
reasonable chance of delayed damage to the receptor or delayed recovery is either discovered or
discounted.
D.3.6. Steady State and Dynamic Critical Loads: Complementary
Information
There is no "correct" choice to be made between steady state and dynamic critical loads analyses.
Both provide estimates of pollutant loads that are intended to avoid "significant harm" to a receptor. Both
are valid scientific expressions of the receptor's sensitivity to the pollutant. They differ primarily in the
time scales implicit in their use. Steady state analyses provide critical load estimates for long-term
sustainable protection, but ignore questions of near-term recovery and avoidance of interim harm.
Dynamic analyses provide critical load estimates that can be used to examine short-term or long-term
options for recovery of damaged systems and avoidance of interim harm, but may ignore the ultimate
long-term sustainability of the estimated deposition, which may evolve over centuries. Clearly, the two
approaches provide complementary information.
The complementary nature of the two critical loads approaches can be exploited in the selection of
a target load estimate for a receptor. Selecting the lower of the two critical load estimates for the receptor
(steady state or dynamic) should result in facilitation of recovery, or avoidance of harm, in the short-term,
as well as long-term sustainability once the receptor has reached equilibrium with the selected target load.
Multiple lines of evidence reflecting multiple critical load values can provide important information that
collectively provides the foundation for management decision-making.
The procedures, data requirements, and computational resources needed for each of the two critical
loads approaches may differ significantly depending on the models (response functions) adopted for the
analyses. Differences in the approaches may also depend on the disturbance, receptor, or indicator being
evaluated. The next two sections discuss the disturbances, receptors, and indicators relevant to deposition
of S and N, and the models used for calculation of critical load estimates by each approach.
D.4. Calculation of Critical Loads
The derivation of quantitative estimates of critical loads requires the development of dose-response
functions (models) for the components of the ecosystem being considered. Models are needed to describe
two different classes of dose-response function. Geochemical models describe the changes in the
D-14
-------
chemical indicators that occur as functions of changes in the pollutant loads. Biological response models
describe the changes in the biological indicators as functions of changes in the chemical variables.
Models for either class of dose-response function can be developed using two general approaches.
Empirical models are based on direct observations of indicator response to pollutant deposition. They are
usually developed using statistical techniques and generally do not contain a mechanistic pathway linking
pollutant deposition to indicator response. Process-based models are based on conceptual representations
of chemical and biological mechanisms, and use mathematical equations to express the inter-relationships
among system components. Whereas process-based models frequently also use observations of receptor
responses to pollutant deposition for calibration and validation, they are fundamentally different from
empirical models in that mechanistic pathways from pollutant deposition to indicator response are
explicitly included in the model structures. In general, the geochemical models used to link S and N
deposition to chemical indicator response are mostly process-based, whereas biological responses to
acidification by S and N are mostly modeled using empirical approaches. Finally, both geochemical
models and biological response models, whether developed using either empirical or process-based
approaches, can be further classified as static or dynamic depending on whether or not time is included
among variables.
D.4.1. Empirical Models
Empirical models can be constructed relating either chemical or biological indicators to pollutant
deposition. The empirical models currently in use for calculating critical loads employ steady state
approaches. This is not a necessary constraint, however, because even with no knowledge of the
underlying mechanisms, there exist many statistical techniques for relating the time-series of outputs and
inputs of ecosystems. The reason empirical critical loads models are usually based on a steady state
approach is primarily because time-series data of long enough duration to parameterize dynamic
empirical models are not generally available. In general, empirical models require less complex datasets,
are more straightforward to implement, and are easier to understand than process-based models. For some
receptors, the lack of conceptual understanding of the mechanisms of indicator response to pollutant
deposition renders the use of process-based models problematic, and the use of empirical models is then
the only viable critical load analysis approach.
D.4.2. Acidification Effects of Sulfur and Nitrogen
Empirical models of critical loads for acidity assign critical loads to soils on the basis of soil
mineralogy and chemistry (UNECE, 2004). For example, at the Critical Loads Workshop at Skokloster
soil forming materials were divided into five classes on the basis of the dominant weatherable minerals. A
critical load range, rather than a single value, was assigned to each of these classes according to the
amount of acidity that could be neutralized by the base cations produced by mineral weathering. The
classification of soil materials developed at Skokloster used a relatively small range of primary silicate
minerals and carbonates. A larger range of minerals was classified by Sverdrup and WarfVinge (1988) and
Sverdrup et al. (1990) for use in the PROFILE model (WarfVinge and Sverdrup, 1992).
D.4.3. Nutrient Effects of Nitrogen
Empirical models of critical loads for nutrient N have been developed in Europe within LRTAP to
set critical loads for atmospheric N deposition (e.g., UNECE, 2004). Empirical critical loads of N for
natural and semi-natural terrestrial ecosystems and wetland ecosystems were first presented in a
D-15
-------
background document for the 1992 workshop on critical loads held under the UNECE LRTAP
Convention at Lokeberg, Sweden (Bobbink et al., 1992b). A number of European expert workshops have
taken place to reach agreement among specialists regarding the impacts of N for various ecosystems and
related critical loads (Achermann and Bobbink, 2003; Bobbink et al., 1992b, 1996.; Hornung et al., 1995;
Nilsson and Grennfelt, 1988). Empirical relationships have also recently been developed in the U.S.,
particularly for western ecosystems (e.g., Baron et al., 1994, 2000; Burns, 2004; Fenn et al., 2003; Nydick
et al., 2004a; Williams and Tonnessen, 2000).
D.4.4. Process-Based Models
A number of process-based models are currently in use for calculating critical loads using both
steady state and dynamic approaches. Developing a process-based modeling approach that includes all
appropriate chemical and biological indicators is a complex task. Some process models incorporate both
geochemical and biological response mechanisms in one program. An alternate approach is to chain
individual process-based models, for example taking the output of a geochemical model and passing it as
input to a biological-response model. In either approach, the level of process complexity varies a great
deal among the various available models. The choice of a particular process-based modeling approach to
be used in a critical load analysis (dynamic or steady state, all-in-one or chained, etc.) will depend on the
scope of the analysis, the quality and quantity of available data, and the availability of resources (time and
money) for the analysis. The following is a brief overview of some of the process-based biogeochemical
models that are commonly used to calculate critical loads.
D.4.5. Steady State Models
The Simple Mass Balance (SMB) model is the standard model for calculating critical loads for
terrestrial ecosystems under the LRTAP Convention (Sverdrup et al., 1990; Sverdrup and De Vries, 1994).
The SMB model is a single-layer model. There also exist multi-layer steady state models for calculating
critical loads in terrestrial ecosystems. Examples are the MACAL model (De Vries, 1988) and the widely
used PROFILE model (WarfVinge and Sverdrup, 1992), which has at its core a model for calculating
weathering rates from total mineral analyses.
The Steady State Water Chemistry (SSWC) model (Henriksen et al., 1992; Henriksen and Posch,
2001; Sverdrup et al., 1990) calculates critical loads of acidity for surface waters, based on the principle
that acid loads should not exceed the balance of non-marine, non-anthropogenic base cation sources and
sinks in a catchment, minus a buffer to protect selected biota from being damaged.
The First-order Acidity Balance (FAB) model for calculating critical loads for surface waters takes
into account sources and sinks within the lake and its terrestrial catchment. The original version of the
FAB model was developed and applied to Finland, Norway, and Sweden by Henriksen et al. (1992) and
Posch et al. (1997). A modified version was first reported in Hindar et al. (2000, 2001) and is described in
more detail by Henriksen and Posch (2001).
D.4.6. Dynamic Models
MAGIC is a lumped-parameter model of intermediate complexity, developed to predict the long-
term effects of acidic deposition on surface water chemistry (Cosby et al., 1985a, 1985b, 2001). The
model simulates soil solution chemistry and surface water chemistry to predict the monthly and annual
average concentrations of the major ions in these waters. MAGIC consists of: a sub-model in which the
D-16
-------
concentrations of major ions are assumed to be governed by simultaneous reactions involving SO42-
adsorption, cation exchange, dissolution-precipitation-speciation of Al, and dissolution-speciation of
inorganic carbon (C); and a mass balance sub-model in which the flux of major ions to and from the soil
is assumed to be controlled by atmospheric inputs, chemical weathering, net uptake and loss in biomass,
and losses to runoff. At the heart of MAGIC is the size of the pool of exchangeable base cations in the
soil. As the fluxes to and from this pool change overtime owing to changes in atmospheric deposition, the
chemical equilibria between soil and soil solution shift to give changes in surface water chemistry. The
degree and rate of change of surface water acidity thus depend both on flux factors and the inherent
characteristics of the affected soils. MAGIC is described in more detail in Annex C.
PnET-BGC is an integrated dynamic biogeochemical model that simulates chemical
transformations of vegetation, soil, and drainage water. It was formulated by adding the sub-model BGC
(biogeochemistry) to PnET-CN, a model of C, water, and N balances (Aber and Federer, 1992; Aber and
Driscoll, 1997; Aber et al., 1997), to expand the model to include vegetation and organic matter
interactions of major elements (i.e., Ca2+, Mg2+, K+, Na+, Si, S, P, Al3+, CI , F ), abiotic soil processes,
solution speciation, and surface water processes (Gbondo-Tugbawa et al., 2001). The model was initially
developed for, and applied to, the northern hardwood forest ecosystem. It was tested extensively at the
Hubbard Brook Experimental Forest, New Hampshire, including a detailed sensitivity analysis of
parameter values. The model has subsequently been applied to intensively-studied watersheds in the
Adirondack and Catskill regions of New York and applied regionally to the Adirondacks (Chen and
Driscoll, 2005b) and northern New England (Chen and Driscoll, 2005a, 2005b). See additional
description in Annex C.
Simulation Model for Acidification's Regional Trends (SMART2) is a soil acidification and
nutrient cycling model and is an extension of the dynamic soil acidification model SMART (Kros et al.,
1995). The original model was a relatively simple simulation of the response of soil and soil water quality
to atmospheric inputs. Improvements in SMART2 include processes of canopy interactions, litter fall, root
decay, mineralization, and root uptake of nutrients. SMART2 has been used primarily in European critical
loads studies.
The Soil Acidification in Forest Ecosystems model (SAFE) was developed at the University of
Lund in Sweden (Alveteg and Sverdrup, 2002; WarfVinge et al., 1993). The main differences between the
SAFE and MAGIC models are: (a) weathering of base cations is not calibrated for SAFE, but it is
modeled with the PROFILE sub-model, using soil mineralogy as input (WarfVinge and Sverdrup, 1992);
(b) SAFE is oriented to soil profiles in which water is assumed to move vertically through several soil
layers, (c) cation exchange between Al, H, and (divalent) base cations is modeled in SAFE with Gapon
exchange reactions rather than Gaines-Thomas reactions, and the exchange between the soil matrix and
soil solution is diffusion-limited. The standard version of SAFE does not include S adsorption although a
version, in which S adsorption is dependent on S042 concentration and pH of soil solution, has recently
been developed (Martinson et al., 2003).
ForSAFE is a mechanistic model that simulates N and C cycling and soil chemistry. Climatic
drivers within the model include temperature, precipitation, radiation, and deposition. ForSAFE combines
three established models (SAFE, PnET-CN, and DECOMP). SAFE simulates soil chemistry (e.g.,
chemical weathering, cation exchange, leaching, and solution equilibrium reactions). PnET-CN (Aber et
al., 1997) is used to predict forest growth within ForSAFE, through the simulation of C fixation, litterfall,
and C and nutrient allocation. DECOMP (Walse et al., 1998) is a dynamic, multi-layered process-oriented
decomposition model that incorporates the influences of temperature, moisture, pH, and Al. Very Simple
Dynamic soil acidification model (VSD) only includes a few key processes, such as cation exchange and
N immobilization, and a mass balance for cations and N (Posch et al., 2003). VSD does not consider
seasonal variations, as the time step in the model is one year. The VSD model is based on mass balance
equations that describe soil input-output fluxes and equations describing the rate-limited (e.g., uptake and
silicate weathering) and equilibrium (e.g., cation exchange) soil processes. Soil solution chemistry is
based solely on the net element input from the atmosphere (i.e., deposition minus net uptake minus net
D-17
-------
immobilization) and geochemical interactions in the soil (i.e., C02 equilibria, weathering of carbonates
and silicates, and cation exchange). VSD simulates a single soil layer with a constant density and a fixed
depth. The concentration of the soil water leaving the compartment is assumed to be equal to the annual
precipitation excess.
D.5. Use of Critical Loads in the U.S. - Current Status
At the Multi-Agency Critical Loads Workshop for Sulfur and Nitrogen Deposition Effects on
Freshwater and Terrestrial Ecosystems, convened by the U.S. EPA, the U.S. Forest Service (USFS), the
National Park Service (NPS), and the USGS in May 2006, approximately 75 scientists, conservation
representatives, and state and federal agency officials gathered to share information, discuss scientific
advances, and develop a broad federal strategy for advancing critical loads in the U.S. (U.S. EPA, 2006c).
The conclusions and recommendations of that workshop are presented below. These conclusions and
recommendations represent the current understanding of critical loads as scientific tool and policy
instrument in the U.S.
The conclusions and recommendations below were reached by the Federal Agencies sponsoring the
workshop. It is worth noting that some state agencies have pursued the use of critical loads independently
to link science and policy in addressing the management of natural resources. For instance, in the State of
Colorado, critical loads for N deposition that were developed for Rocky Mountain National Park (Baron,
2006) are being used to develop goals for N emissions reductions by the State of Colorado, U.S. EPA, and
NPS. (See "Nitrogen Deposition Reduction Plan" at http://www.cdphe.state.co.us/ap/rmnp.html)
D.5.1. Current Recommendations on Critical Loads Uses in the U.S.
The participants in the Multi-Agency Critical Loads Workshop developed a set of findings and
recommendations to help advance critical loads usage in the U.S. The "areas of agreement" published in
the workshop report (U.S. EPA, 2006c) included the following:
¦ A critical load is defined as: a quantitative estimate of the exposure to one or more pollutants
below which significant harmful effects on specific sensitive elements of the environment do not
occur according to present knowledge (Nilsson and Grennfelt, 1988).
¦ Despite reductions in S and N emissions in the U.S., deposition rates still exceed preindustrial
levels and acidification and eutrophication effects remain widespread.
¦ Critical loads can be used to better understand impacts of atmospheric deposition, assess the
effectiveness of emissions programs, and guide natural resource management.
¦ The development of critical loads is a process that is subject to continued development and
improvement as knowledge advances.
¦ Adequate information exists to move forward with the development and limited application of
critical loads in some regions and ecosystems in the U.S.
¦ An intensive research and monitoring agenda should be pursued to support the development and
refinement of critical loads in the U.S.
¦ Critical loads should be based on a matrix of biological and chemical indicators for aquatic and
terrestrial ecosystems that account for acidification, N saturation, and eutrophication effects and
are relevant to the geographic area or ecosystem of concern.
D-18
-------
¦ Adequate information exists to establish harmful effect thresholds for some indicators based on
specific protection and recovery objectives defined by policymakers and managers.
¦ Dynamic models provide the most accurate site-specific information and account for time-
dependent processes, but are generally too data intensive to be applied across large geographic
areas at present. Simple mass balance models can be applied to current conditions in large
geographic areas, but in some instances do not adequately highlight some sensitive areas because
they tend to average conditions across the landscape. Hybrid approaches that link observational
datasets with dynamic and steady state models represent a useful approach for regionalizing site-
specific information.
D.5.2. Questions and Limitations Regarding Critical Loads Uses in the
U.S.
The participants in the Multi-Agency Critical Loads Workshop also developed a set of "Questions
Needing Further Discussion" (U.S. EPA, 2006c):
¦ What are the appropriate applications of critical load estimates to policy and management issues
given current knowledge? For applications where buy-in to an incremental process does not exist,
greater investment in critical loads methods may be needed before this application can be
pursued.
¦ How strong is the relationship between specific indicators, thresholds, and biological responses?
¦ What are the suitable interpretations and uses of existing databases for the development of
national simple mass balance critical load models?
D.5.3. Critical Loads Research and Monitoring Needs
Finally, the participants in the Multi-Agency Critical Loads Workshop presented a list of "Critical
Loads Research and Monitoring Needs" (U.S. EPA, 2006c), which are summarized below.
D.5.3.1. Emissions and Deposition
¦ Update N and S emissions inventories on a state-by-state basis back to the 1900s to correspond
with methods used in current emissions inventories.
¦ Develop NH3 emissions inventory.
¦ Improve dry deposition estimates for S and N.
¦ Improve total S and N deposition estimates.
¦ Measure gaseous NH3 concentrations.
¦ Add NH3 deposition measurements to current networks.
¦ Improve estimates of total deposition in complex terrain.
¦ Develop N and S deposition maps for North America.
D-19
-------
D.5.3.2. Soils
¦ Improve spatial coverage and representativeness of soil chemistry databases, particularly in
sensitive terrain.
¦ Increase soil monitoring.
¦ Improve estimates of mineral weathering rates.
¦ Develop soil archiving and well characterized reference samples to promote cross-laboratory
comparisons.
¦ Expand research on the nature and size of soil nutrient pools.
¦ Conduct research on threshold values of soil quality for biologic responses.
¦ Determine N supply rates in different soil types.
¦ Investigate N soil accumulation rates in arid lands and implications for critical loads.
D.5.3.3. Surface Waters
¦ Incorporate TIME and LTM surface water monitoring programs into a larger network with better
geographic coverage (e.g., the West and Southeast).
¦ Improve spatial coverage and representativeness of surface water chemistry databases,
particularly in sensitive and complex terrain.
¦ Integrate fixed-site monitoring with regional probability monitoring design.
¦ Continue to monitor major drivers of acidity.
¦ Build critical loads considerations (e.g., validation, improvement, regionalization) into
monitoring from the start by combining chemistry, hydrology, deposition and biology, and
integrating site-specific models and measurements into regional contexts.
¦ Expand research to understand what is driving dissolved organic carbon (DOC) changes in the
East.
¦ Analyze the impact of groundwater transport on recovery times.
D.5.3.4. Biological Effects
¦ Develop better understanding of the link between chemical indicators and biological response
(e.g., quantify the minimum N level at which plankton communities shift).
¦ Conduct additional research on the sequential impacts of N and relationship between N
deposition and ecosystem impacts.
¦ Integrate critical load estimates with biodiversity and climate change interactions.
¦ Undertake more research on biological change and "harmful effects" to help establish appropriate
critical loads thresholds (e.g., in arid lands, what level of productivity of exotic invasive species
will cause the reduction versus the extinction of native species?).
¦ Collect sediment cores from lakes that vary in rates of N deposition to track changes in diatom
assemblages.
D-20
-------
D.5.
3.5. Critical Loads Models
¦ Improve representation of N dynamics in models.
¦ Expand models to include NH3.
¦ Improve explicit consideration of changing base cations and DOC.
¦ Conduct ground-truthing of forest sensitivity and other models.
¦ Integrate water flowpaths into nutrient cycling models since lateral and vertically upward
flowpaths are common.
¦ Understand and quantify uncertainties in models.
¦ Conduct site level model comparisons of dynamic and simple mass balance models.
¦ Integrate observational databases with steady state and dynamic models.
¦ Incorporate capacity to understand and evaluate climate change interactions.
Table D-2. Biological indicators for the effects of elevated N deposition and related empirical critical
loads for major ecosystem types (according to the Eunis classification) occurring in Europe.
Ecosystem Type
Biological Effect Indicators
Empirical
Critical Load
(kg N/ha/yr)
GRASSLANDS AND TALL FORB HABITATS (E)
Sub-Atlantic semi-dry calcareous
grassland
Increased mineralization, nitrification and N leaching; increased tall
grasses; decreased diversity
15-25
Non-Mediterranean dry acid and
neutral closed grassland
Increase in nitrophilous graminoids, decline of typical species
10-20
Inland dune grasslands
Decrease in lichens, increase in biomass, accelerated succession
10-20
Low and medium elevation hay
meadows
Increased tall grasses, decreased diversity
20-30
Mountain hay meadows
Increase in nitrophilous graminoids, changes in diversity
10-20
Moist and wet oligotrophic grasslands
Increase in tall graminoids, decreased diversity, decrease in bryophytes
10-25
Alpine and subalpine meadows
Increase in nitrophilous graminoids, changes in diversity
10-15
Moss and lichen dominated mountain
summits
Effects on bryophytes and lichens
5-10
HEATHLAND HABITATS (F)
Northern wet heaths
Decreased heather dominance, transition heather to grass, decline in
lichens and mosses
10-20
Dry heaths
Transition heather to grass, decline in lichens
10-20
Arctic, alpine, and subalpine scrub
habitats
Decline in lichens, mosses, and evergreen shrubs
5-15
D-21
-------
COASTAL HABItAT(B)
Shifting coastal dunes
Increased biomass, increased N leaching
10-20
Coastal stable dune grasslands
Increase in tall grasses, decreased prostrate plants, increased N leaching
10-20
Coastal dune heaths
Increase in plant production, increased N leaching, accelerated succession
10-20
Moist to wet dune slacks
Increase in biomasss and tall graminoids
10-25
MIRE, BOG, AND FEN HABITATS (D)
Raised and blanket bogs
Changed species composition, N saturation of Spagnum
5-10
Poor fens
Increased sedges and vascular plant, negative effects on mosses
10-20
Rich fens
Increase in tall graminoids, decreased diversity, decrease of characteristic
mosses
15-35
Mountain rich fens
Increase in vascular plants, decrease in bryophytes
15-25
FOREST HABITATS (G)
Mycorrhizae
Reduced sporocarp production, reduced below ground species
composition
10-20
Ground vegetation
Changed species composition, increased nitrophilous species; increased
susceptibility to parasites (insects, fungi, virus)
10-15
Lichens and algae
Increase in algae; decrease in lichens
10-15
Source: Achermann and Bobbink (2003). Reprinted with permission.
D-22
-------
Annex E. Effects of NOy, NHx, and SOx on
Structures and Materials
E.1. Introduction
The purpose of this chapter is to summarize the research published since the most recent AQCDs
on materials and structures damage caused by: (1) NO, N02, and their combination (NO + N02 = NOx);
the organic and inorganic reaction products of NOx (denoted as NOz); and the combination of NOx and
NOz (NOx + NOz = NOy), (2) the effects of NH3 and NHX, and of (3) SOx. Materials and structures
exposed to the environment are subject to damage from exposure to sunlight, moisture, salt, windblown
dust, and cycles of temperature and humidity, whether or not air pollutants are present. However, NOY,
NHX, and SOx air pollutants may cause such damage to be greater or occur more rapidly than with natural
environmental factors alone. Damage to materials and structures may be physical, potentially affecting
the durability or maintenance needs of a material or structure, or may be purely aesthetic, affecting only
the outward appearance of the material or structure. In the case of historical buildings, monuments, or
artifacts, aesthetic damage may be a relevant concern.
Note that very extensive work related to materials damage from acidic deposition related to S and
N was conducted in the 1980s as part of the National Acid Precipitation Assessment Program (NAPAP)
(NAPAP, 1991). The results of that work are well known to the U.S. EPA (1991) and so are not discussed
here. In compiling information for this chapter on NOY/NHx/SOx effects, the information presented in the
1993 NOxAQCD (U.S. EPA, 1993a) and the 2004 PM AQCD (U.S. EPA, 2004) was updated by
literature searches reaching back to approximately 1992. This update was based on peer-reviewed
literature, with a focus on studies that were conducted in the U.S., that evaluated effects at realistic
ambient air pollutant levels, and that treated NOY/NHx/SOx as components of a complex mixture of air
pollutants. These latter two factors result in an emphasis on studies done with exposures to ambient
atmospheric pollution, rather than exposures at high levels, e.g., in test chambers. The studies cited in this
chapter were selected from those found in a broad literature search based on criteria that they address
damage caused by exposure to atmospheric contaminants; focus on S and N containing species; provide a
clear link between pollutant concentrations and damage; and give complete information on methods and
data analysis used.
Broadly speaking, the pace of research on NOY, NHX, and SOx materials effects has slowed
considerably since the publication of the previous AQCDs. In particular, although the literature searches
conducted for this update emphasized studies conducted in the U.S., the great majority of the relevant
publications found originated in Europe or Asia. The relative scarcity of recent U.S. studies on structural
and materials damage from NOY/NHx/SOx may be a natural fall-off in research in this area, following the
extensive efforts that were summarized in the previous AQCDs and in the NAPAP report. Certainly the
greater number and age of aesthetically valuable buildings and archeological sites in Europe and Asia,
relative to the U.S., may be a driving force for current research in those geographic areas. In this chapter,
each discussion of the effects of NOY/NHx/SOx on a material type begins with a brief summary of the
state of knowledge as represented in the previous AQCDs, and then continues with a description of recent
research on that type of material.
E-1
-------
E.2. Environmental Exposures of Materials
E.2.1. Mechanisms of Materials Damage
As noted in the introduction to this chapter, materials damage may occur by natural physical
processes without the involvement of NOY/NHx/SOx air pollutants. When those pollutants are involved,
the destructive processes may be chemical, physical, or even biological. Chemical processes include
direct reactions with gaseous pollutants such as N02, SO2, or nitric acid (HN03), reaction with
electrolytes (proton (H+), ammonium (NH/), nitrate (N03 ). S042 . etc.) in water on material surfaces,
and reactions with chemicals in deposited particulate matter. An example of a physical process is the
deterioration of stone that occurs when gypsum (CaS04-2H20) forms from reaction of S02 with the
calcium carbonate (CaCOs) in the stone. The gypsum thus formed occupies a larger volume than the
original stone, causing the surface to deteriorate. Biological degradation can occur when deposited
pollutants are oxidized to acids by fungi or bacteria.
A key factor affecting damage to certain materials, primarily metals and stone, is the frequency and
duration of wetting of the surface. Liquid water on materials surfaces can dissolve deposited pollutants,
producing reactive electrolyte solutions, and can serve as a reaction medium in which S and N oxides are
converted to more damaging acids. Pollutants deposited on surfaces may contain or form hygroscopic
salts, which enhance the formation of liquid water and thereby increase materials damage. As Dubowski
et al. (2004) have shown, the deposition of HN03 onto surfaces can increase the extent of wetting of
surfaces, and promote the damaging effects of both HN03 and other pollutants.
E.2.2. Deposition Processes
Air pollutants come into contact with surfaces through both dry and wet depositional processes.
Dry deposition occurs in the absence of precipitation and is governed by factors such as atmospheric
turbulence, the chemical and physical properties of the pollutant (e.g., water solubility and reactivity for
gases; size, density, and shape for particles), and surface properties (e.g., reactivity, roughness, moisture
level, and pH). The deposition rate of a pollutant is proportional to the atmospheric concentration of that
pollutant. Dry deposition of gases depends primarily on the water solubility of the gas, the moisture level
on the surface, and the pH of the electrolyte formed on the surface of a material. Nitric acid and NH3 are
deposited very efficiently to most surfaces regardless of the surface properties of the material. Particle
size plays an important role in determining the rate of deposition of particles to a surface. For very small
particles, Brownian diffusion is the dominant deposition mechanism. For larger particles, inertial
impaction and gravitational settling are important deposition processes. Particles between 0.05 and 2 |_im.
which include most atmospheric particles containing N03~, S042 , and NH/, may have long atmospheric
lifetimes in the absence of moisture.
Wet deposition occurs when gas or particle species come into contact with moisture (as rain, fog,
snow, or ice). Atmospheric species can be dissolved into moisture and then deposited as the moisture falls
to the ground. Solubility and the chemical reactions of the dissolved species determine the degree of wet
deposition. For acid gases, high dissolution is observed due to the dissociation of the dissolved species in
water. Wet deposition of pollutants occurs at a faster rate than dry deposition, but is only an important
mechanism when moisture is present.
E-2
-------
E.2.3. Chemical Interactions of Nitrogen and Sulfur Oxide Species
N and S oxide species are subject to many atmospheric reactions in both the gaseous and
particulate phase. Emissions of S and N oxides are primarily in the form of gas phase SO2 and NOx. In
the atmosphere, these species can be oxidized by reaction with other atmospheric species to gas and
particle phase product species. On the surface of materials, the oxides are generally oxidized to their acid
forms (nitrous acid (HN02), HN03, sulfiirous acid (H2S03), and sulfuric acid (H2S04)), which then
dissociate to form nitrite, nitrate, sulfite, and sulfate ions. These acids are the primary species responsible
for damage to materials by S and N pollutants. NH3, the primary gaseous basic compound in the
atmosphere, can partly or completely neutralize these acids in particulate matter or in the aqueous phase,
forming NH44" ions.
On the surfaces of materials, N and S species can react to form a variety of degradation products.
On metals and stone, the possible degradation products include nitrite, nitrate, sulfite, and sulfate species
as well as minerals that incorporate nitrate or sulfate into a more complex composition. These degradation
products may be more or less reactive to further degradation than the original material. Degradation
products that are more reactive, or those that are soluble in water, do not have long lifetimes on a material
surface. They undergo further chemical reactions and are transformed to other species, or they are washed
off the surface by precipitation. Products which are less reactive and less soluble in water than the original
material may form a protective layer on the surface of the material which inhibits or prevents further
damage from atmospheric pollutants. Products that are more reactive or water-soluble than the original
material are readily removed, exposing the surface to more damage. The protectiveness of the products
formed depends on the complex mixture of species present and the physical/chemical properties of the
material.
Synergistic effects, which influence the rate of degradation of materials, are possible in
atmospheres containing a complex mixture of pollutants. NOx may enhance the oxidation of sulfite to
sulfate and lead to faster rates of corrosion. The deposition velocity of SO2 and NOx may be influenced
by the presence of HN03 deposited to the surface due to the increased degree of surface wetting.
E.2.4. Materials Damage Experimental Techniques
The NOy/NHx/SOx air pollutants are comprised of numerous distinct chemical species, which may
exist in the gaseous and/or particulate phases in the atmosphere, as well as in dissolved form in
atmospheric precipitation and in condensed water on surfaces. To test the damaging effects of
NOy/NHx/SOx species on man-made materials, it is often necessary to simplify the system by testing
under controlled laboratory conditions, typically with a very limited set of pollutants in a test chamber.
Such tests generally use pollutant concentrations that are greatly elevated relative to ambient atmospheric
levels, and may also use exaggerated temperature, humidity, or wetting, to accelerate the development of
materials damage so that it can be detected. Chamber tests may not accurately mimic the mass transfer of
pollutants in the atmosphere, and efforts in such tests to isolate the effects of one pollutant from the
complex mixture present in the atmosphere are unrealistic. As a result, chamber tests may provide
valuable information on potential effects and mechanisms involving ambient air pollutants, but cannot
accurately predict the corrosion rates or effects of such pollutants in real situations.
Exposing materials of interest to the ambient atmosphere for extended time periods can provide a
realistic look at the effects of air pollutants on materials. However, such ambient exposure tests are
limited by the occurrence of natural (i.e., non-air pollutant) materials damage, and by the complexity of
the NOy/NHx/SOx system. While it is relatively easy to determine which materials suffer more or less
damage during equivalent exposures to ambient air pollution, it is extremely difficult to determine which
air pollutants are responsible for the observed damage. This is due to the co-occurrence of all air
E-3
-------
pollutants simultaneously, complexities in accurately measuring the suite of NOY/NHx/SOx species, and
interconversions among species (e.g., SO) and S042 , NOx and HN03) related to contact with materials or
with moisture. The amount of time that surfaces are wet is a key factor in the extent of materials damage,
and this factor may be difficult to determine in ambient exposures, because the presence of air pollutants
themselves may enhance surface wetness on the microscale beyond that expected based on
meteorological conditions (Dubowski et al., 2004). Ambient exposure tests lead to retrospective analyses,
in which meteorological and air pollutant data, surface analyses, and measurements of chemical and
physical properties are evaluated statistically to estimate the impacts of air pollutants on the exposed
materials.
E.3. Effects on Dyes and Textiles
E.3.1. Fading of Dyes
The fading of dyes by N oxides has long been recognized, and dye manufacturers have worked to
produce products less susceptible to this effect, through both improved dye chemicals and the use of
inhibitors in dye formulations to minimize fading. Fading has been observed with both red and blue dyes,
both on natural fibers (e.g., cotton, silk, wool) and on synthetics (e.g., nylon, rayon, polyester). The fading
effect of N02 is generally reported to be greater than that of NO on various dyes with various fabrics. In
exposures of dyed fabrics to ambient air, test samples must be shielded from sunlight, to avoid the
substantial fading of dyes that results from sunlight exposure. Under such conditions, N02 and ozone (O3)
are often found to be about equally important in the fading of dyed fabrics.
E.3.2. Degradation of Textile Fibers
N oxides can degrade a variety of synthetic fibers, with the greatest effects seen with nylon. With
N02, the damage to nylon occurs due to breaking of the polymer chain (i.e., chain-scissioning). Similar
weakening of nylon has been observed in tests with elevated concentrations of HN03. A synergistic effect
was observed between mechanical stress and NOx in the degradation of oriented nylon-6 fibers (Smith
and DeVries, 1993).
E.4. Effects on Plastics and Elastomers
The group of materials called plastics includes a wide variety of polymeric materials such as
polyethylene, polypropylene, polystyrene, polyurethanes, acrylic polymers, phenolics, and fluorocarbon
polymers, among others. Plastic materials may include other components such as hardeners or
plasticizers, and fillers that may impart properties such as physical strength. Elastomers are polymers that
can stretch to at least their twice their normal dimensions and then return to their original dimensions
when the stress is removed. Examples of elastomers include various rubber formulations and neoprene.
Plastics and elastomers can be damaged by N02, SO2, and O3, as well as by UV radiation in sunlight, and
some studies have been designed to separate the effects of these factors.
Chamber studies at relatively high pollutant concentrations with sunlight or UV light have
generally shown greater damage than from the pollutants alone. N02 is damaging to a variety of polymers
E-4
-------
and elastomers, causing either chain-scissioning or cross-linking (formation of additional bonds between
polymer chains) depending on the polymer. Polypropylene is reported to be damaged more severely by
S02 than by N02. Elastomers are damaged more severely than plastics. In tests where light and N02 have
been present simultaneously, much of the damage observed in chamber tests has been attributed to O3,
produced by the interactions of the pollutants and UV light, rather than to N02 alone. In studies in which
the same pollutant concentrations are present both with and without light, the greater damage observed in
samples exposed to the light is often attributed to the light itself, when in fact chemical processes initiated
by light (such as the formation of O3) undoubtedly also play a part.
Cellulose nitrate can break down through hydrolytic, thermal, and photochemical reactions.
Addition of plasticizer to cellulose nitrate slows the degradation substantially. N02 is of particular interest
with regard to cellulose nitrate because it is not only capable of causing damage, but is also produced as a
result of damage to the material (Shashoua, 2006). N02 is formed when N-0 bonds connecting cellulose
rings are broken. The N02 formed will then further degrade cellulose nitrate, thus the degradation is an
autocatalytic process.
E.5. Effects on Metals
Metals are considered to be the materials most subject to damage from the NOY/NHx/SOx air
pollutants, and have been the subject of a great deal of research. The nature and concentration of the
pollutant, its rate of deposition, and especially the duration of wetting of the surface are key factors in the
corrosion of metals. Numerous studies have indicated corrosion rates of metal surfaces on the order of
1 to several micrometers per year ((.un/yr) under real or simulated atmospheric conditions.
Table E-l summarizes the materials tested, exposure conditions, and findings of recent studies
related to the effects of NOY/NHx/SOx air pollutants on metals. The studies listed in Table E-l are
discussed where applicable in the following sections.
E-5
-------
Table E-1.
Studies on corrosive effects of NOy/NH3/SOx effects on metals.
Materials
Exposure Conditions
Findings
Reference
Zinc
Mild Steel
Samples were exposed to fly-ash in clean air
and in air with S02 and/or HCI (presentation
rates of 27 x 10~6 and 4.7 x 10~6 mg/cm2s,
respectively). A synthetic acid rain solution
was used to model wet deposition.
Corrosion was found to depend on the
surface electrolyte irrespective of the
presence of particles. Inert particles were
found to increase corrosion rates in relatively
unpolluted atmospheres. With higher
pollution levels, species leached from
particulate matter contribute to the
conductance of the surface electrolyte and
thus increase the corrosion rate.
Askey et al.
(1993)
Copper
Copper samples were exposed to 264 ppb
NO2 in a laboratory setting. Exposures were
limited to 72 h to study the initial corrosion
behavior.
The corrosion rate of copper in the presence
of NO2 was much greater than in clean air.
The surface electrolyte was found to contain
predominantly nitrate with only trace levels of
nitrite. After 24 h, the electrolyte had become
sufficiently acidic to dissolve the copper
oxide layer. Once the copper oxide was
dissolved, corrosion proceeded at a
significantly faster rate.
Dante and
Kelly (1993)
Zinc
Zinc samples were exposed to S02 (0.78 ppm)
and/or NO2 (1.06 ppm) for 420 h. Some
samples were treated with NaCI before
exposure.
SO2 slowed the corrosion of zinc with
moderate to high surface concentrations of
NaCI due to the formation of sodium zinc
hydroxychloride sulfate. NO2 (which, alone,
is unreactive toward zinc) accelerated the
corrosion of zinc in the presence of small
amounts of NaCI.
Svensson and
Johansson
(1993b)
Galvanized
Iron
Zinc
Samples were exposed to SO2 or NH3 in the
laboratory. Air was supplied at 5 cm/s and the
pollutant gases at 3 cm/s. The concentrations
of the gases were very high to accelerate the
tests.
Corrosion rates in SO2 were found to be
largely dependent on relative humidity. No
such humidity dependence was observed for
corrosion induced by NH3. Corrosion rates in
both gases decreased sharply with time
(approaching steady state values after 30 h).
Dehri et al.
(1994)
Aluminum
Zinc
Samples were exposed to ambient air for
4 years at 6 sites. SO2 deposition rates ranged
from 10 mg/m2day to non-detectablelevels
across the sites. Time of wetness was also
measured at each site.
Corrosion products that developed in rural
environments were found to be easily
removed from the surface and thus result in
poor protectiveness. Corrosion products
formed in more aggressive environments
were found to be more protective against
continuing corrosion.
Vilche et al.
(1995)
Zinc
Samples were exposed at temperatures of 4,
14, 22, and 30 C with 95% relative humidity.
SO2 was supplied at 500 (± 5) ppb and 107 (±
2) ppb.
SO2 induced corrosion was found to be
inversely dependent on temperature. The
maximum corrosion rate (at 107 ppb SO2) of
11 mg/cm2d was observed at 4 C. The
corrosion rate at 30 C was 6.8 mg/cm2d.
Svensson and
Johansson
(1996)
E-6
-------
Materials
Exposure Conditions
Findings
Reference
Copper
Zinc
Aluminum
Samples were exposed in the laboratory to
SO2 (1.5 ppb, 0.5 ppm, 10 ppm), NO2
(10 ppm), NO (10 ppm), or O3 (10 ppm).
NO was found to have no effect on the
corrosion of copper, zinc, or aluminum.
Copper in the presence of SO2 (10 ppm) and
NO2 led to significant material loss initially
with a slowing of the rate with increasing
time. O3 was found to have the strongest
influence on the corrosion of copper. Only
very slight mass gains were observed for
copper exposed to 0.5 ppm SO2. A small
effect on zinc was observed for NO2 with
SO2 (10 ppm) resulting in the largest weight
gain. SO2 at 0.5 ppm had a much larger
effect on zinc than on copper. For aluminum,
O3 had the largest effect followed by SO2
and NO2.
Oesch and
Faller (1997)
Copper Powdered samples of copper patina
compounds (tenorite, cuprite, brochantite,
antlerite, and atacamite) were exposed to SO2
(476 ppb) alone or in combination with NO2
(450 ppb) or O3 (500 ppb). Some samples
were pretreated with carbon.
Tenorite reacted rapidly with SO2 to form
brochantite and other sulfate containing
products. Cuprite reacted slowly with SO2
alone but addition of O3 formed antlerite and
brochantite. NO2 did not produce the same
effect. For samples with carbon on the
surface, the oxidation reaction was greatly
enhanced. Brochantite and antlerite were
found to be stable in atmospheres with SO2
in combination with O3 or NO2.
Strandberg
(1998)
Copper Samples were exposed to atmospheres
containing approximately 200 ppb of SO2, SO2
and O3, or SO2 and NO2. O3 and NO2 were
introduced at different times in the exposure
scenarios.
Copper sulfite and cuprous oxide formed on
copper surfaces exposed to SO2. With O3
present, an increased rate of mass gain was
measured, and copper sulfite was converted
to copper sulfate. NO2 increased the mass
gain to a lesser extent than O3 and resulted
in the formation of copper nitrate in addition
to copper sulfate.
Aastrup et al.
(2000)
/lild Steel Steel samples were exposed to the
atmosphere in 47 marine atmospheres with
varying levels of chloride and SO2.
Atmospheres were separated for data analysis
based on chloride and SO2 deposition rates.
Samples exposed at sites with moderate
SO2 and chloride deposition rates formed
compact, rounded corrosion structures.
Samples at sites with high SO2 and
moderate chloride exhibited cracking in the
corrosion products. Samples at sites with
high chloride and moderate SO2 exhibited
the highest corrosion rates of the mixed
atmospheres. The 1 site with high chloride
and high SO2 exhibited a lower corrosion
rate than expected.
Almeida et al.
(2000)
E-7
-------
Materials
Exposure Conditions
Findings
Reference
Steel
Two types of steel were exposed to urban-
industrial and rural atmospheres for 20 years.
Avg SO2 concentrations were 90 and
<10 |jg/m3 (34 and <4 ppb) for urban and rural
environments, respectively.
The corrosion rates of the two grades of
steel were similar with values of 0.1 and 0.08
mm/yr for the urban and rural environments,
respectively. The initial rate of corrosion was
significantly faster and steady state values
were approached after 4000 days exposure.
The similar corrosion rates measured for the
samples were thought to be due to similar
time of wetness at both sites.
Damian and
Fako (2000)
Copper
Zinc
Samples were exposed in the field at 8 sites.
SO2, NO2, and O3 concentrations were
monitored over a 4-yr exposure duration at
each of the sites.
The highest corrosion losses for copper were
observed at the site with the highest
combination of SO2 and O3. For zinc, the
highest corrosion losses were observed at
the site with the highest SO2 concentration.
Both metals showed a decrease in corrosion
rate with time. Runoff rates from copper were
much smaller than from zinc. 90% of the
corrosion products remained on the copper
surface after 4 years; only 40% of the zinc
corrosion products remained after 4 years.
Leuenberger-
Minger et al.
(2002)
Nickel
Nickel samples were exposed in the field at 3
sites (urban, industrial, and rural) for 1 year.
Concentrations of NO (41.1, 9.7, and
2.9 |jg/m3) (33, 8, and 2 ppb), NO2 (50.1, 24.2,
and 8.7 pg/m3) (26,13, and 5 ppb), SO2 (22.3,
29.0, and 12.2 pg/m3) (8,11, and 5 ppb), and
03 (25.8, 47.1, and 60.1 pg/m3) (13, 24, and
30 ppb) were measured at the urban,
industrial, and rural sites, respectively.
Mass loss rates of 320, 570, and 200
|jg/cm2y were determined for urban,
industrial, and rural environments,
respectively. Mass loss was found to
increase with increasing SO2 concentration.
Soluble corrosion products were formed on
the surface and then removed by rainfall
events. Hydrated nickel sulfates were the
main corrosion products formed on the nickel
surface.
Jouen et al.
(2004)
Iron
Samples were exposed to humidified air in the
laboratory. Samples were exposed to clean
air, SO2 (200 ppb), SO2 and NO2 (each
200 ppb), or SO2 and O3 (each 200 ppb). The
same exposure conditions were used for iron
samples with NaCI deposited on the surface.
No corrosion products were detected on
samples exposed to humidified air alone.
The addition of SO2 alone was not enough to
initiate a change in corrosion behavior of the
samples. When an oxidant (NO2 or O3) was
added to the humidified air/S02 system, a
significant increase in corrosion rate was
observed. SO2 was found to inhibit the NaCI
induced corrosion of iron, but the
combination of SO2 and NO2 was found to
accelerate NaCI induced corrosion.
Weissenrieder
etal. (2004)
E-8
-------
Materials
Exposure Conditions
Findings
Reference
Aluminum Aluminum samples were exposed to SO2
(96 ppb) either alone or in the presence of
other pollutants (NaCI, NO2, or O3).
SO2 alone resulted in the loss of metallic
luster. 50% of the surface had developed
corrosion products after 672 h. Samples
exposed to NaCI alone showed significantly
larger mass gain than samples exposed to
only SO2. The combination of SO2 and NaCI
resulted in the largest mass gain (fastest
corrosion rate). While the rate of mass gain
was highest with a combination of SO2 and
NaCI, the pitting observed in the presence of
NaCI alone was significantly reduced. O3
was found to slightly increase the deposition
rate of SO2, no effect on SO2 deposition rate
was observed for NO2.
Blucher et al.
(2005)
Copper
Zinc
Steel
Samples were exposed to HNO3 (50-180 ppb)
in a laboratory exposure chamber. Tests were
conducted at 65% and 85% relative humidity.
The corrosion effects of HNO3 on carbon
steel were larger than on zinc or copper. The
corrosion effect of HNO3 was found to be
larger than corrosion from SO2 alone or a
mixture of SO2 with O3 or NO2. No increase
in corrosion was observed at 85% relative
humidity compared to 65% relative humidity.
Samie et al.
(2007)
E.5.1. Role of NOy, NHx, and SOx in the Corrosion Process
In the atmosphere the NOY/NHx/SOx pollutants occur together, along with other pollutants such as
03 or chloride salts. While wetting of metals surfaces is the single greatest factor promoting corrosion, an
important observation is the enhanced damage that occurs due to interactions among this mixture of
pollutants. It must be noted that in many studies the various NOY/NHx/SOx species have not been
adequately separated or quantified, and this may be the cause of conflicting observations from some
studies. However, some generalizations can be made. Sulfur and chloride pollutants are generally more
important at causing metals corrosion than N pollutants, however NOx (or NOY) and SO2 together have
been shown to be more damaging than SO2 alone. The combination of N02 and SO2 has been shown to
result in a synergistic effect where the total damage from the mixture is greater than the additive damage
from the two pollutants separately (Svensson and Johansson, 1993a). This effect may be due to enhanced
wetting of the surfaces caused by NOY pollutants, resulting in corrosion at lower relative humidities than
would otherwise be the case. This enhancement has been attributed to the formation of hygroscopic
nitrate salts, but may also be caused directly by the deposition of gaseous HN03 onto the surface
(Dubowski et al., 2004). The corrosion effect of HN03 on zinc, copper, and steel is larger than that of S02
alone or a mixture of SO2 and N02 (Samie et al., 2007).
Although deposition of NOY/NHx/SOx species in particulate matter can soil metal surfaces, such
deposition does not directly result in substantial metals damage. However, under wet conditions these
soluble species form an electrolytic solution that can cause corrosion. Corrosion of steel and zinc has been
found to depend on the surface electrolyte irrespective of the presence of particles (Askey et al., 1993).
Temperature has been found to have a complex effect on metals corrosion. Lower temperatures
tend to increase surface wetness, but decrease the diffusivity of gaseous pollutants, and may reduce the
E-9
-------
rates of some reactions that convert SO2 and NOx to sulfuric and nitric acids. Thus, the effect of
temperature changes on long-term corrosion rates can be hard to predict.
E.5.2. Effect on Economically Important Metals
Steel is the most common and economically important structural metal, and is often used in
galvanized form (i.e., with a protective coating of zinc). SO2 is generally reported to be more corrosive
than NOx, however for well-protected steel the effects of ambient air pollutants are usually a small
increment on top of the natural weathering process. The N pollutants can have an enhancing effect on the
corrosion caused by the S pollutants. This is attributed to the increased wetting that can result from the
presence of hygroscopic N03 salts. Relative humidity has been shown to be very important in the
corrosion of steels by SO2 with much slower corrosion rates observed when the relative humidity is below
70% (Dehri et al., 1994). The presence of SO2 has been shown to reduce the corrosion pitting of iron
induced by sodium chloride (NaCl) but there may be an overall synergistic effect among S02, N02, and
NaCl (Weissenrieder et al., 2004). Steel corrosion rates have been shown to decrease over time (Almeida
et al., 2000; Damian and Fako, 2000) approaching steady state rates after approximately 4000 days
(Damian and Fako, 2000).
Zinc corrosion has been shown to be inversely dependent on temperature (Svensson and Johansson,
1996). Corrosion products formed on zinc in polluted environments are less water soluble, and therefore
more protective against further corrosion, than corrosion products formed in clean environments (Vilche
et al., 1995). The combination of SO2 and N02 showed synergistic (i.e., greater than simply additive)
corrosive effects on zinc (Svensson and Johansson, 1993a). SO2 slowed the NaCl induced corrosion of
zinc while N02 accelerated that corrosion (Svensson and Johansson, 1993b).
Aluminum is naturally protected from corrosion by a formation of a durable surface film, but some
effects of the NOY/NHx/SOx pollutants have been observed. Minimal damage is caused to Al by NOx.
The mixture of SO2 and NOx is variously said to be either more or less corrosive to Al than S02 alone.
The deposition rate of SO2 to Al was shown to increase in the presence of O3, but no effect on SO2
deposition rate was found for N02 (Blucher et al., 2005). Interaction between SO2 and NaCl results in an
increased corrosion rate but decreased pitting of Al compared to NaCl alone (Blucher et al., 2005). Oesch
and Faller (1997) found that SO2 is more corrosive to Al than N02 and that there is no difference in Al
corrosion rate when exposed to NO or clean air.
Mixtures of NOx and SO2 are more corrosive to copper than either pollutant alone. When hydrogen
sulfide (H2S) and 03 were also evaluated for damage to copper, they also were found to be more
damaging to copper than NOx. The corrosion rate of copper exposed to SO2 or N02 has been shown to
slow over time (Leuenberger-Minger et al., 2002; Oesch and Faller, 1997). The corrosion rate of copper in
the presence of N02 is greater than in clean air. In the first 24 h of exposure to N02, an acidic electrolyte
is formed on the surface that dissolves copper oxides and results in an increased corrosion rate (Dante and
Kelly, 1993). Synergistic effects have been seen between SO2 and O3 (strong) and N02 (weak) (Aastrup
et al., 2000). High corrosion rates were observed for field exposures of copper at sites with a combination
of SO2 and O3. After four years of exposure, 90% of the corrosion products formed on copper remained
on the surface (Leuenberger-Minger et al., 2002). The copper hydroxy sulfates brochantite and antlerite
are stable copper corrosion products formed in the presence of SO2, O3, and N02 (Strandberg, 1998).
Nickel is also damaged more severely by SO2 or chloride salts than by NOx. Nickel samples
deployed at urban, industrial, and rural sites showed that corrosion rates increase with SO2
concentrations. Soluble hydrated nickel sulfates were the main corrosion products and are easily removed
from the surface by rainfall events, thereby exposing the underlying surface (Jouen et al., 2004).
E-10
-------
Kim et al. (2004) conducted a study of the effects of ambient SO2 and N02 on steel, bronze,
copper, and marble at sites in China, Korea, and Japan. Both sheltered and unsheltered samples were
exposed with the corrosion rates of the unsheltered samples higher in all cases. The corrosion rate of steel
was the highest, followed by marble, bronze, and copper. Higher corrosion rates (especially for
unsheltered samples) were found to be correlated with high S02 concentrations.
E.5.3. Effects on Electronics
The increasingly wide penetration of electronic devices into daily life offers greater opportunities
for environmental damage to sensitive components. The hardware of communication systems may be
exposed to pollutants in outdoor air, and the ubiquitous cell phones may be exposed in both indoor and
outdoor environments. Sulfur and N oxides have been shown to corrode the metallic contacts in electronic
equipment, which are often made of copper or brass coated with a precious metal such as gold, palladium,
or nickel. Such materials are corroded more by N02 than by S02, but a mixture of these two pollutants is
more corrosive than either alone. The combination of SO2 and H2S is also less damaging than either N02
alone or a combination of N02 and these pollutants. N02 is also moderately corrosive to solder in
electronic components.
E.6. Effects on Paints
Painted surfaces are extremely common as a means of preventing damage to other materials, and
may be categorized as architectural coatings (e.g., house paint), product coatings (e.g., automobile
finishes), and special-purpose coatings (e.g., bridge paint). Environmental damage to painted surfaces is
expected, and periodic repainting is normal, but any factor that causes more rapid degradation or
discoloration of paints will require more frequent repainting and thus result in higher costs. Paint
formulations may differ widely for different applications, so the extent of air pollution damage in a given
application cannot necessarily be predicted from published information. Previous work at elevated
pollutant concentrations has shown that oil-based house paint is readily damaged by SO2 and moisture,
and is more subject to damage by S02 than by N02. Sample weights increased with increasing N02, but it
is not clear if this indicates direct reaction of N02 or an enhancement of the effects of SO2 and moisture
by N02. The effect of SO2 may be due to reaction with CaCC>3 and zinc oxide (ZnO) present in the paint.
Tests with various paints showed that NOx becomes incorporated into the paint surface upon long
exposure, apparently by reaction with polymers that make up the cured paint. In other tests HN03 was
found to produce substantially more damage to paints containing both low and high levels of carbonate
(CO3 ) than did an equal mixing ratio of N02.
Grosjean et al. (1994) studied the fading of colorants on cellulose paper. Twelve colorants were
exposed to various atmospheres including purified air, N02, SO2, and a mixture of oxidants (O3, N02, and
peroxyacetyl nitrate [PAN]). Not all colorants were tested in each atmosphere. The mixture of oxidants
resulted in the largest color change for each tested colorant. Only one colorant was exposed to SO2, N02,
and the mixture. For that colorant, the color change induced by the mixture of oxidants was
approximately three times the color change with N02 alone; the color change with SO2 was
approximately 70% of the color change with N02. An increase in relative humidity results in increased
fading of colorants (Grosjean et al., 1993, 1994). Of 35 colorants exposed to a mixture of O3, N02, and
PAN, nine exhibited substantial color changes and three exhibited moderate color changes (Grosjean
et al., 1993).
E-11
-------
Paint samples were exposed to UV light, N0X, SO2, and a combination of the treatments by
Colombini et al. (2002). The exposure conditions were chosen to produce accelerated aging of paint
samples. Non-pigmented paints were chosen to isolate degradation of the paint binder from synergistic
effects with pigments. Exposure to the combination of treatments resulted in increased cross-linking in
the paint binder as well as formation of organic acids. An examination of paint samples taken from
naturally aged paintings confirmed the presence of organic acids as degradation products.
E.7. Effects on Stone and Concrete
The effects of NOY/NHx/SOx air pollutants on stone and concrete are undoubtedly the most widely
studied because of their impact on historic buildings, monuments, and archeological treasures. Table E-2
summarizes studies of the effects of NOY/NHx/SOx on stone, concrete, and mortars. Calcareous stone
(i.e., that consisting of CaCC>3, such as marble, limestone, and cement) is most susceptible to damage.
Mortar used in stone construction is often more porous than calcareous stone, and therefore more subject
to damage. The damage to such materials is attributed primarily to the effect of SO2 in forming gypsum
(CaS04=2H20):
CaC03 + S02 +2 H20 >CaSO42H20 + CO2
The gypsum thus formed occupies a larger volume than the original carbonate, so the stone surface
becomes pitted and damaged. Gypsum is also more soluble than the carbonate, so it can be removed by
precipitation, exposing the surface to further reaction and damage. As a result, dry deposition of SO2 to
the stone surface between rain events is important, as it causes continued damage. The reaction of SO2
with calcareous stone is more energetically favorable than the reaction of N oxides, and thus SO2 is the
primary cause of damage to stone, however the combination of SO2 and NOx is more damaging than SO2
alone. This effect may be due to enhanced wetting of the stone, to oxidation of SO2 by the N oxides, or to
formation of calcium nitrate (Ca(N03)2), which is much more soluble than CaCC>3 and is easily washed
off the stone surface by precipitation. Removal of the nitrate salts in this way may result in
underestimation of the role of N oxides in stone damage when surface layers are analyzed for chemical
composition.
E-12
-------
Table E-2.
Studies on corrosive effects of NOy/NH3/SOx on stone.
Materials
Exposure Conditions
Comments
Reference
Marble Laboratory exposure to HNO3 ranging from 54 to Marble was found to be a very good sink for HNO3. The
4174 |jg/m3 (21 to 1603 ppb). Field exposures extent of corrosion by sulfates and nitrates were found to be
were conducted in several stages in Greece. Avg of the same order of magnitude. Corrosion was found to be
concentrations for field exposures were 1.41 |jg/m3 caused by acid and salt species (HNO3, H2SO4, etc.) on the
(0.5 ppb) HNO3,2.39 |jg/m3 (3 ppb) NH3,4.84 surface rather than the oxides (SO2, NOx).
|jg/m3 NO3", 14.61 |jg/m3 S042-, and 5.01 |jg/m3
NH4+.
Sikiotis and
Kirkitsos
(1995)
Limestone Samples of differing thickness were exposed in the
field and runoff water collected for 5 mos. The avg
SO2 concentration was 60 |jg/m3 (23 ppb).
Damage functions were developed to try to determine the
ionic sulfate content in runoff water from the ambient SO2
concentration. It was determined that the ambient SO2 con-
centration alone does not determine the sulfate concentra-
tion in runoff water.
Torfs and van
Grieken (1996)
Marble Laboratory exposure to 10 ppm of SO2 and NO2.
Field exposure to either dry deposition or dry and
wet deposition in Louisville, Kentucky. Avg concen-
trations of 10 ppb SO2 and 25 ppb NO2 for field
exposures.
Gypsum crust thickness of 1.9 |jm for rain sheltered sam-
ples after 1 year of exposure. SO2 was found to be the
dominant factor in crust formation. Surface recession due to
rain washing was 14.5 pm/yr and due to dissolution of the
gypsum crust as well as dissolution of the original marble.
Yerrapragada
etal. (1996)
Marble
Brick
Samples exposed in the field for 8 days with multi-
ple fog episodes. Over 3 measurement campaigns,
the mean pollutant concentrations were 17.2 |jg/m3
(7 ppb) SO2, 265 |jg/m3 (139 ppb as NO2) NOx,
and 131 |jg/m3 suspended particles.
For all samples, gypsum was the only stable mineral formed
following exposure to fog water in a polluted environment.
Exposure to fog water may be a significant cause of corro-
sion for materials sheltered from rainwater, but is of lesser
importance if a material is exposed to rain.
Del Monte and
Rossi (1997)
Marble
Limestone
Mortars
Samples were exposed to urban environments in
both sheltered and unsheltered configurations.
Pollutant concentrations were not reported.
Sulfation was the primary damage mechanism and was
more intense on mortars than on stones due to higher
porosity. Higher concentrations of degradation products
were found on samples sheltered from rain than on samples
exposed to rain.
Zappia et al.
(1998)
Jaumont Samples were exposed to 340 |jg/m3 (125 ppb)
limestone SO2 and 98 |jg/m3 (50 ppb) NO2 in the laboratory.
Samples were either exposed naked or sprinkled
with fly-ash particles. Field exposure was con-
ducted for 1 year with samples sheltered from
rainwater. Avg SO2 concentration during the field
exposure was 107 |jg/m3 (40 ppb).
Sulfation (gypsum formation) was found to proceed with Ausset et al.
greater intensity for samples sprinkled with fly-ash than for (1999)
naked samples. The growth of gypsum crystals fixed the fly-
ash to the surface of the limestone. Fly-ash was found to be
an important factor in crust formation by facilitating gypsum
crystal formation. Fly-ash particles also darken gypsum
crusts from gray to black.
Calcium Powdered calcium carbonate was exposed to
carbonate 10 ppm SO2 and 90% relative humidity for
124 days.
Reaction between calcium carbonate and SO2 was found to
take place in a liquid film on the calcium carbonate surface.
The presence of several different types of airborne particles
was found to increase the extent of sulfation by 20%. The
SO2 concentration used is unrealistic for ambient conditions.
Boke et al.
(1999)
Mortars A wide range of mortar and plaster samples were
collected from sites throughout Europe. Pollutant
concentrations over the life of the buildings were
not reported.
Sulfation was the primary damage mechanism observed. Sabbioni et al.
Sulfite was found as an intermediate damage product in the (2001)
sulfation process. Ettringite was also found as a secondary
damage product due to a reaction between gypsum and
calcium aluminum hydrates.
E-13
-------
Materials
Exposure Conditions
Comments
Reference
Mortars Plaster and mortar samples were collected from
buildings in the Old Venice Arsenal. Pollutant con-
centrations over the life of the building were not
reported.
Gypsum was found to be the primary damage product on all
of the mortars sampled. A secondary damage mechanism
was found where gypsum reacts with calcium aluminum
hydrates to form ettringite, an insoluble sulfate. The pres-
ence of sulfur in the damage products indicates SO2 as the
most aggressive atmospheric pollutant toward mortars.
Sabbioni et al.
(2002)
Concrete Samples were collected from the interior of a tunnel
in Italy. The tunnel formerly held a railway and
currently houses a road with heavy automobile
traffic. SO2 levels have declined from 350 |jg/m3
(132 ppb) in 1970 to -10 pg/m3 (4 ppb) in 2002.
NOx concentrations have remained relatively con-
stant around 100 |jg/m3 (52 ppb as NO2) over the
same time period.
The urban mixture of pollutants (SO2, NOx, CO2, and parti-
cles) results in formation of dendritic crusts on concrete.
Nitrates were found to be present in the largely gypsum
crusts. Soot particles were found embedded in the crusts as
well. Low quality starting materials provide a more porous
media that is more susceptible to degradation by atmos-
pheric pollutants. The degradation of concrete is more simi-
lar to that of sandstone than of limestone.
Marinoni et al.
(2003)
Marble Samples exposed to atmosphere and sheltered
from rain at 4 sites. SO2 concentrations ranged
from ~2 to 20 ppb across the sites.
Sulfation was the primary damage mechanism observed.
Marble containing dolomite was less sensitive to SO2 than
calcite marble. For relative humidity greater than 72%,
humidity was an important factor in determining the sulfation
rate.
Lan et al.
(2005)
Limestone
Sandstone
Samples were collected from 3 facades of a histori-
cal building in Spain. Samples were collected from
the surface as well as 5 mm below the surface to
determine degradation and original compositions,
respectively. Pollutant concentrations over the life
of the building were not reported.
The main decay products on the surface were found to be
nitrate compounds. Samples with black crusts on the sur-
face were found to have predominantly gypsum and soot,
but nitrate compounds were identified in the crusts as well.
Sandstone samples were much more damaged than lime-
stone samples due to their higher porosity.
Martinez-Ark-
arazo et al.
(2007)
Concrete is more susceptible to damage from N oxides than are the calcareous stones, because
concrete contains calcium hydroxide (Ca(OH)2), which can react to form calcium nitrate (Ca(N03)2). This
product is soluble and can be washed out of the concrete, weakening the material.
Deposition of particulate matter onto stone primarily results in soiling of the stone, due to the
elemental carbon and organic compound content of the deposited particles. The orientation of the surfaces
and size of the particles affect deposition: vertical surfaces are more affected by deposition of fine
particles, whereas horizontal surfaces are more affected by large particles. The fine particles carry the
bulk of the carbon and organic material. Metal oxides present in deposited particles may enhance the
reaction of S02 to form gypsum.
Gypsum crusts form more readily in rain-sheltered environments than on rain-washed stone
surfaces (Zappia et al., 1998). The presence of fog water has been shown to increase the rate of gypsum
formation on surfaces sheltered from rain washing (del Monte and Rossi, 1997). Gypsum crust formation
has been shown to proceed at a faster rate when the stone surface is sprinkled with fly-ash particles. In
addition, fly-ash (or other carbonaceous particles) can become entrained in the gypsum matrix and affixed
to the stone surface. Normally, gypsum crusts are gray in color, but when carbon containing particles are
entrained, they become black (Ausset et al., 1999). While gypsum crusts are composed primarily of
sulfates, they have been found to contain nitrate compounds as well (Marinoni et al., 2003; Martinez-
Arkarazo et al., 1999). The inclusion of nitrates in gypsum crusts suggests that N oxides, as well as S02,
play a role in the degradation of stone exposed to the atmosphere. The presence of particulate matter, in
addition to S02, has been shown to increase gypsum formation by 20% (Boke et al., 1999). Ambient S02
concentrations alone are not adequate to predict the degree of damage to stone samples (Torfs and Van
Grieken, 1996).
E-14
-------
Dolomite has been shown to be less sensitive to sulfation than calcite (Lan et al., 2005). Corrosion
of marble due to S species has been found to be of the same order of magnitude as that caused by N
species. Damage is caused not by the gas phase oxides (S02, N02, NO) but by acid (H2S04, HN03) and
salt (S042 , N03 ) species present in the electrolyte which forms on the marble surface (Sikiotis and
Kirkitsos, 1995). The rate of marble surface recession by rain washing is faster than the rate of gypsum
crust formation due to dry deposition of S and N containing pollutants. Marble is damaged by rain
washing through two mechanisms, dissolution of the gypsum crust and dissolution of the underlying
marble. The gypsum crust is more soluble in water than marble and is rapidly dissolved in rain. The
naturally occurring acidity in rainwater from the dissolution of carbon dioxide (C02) is an important
mechanism by which stone samples are degraded, but additional acidity from the dissolution of SO2 and
N02 in rainwater does not greatly increase the solubility of marble in rainwater (Yerrapragada et al.,
1996).
While gypsum is the primary degradation product found on stone, mortar, and concrete samples,
other damage products do occur. Sulfite species have been found on mortars as intermediate damage
products. On mortars, a secondary damage mechanism exists in which gypsum reacts with calcium Al
hydrates present in the mortar to produce ettringite (3CaO = AI2O3 = 3CaS04 = 32H20). Ettringite is an
insoluble sulfate that may cause damage by expansion and lead to cracking of mortars (Sabbioni et al.,
2001,2002).
E.8. Effects of NOx on Paper and Archival Materials
The cellulose fibers that make up paper are reactive with N02 and other NOY species, and storage
condition standards have been set regarding accepTable levels of NOx for archives, libraries, and
museums. Exposure of archival materials to NOY species in such facilities can arise from normal outdoor
or indoor sources, but also from generation of such NOY species from the materials themselves.
Specifically, stored materials that include cellulose nitrate, e.g., in the form of photographic film,
adhesives, or recording media, can slowly decompose to release NOx and product species such as HN03.
These emissions can degrade archival materials, and even be a safety hazard if allowed to accumulate. In
terms of outdoor air pollutants, it is likely that HN03 is a key reactant in the degradation of paper
archives. The rapid deposition velocity of HN03 and the numerous surfaces in archival facilities provide
opportunity for attack by HN03, and probably result in the effects of HN03 being underestimated, relative
to those of NOx, based on indoor air measurements. Artists' pigments can also be damaged by extended
exposure to ambient atmospheric N02.
The effects of SO2 and O3 on paper were studied by Johansson and Lennholm (2000). The
deposition rate of SO2 to fresh paper was found to decrease rapidly with time and approached steady state
after ten hours. The deposition rate of S02 to fresh paper in the presence of 03 was found to be elevated
compared to SO2 alone. The deposition rates to aged paper were much lower and there was no effect on
the SO2 deposition rate observed in the presence of O3. The decrease in deposition rate with time is
thought to be due to protonation of all available carboxylate ions to carboxylic acid.
Pigments in works of art can be degraded or discolored by atmospheric pollutants. H2S has been
shown to react with both copper and lead pigments, but only lead white has been seen to darken over time
(Smith and Clark, 2002). A synergistic effect has been detected between N02 and both benzene and
toluene resulting in an increased rate of attack on pigment oxides (Agelakopoulou et al., 2007).
Deposition of S to the surface of paintings, either as SO2 or ammonium sulfate ((NH4)2S04) particles, can
damage, varnish, or cause discoloring of paint (Gysels et al., 2004). Paint models subjected to accelerated
aging in SO2 (10 ppm) and NOx (10 ppm) as well as UV radiation for 15 days exhibited a variety of
E-15
-------
damage markers. Both nitrate and sulfate damage mechanisms were observed with sulfation sometimes
masking other processes (Arbizzani et al., 2004).
E.9. Costs of Materials Damage from NOy, NHx, and SOx
Materials exposed to the ambient atmosphere are degraded and damaged through a number of
mechanisms. Damage associated with air pollutants result in effects such as decreased usable lifetime,
increased maintenance frequency, and loss of aesthetic appeal. It is difficult to separate the costs
associated with air pollutants from costs associated with other damage mechanisms. Some estimates of
cost have been based on empirically derived dose-response functions for specific materials. Other
estimates have been developed using inspection of actual materials damage and maintenance guidelines
for the materials (Cowell and Apsimon, 1996). Estimation of costs over large geographic areas is subject
to considerable uncertainty due to unknown distribution of materials at risk and spatial variations in
pollutant concentrations. A cost estimate for material cost savings from SO2 emission reductions in
Europe was performed by Cowell and Apsimon (1996). In this study, the cost savings of theoretical future
emission reductions across Europe was modeled using cost data extrapolated from a study conducted in
three Norwegian cities. Total theoretical S02 reductions of 15,904 kilotons per year resulted in modeled
annual cost savings of $9,504 million total for Europe (Apsimon and Cowell, 1996).
E.10. Summary
Many types of materials, including metals, electronics, plastics, paints, stone, and paper may be
damaged by atmospheric NOY/NHx/SOx species. Damage occurs due to dry and/or wet deposition of the
pollutants onto the surface of a material and subsequent formation of an electrolytic solution in water
present on the surface. At low relative humidity, when little water is present on surfaces, damage rates
have been observed to be much lower, and in some cases, no damage has been observed. Both S02 and
N02 have been implicated in damage processes for different materials. In general, damage to materials by
S02 is greater than by N02. Little work has been conducted to investigate the effects of NO on material
damage. What work has been conducted shows no damage, or very minor damage for NO containing
environments compared to clean air. Synergistic effects between S02 and N02 lead to increased damage
rates for the gases in combination. Other species such as O3, NaCl, organics, or particulate matter have
also been shown to have synergistic effects with S02 and N02. The corrosive effects of nitric acid have
been found to be stronger than effects of other NOY/NHx/SOx species. Costs associated with damage to
materials by atmospheric pollutants are difficult to estimate because of the many sources of uncertainty in
the estimation process. For heavily polluted environments, the cost savings due to decreased rates of
material degradation could offset a significant portion of the costs to reduce emissions. In general, for
polluted environments, reductions in S02 or HN03 concentrations will reduce damage rates more than
reductions in N02, NO, or NH3. In areas with low S02 concentrations, reductions in N02, O3, or
particulate matter concentrations may reduce damage rates.
E-16
-------
Annex F. Valuation of the Environmental
Effects of N and S (non-materials)
F.1. Introduction
The monetary valuation of ecological effects associated with NOx and SOx emissions starts with
natural science endpoints. These may be things that people value directly, such as loss of a particular
species, or some remote effect on a resource that is not clearly understood by the general public or not
valued by the public for itself, such as forest soils. Of course, damage to forest soils will affect the
terrestrial ecosystem in ways that may be valuable to humans, such as tree growth, habitat, and even the
aesthetics of the forest. This Annex is a review of the literature that estimates such values for various
ecosystem endpoints or that provide values for effects that can be reasonably inferred from what is
provided.
The purpose of this Annex is to provide an assessment of the economics literature on the effects of
NOx and SOx emissions on terrestrial, transitional, and aquatic ecosystems.
F.1.1. Valuation in the Context of NOx and SOx
Figure F-l provides a schematic representation of how economic valuation is derived from changes
to NOx and SOx secondary standards. Starting at the upper left-hand side, the NOx and SOx standards are
set and emissions reductions occur to change the ambient concentrations ofNOx and SOx. Reading down
from "Change in Ambient Concentrations," these reductions will lead to changes in a variety of ecological
endpoints (as identified in the ISA) in terrestrial, transitional, and aquatic ecosystems. The box below,
"Change in Economic Endpoints," refers to physical endpoints that people care about, in which changes
can be valued (at least in principle) in monetary terms. Many times, these are referred to as ecosystem
services. In a few cases, such as agricultural crop growth and yield, ecological and economic endpoints
are nearly the same. Finally, at the bottom of this diagram is a box labeled "Valuation Methods," which
notes alternative approaches for placing monetary values on these economic endpoints. As endpoints are
discussed in detail in the ISA, this Annex focuses solely on valuation.
F.1.1.1. Ecosystem Services
Broadly defined, ecosystem services are the benefits that people obtain from ecosystems
(Millennium Ecosystem Assessment, 2003). In the Millennium Ecosystem Assessment (MA), ecosystem
services are classified into provisioning, regulating, supporting, and cultural services. Provisioning
services denote the products people obtain from ecosystems; regulating services are associated with the
ecosystem functions that regulate climate, nutrient cycle, water filtration, and so forth; supporting services
are ecosystem functions, such as primary productivity and production of 02, that support the provision of
ecosystem services; and cultural services are the non-material benefits ecosystems provide to people
through spiritual enrichment, cognitive development, reflection, recreation, and aesthetic experiences.
Ecosystems are productive systems in which various biological and physical factors, as well as
their interactions, serve various functions in the production of ecosystem services. However, economic
valuation of the environment has focused mostly on the contributions of individual goods and services to
F-1
-------
human well-being. Alternately, ecosystem services valuation is based on the various benefits generated by
the ecosystem (Polasky et al., 2005). In this case, benefits include both marketed and non-marketed
services, and their valuation considers the environment as a natural capital asset that generates returns on
investment in ecosystem protection and management.
Change in Ambient Concentration
(NO,, SO,, and Ozone*)
NO./SO, Standards
• forest growth and
appearance
• crop growth
* soil chemistry
* species composition
and richness
«population abundance
Change In Ecological End point*
¦ water chemistry
and quality
• species composition
and richness
«population abundance
• water chemistry
and quality
• species composition
and richness
• population abundance
T«ms!nal.EsQsy8tera
¦ crop yields
* timber yields
* recreational activities
¦ aesthetics
* other non-consumptive
Chanfjfi in Economic Enripoint*
Transitional Ecofiysiem AauaiisUi&asystem
¦ recreational activities «recreational activities
• non-consumptive uses * commercial fishery
yields
«rio n-con sumptrve uses
Valuation Methods
• Stated Preference
• Revealed Preference
• Avoided Cost
• Replacement Cost
• Benefit Transfer
Figure F-1. Illustration chart of the assessment.
For example, wetlands constitute a form of natural capital. They serve as flood barriers, soaking up
excess water and slowing and preventing floodwaters from spreading uncontrollably. Wetlands help
replenish groundwater and improve both ground and surface water quality by slowing down the flow of
water, and absorbing and filtering out sediments and contaminants. They also provide spawning habitat
for fish, supporting the regeneration of fisheries. In addition, wetlands provide habitat for many wildlife
species and support commercial and sport fishing, as well as hunting and other forms of recreation.
Though different functions and processes of ecosystems, such as water filtration, may be
economically important, they need to be viewed as inputs of or mechanisms for the production of
economically valuable services, such as drinking water, timber, or recreational benefits. The end products,
not the elements of the production process, ultimately generate economic well-being. Along these lines,
Boyd and Banzhaf (2007) advocate defining ecosystem services as "components of nature, directly
enjoyed, consumed, or used to yield human well-being." In other words, ecosystem services are the end
products of nature to which ecosystems contribute as intermediate inputs or production technologies.
Though this distinction may at first seem unimportant, it is crucial for the accurate valuation of ecosystem
F-2
-------
services. Regarding the incorporation of ecosystem services into the measurements of national income
and the value of goods and services produced in an economy, such as gross domestic product (GDP)
accounts, Boyd and Banzhaf (2007) note that if intermediate and final goods are not distinguished, the
value of intermediate goods is double-counted because the value of intermediate goods is embodied in the
value of final goods. For example, clean drinking water, which is consumed directly by a household, is
dependent on a range of intermediate ecological goods, but these intermediate goods should not be
counted in an ecosystem service welfare account. Also important is that ecosystem services are attributed
only the incremental value they contribute to the production of valuable end products. Using the above
example, the value of ecosystem services associated with drinking water denotes the marginal
contribution of ecosystems in the production of drinking water, not the full value of the final product.
If intermediate and final goods are not distinguished, the value of intermediate goods is double-
counted because the value of intermediate goods is embodied in the value of final goods. For example,
clean drinking water, which is consumed directly by a household, is dependent on a range of intermediate
ecological goods, but these intermediate goods should not be counted in an ecosystem service welfare
account. Also important is that ecosystem services are attributed only the incremental value they
contribute to the production of valuable end products. Using the above example, the value of ecosystem
services associated with drinking water denotes the marginal contribution of ecosystems in the production
of drinking water, not the full value of the final product.
Given the complexity and variety of ecosystems and their services, their valuation poses several
challenges. According to the National Academy of Sciences' Committee on the Valuation of Ecosystem
Services, the importance of ecosystem functions and services is often taken for granted and overlooked in
environmental decision making. Moreover, the key challenge in the valuation of ecosystem services lies
in the difficult integration of economic valuation and ecological production theory. This is no
straightforward task, because many ecosystem goods and services are not quantifiable using available
methods, and the application of economic valuation methods may be subject to judgment, uncertainty, and
bias (Heal et al., 2005).
A study by Costanza et al. (1997), seeking to determine the value of global ecosystem services,
exemplifies the problems and pitfalls in the valuation of ecosystem services. Deriving and summing value
estimates from the existing literature for a wide range of ecosystem attributes and services, this study
suggested that the total value of global ecosystem services likely ranges from $16 to $54 trillion annually,
or roughly one to three times global GDP. The study has been influential and widely quoted and used,
especially among scientists and environmentalists. Economists consider it fundamentally problematic
both conceptually and methodologically, preferring to focus on the value of changes to ecosystem
services, which is relevant for policy, or what is termed the marginal value of ecosystem services. The
estimate of the value of global ecosystem services by Costanza et al. (1997) has therefore been
characterized as a "serious underestimate of infinity" (cf. Smith, 2007; Toman, 1998).
F.1.1.2. Use of the Valuation Literature to Define Adversity
A secondary standard, as defined in Section 109(b)(2) of the CAA, must "specify a level of air
quality the attainment and maintenance of which, in the judgment of the Administrator, based on such
criteria, is required to protect the public welfare from any known or anticipated adverse effects associated
with the presence of [the] pollutant in the ambient air." One way to quantify adverse effects is through
monetary valuation.
Adversity is difficult to quantify and measure, and there are several challenges to using a monetary
valuation approach. A major effect that is geographically extensive might be considered to be more
adverse than a more severe effect limited to one geographic location. Another problem is aggregation.
Any change in pollution may have multiple effects (i.e. effects on many types of ecosystem services)
leading to difficulty in aggregating in a consistent way.
F-3
-------
Monetary values on any service or resource degradation reflect human preferences about what is a
severe effect. Larger unit values correlate with more severe effects, other things equal. Also, more
extensive effects, will contribute to larger welfare loss (or gain). In addition, since monetary units can be
added, the aggregation issue can be addressed by "simply" summing the welfare losses (or gains).
Although this is not strictly true (e.g., values for improvements in water quality and fish populations may
not be additive), in principle the differences in how people conceptualize ecosystem improvements can be
captured in the way resource improvements are valued in monetary terms.
Clearly, there are many practical problems associated with using monetary value as a way of
defining adversity. First, many resources and services have not been valued and efforts to credibly transfer
the results of valuation studies to other areas and resources have been minimal. Second, studies
addressing multiple effects are particularly difficult to transfer and few in number. Finally, even with the
first two problems addressed, a judgment would still need to be made on whether the air quality standard
was the contributing factor for eliminating adverse (highly monetarily valued) effects.
F.1.1.3. Methods for Selecting Literature for this Assessment
Assessing the economics literature on the effects of NOx and SOx emissions on terrestrial,
transitional, and aquatic ecosystems requires identifying and reviewing relevant studies addressing these
effects. Multiple methods were used for this Annex: searching existing databases of this valuation
literature; conducting systematic searches of the economics literature; reviewing a large number of key
articles, reports, authors, and journals; and identifying studies based on the expertise and familiarity with
the relevant literature of lead researchers.
Two existing databases on environmental valuation studies - the Environmental Valuation
Reference Inventory (EVRI) and the Beneficial Use Values Database (BUVD) - were particularly useful
for this assessment. The EVRI database, which includes nearly 1,900 valuation articles/studies on
environmental and human health effects, was screened according to criteria regarding the potential
relevancy of geographical location (U.S.), types of environmental goods and services valued (ecological
functions, extractive uses, non-extractive uses, passive uses); and environmental stressor. This resulted in
over 200 articles/studies of interest. BUVD is a relatively small database (131 articles/studies), so it was
imported into the literature review database in its entirety, with unrelated articles/studies later excluded on
an individual basis.
A large number of additional journals and literature databases were identified that publish and
cover research potentially relevant to this assessment. The selected peer-reviewed journals1 and library
databases2 were then reviewed using search engines and a range of key words developed to find studies
addressing relevant ecological endpoints (aquatic, transitional, terrestrial) and their economic values. The
tables of contents of those journals that could not be searched electronically were reviewed in hard copy
and relevant articles were added to the literature review database. These searches were augmented by
reviews of the bibliographies of the following EPA reports: EPA Report to Congress: The Benefits and
Costs of the CAA 1990-2010 (U.S. EPA 1999b); Air Quality Criteria for Ozone and Related
Photochemical Oxidants (U.S. EPA 2006b): Air Quality Criteria for Particulate Matter (U.S. EPA 2004).
1 American Economic Review, American Journal of Agricultural Economics, Canadian Journal of Economics, Canadian Journal of Forestry,
Contemporary Economic Policy, Ecological Economics, Environment and Development Economics, Environmental and Resource Economics,
Environmental Science and Technology, Forest Science, Forestry Chronicle, Journal of Agricultural and Applied Economics, Journal of Applied
Econometrics, Journal of Agricultural and Resource Economics, Journal of Agricultural Economics, Journal of Environmental Economics and
Management, Journal of Forest Economics, Journal of Forestry, Journal of Political Economy, Journal of Risk and Uncertainty, Land Economics,
Marine Resource Economics, Resource and Energy Economics, Review of Agricultural Economics, Review of Economics and Statistics, Water
Resources Research.
2 AgEcon Search, Agricola, BioOne, CSA Illumina, EconLit, GeoRef, Google Scholar, SciSearch/Science Citation Index (Web of Science),
SCOPUS, Sportfishing Values Database, SSRN.
F-4
-------
The results from these searches were checked for duplicates and clearly irrelevant studies, after
which over 500 potentially relevant articles/studies were identified for initial assessment. Relevancy of
each study for this assessment was determined according to the following key criteria.
¦ Does the study address an ecological endpoint sensitive to reductions in NOx and SOx emissions?
¦ Does the study value quality changes in the ecological endpoint, which is actually or potentially
attributed to reductions in NOx and SOx emissions?
¦ Is the study peer-reviewed and preferably, published in an academic journal?
¦ Very few, if any studies fully satisfy all above criteria. For this reason, studies that at least
partially satisfy these criteria were deemed potentially relevant for this project. Finally, reviews
and meta-analyses were included in the assessment whenever they were available and dealt with
potentially relevant ecological endpoints.
In the initial assessment, each record's potential relevancy to the assessment was rated on a scale of
1 to 4, with 1 indicating that the record appears directly relevant to this assessment (the study addresses
quality change of an ecological endpoint, which is actually or potentially attributable to NOx/SOx).
Records rated 2 only partially satisfied the "relevancy criteria," but were considered important to be
referenced in this report. Records rated 3 were to be reviewed more closely to determine their usefulness,
and those rated 4 were found not relevant for the purposes of this assessment. Because the goal in the
initial assessment was to avoid missing potentially relevant studies, borderline cases have been classified
to the lower number category.
Figure F-2. Reviewed studies by ecosystem addressed.
All studies rated 1 through 3 were next reviewed using several attributes, including ecological
endpoints, valuation techniques, geographical area, use vs. non-use value category, and other details of
interest. Of those studies, about half addressed aquatic ecosystems (Figure F-2). The reviewed studies
addressed many different ecological endpoints, such as sport fishing, commercial fisheries, aquatic
recreation (e.g., swimming and boating), general water quality, ecosystems services provided by aquatic
ecosystems, and coral reefs (Figure F-3). Nearly one third of the studies addressed terrestrial ecosystems
(e.g., forestry/commercial timber, outdoor recreation, and agriculture); the rest dealt with transitional
ecosystems (e.g., ecosystems services provided by wetlands and wetlands recreation).
Aquatic
ecosystems
Transitional
ecosystems
Terrestrial
ecosystems
F-5
-------
F.2. Conceptual Framework
F.2.1. Taxonomy of Values for Environmental Goods and Services
The economic value derived by society from these ecological goods and services can be
categorized as either "use" or "non-use" values. Use values comprise values for those goods and services
that can be used either directly or indirectly. Non-use values denote the characteristics of the ecological
goods and services that are not used at all but still hold economic value. The schematic in Figure F-4
illustrates these divisions of values for environmental goods and services.
Direct use values are held for those goods and services which can be directly consumed or utilized
by individuals or society. Some direct use goods, such as fish caught or timber cut are sold in markets and
thus valued using market data. Other direct uses, such as recreational use of ecosystems for fishing,
hunting, and sightseeing, are usually not bought or sold through a market and are therefore more difficult
to value. Similarly, changes in their value are more difficult to quantify. Thus, direct use can be
subdivided into directly used market and non-market goods or services.
50
T>
cu
40
>
DC
£ 30
T>
=5
n
1
n
at
X}
E
3
10 -
o c c
•j= o 0
en = c ¦—
^ CO CO
cr a) go a)
< 5 «= o
£i £
cr
CO
O
c
00 d)
w
CD P
CO
CD
CD
JZ
00
CD
<
CD
r~
M—
o
00
{/>
00
CD
o
-O
"O
SZ
c
(D
CD
o
CD
$
£
Cl
CO
CO
CO
5
£
CO
CO
3
cr
CD
CD
O
O
Figure F-3. Reviewed studies by ecological endpoint.
F-6
-------
Good or Service
Non-SJse Value
Use Value
Direct Use
Indirect Use
Market Use
Non-Market Use
Nan- Market Use
Source: U.S. EPA (2002a).
Figure F-4. Taxonomy of values for environmental goods and services.
Examples of direct, market uses of ecological goods and services include commercially sold food
sources (fish, crops, other animals); building materials (wood, stone); fuel sources (wood, coal, oil);
drinking water (groundwater, surface water); chemicals; and minerals (U.S. EPA, 2002a). Examples of
direct, non-market uses of ecological goods and services include recreational fishing and hunting; beach
use (sunbathing, swimming, and walking); boating; hiking; camping; wildlife watching; and sightseeing
(U.S. EPA, 2002a).
Indirect use values capture those ecological services that are not used directly but still provide
benefits of economic value to society. These services include flood control, storm water treatment, ground
water recharge, climate control, pollution mitigation, wave buffering, soil generation, nutrient cycling,
habitat value, and biodiversity (U.S. EPA, 2002a).
Non-use values denote the characteristics of the ecological goods and services that are not used at
all but still hold economic value. These non-use values include what are known as existence and bequest
values-the value of simply knowing that certain ecosystems exist and of ensuring that they continue to
exist for future generations, respectively. These services are also not traded in a market and quantifying
their value is a challenge.
Importantly, empirical methods for addressing non-use value generally estimate the total value of a
resource. Distinguishing between the use and non-use values often is not possible, and the valuation
results including non-use values should therefore generally be considered total valuation rather than non-
use valuation studies. Examples of non-use values of ecological goods and services include existence
value, cultural/historical value, intrinsic value, bequest value, and altruistic value (U.S. EPA, 2002a).
In this Annex, the focus is on the potential incremental benefit that might be realized from reducing
levels of SOx or NOx by a relatively small amount, rather than the total value of the affected ecosystems.
Therefore, the term "total value" in this Annex generally denotes total marginal benefits that contain both
use and non-use values, not the total value of the entire resource. This notation is consistent with the
principles of economic valuation of the environment, which generally focus on predicting damages or
benefits from marginal changes in environmental conditions.
Any environmental goods or services that somehow contribute to human well-being are
economically valuable. Their economic value reflects the capacity of the environment to satisfy different
human needs, which is related to the direct consumption of different goods and services derived from the
F.2.2. Welfare Economics
F-7
-------
environment. It also includes different indirect, passive, and non-use values for environmental goods and
services, such as recreational benefits, enjoyment of natural landscape, purity of air and water, and
provision of habitat for species other than humans.
Economic valuation is rooted in the basic principle of consumer sovereignty. Rather than judging
whether an individual's choices are right or wrong, each person is considered able to make rational
choices that advance his or her well-being, given the possibilities available. The principle of consumer
sovereignty extends also to the valuation of environmental goods and services. Even in the absence of
markets for environmental benefits, each individual is considered able to assess the importance of
changes in environmental quality on personal well-being or, as economists commonly refer to it, utility.
Consumer (CS) and producer surpluses (PS) are the basic monetary measures of well-being (or
welfare) in economics. They denote the "excess utility" consumers and producers enjoy when consuming
or producing specific goods or services after paying for them. Producer surplus is measured by profit, the
value of production after accounting for all costs. Consumer surplus, which is a more subtle concept, can
be thought of as the difference between the price and the maximum value that an individual holds for a
good or service. However, it generally is not an exact measure of changes in welfare because CS does not
fix the baseline level of utility, thereby ignoring the income effects of a changing baseline
CS has two exact (Hicksian) measures: willingness-to-pay (WTP) and willingness-to-accept
(WTA). In the context of environmental valuation, WTP denotes an individual's maximum willingness to
pay for an environmental good or service. WTA, on the other hand, stands for the minimum compensation
an individual is willing to accept to forgo an environmental good or service. Though WTP and WTA both
are exact measures of CS, they generally are not equal, meaning that the CS in general does not have a
unique measure. WTP and WTA for usual marketed goods and services are similar and within narrow
bounds of income effects (Willig, 1976). Similar findings can be expected with environmental goods and
services only when they have close substitutes (Hanemann, 1991; Shogren et al., 1994).
Since many environmental goods and services cannot be easily substituted, WTP and WTA are
expected to differ, sometimes substantially. In fact, the WTP-WTA divergence can range from zero to
infinity, depending on the substitutability of an environmental good and other market or non-market
goods (Hanley et al., 1997). Using WTA to measure welfare changes might often be justified on the
grounds of economic theory and property rights, but many studies choose to estimate WTP for its
practicality. In addition, WTP generally is lower than WTA and therefore is conservative, providing
another justification for using WTP for valuing the environment. Finally, both WTP and WTA can be
difficult to measure, and valuation studies therefore sometimes estimate the ordinary CS. It approximates
WTP and WTA and is between these two exact measures of welfare. In the special case of no income
effects, all different measures of consumer welfare coincide.
Various methods have been developed to determine the value of different ecological goods and
services by estimating the change in social welfare or WTP for changes in the quantity or quality of a
given environmental resource (see Table F-l). Some valuation techniques obtain WTP from observing
people's actions (revealed preferences or RP) while others rely on people's responses to hypothetical
situations (stated preferences or SP). Yet another set of valuation methods relies on other studies, either by
transferring their estimates into another context (benefits transfers) or by conducting statistical meta-
analyses of earlier studies to examine their systematic findings. See Section F.2.3 for details about these
approaches.
As noted by Kramer et al. (2003), several forest protection valuation studies also considered the
sensitivity of contingent valuation estimates to various preference elicitation methods, including
dichotomous choice, payment card, and open-ended techniques. Haefele et al. (1992) used payment card
and dichotomous choice techniques in a contingent valuation survey measuring the WTP of Southern
Appalachian residents for protecting high-elevation spruce-fir trees from exotic insects and air pollution.
Estimates for mean WTP were $20.86 per year using a payment card method, and $99.57 per year using a
discrete choice method. Sample sizes for each method were relatively small (232 and 236 respondents,
respectively). Another limitation concerns the wide span between the two valuation statistics, which
F-8
-------
despite a large difference in the mean estimates showed no overlap of the 95% confidence intervals.
Similar studies have been performed by Loomis et al. (1996) and Kramer and Mercer (1997).
F.2.3. Benefit Estimation Approaches
Table F-l introduces different valuation techniques and indicated which type of value (direct use,
indirect use, total value) each can be used to estimate. Studies addressing non-use values are referred to in
this Annex as total valuation studies. This is because non-use valuation methods in fact generally estimate
the total value of a resource, including both its use and non-use components.
Travel cost (including site choice models) and hedonic pricing methods are perhaps the most
regularly applied revealed preference (RP) methods for valuing the environment, whereas contingent
valuation and choice experiments are the most popular stated preference (SP) methods. One of the key
differences between the RP and SP methods is that RP methods can only fully address direct and indirect
use values, whereas SP methods are required for the estimation of total values of environmental resources,
including their non-use value component.
Travel cost studies predict use values for ecological resources, such as natural parks, by examining
individuals' travel expenditures to utilize that resource (most often at a park or some other recreational
site), including the opportunity cost of work time missed while traveling to and utilizing the resource.
Travel cost studies commonly use a random utility framework, which infers individuals' WTP for an
ecological resource (again usually a recreational site) by observing their choice from among one or more
alternatives. While the travel cost method uses changes in the quality of one resource to ascertain its
value, the random utility model uses individuals' choices among various options of various qualities at
various prices to do the same.
Hedonic price studies predict the value of ecological resources by examining their effect on
property values. Assuming that all the benefits from living in a specific location and house are capitalized
into the market value of the property, hedonic models estimate the independent effects of different
housing characteristics on housing prices. Controlling for all observable housing and location
characteristics, hedonic pricing models examine environmental values for, for example, proximity to
forests or particular watersheds by estimating the implicit incremental price people are willing to pay for
that proximity. Hedonic pricing relies on assumptions such as efficiently functioning housing market and
perfect information and mobility by individuals. However, because WTP is not necessarily tied to
ecosystem changes, this Annex does not consider property values as a method of valuation.
Other market or RP-based approaches to valuing changes in ecological goods and services include
the alternative/replacement cost method; avoidance expenditures/averting behavior method; referendum
method; user fee method; market price and market simulation method; and a host of different variations of
these and other valuation approaches. See, for example, Freeman (2003) for the theory and applications of
different methods for valuing environmental quality.
The RP methods have the advantage of gleaning their value estimates from individuals' real world
actions. However, because they do not include the non-use value of ecological resources, none of them
capture total value. This problem has given rise to the development of a variety of non-market valuation
methods that use surveys to elicit preferences for public goods. Because these methods are generally
based on eliciting "stated" rather than "revealed" preferences, they are broadly categorized as SP
methods.
The contingent valuation method is the most common SP method. It involves developing and
administering surveys, in which respondents are presented with a scenario or a program with specified
environmental outcomes and costs. Each respondent is asked to indicate approval or disapproval of the
proposed environmental scenario and its monetary cost. Researchers vary the proposed costs across
different survey respondents and use their choices to estimate how much people on average are willing to
pay for different scenarios to improve the environment. Because some of the respondents may use a
F-9
-------
public good also for direct enjoyment, say viewing an endangered bird, the surveys actually capture total
value for the improvements, rather than just their non-use value.
Choice experiment methods (including contingent ranking, contingent choice, and conjoint
analysis) separate an environmental good into its constituent attributes, recombine those attributes into
different bundles, and elicit respondents' preferences for those bundles. Often a monetary value can be
assigned to those attributes and thus the process allows researchers to determine WTP for the bundle and
each attribute. Conjoint analysis usually is performed by choosing the most preferred attribute bundle
from a group (choice experiment, contingent choice) or via ranking a series of attribute bundles
(contingent ranking). Using conjoint analysis, researchers may be able to simultaneously value various
relevant goods or services that an environmental resource provides. For example, improving a public
body of water provides improved recreational opportunities, drinking water, and support of aquatic
ecosystems.
Valuing the environment as a factor of production monetizes the incremental benefits from using the
environment as an input to production. In other words, this method treats the environment as a production
input comparable to other raw materials and infrastructure, such as land, capital, labor, and so forth. This
method is appropriate for valuing environmental effects that have direct value as a factor of production.
Examples of such cases include the effects of water quality on the productivity of commercial fisheries, or
the impact of soil characteristics on agricultural productivity. However, this method is limited to market
goods and services and can only address their role as part of the production process. This method does not
address non-market environmental goods and services or non-use values.
Benefits transfer approaches translate the entire estimated demand function from one application to
another. Sometimes the function is adjusted to meet the specific criteria of the target site, and then new
WTP value estimates are generated for the environmental good/service at the new site using the demand
function. Using the transferred demand function, both changes in the level of use and the unit value
benefits for the new site can be estimated (U.S. EPA, 2002a). There are four general types of benefits
transfer technique: mean unit value transfer/ adjusted unit value transfer, benefit/demand function or
model transfer, meta-function transfer, and structural benefits transfer. The first three methods dominate
the literature (Smith et al., 2006b).
Mean unit value transfer/adjusted unit value transfer entails taking the value of a specific
environmental good or service (such as recreational hunting), sometimes from a single study of the same
good and sometimes estimated by averaging a range of value estimates from various primary studies, and
transferring that value to the same good or service at a new site.
Benefit/demand function or model transfer is the translation of the entire estimated demand function
from one site to another. Sometimes the function is adjusted to meet the specific criteria of the target site,
and then new WTP value estimates are generated for the environmental good/service at the new site using
the demand function. Using the transferred demand function, both changes in the level of use and the unit
value benefits for the new site can be estimated (U.S. EPA, 2002a). Meta-function transfer involves the
use of meta-regressions to combine the results of numerous valuation studies and allows researchers to
account for influencing factors, thus enabling them to create value estimates for new policy sites.
Structural benefits transfer, also known as preference calibration, requires selection of a preference model
which can describe individual choices over a set of market and associated non-market goods to maximize
utility when faced with budget constraints (Rosenberger and Loomis, 2001; Smith et al., 2006b).
Meta-study review provides a useful way of summarizing the literature on the valuation of
ecological endpoints from reductions in NOx and SOx, though there are few studies exactly on this topic.
But there are studies summarizing monetary valuation efforts for particular sets of endpoints, such as the
economic valuation of fresh water ecosystem services. Where such studies are available, they are
summarized in the appropriate section of this annex.
Even more valuable are meta-analyses, which perform statistical analyses of the results of original
studies. Such studies explain variation in monetary value estimated for various endpoints using features of
the original studies' methodologies as well as the characteristics of the site being studied and other factors.
F-10
-------
Smith and Pattanayak (2002) defined meta-analysis as the practice of using a collection of formal and
informal statistical methods to synthesize the results found in a well-defined class of empirical studies.
In general, the three uses of meta-analysis are: synthesize or "take stock" of the literature on a
particular valuation topic; test hypotheses with respect to the effects of explanatory variables on the value
construct of interest; and use the estimated meta-analysis model to predict estimates of the value construct
across time and space (Bergstrom and Taylor, 2006). Bergstrom and Taylor (2006) provided a review of
the techniques and theory behind the use of meta-analysis for benefits transfer. They noted that to conduct
a successful meta-analysis for benefits transfer, it is important to be as comprehensive as possible in terms
of the studies to be included. Excluding a study would be equivalent to applying a zero weight to the
information in that study. The authors list some additional criteria to be considered when identifying
studies for inclusion in a meta-analysis, including controlling for the valuation method and estimated
welfare measure, as well as addressing the temporal and spatial scales of the valued commodity.
Extensive evaluation of the relative merits of, and issues with, different environmental valuation
methods is not within the scope of this Annex. Vast amounts of research have been conducted to develop
and evaluate alternative environmental valuation methods. For example, the Handbook of Environmental
Economics recently dedicated an entire 1,100-page volume to addressing methods for valuing
environmental changes (Maler and Vincent, 2005).
The validity of environmental valuation methods is sometimes questioned, in particular that of SP
methods. Because survey-based valuation methods are based on what people say rather than what they do,
there is a tendency to question the credibility of the results. For this reason, nonmarket valuation
researchers, following the lead of the National Oceanic and Atmospheric Administration (NOAA) Expert
Panel that reviewed the highly publicized studies valuing damages from the Exxon Valdez oil spill, often
build into their surveys a series of validity tests, such as testing for the sensitivity of WTP to the scope of
resource being valued (Federal Register, 1993). Additionally, SP surveys are vulnerable to a variety of
issues dealing with the design and administration of surveys, as well as analyzing their data (e.g., Carson
and Hanemann, 2005).
Though examining actual choices lends credence to RP methods, they are not free of problems. For
example, hedonic pricing studies of housing markets rest on the assumptions that the housing market is at
equilibrium and that housing choices accurately reflect the attributes of interest, such as air pollution or
environmental amenities associated with the residential location. Hedonic studies and other RP studies,
such as recreation trip demand analysis, are also susceptible to potential biases. These include the
omission of important variables, which may thwart the efforts to accurately value environmental quality.
A number of studies have compared SP analyses with RP analyses, such as hedonic property value
studies. Generally, these comparisons have suggested that when similar environmental values are
examined, RP methods generally yield somewhat higher value estimates than SP methods. For example,
Carson et al. (1996) reviewed over 80 studies which included comparisons of SP and RP methods, and
concluded that SP methods are on average about 75% to 90% of corresponding RP values.
F.3. Valuation of Forests and Terrestrial Ecosystems
F.3.1. Use Values
The impacts of NOx and SOx can occur over many different terrestrial ecosystems and use-value
efforts must look at each of these ecosystems individually. One serious threat to agriculture from NOx
emissions comes from ambient ozone (03), which is a byproduct of atmospheric reactions between
Volatile Organic Compounds (VOCs) and NOx. In commercial forests, air pollution effects have not been
addressed in economics beyond evaluating the potential effects of 03. Unlike in agriculture, the scientific
F-11
-------
understanding of 03 effects on trees is more limited and related mostly to visual injury to leaves in young
saplings, an indicator that is difficult to link to tree growth in mature forests. However, the valuation of 03
has already been evaluated in detail in the 2006 03 AQCD (U.S. EPA, 2006b), and will not be covered in
this Annex or the ISA (see also Tables F-2 and F-3).
Some studies have attempted to specifically measure use values associated with forest quality,
without regard to the factor that is altering forest quality. For example, an early study by Leuschner and
Young (1978) considered the effects of crown density changes due to insect infestation around 19 lakeside
campgrounds in east Texas. A travel-cost model estimated the CS losses from a 10% reduction in crown
density to fall between 0.69% and 6.5%, depending on the number of substitute recreation sites available.
More studies that look at how forest quality affects various uses are listed in Table F-4. Also, Table F-5
lists value estimates from a meta-analysis but does not reflect the considerable variation of value
estimates across underlying studies.
Figure F-5 depicts the linkages between aesthetic welfare benefits and air pollution. One such
linkage that is shown is the visual quality of forests. Changes in woodland appearances are monitored
using scientific indicators of ecosystem change, such as crown condition, mortality, foliar damage,
vegetation structure, and plant diversity (McLaughlin and Percy, 1999). However, in comparison to
valuation of changes in commercial timber, measuring the economic value of aesthetic changes in forests
and natural ecosystems can be more directly based on scientific information regarding the effects of
pollution on forest health. Although the scientific linkages between air pollution and visual forest quality
include large degrees of uncertainty, historic examples of air pollutant effects on forest aesthetics facilitate
their empirical valuation. Figure F-6 shows the regions and species that have been identified as
historically affected by air pollutants.
F-12
-------
Emissions Data
Emissions Scenario
Model
Amounts emitted
Fate and Transport
Model
Deposition predictions
'
Visual injuries
Response Model
Change in visual injuries
Forest Aesthetics
Model
Economic Model
Change in forest aesthetics
Value of the impact on the ecosystem
Figure F-5. Linkages from emissions to forest aesthetics.
Source: Exhibit 1 in lEc (1999c).
F-13
-------
U.S. Major Forest Types Affected by Air Pollution-Induced Visual Injuries
Whfce Mountains (NH)
Green Mountains (VT)
Adirondacks (NY) ]
Maine
Sierra.
Nevada
-------
the highest forest quality (125 to 175 trees per acre). Overall mean WTP for changes in forest quality was
estimated at $47 per household. In addition, respondents decomposed total value into four categories of
value (recreational use, option, existence, and bequest). Results showed that use values accounted for
27.4% of total value, and non-use values accounted for 72.6% of total value. The study was limited by
small sample size (200 respondents) and possible bias from framing the scenario as "one of the most
important issues facing Colorado residents."
A study by Haefele et al. (1992) that measured WTP to protect Southern Appalachian spruce-fir
forests, used a decomposition approach to determine dominate influences on total values. The authors
found that non-use values (bequest and existence values) overshadowed use values as reasons to protect
forests. Holmes and Kramer (1996) relied on results from Haefele et al. (1992) to compare WTP between
recreational forest users and members of the general public. Mean household estimates for forest-users
($36.22) were substantially larger than estimates for nonusers ($10.37).
Kramer et al. (2003) used a dichotomous-choice, contingent-valuation format to determine WTP of
Southern Appalachian residents for protecting spruce-fir forests from insects and air pollution. The study
measured incremental levels of forest protection using two scenarios: the first increment occurred along
road and trail corridors, a scenario that would appeal to people valuing the ecosystem primarily for
recreational use; the second level was for the entire ecosystem, a scenario that would appeal to people
valuing the continued existence of the entire threatened ecosystem. Using randomly assigned values
between $2 and $500 dollars, respondents were asked if they would pay a certain tax amount for each
scenario. Results suggested that preferences for forest ecosystem protection created a well-behaved
demand curve, with incremental WTP increasing at a decreasing rate.
F.3.3. National-Scale Valuation
An EPA-contracted valuation study by IEc (1999b) developed a benefits-transfer model using
results from Holmes and Kramer (1996), Peterson et al. (1987), and Walsh et al. (1990) to estimate
national benefits from the 1990 CAAA. Calculations applied WTP values from these previous studies
across all households in the affected states. (See Table F-6 for more information on the underlying
original studies, and Table F-7 for benefits transfer results.) Estimates for the total value of avoiding air
pollution-induced damage to forest aesthetics in the U.S. during the period 1990-2010 were found to
range from $3 to $17 billion. Due to limitations in the original studies and the simplicity of the benefits-
transfer model, the authors warned that the estimates should only be used to consider the general
magnitude of benefits to forest aesthetics (in the range of billions of dollars) rather than as precise values.
F.3.4. Valuation of Degrees of Injury
A few studies—Hollenhorst et al. (1993); Ruddell et al. (1989); Hammitt et al. (1994); Buhyoff
et al. (1982)—measured general preferences for forest aesthetics without estimating changes in welfare.
Hollenhorst et al. (1993) considered the effect of tree mortality on perceived forest aesthetics. On a 1 to
10 scale for visual and recreational appeal, 400 respondents ranked pictures of 25 sites with tree mortality
ranging from 6% to 98% from gypsy moth (Lymantria dispar) damage. The results produced a hill-
shaped function, whereby site appeal increased with mortality at levels as high as 40% but then declined.
The authors speculated that respondents had a general distaste for tree mortality but valued light
penetration to the forest understory, which allows for the growth of wild flowers and lower-level
vegetation.
Other studies confirmed this visual preference for light penetration, including Ruddell et al. (1989)
and Hammitt et al. (1994), who observed positive responses to forest edges and open middle grounds
without light-obstructing tree canopies. Buhyoff et al. (1982) tested how awareness of environmental
F-15
-------
damages affects perception of aesthetic losses. The study compared aesthetic rankings of photographs
between a control group and a group that is informed about the cause of forest damage before the ranking
session. Results show a heightened level of sensitivity to forest damage by the "informed" group of
subjects.
Several other studies investigated the notion of aesthetic thresholds, or discontinuous jumps in
aesthetic preferences across small changes in visual injuries. Of particular interest to early studies were
thresholds at the lower limit of visual perceptibility.
Contingent valuation studies by Vaux et al. (1984) and Flowers et al. (1985) assessed aesthetic
preferences at recreational areas with fire damage and found that small differences in site appearance can
produce large changes in recreation preferences. This finding is generally consistent with the results from
two early studies by Buhyoff et al. (1979) and Buhyoff and Wellman (1980), which asked respondents to
rate various levels of insect damage. The authors found that preferences seemed to be most affected by
the presence or absence of insect damage, as opposed to the degree of damage.
Crocker (1985) surveyed 100 recreationists for their WTP for recreational experiences at forested
sites with slight injury, moderate injury, or severe injury. Mean WTP estimates for the environment with
slight injury were three times higher than WTP for environments with moderate and severe injury,
suggesting that people are willing to pay a premium for recreational access below the lower limit of
perceptibility. As a group, respondents did not indicate a clear preference ordering between the moderate
and severe injury environments.
Holmes et al. (2006) use a hedonic method to predict the impact of forest damage due to the
hemlock woolly adelgid (Adelges tsugae), an invasive insect, on the value of residential properties.
Examining 3,379 residential property sales in New Jersey between 1992 and 2002, the study analyzed
how the appearance/health of the forest on the home's parcel and within 0.1, 0.5, and 1 km buffers around
it affects the housing market. Controlling for other relevant variables, the study estimated that hemlock
health status had a statistically significant effect on property values. The estimation results suggest, for
example, that a 1-point increase (e.g., from 10% to 11%) in the percentage of healthy hemlocks (less than
25% foliar damage) of all forests on the home's parcel was associated with a 0.66% sales price increase.
Similar changes in the hemlock forests and their health status in the home's near proximity are predicted
to be associated with yet larger increments in the sales price.
At least one study seems to provide evidence against the concept of thresholds at the lower limit of
perceptibility. A referendum-type contingent valuation survey by Jenkins et al. (2002) used two sample
populations to value a forest protection program. The initial forest condition was described as "pristine"
for one group and already "somewhat damaged" for another group. The degree of forest damage incurred
in the absence of a forest protection program was the same for both sample groups. Regressions for the
entire sample population showed no statistically (Jenkins et al., 2002) significant difference in WTP for
forest protection between the two groups.
Further analysis of Jenkins et al. (2002) suggested that aesthetic preferences and thresholds differ
between recreational groups. Comparisons among recreational groups revealed that consumptive forest
users (hunters and anglers) held values that were sensitive to change in forest condition, while non-
consumptive forest users (hikers and campers) held values that were insensitive to the same amount of
change. Overall, however, non-consumptive forest users expressed higher values than consumptive forest
users. Aesthetic thresholds also seem apparent for recovering forests. Paquet and Belanger (1997) found
that the aesthetic effects of clear cutting were largely removed once re-growth reached a height of 4 m.
Finally, there is substantial literature related to the amenity value of urban/suburban forests and
open space. McConnell and Walls (2005) recently reviewed this literature, evaluating more than 60
published articles that have attempted to estimate the value of different types of open space. Two lines of
research emerge in this literature: studies that estimate the hedonic value of open space proximity to
residential properties, and studies that use SP methods to value preservation of open space. Unfortunately,
neither line of research generally comprises studies that provide information on the WTP for quality
changes to open space that might be caused by changes in NOx and SOx emissions. In most cases, the
proximity or preservation of open space or urban/suburban forest is valued without reference to the
F-16
-------
quality or even the type of open space or forest. Although this literature demonstrates the value of the
availability and preservation of open space, its relevancy for the purposes of this assessment is at best
limited.
F.3.5. Limitations and Uncertainties
Shortcomings in the air pollution terrestrial valuation literature have been well documented in
recent reviews (Adams et al., 2003; U.S. EPA, 2006b). In existing economics literature, the effects of 03
on terrestrial endpoints overshadow assessments of other effects of NOx and SOx pollution, such as acid
deposition and N fertilization. The latter effects have been well documented in the scientific literature, but
the lack of valuation studies related to them limits full assessments of terrestrial damages from NOx and
SOx pollution. Incomplete scientific understanding of the effects of air pollutants on ecosystems and
economic endpoints extends to many valuation studies. Even in the valuation of 03 effects, which have
been relatively thoroughly studied, improvements in noneconomic input data could play an important role
in evaluating the magnitude of vegetation damages from O3.
Economic models should incorporate more specific temporal (dynamic) and spatial data regarding
terrestrial effects of air pollution to reflect more realistic situations. For example, the O3 effects valuation
literature analyzes air pollution effects across a period of time using only two or three scenarios that
represent large, static changes. In the real world, however, these changes would occur gradually and
incrementally. Future studies should try to consider more dynamic models that can describe effects of
marginal air quality changes. From a spatial perspective, previous studies often assumed that producer
responses are similar across large geographic areas. However, regional factors may be important,
suggesting a possible need for finer-scale agricultural and forestry data that would allow models to
consider micro-level physical and economic factors (Adams et al., 2003).
With regards to the effects of air pollution on agricultural crops, most economics studies date from
the 1980s and many focus on O3. Therefore, there is a general need for updated valuation studies that
consider agricultural damages from air pollution, especially with regards to the roles played by NOx and
SOx. Specific issues include the need to develop new exposure-response functions for sensitive crops and
the effects of air pollution on crop yields under actual farm conditions.
Assessments of welfare losses in commercial forestry from air pollution are mainly limited by
scientific uncertainty regarding the extent of ecological damage. Restrained by the large size and slow
growth of trees, scientific research has primarily considered 03 effects on seedlings and has little
transferability to mature trees.
Existing literature on valuation of changes in forest aesthetics has focused primarily on historical
cases of acute air pollution or insect infestation and may be overly simplistic. Although these valuations
are useful, no existing assessments have examined the effects of reduced air pollution on forest aesthetics.
Understanding the effects of marginal changes in air pollution would require the collection of long-term
high-quality data about forest health, as well as improved casual linkages between air pollutants and
visible injuries to trees. Scientific advancements at this scale would require a sophisticated monitoring
network operating over several decades. Until the impacts of pollution on the foundational services of
terrestrial ecosystems are better understood, valuation assessments may be premature. Additionally, most
studies also fail to distinguish between marginal values of forest health and average values of forest
health, an oversight that may create bias in final estimates.
A major limitation of both non-economic and economic forest health assessments concerns the
limited extent of documented ecosystem-level changes. Most forest health surveys focus on average forest
conditions that allow for modeling of near-term trends in economic value, but fail to detect fundamental
changes in ecosystem processes that sustain natural capital in the long run (McLaughlin and Percy, 1999).
Quantifications of impacts on species effectively conform to existing valuation methods; however, these
effects may be overshadowed by long-term or irreversible reductions in ecosystem structure and function
F-17
-------
(U.S. EPA, 1999b). Assessing ecosystem-level changes presents new challenges to both science and
economics as they include a large degree of uncertainty in the scale and nature of effects. In particular,
economic assessments of ecosystem impacts may require regional- to national-scale modeling of
numerous ecosystem functions rather than analyses of specific service flows that directly contribute to
human welfare (U.S. EPA, 1999b). In spite of these challenges, the relationship between air pollution and
forest health has enormous implications for policy development and should be addressed in future
research.
More advanced methods to address uncertainty and irreversibility are essential to future modeling
of complex ecosystem-level impacts. Several alternative methods to address ecological uncertainties are
currently being developed and should play a central role in future models (U.S. EPA, 1999b).
F.4. Valuation of Transitional Ecosystems
The economic analysis of the benefits on wetlands (transitional ecosystems) from reduced
emissions are limited in this Annex to three recent and comprehensive meta-analyses. Brander et al.
(2006) provided a comprehensive review and meta-analysis of the wetland valuation literature. This study
improved upon previous similar studies (Brouwer et al., 1999; Woodward and Wui, 2001) by including
tropical wetlands estimates from other valuation methodologies, other wetland services, and estimates
from more countries.
It is well known that wetlands serve a variety of potentially valuable ecological functions,
including flood and flow control, storm buffering, sediment retention, groundwater discharge, and habitat
for plant and animal species. Wetlands also contribute to climate stabilization, carbon sequestration, and
the overall quality of the natural environment. Brander et al. (2006) (Table F-8) described these functions
and their associated goods and services, as well as their values as determined by common valuation
methods.
F.4.1. Use and Non-Use Values
The Brander et al. (2006) meta-analysis included 191 different wetland valuation studies from 25
different countries. Eighty of those studies, half of which were from North America, provided useful
empirical data for the meta-analysis. A total of 215 individual observations were gleaned from those
studies for use in meta-regressions. The study reviewed the literature, discussed valuation techniques,
calculated average wetland values using the results from the combined data set, and then performed a
meta-regression of the data to determine which explanatory variables had the greatest effect on wetland
value. Non-market use values of wetlands, as well as non-use/existence values, were also included in the
analysis. Because not all of the studies generated WTP value estimates, wetland size and population
statistics were used to convert estimates from the diverse valuation methodologies to 1995 dollar values
per hectare per year, following the example of Woodward and Wui (2001), which is described below.
The results of the Brander et al. (2006) meta-analysis showed an average value for wetlands of
$2800/ha/yr (1995 dollars), and median value of $150/ha/yr. Individual values were calculated by wetland
type, the wetland service provided, contingent, and valuation method used. The meta-regression also
revealed that studies using the contingent valuation method consistently returned the highest wetland
values. It is noted that this may have been due to the type of wetland values that this method was applied
to, rather than something intrinsic to the methodology.
To look at the value differences from a reduction specifically in N emissions, there is a focus on
Brander et al.'s (2006) analysis of wetland value changes that resulted from a small change in water
quality or some other ecosystem service. These results are valuable because they may be related to
F-18
-------
changes in acid deposition as a result of changes in the concentrations of NOx or SOx. In the meta-
regression results of Brander et al. (2006), the only such ecosystem service which was shown to have a
statistically significant effect on wetland value was hunting. Hunting had an unexpected negative
influence on wetland value. Although not statistically significant, water quality, habitat and nursery
service (specifically support for commercial fisheries and hunting), fishing, and biodiversity all had
positive effects on wetland value. The accuracy of the value transfer exercise in Brander et al. (2006)
seems to be in line with other similar efforts, with an average transfer error of 74%. Overall, the value
transfer function from Brander et al. (2006) produced higher-value transfer error for less valuable
wetlands (and systematically over-predicted their value) and lower value transfer error for more valuable
wetlands (and slightly under-predicted their value).
Woodward and Wui (2001) found 39 studies that contained sufficient data for inter-study
comparison, including peer-reviewed and gray literature. Similar to the approach of Brander et al. (2006),
Woodward and Wui (2001) assumed that a wetland's value is a function of its ecological characteristics
and socio-economic environment, and that there exists a true public WTP at a given moment for a
particular wetland. With these assumptions, they attributed variability in wetland value to two principal
sources: differing characteristics of the wetland and error in the estimation of the true value (Woodward
and Wui, 2001).
Brouwer et al. (1999) analyzed the WTP of 30 CV studies addressing ecosystem functions of
wetlands, providing a total of 103 estimates. The studies were mostly conducted in the U.S., but were not
in any way comprehensive spatially. They ranged from 1981 to 1997 and covered a wide range of
commodity definitions, including preserving wetlands threatened by flooding, maintaining or improving
current catch levels and fish populations, saving the bald eagle (Haliaeetus leucocephalus),
improving/maintaining habitat, water quality ladder changes, preventing shellfish bed closures, increasing
the number of protected rivers, and explicitly preserving a measure of water quality. No attempt was made
to standardize the degree of change being valued. Instead, the authors attempted to address differences in
the studies through the use of dummy variables and some degree of sub-setting.
Based on all of the included studies, Brouwer et al. (1999) estimated that the average WTP for the
addressed ecosystem functions was about $90 per household per year. These studies covered types of
wetlands (primarily fresh water systems), ecosystem functions (water quality and biodiversity, and to a
lesser extent water quantity and flood control), and wetlands size. Cutting across types of systems,
location, and function enhanced, the WTP for salt and fresh water improvements was about the same.
Within fresh water wetland systems, riverine wetland improvements appeared to be more highly valued
than palustrine wetland improvements. Across ecosystem function, flood control was more highly valued
than biodiversity, water quality and water generation, in that order. Improvements to larger wetlands
appeared to be somewhat more highly valued than to improvements to smaller wetlands. Improvements in
use values averaged slightly more than $100 per household per year, whereas improvements in non-use
values only showed values about half this size. Spatially, values were higher in California, followed by
Georgia and Louisiana.
The review of outdoor recreation valuation studies by Rosenberger and Loomis (2001) included 13
studies and 59 value estimates associated with waterfowl hunting, a popular activity in transitional
ecosystems. On average, these studies estimated that the value of waterfowl hunting is about $32 per
person per day (1996 dollars). However, different value estimates ranged from about $2 to over $160 per
person per day. In their meta-analysis, Rosenberger and Loomis (2001) predicted a $40 value per person
for a day for waterfowl hunting (Table F-4). These results demonstrate potential benefits from waterfowl
hunting, but do not address changes in the quality of transitional ecosystems that may possibly be
associated with reduced NOx/SOx emissions.
F-19
-------
F.4.2. Limitations and Uncertainties
Meta-analyses of wetland value (and the underlying original studies) are not directly useful because
they do not measure changes in ecosystem services that would follow from some tightening or loosening
of the standard for NOx or SOx. Nevertheless, they have indirect use by pointing out the importance of
various methodologies, the overall values commonly assigned to wetland improvement, and types of
ecosystem services that might be affected by air pollutants. For example, both Brander et al. (2006) and
Woodward and Wui (2001) found that CV studies generally yielded greater value per acre than travel cost
studies. This is not surprising, as the former would generally include non-use values. Furthermore,
Woodward and Wui's (2001) meta-analysis showed that the availability of bird hunting, bird watching,
and amenity services affect wetland value. For example, bird watching (which could be affected by bird
populations which, in turn, could be affected by nutrient enrichment) contributes more value to an
ecosystem than any of the other ecosystem services. This is mainly due to the popularity of bird watching
and the large numbers of people who engage in this recreational activity.
F.5. Valuation of Aquatic Ecosystems
As discussed in the ISA and the preceding annexes, acidification and eutrophication are the two
main effects of NOx and SOx on aquatic ecosystems. In economic valuation of the effects of NOx and
SOx deposition on aquatic ecosystems, these effects are reflected partly in the WTP for recreational
fishing (with effects on the catch rate and fishing quality in a particular aquatic ecosystem) and partly on
aesthetic and non-use (and total) values. Valuation methods used to assess the value of aquatic ecosystems
include contingent valuation (CV) (open-ended, discrete choice, etc.), choice experiments, hedonic -
pricing valuation, travel cost (TC) (hedonic travel cost and traditional travel cost), cost-effectiveness
analysis, avoided-damage cost, replacement cost, market analysis, and so forth. The CV method has
gained popularity for total (use plus non-use) value estimation, whereas the travel cost method plays an
important role in deriving use values. Currently, there is also a tendency to use stated and RP techniques
together for the valuation of recreational activities. Supplemental to the below discussions about valuation
of acidification and eutrophication on aquatic ecosystems, Table F-9 describes different aquatic valuation
studies according to the ecological effect examined (acidification, eutrophication) and lists their relevant
details.
An extensive literature review by Wilson and Carpenter (1999) provides an excellent starting point
to the valuation literature on aquatic ecosystems. This paper examined 30 studies published between 1971
and 1997 that estimated the value of nonmarket ecosystem services provided by fresh water bodies in the
U.S. Sixteen were SP studies, with the rest split between travel cost and hedonic pricing approaches. The
travel cost studies covered very specific water bodies and stressors, such as a boaTable to swimmable
change in water quality for 13 recreation sites along the Monongahela River in Pennsylvania (Smith and
Desvousges, 1986), and a change in water quality measured by the Uttormark's Lake Condition Index for
recreational use of Pike Lake in Wisconsin (Bouwes and Schneider, 1979). Beyond spatial and stressor
differences, the studies differ in their units for expressing value. The Monongahela River study used
benefits per household, with the largest value expressed for a change from boatable to swimmable ($51).
The Bouwes and Schneider (1979) study concluded that the total (aggregate) annual mean CS was almost
$86,000.
There are several additional meta-analyses that describe methods of valuation for aquatic
ecosystems, but not specifically relating to effects caused by NOx or SOx. Van Houtven et al. (2007)
analyzed WTP for fresh water quality improvements, covering 131 valuation estimates from 18 SP studies
that used or could be transformed to use the water quality ladder (boatable, fishable, swimmable)
modified into a 10-point scale. The analysis started with 90 studies but found that most were not
F-20
-------
sufficiently comparable for statistical analyses. The studies ranged from 1977 to 2003 and covered rivers,
lakes and estuaries.
A meta-analysis by Johnston et al. (2005) examined the value of fish catch based on a total of 48
studies published in the U.S. and Canada between 1977 and 2001. The study statistically examined WTP
for fish relative to variations in resource, context, angler, and policy attributes, as well as methodological
attributes of the studies themselves. The attributes examined included species targeted, geographic region,
water body type, catch rate, angler demographics, and fishing method. Among 391 WTP observations,
122 were estimated using CV methods, 59 using TC methods and the remaining from discrete choice
models.
Johnston et al. (2005) found that SP studies have generally resulted in lower WTP estimates than
RP studies, which is consistent with earlier multiple study reviews and meta-analyses. Dichotomous
choice CV also has produced lower WTP estimates than choice experiment or conjoint surveys, compared
to the default of open-ended surveys (including payment cards and iterative bidding). However, these
findings were not robust across all regression model specifications. The study supported previous findings
(Poe et al., 2000) that WTP is systematically associated with variations in resource, context, and angler
attributes. The study also concluded that WTP per fish varies systematically across study methods, but the
variation due to methodology accounted for a relatively small proportion of total WTP variation. Again,
these results demonstrate the value of recreational fishing, but do not address changes in water quality
that may be associated with reduced NOx/SOx emissions.
F.5.1. Acidification
The Adirondack Park in New York is the best documented of all areas affected by acidic deposition
in the U.S. Due to air pollution that largely originated from power plants and other sources to the west
and southwest of the park, many watersheds in the park have experienced slow acidification of water and
soil since the late 18th century. This park has been the subject of numerous valuation studies in recent
decades.
Morey and Shaw (1990) applied the travel cost model and estimated recreational fishing values
associated with water quality change resulting from acidic deposition in the Adirondack Park area. The
results showed that the aggregate expected conditional compensating variation for a 50% increase in catch
rate of two trout species at four sites was $3,863 (1977 dollars; $13,175 in 2007 dollars), based on 607
survey respondents with an estimated standard deviation of $89 ($304 in 2007 dollars).
A travel cost model was also applied by Mullen and Menz (1985) to link acidic deposition and
recreational fishing in the Adirondack Park. A net average economic value per angler day was calculated
to be $20 (1976 dollars; $73 in 2007 dollars) for lakes, ponds, and streams in the Adirondacks, with an
estimated total loss in net economic value of $1.07 million (1976 dollars; $3.89 million in 2007 dollars)
per year for a 5% reduction in fishable area. These estimates excluded streams because limited knowledge
existed to assess the effect of pH fluctuations on aquatic life in streams. Shaw (1989) questioned the
reliability of these results, arguing that the demand equation was not compatible with economic theory
and that aggregating separate CS measures from different equations was inappropriate.
Englin et al. (1991) assessed the economic impact on recreational fishing in four upper northeastern
states resulting from acidic deposition control. Biological effects of acidification were linked with
hedonic travel cost and random utility models through the acid stress index (ASI) and fish catch per unit
effort (CPUE). The analysis was based on multiple data sources and yielded positive social welfare for
2030 under a scenario of reducing by 50% the ambient deposition in 1989. This level of decline in acidic
deposition was expected to occur on the basis of analyses conducted by the National Acid Precipitation
Assessment Program (NAPAP).
Cameron and Englin (1997) measured welfare under an assumption that people are not certain
about participation in fishing. Based on WTP results from a survey and a random utility model, surpluses
F-21
-------
were estimated for preventing a 20% loss in the availability of high-elevation lakes for fishing in
northeastern states.
Randall and Kriesel (1990) employed a SP method to value a National Pollution Control Program,
which led to improved air and water quality by 25% within 5 years. A nationwide survey conducted in
multiple modes found an estimated annual willingness to pay of $694 ($1,098 in 2007 dollars) per
household. The researchers concluded that the estimated value was lower than their multimarket hedonic
estimates.
Banzhaf et al. (2006) conducted a CV survey, which was designed to estimate the total value (use
plus non-use values) of reducing acidification by liming the lakes and forests in the Adirondack Park to a
degree matching recent SOx and NOx reductions under emissions trading programs. The mean WTP for
improving 600 lakes "of concern," and small increases in populations of two bird species and one tree
species was $48 (2003 dollars; $54 in 2007 dollars) per year per New York household. The mean WTP for
improving 900 lakes "of concern," four bird species and three tree species was $54 (2003 dollars; $61 in
2007 dollars) per year per New York household.
Sequential implementation of several regulations, including in particular the CAA, has
substantially reduced acidic deposition, especially in the eastern part of the U.S. Burtraw et al. (1997)
conducted a benefit-cost analysis for the NAPAP program. Besides improving public health and visibility,
emissions controls contributed to decreased lake acidification which was projected to have economic
benefit associated with improved recreational fishing in the Adirondack Park. The ASI and CPUE
approaches were used in the Tracking and Analysis Framework (TAF) model to capture the response of
three recreationally fished species to water quality improvement in 33 statistically selected Adirondack
lakes. The per capita recreational fishing benefits in 2010 were estimated to be $0.62 (1990 dollars; $0.98
in 2007 dollars) annually per angler fishing in the Adirondack area.
F.5.2. Eutrophication
As is the case for acidification, the endpoints most studied in the valuation literature for nutrient
enrichment concern fishing and non-use (or total) values. Change in aquatic recreational behavior is
another endpoint for looking at valuation of such ecosystems, although it is difficult to link those changes
due to ambient NOx or SOx concentrations.
F.5.2.1. Recreation
Bockstael et al. (1989b) measured the benefits of improvements in water quality of the Chesapeake
Bay using travel cost and SP methods. A telephone survey with 959 respondents revealed an aggregate
WTP for water quality improvement from current level to a level acceptable for swimming and/or other
water activity of about $91 million (1987 dollars; $166 million in 2007 dollars) for households in
Maryland. Based on three other surveys, a travel cost model yielded aggregate use values of about $34.7,
$4.7, and $1.4 million ($63.1, $8.6, and $2.6 million in 2007 dollars) for a 20% water quality
improvement for beach use, boating, and sport fishing, respectively.
Morgan and Owens (2001) simulated water quality change under a baseline (without additional
pollution control) and a scenario with pollution control in the Chesapeake Bay watershed and then
calculated aggregate benefits for beach use, boating, and striped bass sport fishing by transferring the
benefits from Bockstael et al. (1989b). Lower bound aggregate benefits for beach use, boating, and sport
fishing were $288.8, $6.7, and $288.8 million (1996 dollars; $380.4, $8.8, and $380.4 million in 2007
dollars), respectively for a 60% improvement in Chesapeake Bay water quality.
Lipton (2003) mailed out an open-ended contingent valuation survey to 2,510 randomly selected
Maryland boaters regarding willingness to pay for one unit improvement in the water quality of the
Chesapeake Bay. The boaters ranked their perception of water quality on a scale of one to five. The
F-22
-------
median WTP for a one step improvement in water quality was $17.50 ($19.66 in 2007 dollars) per year
and the mean was $63 ($71 in 2007 dollars), with 38% of respondents expressing a zero WTP. It was
found that the boaters who keep their boats in the water would pay more than those who keep them on
trailers. Individuals who ranked ambient water quality lower and had more concern for the health effects
from water quality degradation would likely pay more for water quality improvement. The aggregated
WTP for the Chesapeake Bay boaters in Maryland was approximately $7.3 million ($8.2 million in 2007
dollars) per year with a $146 million ($164 million in 2007 dollars) present value (a 5% discount rate).
Smith et al. (1991) estimated the recreational fishing value of Albemarle-Pamlico Estuary in North
Carolina using a hedonic travel cost model. The analysis was based on an intercept survey with 1,012
interviews at 35 boat sites and 44 bank sites in 1981 and 1982. The estimated benefit derived from the
application of a conventional demand model to the boat site sample ranged from $1.49 to $2.58 (1982
dollars; $3.19 to $5.53 in 2007 dollar), and for the bank site sample ranged from $0.65 to $1.11 ($1.39 to
$2.38 in 2007 dollars) for an increase in catch rate of one fish per hour per person.
Four years later, Kaoru (1995) and Kaoru et al. (1995) used the information collected from the
same survey to estimate the recreation value from improvements in Albemarle and Pamlico Sounds using
a household production model and a random utility model. They linked the effluent loading and quality of
sport fishing with the fisherman's decision on site selection. Instead of fish availability being estimated as
the average number of fish caught per person per hour for each entry point, Kaoru et al. (1995) used total
catch per trip per person as their key measure of fishing success. They included not only the estimated N
loadings but also the effect on biochemical 02 demand as an influence on fishing quality. The results
suggested that depending on aggregation methods, CS ranged between $7.05 and $36.19 ($9.56 and
$49.08 in 2007 dollars) for a 5% increase in total catch. The CS was $1.27 to $6.52 ($1.72 to $8.84 in
2007 dollars) for a 36% decrease in N loadings at all sites.
Whitehead et al. (2000) used a joint stated and RP model to estimate recreation value of
improvements to Albemarle and Pamlico Sounds in terms of CS per trip and per season. A telephone
survey was conducted in 1995 to learn how North Carolina residents would value improvements of water
quality resulting in a 60% increase in fish catch and a 25% increase in open shellfish beds. The analysis
was based on 765 completed responses. The CS per trip was estimated at $64 (1995 dollars; $87 in 2007
dollars) for current quality and $85 ($115 in 2007 dollars) for improved quality and the CS per season
was $121 ($164 in 2007 dollars) for current and $155 ($210 in 2007 dollars) for improved quality.
The Catawba River in North and South Carolina is used for electric power generation, recreation,
drinking water, and wastewater assimilation. With population growth and land use change, the health and
vitality of the river system are declining as the river flows downstream. Through a mail-phone survey of
1,085 residents randomly selected from 16 counties of Catawba River Basin, Kramer and Eisen-Hecht
(2002) presented a management plan designed to maintain the water quality at the current level over time.
The water quality in the entire watershed was classified as good, fair or poor on maps to illustrate the
distribution of the tributaries with different water quality. The mean WTP was estimated to be $139 ($160
in 2007 dollars) for protecting current water quality from deterioration.
A site choice model applied jointly with a trip frequency model was used by Needelman and Kealy
(1995) to assess the relative benefits of eliminating eutrophication, fecal bacteria, and oil and grease in
New Hampshire lakes. The site choices analysis was based on 53 individuals and 1,021 trips, while the
trip frequency model was based on a total of 519 individuals including the responses for the site choices
model. The mean seasonal benefits were valued at $1.40 ($1.90 in 2007 dollars) for eliminating
eutrophication from all sources ($1.33; $1.80 in 2007 dollars) for eliminating nonpoint source pollution
alone and $0.09 ($0.12 in 2007 dollars) for eliminating point source pollution alone) with an aggregate
seasonal benefit of $1.16 million ($1.57 million in 2007 dollars), with $1.11 million for nonpoint source
and $0.08 million for point source pollution. However, the economic benefit estimates were exclusively
for swimming and day trips. The measures of water quality were not from the year of the survey (1989)
but from a range of years from 1976 to 1991.
The Tar-Pamlico River has experienced declining fish catches, disease in fish and shellfish beds,
algae blooms, and aquatic grass losses. More than half of the pollution impairing one third of the river
F-23
-------
was estimated to come from agricultural nonpoint sources. Whitehead and Groothuis (1992) proposed a
management program in which farmers used best management practices (BMPs) to significantly improve
the water quality of the Tar-Pamlico River such that recreational anglers would be able to catch twice as
many fish per trip. A mail survey was sent to 179 households in four counties in Tar-Pamlico River basin
(with a 61% response rate). The mean WTP for doubling fish catch was estimated at $25 ($38 in 2007
dollars), which was bounded from above by a $35 use value ($53 in 2007 dollars) and from below by a
$21 ($32 in 2007 dollars) non-use value. The benefits of water quality improvement in the study area
were aggregated at $1.62 million (1991 dollars; $2.46 million in 2007 dollars) each year. The researchers
believed that non-use value could account for as high as 84% of the estimate. However, the study suffered
from small sample size and relatively high non-response rate (11%) to the WTP questionnaire.
Some portions of the coastal coral reefs in the Florida Keys are projected to disappear within 10 to
25 years. In 2000, the U.S. government announced a long-term plan to save coral reefs proposing that
20% of all coral reefs in American-controlled waters would become ecological preserves by 2010. Park
et al. (2002) investigated the WTP to preserve current water quality and health of the coral reefs in the
Florida Keys. Based on 460 responses to a CV survey and 4,035 respondents to a travel cost survey (460
responses used in the analysis), the annual use value was estimated at $481 per person per snorkeling trip
($553 in 2007 dollars). The mean predicted WTP per trip from a Tobit model was $735 ($844 in 2007
dollars). Over 85% of the predicted WTP value was within plus or minus $50 ($57 in 2007 dollars) of the
total trip expenses from the contingent valuation scenario.
F.5.2.2. Commercial Fisheries
As one of six background studies for the National Science and Technology Council (NSTC), Diaz
and Solow (1992) used a time series estimation approach to examine the effects from hypoxia1 in the Gulf
of Mexico. The study failed to confirm the relationship between the annual occurrence of hypoxia and
commercial fishery health, based on catch rate per unit effort of three major species from the 1960s to
1990s. Although the benefit assessment did not detect effects attributable to hypoxia, this does not
necessarily mean that the economic effects would not occur. NSTC (2000) identified alternatives for
reducing the adverse effects of hypoxia and examined the costs associated with reduction of N and P
inputs. A net cost estimated by the U.S. Mathematical Programming Model for Agriculture was about
$0.8 per kilogram ($0.36 per pound; $0.96 per kilogram/$0.36 per pound in 2007 dollars) for a 20% edge-
of-field N loss reduction on agricultural lands, whereas restoring 5 million acres of wetland would have a
net cost of $1.00 per kilogram ($0.45 per pound) ($1.2 per kilogram/$0.54 per pound in 2007 dollars) of
N removal.
The Neuse River in North Carolina is important to the commercial blue crab (Callinectes sapidus)
fishery in the eastern U.S. It accounted for about a quarter of the blue crab harvest from 1994 to 2002.
Smith and Crowder (2005) simulated the progress of eutrophication in the Neuse River using a series of
ordinary differential equations, which linked changes in the quantity of nutrients, algal growth, spatial
population distribution of blue crab and its prey species with fishing efforts. Results suggested that a 30%
reduction in N loading in the Neuse River watershed over a 50-yr period would result in about a $2.56
million ($2.71 million in 2007 dollars) discounted present value generated in fishery rent (the difference
between fishing revenues and costs including fixed and opportunity cost).
1 Hypoxia (DO depletion) is a phenomenon resulting from the overloading of nutrients (usually N or P) in water.
Most fresh waters are P-limited, and therefore added N from atmospheric deposition does not have a substantial
effect on productivity. In contrast, marine and estuarine waters tend to be N-limited, and are therefore expected to
respond to additional N inputs by increasing algal productivity. Excessive production of algae can deplete the water
of oxygen when those algae die and are decomposed by oxygen-consuming microorganisms. If the concentration of
DO decreases to very low levels, fish, and other life forms can die.
F-24
-------
F.5.2.3. Water Clarity
In addition to effects on fish populations, eutrophication reduces water clarity due to excessive
growth of algae. Boyle et al. (1999) studied the social welfare related to water clarity of four lakes in
Maine employing a two-stage hedonic demand model. Based on data from 1990 to 1995 on property sales
for 25 lakes, tax records, mail survey, and water clarity data from the Maine Department of
Environmental Protection (DEP), the CS for water clarity improvement from the average ambient level
(3.78 m) to 5.15 m was estimated to be $3,677, $3,765, and $12,870 ($4,562, $4,671, and $15,967 in
2007 dollars) for differently specified demand systems (Cobb-Douglas, semilog, and linear demand
models, respectively). Social welfare loss for visibility deterioration from an average level to 2.41 m
ranged from $25,388 to $46,750 ($31,497 to $57,989 in 2007 dollars). One of the interesting findings in
the study was that the slope of the Cobb-Douglas demand model increased to 3.0 m, a threshold used by
the Maine Department of Environmental Protection to indicate improved water quality and public
preferences for it.
Poor et al. (2006) evaluated water quality in a small watershed of Maryland using a hedonic
property valuation model. The watershed is located in a peninsula surrounded by the Potomac and
Patuxent Rivers and the Chesapeake Bay. Due to nonpoint source pollution runoff, water quality has
deteriorated, which might have negative impacts on residential housing prices, especially for those close
to the river. To estimate the possible extent of the impact, total suspended solids (TSS) and dissolved
inorganic nitrogen (DIN) were averaged by year and included in the model as indicators of ambient water
quality. The estimates showed that a marginal increase in TSS reduced average housing prices by $1,086
($1,113 in 2007 dollars), and a marginal increase in DIN decreased housing prices by $17,642 ($18,087 in
2007 dollars).
F.5.3. Avoided Costs
Based on a model projection, IEc (1999c) estimated the benefits of decreased N deposition to the
estuaries in the eastern U.S. using the avoided cost method. According to their results, the annual avoided
cost in 2010 would range from $26 to $102 million if 12.8 million pounds of N loading was reduced
annually in Long Island Sound. An annual reduction of 58 million pounds of N loading into Chesapeake
Bay would avoid an annual cost of $349 to $1,278 million. Uncertainty associated with model
assumptions and the inappropriate use of avoided cost estimates to value benefits or damages are two
major sources of potential error in generating such estimates. In addition, the avoided costs method
generally does not measure values conceptually accurately, and these values also are not linked to benefits
from reduced pollution.
F.5.4. Limitations and Uncertainties
Several general limitations apply to the valuation of water quality changes. First, the definition of
water quality is too ambiguous to quantify comparisons across studies. Second, the degree of water
quality improvement is not often clear. A common obstacle in any environmental economic valuation is
the availability of data. Problems stemming from lack of data extend from biological data on the
populations of target species to limitations of the available economic data on the value of commercial and
recreational fisheries to small samples of survey respondents (Smith and Crowder, 2005).
The geographical focus of valuation studies is another limitation. The studies reviewed here largely
focused on the eastern part of the U.S., represented by the Adirondacks, Chesapeake Bay, and Albemarle
and Pamlico Sounds. Several studies investigated the total value of aquatic ecosystems and most the
studies calculated recreational values of water quality improvement. Only a few studies addressed the
F-25
-------
economic value of commercial fisheries. There is considerable uncertainty in the estimates of benefits as
they vary significantly even for studies in the same area, with similar changes in the commodity, and use
of the same valuation methodology.
Almost every study reviewed here mentioned the problem of uncertainties about the natural science
(Diaz and Solow, 1992; National Science and Technology Council, 2000). For example, the processes
involved in the development of hypoxia are not fully known. There is also no clear connection between a
decrease in pollution level and an increase in catch rate (Bockstael et al., 1989b). Additionally,
uncertainty arises from the selection of parameter values (Smith and Crowder, 2005) and models (Burtraw
et al., 1997). Due to the uncertainty about water quality improvement in tributaries of the Chesapeake
Bay, the estimates provided by Morgan and Owens (2001) excluded benefits from water recovery in the
tributaries.
Study coverage also affects benefit estimates. For example, Morey and Shaw (1990) only included
those fishermen who could make day trips to one or more of the study sites, fishermen with complete
records, and fishermen whose distance to the farthest site was less than 200 miles. Morey and Shaw
(1990) evaluated four sites and two kinds of trout in the Adirondacks while Englin et al. (1991) valued
only one trout species and lakes in four states.
Many studies relied on information from surveys, which may suffer from various biases. For
example, several studies reviewed here used intercept surveys, which are not necessarily representative of
the target population (Bockstael et al., 1989b; Smith et al., 1991). Double counting of values may also
introduce errors into value estimation. For example, people who are boating may also go fishing
(Bockstael et al., 1989b).
F.6. Summary
Previous ecosystem valuations presented in AQCDs were very limited because they considered
only studies that could directly attribute monetary values to changes in emissions. The assessment of the
literature in this Annex has a different approach: the voluminous literature that values various ecosystem
services affected by emissions is included, whether or not the actual linkages all the way from emissions
to those effects have been quantified. This approach nevertheless necessitates that the natural science
underpinnings be examined in the context of preferences, that is, descriptions of damages (or benefits)
from NOx and SOx emissions (or reductions in those emissions) from the natural sciences that translate
into things that the public cares about. These include, for example, whether the water is boatable or
swimmable, the marketable yield associated with changes in a forest or a crop, and effects on aesthetics of
wetlands.
The physical endpoints and their corresponding valuation studies are divided into different
ecological and value endpoints applicable to terrestrial, transitional, and aquatic ecosystems. There is
significant valuation literature on the effects of O3 on crops (and to a far lesser degree on forest yield), but
this topic is beyond of the scope of this assessment. Beyond the O3 effects, there is little quantification of
the science describing the effects of pollution related to NOx and SOx on ecosystems. Valuation studies
are themselves classified into meta-analyses and original studies, the latter into market studies (e.g.,
commercial fishing), RP studies (e.g., those related to recreation behavior, property values, etc.), and SP
studies (those that ask people survey questions about their WTP for hypothetical ecosystem
improvements).
For valuation of terrestrial effects, survey methods are most common. Supplemented by travel cost
approaches, this literature leads to a variety of estimates of WTP for improved forest quality that could
prove useful for estimating benefits of N and S reductions (provided a number of linkages can be made),
even though some of the endpoints valued are related to insect damage. One meta-analysis is available
that summarizes this valuation literature.
F-26
-------
For wetland valuation, three meta-analyses are available that summarized the valuation literature.
This literature is problematic because it focuses on values per wetland acre rather than WTP for changes
in services provided by a wetland affected by deposition of N and S pollutants. However, some of the
available studies provide WTP for discreet ecosystem services, which is helpful in matching these values
to services affected by reductions in N and S deposition. Nevertheless, because there is a very poor
understanding of the scientific basis for linking NOx and SOx emissions to ecosystem health endpoints,
the ability to make the necessary linkages to estimate benefits of pollution reductions is very limited.
The aquatic service valuation literature is the most voluminous of the three categories reviewed. It
contains many recreation value studies, a number of property value studies, some total (use plus non-use)
value studies, some studies on commercial fishery damages and several good meta-analyses of primary
valuation studies, which provide estimates of the WTP of households for improvements in the aquatic
ecosystem related to N and S changes. Again, however, there is little focus on NOx or SOx as the cause of
alterations to the aquatic ecosystem.
Overall, there is a robust literature valuing a variety of ecosystem services that could be related to
reductions in N and S emissions. Therein addition, issues affecting the credibility of any individual study
and even the studies grouped by technique, such studies can only be used for general valuation purposes.
The most important limitation is establishment of the linkages between the physical, chemical, and
biological effects of air pollutants on natural ecosystems and changes in exposure to NOx and SOx.
Table F-1. Commonly adopted environmental valuation methods based on revealed or state preferences.
Methods
Description
Direct
Use
Indirect Total
Use Value
Revealed Preference Observing individual choices and behavior to predict their preferences for environmental
Methods goods and services.
Avoided expenditure
method
Predicting the cost of mitigating the effects of reduced environmental quality.
X
X
Derived demand
functions
Estimate the value of environmental goods and services by deriving the demand
functions of households or firms for them (e.g., water use)
X
X
Hedonic valuation
method
Estimating an implicit price for the environmental quality attributes of marketable goods,
such as housing.
X
X
Market analysis
Used for valuing market goods using data on prices and quantities of outputs and
inputs. May use prices of close substitutes, methods of deriving shadow prices, or
simulation of changes in market conditions.
X
X
Referenda
Examining voting results related to environmental resources to predict values for them.
X
X
Replacement cost
Measures expenditures incurred in replacing or restoring environmental good or service
lost (e.g., water filtration). Provides accurate valuation only under strict assumptions
usually not met.
X
X
Travel cost method
Values the environmental attributes of recreational sites by examining visitation
frequency and cost differential incurred in reaching site with different attributes.
X
X
User fee
Examine user fees paid to gain access to an ecological resource such as a park to
estimate the lower bound of society's value for that resource.
X
X
Stated Preferences
Approach
Directly surveying individuals to predict their preferences for environmental goods and
services.
F-27
-------
Methods
Description
Direct
Use
Indirect
Use
Total
Value
Choice experiments,
conjoint analysis
Elicits individuals' choices from several alternatives associated with different
environmental and cost outcomes. Elicits data on WTP indirectly (by eliciting cost-
environmental outcome tradeoffs, similar to contingent ranking).
X
X
X
Contingent valuation
Elicits data on individuals' willingness to pay for environmental goods and services
(hypothetical setting, limited by potential biases). Often implemented using a
hypothetical referenda.
X
X
X
Contingent ranking
Elicits data on the ranking of several alternatives associated with different
environmental and cost outcomes. Elicits data on WTP indirectly (by eliciting cost-
environmental outcome tradeoffs, similar to choice experiments, conjoint analysis).
x
X
X
Table F-2. Economic effects of ozone and other pollutants on agriculture, as reported in the 1996 ozone
criteria document.
Model Features
Results (Annual 1980 U.S. Dollar)
Study Region
Pollutant and
Concentration
Price
Changes
Output
Sub-
stitu-
tions
Input
Substi-
tutions
Quality
Changes
Crops
Consumer
Benefits
Producer
Benefits
Total
Benefits
(Costs)
Adams U.S.
et al.
(1986a,
1986b)
Ozone, 25% reduc- Yes
tion from 1980 level
for each statec
Yes
Yes
No
Corn, soy-
beans, cot-
ton, wheat,
sorghum,
barley
1160x106 550x106 1700x106
Kopp U.S. Ozone, universal
etal. reduction from 53 ppb
(1985)b to 40 ppba
Yes
Yes
Yes
No
Corn, soy-
beans,
wheat, cot-
ton
Not reported Not re- Not re-
ported ported
Adams U.S.
et al.
(1986b)
Acid deposition, 50%
reduction in wet
acidic deposition
Yes
Yes
Yes
No
Soybeans 172x106
-30x106 142x106
Adams U.S.
and
Crocker
(1989)
Ozone, seasonal
standard of 50 ppb
with 95% compli-
anced; Includes
adjustments for 1985
Farm Bill
Yes
Yes
Yes
No
Corn, soy-
beans, cot-
ton, wheat,
sorghum,
rice, hay,
barley
905x106
769x106 1674x106
8 Seven hour growing season geometric mean. Given a long-normal distribution of air pollution events, a 7-h seasonal ozone level of 40 ppb is approximately equal to an hourly
standard °f 80 ppb, not to be exceeded more than once a year (Heck et al., 1983).b Reported in 1986 criteria document.c Reported All studies except Garcia et al. use NCLAN data to
generate yield changes due to ozone.d Seven- and twelve-h growing season geometric mean. Analysis includes both fixed rollbacks (e.g., 25% and seasonal standards (with variable
compliance rates). Source: Adams and Horst (2003).
F-28
-------
Table F-3.
Economic effects of ozone on marketable benefits from forests.
Reference
Pollutant/Coverage
Response and Air
Quality Data
Economic Model
Annual Damages or
Benefits of Control
Callaway
etal. (1986a)
All pollutants. Forest
products (hardwood and
softwood) in the eastern U.S.
Assumes three arbitrary growth
reductions (10%, 15%, and 20%)
for hardwood and softwood tree
species.
Spatial equilibrium
models of softwood
and hardwood
stumpage and forest
products industries in
the U.S.
$340 to $510 million;
damage in 1984 dollars for
assumed reductions in
growth levels.
Crocker
(1985)
Acid deposition.
Forest products and forest
ecosystem service flows for
eastern U.S.
Assumes a 5% reduction in
products due to acid deposition:
assumes a pristine background pH
of approximately 5.2.
Naive; assumed
changes in output
multiplied by avg
value of those goods
or services.
$1.75 billion damage in 1978
dollars from current levels of
acid deposition.
Crocker and
Forster
(1986)
Acid deposition. Forest
products and forest
ecosystem services for
eastern Canada.
Assumes 5% reduction in forest
productivity for all eastern
Canadian forests receiving $10
kg/ha/yr sulphate deposition.
Naive; assumed
changes in output
multiplied by avg
value of goods or
services.
$1.5 billion damage in 1981
Canadian dollars from
current levels of acid
deposition.
Haynes and
Adams
(1992)
Air pollutants, including acid
precipitation. Losses
estimated for eastern U.S.
softwoods.
None; paper demonstrates a
methodology for assessing
economic effects of yield (growth
and inventory) reductions due to
any course. Assumes losses from
6% to 21 % for softwoods.
Econometric model
of U.S. timber sector
(TAMM).
Damages of $1.5 to $9.0
billion in 1988 dollars.b
Bentley and
Horst (1998)
Ozone. All hardwoods and
softwoods except western
hardwoods.
Dose-response based on survey of
experts scaled to partial attainment
of secondary standard relative to
the current primary standard.
SUM06 exposure metric is based
on a cumulative daytime exposure
during the growing season.
Econometric model
of U.S. timber sector
(TAMM).
Benefits of $14 million in
2010 (1990 dollars) for
partial attainment of 0.08
ppb 3rd max secondary
standard. This standard was
considered during the
standard development
process.
aWe used the updated version of Crocker (1986), while Adams and Horst (2003) cited a prior version, from Crocker (1985).
bWe drew different numbers from the Crocker and Forster (1986) study than those reported by Adams and Horst (2003).
Source: Adams and Horst (2003), exceptions: Table 2 in Bentley and Horst (1998).
F-29
-------
Table F-4.
Forecasted average values for select activities, per day per person in 1996.
Activity
Northeast
Southeast
Intermountain
Pacific Coast
Alaska
USA
Swimming
14
9
19
9
14
14
General recreation*
30
25
34
25
30
30
Fishing
37
32
41
32
37
37
Waterfowl hunting
40
35
44
34
40
39
Big game hunting
45
40
50
40
45
45
*The activity category includes camping, picnicking, sightseeing, hiking, small game hunting, wildlife viewing, and other general recreation. In the meta-analysis, values for these activity
categories do not differ statistically significantly. Source: from the meta-analysis benefit function by Rosenbergerand Loomis (2001), Table 6.
Table F-5. Typical impacts of specific pollutants on the visual quality of forests.
Pollutant
Geographic
Extent
Injury
Type
Major Types of Visual Injuries
Notes
Ozone
Area or
regional effects
Direct
injuries
Foliar injuries (e.g., pigmented stipple), increased
needle/leaf abscission, premature senescence of
leaves. Pattern, size, location, and shape of foliar
injuries to indicator species can be specific for ozone.
Indirect
Injuries
Increased susceptibility to visual injuries may result
from other adverse environmental factors, such as
insect attacks. For example, increased needle/leaf
abscission, elevated mortality rates, and/or changes in
species composition.
Acidic
Deposition
Area or
regional effects
Indirect
Injuries
Increased susceptibility to visual injuries may result
from other adverse environmental factors, such as
climatic factors. For example, increased needle/leaf
abscission, elevated mortality rates, and/or changes in
species composition.
Acidic deposition can also cause
direct foliar injuries. Acids are,
however, more likely to indirectly
affect the visual appearance of forest
trees, unless exposure levels are
very high.
Sulfur
Dioxide
Point source
pollution
Direct
Injuries
Foliar injuries including leaf/needle discoloration and
necrosis. Pattern, size, location, and shape of foliar
injuries to indicator species can be specific for sulfur
dioxide. At high concentrations, elevated mortality rates
of sensitive species and changes in species
composition may occur.
Sulfur dioxide may also cause
indirect injuries. Indirect injuries,
however, are not well documented.
Hydrogen
Fluoride
Point source
pollution
Direct
Injuries
Foliar injuries including leaf/needle discoloration and
necrosis. Pattern, size, location, and shape of foliar
injuries to indicator species can be specific for sulfur
dioxide. At high concentrations, elevated mortality rates
of sensitive species and changes in species
composition may occur.
Hydrogen fluoride may also cause
indirect injuries. Indirect injuries,
however, are not well documented.
Source: Exhibit 2 Industrial Economics (lEc) (1999a)
F-30
-------
Table F-6.
Economic valuation studies related to forest aesthetics.
Study
Method
Study Area
Description
Economic
Endpoint
Economic Value
Crocker
(1985)
CV (open
ended)
Southern
California
Evaluating visits to
recreational sites with
slight, moderate, and
severe O3 induced
damages to Ponderosa
and Jeffery pine stands.
Household
WTP
(per trip)
Slight damage: $2.09
Moderate damage: $0.66
Severe damage: $0.74
Haefele
etal. (1992)
CV (payment
card, dicho-
tomous choice)
Southern
Appalachian
Mountains
Protect high-elevation
spruce firs from insect and
air pollution
Household
WTP (per
year)
PC: $21
DC: $100
Holmes and
Kramer
(1996)
CV (payment
card,
dichotomous
choice)
Southern
Appalachian
Mountains
Protect high-elevation
spruce firs from insect and
air pollution. Analysis
compared forest users and
non-users.
Household
WTP (per
year)
Forest users: $36
Non-users: $10
Holmes
etal. (1996)
Hedonic pricing
New Jersey
Evaluate the effects of
hemlock forest health
status on housing prices.
Hemlock health status is
potentially deprived by
hemlock woolly adelgid; an
exotic forest insect.
Housing
price
Hemlock forest health status is positively
associated with housing prices. For
example, a 1—point increment in (e.g.,
from 10% to 11 %) in the % of healthy
hemlocks of all forests on the home
parcel was associated with a 0.66% sales
price increase. Similar changes in home's
near proximity are associated with yet
larger price increments.
lEc
(1999b)b
Benefits
transfer
National
Protect trees from various
different types of damage
(see description of source
studies)
Welfare Loss
(1990-2010)
$3 to $17 billion
Jakus and
Smith
(1991)
CV
(dichotomous
choice)
Maryland and
Pennsylvania
Protect private homeowner
property and surrounding
public lands from gypsy
moth damages (reduce
tree defoliation by 25%)
Household
WTP
(for entire
program)
Private Property Prgm (only): $254 to
$420
Private and Public Prgm:
$314 to $527
Jenkins
(2002)
CV (open
ended)
Southern
Appalachian
Mountains
Protect high-elevation
spruce firs from insects
and air pollution along
roads and throughout
ecosystem
Household
WTP (per
year)
$153
Kramer
etal. (2003)
CV
(dichotomous
choice)
Southern
Appalachian
Mountains
Protect high-elevation
spruce firs from insects
and air pollution along
roads and throughout the
entire ecosystem
Household
WTP (per
year)
Road side only: $18
Entire ecosystem: $28
Non-use: 87% of total value
Use: 13% of total value
Kramer and
Mercer
(1997)
CV (payment
card and dichot-
omous choice)
National (USA)
Protect 10% of tropical
rainforests as national
parks or forest reserves.
Household
WTP (per
year)
PC: $31
DC: $21
F-31
-------
Study
Method
Study Area
Description
Economic
Endpoint
Economic Value
Loom is
etal. (1996)
CV (dicho-
tomous choice,
open ended)
Oregon
Protect old growth forests
of Pacific Northwest from
fires.
Household
WTP (per
year)
DC: $98
OE: $33
Leuschner
and Young
(1978)
TCM
Texas
Reduce crown density in
recreational areas.
Consumer
Surplus
(losses)
-0.69 to -6.5% (depending on level of
substitute sites)
Miller and
Lindsey
(1993)
CV
(dichotomous
choice)
New
Hampshire
Protect private homeowner
property from gypsy moth
damages (reduce tree
defoliation by various
amounts)
Household
WTP (per
year)
$55 to $86 (depending on level of
reduction in tree defoliation)
Peterson
etal. (1987)
CV, hedonic
property
Los Angeles
area
Avoid O3 induced
damages to trees in local
national forests for
recreationists (in greater
LA area) and homeowners
(with property bordering
forest).
Household
WTP
(per year)
Consumer
Welfare
Recreationist: $43
Homeowner: $131
$31 to $161 million
Treiman
(2006)
CV
(dichotomous
choice)
Missouri
Improve residential tree
care and maintenance in
different sized cities.
Household
WTP (per
year)
Urban areas: $14 to $16
Walsh et al.
(1989)
TCM
Colorado
Rockies
Reduce tree density in
recreational areas.
Consumer
Surplus
(losses)
-8.5 to -23.2% (for reductions in tree
density ranging between 10 to 30%)
Walsh et al.
(1990)
CV (iterative
bidding)
Colorado
Protect Ponderosa pine
forests from damages
caused by the mountain
pine beetle.
Household
WTP (per
year)
$47
Non-use: 73% of total value
Use: 27% of total value
aHolmes and Kramer study applies same results found in Haefele et al. (1991). bBenefits transfer was based on results from Holmes and Kramer (1996), Peterson et al. (1987), Walsh
et al. (1990).
F-32
-------
Table F-7.
Summary of the monetized estimates of the annual value of forest quality changes
Reference
Aesthetic Change
Valued
Value of Change
per Household
(Current Dollars)
Value of Change
per Household
(1990 Dollars)!
Total Value of
Change for
Region (Current
Dollars)
Total Value of
Change for
Region (1990
Dollars)!
Peterson et al.
(1987a)
Ozone damage to San
Bernardino and
Angeles National
Forests
O
Is-.
csi
CO
CO
CO
$7.26 - $37.62
$27 - $140 million
$31 - $161million
Walsh et al.
(1990)
Visual damage to
Colorado's Front
Range
$47
$61.68
$55.7 million
$73.09 million
Holmes and
Kramer (1996)
Visual damage to
spruce-fir forests in
southern Appalachia
$10.81 nonusers
$36.22 users
$10.37 nonusers
$34.76 users
NA
NA
Source: Exhibit 4 Industrial Economics (lEc) (1999c).
Table F-8. Estimated value of avoiding forest damage in the U.S.
Affected System
States Included
Value per
Household
Householdsi
Total Annual
Value
Cumulative
Value
(1990—2010)ii
Sierra Nevada and Los
CA
$7.26 - $37.62
10.4 million
$75.5 million -
$1.02 billion-
Angeles Basin
$391.2 million
$5.27 billion
Eastern Spruce Fir and
ME, VT, NH, MA, NY, PA,
$7.26 - $37.62
23.2 million
$168 million -
$2.27 billion -
Selected Eastern Hardwood
VW, TN, KY, NC, VA
$872.8 million
$11.75 billion
Source: Exhibit 5 Industrial Economics (lEc) (1999c)
F-33
-------
Table F-9. Ecological wetland functions, economic goods and services, types of value, and applicable
valuation methods.
Ecological Function
Economic Goods and Services
Value
Type
Commonly Used Valuation
Method (s)a
Flood and flow control
Flood protection
Indirect use
Replacement cost
Market prices
Opportunity cost
Storm buffering
Storm protection
Indirect use
Replacement cost
Production function
Sediment retention
Storm protection
Indirect use
Replacement cost
Production function
Groundwater recharge/discharge
Water supply
Indirect use
Production function
NFI Replacement cost
Maintenance/Nutrient retention
Improved water quality
Indirect use
CVM
Waste disposal
Direct use
Replacement cost
Habitat and nursery for plant and animal
Commercial fishing and hunting
Direct use
Market prices, NFI
species
Recreational fishing and hunting
Direct use
TCM,CVM
Harvesting of natural materials
Energy resources
Direct use
Direct use
Market prices
Market prices
Biological diversity
Appreciation of species existence
Non-use
CVM
Micro-climate stabilization
Climate stabilization
Indirect use
Production function
Carbon sequestration
Reduced global warming
Indirect use
Replacement cost
Ecological Function
Economic Goods and Services
Value Type
Commonly Used Valuation Method(s)a
Natural environment
Amenity
Direct use
HP, CVM
Recreational activities
Direct use
CVM, TCM
Appreciation of uniqueness to culture/
heritage
Non-use
CVM
8 Acronyms refer to the contingent valuation method (CVM), hedonic pricing (HP), net factor income (NFI), and the travel cost method (TCM).
Source: Table 1, Branderet al. (2006). Modifications adapted from Barbier (1991); Barbieret al. (1997); Brouweret al. (1999); and Woodward and Wui (2001).
F-34
-------
Table F-10.
Economic valuation studies related to acidification and eutrophication in aquatic ecosystems.
Reference Study Area
Method
Ecological
Endpoint
Valuation
Endpoint
Economic Value
Note
ACIDIFICATION
Banzhaf
etal. (2006)
Injured lakes
in
Adirondacks,
NY
CV
Fish, bird WTP (total 1. The mean WTP for
species, value) base version was $48
and tree per year per household
species in New York State while
the mean WTP for
scope version was $54
per year per household
in New York State
(discounting rate =3%)
1. Base version:
improvement in fish
population of 600
lakes, small
improvements in the
populations of two
bird and one tree
species.
2. Scope version:
improvement in fish
population of 900
lakes, improvement
in the population of
four bird and three
tree species.
Bockstael
Chesapeake CV and TC Fish
Aggregate
1.20%/vater quality
1. TNP, product of N
et al.
Bay, MA and
benefit/
improvement resulting in
and P, was included
(1989b)
DC
consumer
aggregate benefits for
in the model.
surplus
beach use, boating, and
2 WTP was
(recreational
sportfishing were about
aggregated for
households in
Maryland.
value)
$34.7,4.7, and 1.4
millions.
2. aggregate WTP for
water quality
improvement from
current to a level
acceptable for water
activity was about $91
million (1987 dollars)
Cameron
and Englin
(1997)
Northeast
U.S.
CV and RUM
Trout fishing Consumer
(uncertain surplus
about use)
(recreational
value,
existence
value)
1. For passive user,
surplus for preventing a
20% loss in currently
fishable high altitude
lakes in the Northeast
was $218 for S model
and $215 for OP model.
2. For active user,
surplus for preventing a
20% loss in currently
fishable high altitude
lakes in the Northeast
was $283 for S model
and $439 for OP-I model
and $436
for OP-D.
1. S model: Surplus
interpretation of the
WTP responses
2. OP model: Option
price interpretation
of the WTP
responses.
3.1: complete
independence
between year-to-
year decisions
4. D: complete
dependence on
previous decisions
F-35
-------
Reference Study Area
Method
Ecological
Endpoint
Valuation
Endpoint
Economic Value
Note
Englin et al.
(1991)
ME, NH, NY
(excluding
New York
City) and VT
HTC and RUM
Catch per
unit effort
(number of
fish caught
in an h)
Consumer
surplus
(recreational
value)
1. Under baseline
scenario (1989), welfare
loss was estimated at
$0.27 m (HTC) and
$1.75 (RUM)
2. Under NAPAP
scenario 1, welfare loss
was occurred in 2010 at
$13.7 m (HTC) and $1.2
m (RUM) but social
welfare would gain in
2030 at $3.5 m (HTC)
and $5.5 m (TCM)
3. Under NAPAP
scenario 4, welfare
would gain in 2010 at $3
m (HTC) and $7.4 m
(TCM) and also gain in
2030 at $4.4 m (HTC)
and $9.7 m (TCM)
1. Acidic stress
index (ASI)
measuring fish
biological tolerance
to acidity was
included in the
model.
2. The benefits were
valued in 1989
dollars
3. Multiple datasets
were used in the
analysis.
4. Only trout species
fishing was
included.
5. The benefit
estimates were
aggregated for four
states by accounting
income increase
and population
change resulting
from baby boom.
Kaoru et al.
(1995)
Albemarle
and Pamlico
Sounds, NC
CV and RUM
Fish
Consumer
surplus
(recreational
value)
1.35-site model: $36.19
for 5% increase of total
catch at all sites (full-
wage)
2.35-site model: $6.52
for 36% decrease in N
loadings at all sites (full-
wage)
1. Estimated N
loadings and
biochemical O2
demand, were used
in household
production models
to measure fishing
quality.
2. The paper also
estimated benefit
based on
opportunity cost at
1/3 wage for 23-site
and 11-site models.
Morey and
Four fishing CV and TC
Brook and
Consumer
1. $475.87 for 5%
1. Economic values
Shaw (1990)
sites in
Lake Trout
surplus
increase in catch rates
were aggregate
Adirondacks,
(conditional
for trout (1977 dollars)
CCV for 607 survey
NY
compensating
variations)
(recreational
value)
2. $2162.04 for 25%
increase in catch rates
for trout (1977 dollars)
3. $3862.94 for 50%
increase in catch rates
for trout (1977 dollars)
respondents.
2. Catch rate, as an
indicator of acid
deposition, was
included as a
variable in the
model.
F-36
-------
Reference Study Area
Method
Ecological
Endpoint
Valuation
Endpoint
Economic Value
Note
Morgan and
Owens
(Morgan and
Owens)
Chesapeake
Bay
Water quality
model and
Benefit transfer
Water
quality
Aggregate
benefit
(recreational
value)
1. Lower bound total
benefit for beach use
was $288.8m from a
60% improvement in
Chesapeake Bay water
quality.
2. Lower bound total
benefit for trailered
boating was $6.7m from
a 60% improvement in
Chesapeake Bay water
quality.
3. Lower bound total
benefit for striped bass
sport fishing was
$288.8m from a 60%
improvement in
Chesapeake Bay water
quality.
1. The study also
provided aggregate
benefits at avg and
high level for three
recreational uses.
2. Benefit transfer
was based on the
results in Bockstael
etal. (1988).
Mullen and
Adirondack, TC
Fish
Net economic
1. Net economic value
1. The values were
Menz (1985)
NY
value per
angler day/
per angler day was avgd
at $19.90 for entire
in 1976 dollars.
2. Total value was
Consumer
waterbody including
aggregated by
number of trips.
surplus
lakes, ponds, and
(recreational
streams.
value)
2. Total angler value was
estimated $31.3 million
for entire water body.
3. The loss in net
economic value was
estimated at $1.07
million for lakes and
ponds due to reduction
in angler visitation.
Randall and
Nationwide, CV
Water/air
WTP (total
1. The estimated annual
1. The study valued
Kriesel
U.S.
quality
value)
willingness to pay was
a National Pollution
(1990)
$694.42 per household.
Control Program,
which improved air
and water quality by
25 percent in 5
years.
F-37
-------
Reference
Study Area
Method
Ecological
Endpoint
Valuation
Endpoint
Economic Value
Note
Smith et al.
Albemarle
HTC and
Fish
Benefit/
1. Conventional demand
1. Benefit was
(1991)
and Pamlico
demand
consumer
model: $2.58 for boat,
calculated for an
Sounds, NC
function
surplus
(recreational
value)
$1.11 for bank fishing
2. Simple inverse
demand model: $0.79
for boat, $0.62 for bank
fishing
3. Detailed inverse
demand model: $0.88
for boat, $0.79 for bank
fishing
one-fish increase in
catch rate per hour
per person
2. Catch rate was
included in the
model.
3. The paper also
estimated benefit
based on
opportunity cost at
1/3 wage.
Whitehead
Albemarle
CVandTCM
Fish
Consumer
1. The consumer surplus
1. The consumer
etal. (2000)
and Pamlico
catches and
surplus per
per trip is $64 for current
surplus per trip
Sounds, NC
shellfish
season/ per trip
quality and $85 for 60%
difference between
beds
(recreational
value)
increase in fish catch
and 25% more open
shellfish beds.
2. The consumer surplus
per season is $121 for
current quality and $155
for 60% increase in fish
catch and 25% more
open shellfish beds.
two baseline and
scenario was not
significant at the
0.10 level while the
consumer surplus
per season
difference between
two situations was.
2. The studies also
estimated annual
aggregate
consumer surplus
per season for 41
counties within the
watershed.
EUTROPHICATION
Boyle, et al.
Four lakes,
HTC (two
Water clarity
Consumer
1. The consumer surplus
1. The welfare was
(1999)
ME
stages, hedonic
(visibility)
surplus
(+) for an avg visibility
also measured in
demand model
(recreational
value)
improvement from
3.78m to 5.15m was
$3,765 and $3,677 for
semi-log and Cobb-
Douglas model.
2. The welfare losses (-)
for an avg visibility
decrease from 3.78m to
2.41m was $25,388 and
$46,750 for semi-log
and Cobb-Douglas
model.
linear demand
model. However, the
own price of water
clarity is not
significant.
Diaz and
Gulf of
Time series
Brown
Mean catch
1. The study failed to
1. Hypoxia was
So low
Mexico
shrimp,
rate per unit
quantify economic
measured in terms
(1992)
white
shrimp, and
Menhaden
effort
effects of hypoxia based
on past data.
of area (zone) and
index in the model
to calculate
correlation.
F-38
-------
Reference
Study Area
Method
Ecological
Endpoint
Valuation
Endpoint
Economic Value
Note
Kramer and
Eisen-Hecht
(2002)
Catawba
River, NC and
SC
CV
Water
quality
WTP (total
value)
1. A mean willingness to
pay was revealed at
$139 for maintaining
current water quality
over time.
1. Phone-mail-
phone and mail-
phone survey
formats applied.
Lipton
(2003)
Chesapeake
Bay, MA
cv
Water
quality
WTP (use
value)
1. The median WTP for
a one scale
improvement in water
quality was $17.50 per
boater per year and the
mean was $63, with
38% expressing a zero
WTP.
1. The aggregate
WTP for the
Chesapeake Bay
boaters in Maryland
was about $7.3
million per year,
total improvement
gets a $146 million
present value.
Needelman
and Kealy
(1995)
Lakes, NH
Site choice
model and a
trip frequency
model
Water
quality
Benefit
(recreational
value)
1. The mean seasonal
benefits estimates were
$1.4 for eliminating
eutrophication from all
sources ($1.33 for
nonpoint source and
$0.09 for point source)
and the aggregate
seasonal benefits were
estimated at $1,163,000
for eliminating
eutrophication from all
sources ($1,105,000 for
nonpoint source and
$75,000 for point
source).
1. The study also
reported mean and
aggregate seasonal
benefits for
eliminating bacteria,
oil and grease,
eutrophication
problems, and all
pollution from all
sources.
2. The benefit
estimates address
exclusively
swimming and day
trips.
vNSTC
(2000)
Gulf of
Mexico,
Mississippi-
Atchafalay
River Basin
Cost effective,
simulation
model (US
Mathematic
Programming
Model in
Agriculture)
Fish,
shrimp, and
marine
ecosystem
Social cost for
reduction/
restoration
1 For 20% edge-of-field
N-loss reduction, the net
cost was estimated at
$0.8/kg ($0.36/pound),
while restoring 5 million
acres of wetland would
have net cost of
$1.00/kg ($0.45/pound).
1. The benefit to
reduce N loadings
to the Gulf was
difficult to calculate
because economic
analysis failed to
show direct effects
on Gulf fisheries,
which was a
background study
conducted by Diaz
and Solow (1992).
Park et al.
(2002)
Florida Keys,
FL
CV
Water
quality and
health of
coral reefs
WTP per trip
expenses
(recreational
value)
1. The annual avg use
value was $481.15 per
person per snorkeling
trip.
F-39
-------
Reference Study Area Method Ecological Valuation Economic Value
' EnHnAiht EnHnAiht
Note
Poor et al.
St. Mary's
HPVM
Ambient
Avg housing
1. One unit (mg/L)
1. Total suspended
(2006)
River
water
price within the
increase in TSS resulted
solids and dissolved
watershed,
quality
watershed
in a $1,086 loss on avg
inorganic N included
MA
(commercial
value)
housing prices within the
watershed.
2. One unit increase in
the dissolved inorganic
N resulted in a $17,642
loss on avg housing
prices.
as indicators of
ambient water
quality.
Smith and
Neuse River,
Bio-economic
Blue crab
Fish rent
1. A 30% reduction in N
1. Discounting rate
Crowder
NC
model
(commercial
loadings over a 50—yr
was 2.5%.
(2005)
value)
time period increases
present value rents by
2.56 million, total catch
by 12.4 million pounds,
and total trips by 91,000.
Whitehead
Tar-Pamlico
CV
Catch rate
WTP (total
1. The mean WTP
1. The study
and
River, NC
value, use, and
resulting from a 61%
proposed a program
Groothuis
non-use value)
response rate was
in which farmers are
(1992)
estimated at $25, which
was bounded from
above by a $35 use
value, and from below a
$21 nonuse value.
2. The aggregate
benefits from water
quality improvement
estimated at $1.62
million each year.
required to use
BMPs to improve
the water quality of
the Tar-Pamlico
River so that
anglers would catch
twice as many fish
per trip.
2. A mail survey was
sent to 179
households in four
counties in Tar-
Pamlico River basin.
VARIOUS EFFECTS
Johnston
U.S. and
Meta-analysis
Catch rate
WTP
1. WTP per fish over the
1. This study
etal. (2005)
Canada
(recreational
use)
sample ranged from
$.048 to $612.79, with a
mean of $16.82
supports previous
findings that WTP is
systematically
associated with
resource, context,
and angler
attributes.
F-40
-------
Glossary
Acid Neutralizing Capacity (ANC)
A key indicator of the ability of water to neutralize acidifying inputs. This ability depends largely on
associated biogeochemical characteristics.
Acidification
The process of decreasing the pH of a system. Systems that can be acidified by atmospheric deposition of
acidic or acidifying compounds include lakes, streams, and forest soils.
Algae
Photosynthetic, often microscopic and planktonic, organisms occurring in aquatic ecosystems.
Algal bloom
A rapid and extreme increase of an algae population in a lake, river, or ocean.
Alpine
The biogeographic zone made up of slopes above the tree line, and characterized by the presence of rosette-
forming herbaceous plants and low, shrubby, slow-growing woody plants.
Anthropogenic
Resulting from human activity or produced by human beings.
Arid region
An area receiving <250 mm precipitation per year.
Atmosphere
The gaseous envelope surrounding the Earth. The dry atmosphere consists almost entirely of nitrogen and
02, together with trace gases including carbon dioxide and ozone.
Base cation saturation
The degree to which soil cation exchange sites are occupied with base cations (e.g., Ca2+, Mg2+, K+) as
opposed to Al3+ and H+. Base cation saturation is a measure of soil acidification, with lower values being
more acidic. A marked increase in the sensitivity of soils to changes in base saturation occurs at a threshold
of approximately 20%.
Bioaccumulation
The gradual increase in accumulation of some compounds in organisms with increasingly higher trophic
levels.
Biodiversity
The total diversity of all organisms and ecosystems at various spatial scales (from genes to biomes).
Buffering capacity
The ability of a body of water and its watershed to resist changes in pH.
Carbon sequestration
The process of increasing the carbon content of a reservoir other than the atmosphere.
G-1
-------
Catchment
An area that collects and drains rainwater.
Climate
Climate in a narrow sense is usually defined as the 'average weather', or more rigorously, as the statistical
description in terms of the mean and variability of relevant quantities over a period of time ranging from
months to thousands or millions of years. These quantities are most often surface variables such as
temperature, precipitation, and wind. Climate in a wider sense is the state, including a statistical
description, of the climate system. The generally accepted period of time is 30 years, as defined by the
World Meteorological Organization (WMO).
Critical load
A quantitative estimate of an exposure to one or more pollutants below which significant harmful effects on
specified sensitive elements of the environment do not occur according to present knowledge.
Denitrification
The anaerobic reduction of oxidized nitrogen (e.g., nitrate or nitrite) to gaseous nitrogen (e.g., N20 or N2),
normally accomplished by denitrifying bacteria.
Dry deposition
The movement of gases and particles from the atmosphere to surfaces in the absence of precipitation (e.g.,
rain or snow) or occult deposition.
Ecological community
An assemblage of populations of different species, interacting with one another
Ecosystem services
Ecological processes or functions having monetary or non-monetary value to individuals or society at large.
They may be classified as (i) supporting services such as productivity or biodiversity maintenance; ii)
provisioning services such as food, fibre, or fish; iii) regulating services such as climate regulation or
carbon sequestration; and (iv) cultural services such as tourism or spiritual and aesthetic appreciation.
Ecosystem
The interactive system formed from all living organisms and their abiotic (physical and chemical)
environment within a given area. Ecosystems cover a hierarchy of spatial scales and can comprise the entire
globe, biomes at the continental scale, or small, well-circumscribed systems such as a small pond.
Eutrophication
The enrichment of a waterbody with nutrients, resulting in increased productivity (of algae or aquatic
plants), and sometimes also decreased dissolved 02 levels.
Eutrophy
Eutrophy generally refers to a state of nutrient enrichment, but it is commonly used to refer to condition of
increased algal biomass and productivity, presence of nuisance algal populations, and a decrease in
dissolved 02 concentrations.
Evapotranspiration
The combined process of water evaporation from the Earth's surface and transpiration from vegetation.
Fen
A phase in the development of the natural succession from open lake, through reedbed, fen and carr, to
woodland as the peat develops and its surface rises.
G-2
-------
Freshet
A great rise or overflowing of a stream caused by heavy rains or melted snow.
Greenhouse gas
Those atmospheric gasses that absorb and emit radiation emitted by the Earth's surface, the atmosphere,
and clouds within the infrared portion of the spectrum. This property causes the greenhouse effect. Water
vapour (H20), carbon dioxide (C02), nitrous oxide (N20), methane (CH4), and ozone (03) are the primary
greenhouse gases in the Earth's atmosphere. Besides these, the Kyoto Protocol also deals with the
greenhouse gases sulphur hexafluoride (SF6), hydrofluorocarbons (HFCs), and perfluorocarbons (PFCs).
Gross primary production
The total carbon fixfixed by plant through photosynthesis.
Heathland
A wide-open landscape dominated by low-growing woody vegetations such as heathers and heathland
grasses. Heathlands generally occur on acidic, nutrient-poor, and often sandy and well-draining soils.
Hypoxic
Events that lead to a deficiency of 02.
Invasive species and invasive alien species
A species aggressively expanding its range and population density into a region in which it is not native,
often through outcompeting or otherwise dominating native species.
Leaching
The removal of soil elements or chemicals by water movement through the soil.
Lowland
In physical geography, lowland is any relatively flat area in the lower levels of regional elevation. The term
can be applied to the landward portion of the upward slope from oceanic depths to continental highlands, to
a region of depression in the interior of a mountainous region, to a plain of denudation, or to any region in
contrast to a highland.
Net ecosystem exchange (NEE)
The net flux of carbon between the land and the atmosphere, typically measured using eddy covariance
techniques. Positive values of NEE usually refer to carbon released to the atmosphere (i.e., a source), and
negative values refer to carbon uptake (i.e., a sink)
Net ecosystem production (NEP)
The difference between net primary production (NPP) and heterotrophic respiration (mostly decomposition
of dead organic matter) of that ecosystem over the same area. NEP = -NEE, with positive values indicating
a sequestration of atmospheric carbon in to biosphere.
Net primary production (NPP)
The gross primary production minus autotrophic respiration, i.e., the sum of metabolic processes for plant
growth and maintenance, over the same area.
Nitrification
The biological oxidation of ammonia to nitrite and then to nitrate. This process is primarily accomplished
by autotrophic nitrifying bacteria that obtain energy by reducing ammonium and/or nitrite to nitrate.
G-3
-------
Nitrogen mineralization
The conversion of organic nitrogen into plant-available inorganic forms (e.g., NH3 or NH4NH4) by
microorganisms.
Nitrogen-retention capacity
The length of time that an ecosystem can retain nitrogen in plants, microbes and soil-organic matter.
Nitrogen-retention capacity is highly affected by soil, vegetative, topographic, and land-use factors.
Nitrogen saturation
The condition in which nitrogen inputs from atmospheric deposition and other sources exceed the
biological requirements of the ecosystem.
Occult deposition
The transmission of gases and particles from the atmosphere to surfaces by fog or mist.
Ombrotrophic bog
An acidic peat-accumulating wetland that is fed by rainwater (instead of groundwater) and, thus, especially
poor in nutrients.
pH
A measure of the relative concentration of hydrogen ions in a solution. The formula for calculating pH is:
pH = -logio[H+], where [H+] represents the hydrogen ion concentration in moles per liter. The pH scale
ranges from 0 to 14. A pH of 7 is neutral. A pH less than 7 is acidic and a pH greater than 7 is basic.
Phytoplankton
The plant forms of plankton. Phytoplankton are the dominant plants in the sea and are the basis of the
marine food web. These single-celled organisms are the principal agents of photosynthetic carbon
fixfixation in the ocean.
Primary Production
All forms of production accomplished by plants, also called primary producers. See GPP, NPP, and NEP.
Semi-arid regions
Regions of moderately low rainfall (100- and 250-mm precipitation per year), which are not highly
productive and are usually classified as rangelands.
Sensitivity
The degree to which a system responds to pollution (e.g., acidification, n-nutrient enrichment, etc.). The
response may be direct (e.g., a change in growth following a change in the mean, range, or variability of N
deposition) or indirect (e.g., changes in growth due to alterations in competitive dynamics between species
or decreased biodiversity , themselves following N deposition).
Streamflow
Water flow within a river channel. A synonym for river discharge.
Surface runoff
The water that travels over the land surface to the nearest surface stream; runoff of a drainage basin that has
not passed beneath the surface since precipitation.
Throughfall
The precipitation falling through the canopy of a forest and reaching the forest floor.
G-4
-------
Trophic level
The position that an organism occupies in a food web.
Tundra
A treeless, level, or gently undulating plain characteristic of the Arctic, sub-Arctic regions and some alpine
regions characterized by low temperatures and short growing seasons.
Upland terrestrial ecosystem
Generally considered to be the ecosystems located at higher elevations directly above riparian zones and
wetlands. Vegetation in an upland ecosystem is not in contact with groundwater or other permanent water
sources.
Valuation
The economic or non-economic process of determining either the value of maintaining a given ecosystem
type, state, or condition or the value of a change in an ecosystem, its components, or the services it
provides.
Vulnerability
Susceptibility to degradation or damage from adverse factors or influences. Vulnerability is a function the
exposure and its sensitivity.
Welfare effects
Effects on soils, water, crops, vegetation, man-made materials, animals, wildlife, weather, visibility and
climate, damage to and deterioration of property, and hazards to transportation, as well as effects on
economic values and on personal comfort and well-being, whether caused by transformation, conversion,
or combination with other air pollutants (CAA 302(h)).
Wet deposition
The transmission of gases and particles from the atmosphere to surfaces by rain or other precipitation.
Wetland
Those areas that are inundated or saturated by surface or ground water at a frequency and duration
sufficient to support a prevalence of vegetation adapted to water-saturated soil conditions. Wetlands
include swamps, marshes, bogs, and similar areas.
Zooplankton
The animal forms of plankton. They consume phytoplankton or other zooplankton.
G-5
-------
References
Aastrup T; Wadsak M; Leygraf C; Schreiner M. (2000). In situ studies of the initial atmospheric corrosion of copper influence of humidity, sulfur
dioxide, ozone, and nitrogen dioxide. J Electrochem Soc, 147,2543-2551.
Abbot DS; Palmer PI; Martin RV; Chance KV; Jacob DJ; Guenther A. (2003). Seasonal and interannual variability of North American isoprene
emissions as determined by formaldehyde column measurements from space. Geophys Res Lett, 30,1886.
doi:10.1029/2003GL017336.
Aber JD; Nadelhoffer KJ; Steudler P; Mellilo JM. (1989). Nitrogen saturation in northern forest ecosystems. Excess nitrogen from fossil fuel
combustion may stress the biosphere. Bioscience, 39,378-386.
Aber JD; Mellilo JM; Nadelhoffer KJ; Pastor J; Boone RD. (1991). Factors controlling nitrogen cycling and nitrogen saturation in northern
temperate forest ecosystems. Ecol Appl, 1,303-315.
Aber JD. (1992). Nitrogen cycling and nitrogen saturation in temperate forest ecosystems. Trends Ecol Evol, 7,220-224.
Aber JD; Federer CA. (1992). A generalized, lumped-parameter model of photosynthesis, evapotranspiration and net primary production in
temperate and boreal forest ecosystems. Oecologia, 92,463-474.
Aber JD; Driscoll CT. (1997). Effects of land use, climate variation, and N deposition on N cycling and C storage in northern hardwood forests.
Global Biogeochem Cycles, 11,639-648.
Aber JD; Ollinger S; Driscoll C. (1997). Modeling nitrogen saturation in forest ecosystems in response to land use and atmospheric deposition.
Ecol Modell, 101, 61-78.
Aber JD; McDowell W; Nadelhoffer K; Magill A; Berntson G; Kamakea M; McNulty S; Currie W; Rustad L; Frenandez I. (1998). Nitrogen
saturation in temperate forest ecosystems: Hypotheses revisited. Bioscience, 48,921-934.
Aber JD; Ollinger SV; Driscoll CT; Likens GE; Holmes RT; Freuder RJ; Goodale CL. (2002). Inorganic nitrogen losses from a forested
ecosystem in response to physical, chemical, biotic, and climatic perturbations. Ecosystems, 5,648-658.
Aber JD; Goodale CL; Ollinger SV; Smith ML; Magill AH; Martin ME; Hallett RA; Stoddard JL. (2003). Is nitrogen deposition altering the
nitrogen status of northeastern forests? Bioscience, 53,375-389.
Aber JD; Magill AH. (2004). Chronic nitrogen additions at the Harvard Forest (USA): The first 15 years of a nitrogen saturation experiment. For
Ecol and Manage, 196,1-5.
Achermann B; Bobbink R. (2003). Empirical critical loads for nitrogen. Proceedings of an expert workshop held under the UNECE Convention
on Long-range Transboundary Air Pollution held in Berne, Switzerland; November, 2002. (Environmental Documentation no. 164).
Berne, Switzerland; Swiss Agency for the Environment, Forests, and Landscape (SAEFL).
Adams AB; Harrison RB; Sletten RS; Strahm BD; Turnblom EC; Jensen CM. (2005). Nitrogen-fertilization impacts on carbon sequestration and
flux in managed coastal Douglas-fir stands of the Pacific Northwest. For Ecol Manage, 220,313-325.
Adams MA; Iser J; Keleher AD; Cheal DC. (1994). Nitrogen and phosphorus availability and the role of fire in heathlands at Wilsons
Promontory. Aust J Bot, 42,269-281.
Adams MB; Edwards PJ; Wood F; Kochenderfer JN. (1993). Artificial watershed acidification on the Fernow Experimental Forest, USA. J
Hydro1,150,505-519.
Adams MB; Angradi TR; Kochenderfer JN. (1997). Stream water and soil solution responses to 5 years of nitrogen and sulfur additions at the
Fernow Experimental Forest, West Virginia. For Ecol Manage, 95,79-91.
Adams MB. (1999). Acidic deposition and sustainable forest management in the central Appalachians, USA. For Ecol Manage, 122,17-28.
Adams MB; Burger JA; Jenkins AB; Zelazny L. (2000). Impact of harvesting and atmospheric pollution on nutrient depletion of eastern U.S.
hardwood forests. For Ecol Manage, 138,301-319.
Adams MB. (2003). Ecological issues related to N deposition to natural ecosystems: Research needs. Environ Int, 29,189-199.
Adams RM; Hamilton SA; McCarl BA. (1986a). The benefits of pollution control: The case of ozone and U.S. agriculture. Am J Agric Econ, 68,
886-893.
Adams RM; Callaway JM; McCarl BA. (1986b). Pollution, agriculture and social welfare: The case of acid deposition. Can J Agric Econ, 34,3-
19.
Adams RM; Crocker TD. (1989). The agricultural economics of environmental change: Some lessons from air pollution. J Environ Manage, 28,
295-307.
Adams RM; Horst RL Jr. (2003). Future directions in air quality research: Economic issues. Environ Int, 29, 289-302.
Adams SM; Greeley MS; Law JM; Noga EJ; Zelikoff JT. (2003). Application of multiple sublethal stress indicators to assess the health of fish in
Pamlico Sound following extensive flooding. Estuaries, 26,1365-1382.
Adamsen APS; King GM. (1993) Methane consumption in temperate and sub-arctic forest soils: Rates, vertical zonation, and responses to
water and nitrtogen. Appl Environ Microbiol, 59,485-490.
Aeschlimann U; Nosberger J; Edwards PJ; Schneider MK; Richter M; Blum H. (2005). Responses of net ecosystem CO2 exchange in managed
grassland to long-term CO2 enrichment, N fertilization and plant species. Plant Cell Environ, 28,823-833.
Aerts R; Berendse F. (1988). The effect of increased nutrient availability on vegetation dynamics in wet heathlands. Plant Ecology, 76,63-69.
Aerts R. (1990). Nutrient use efficiency in evergreen and deciduous species from heathland. Oecologia, 84,391-397.
Aerts R; Berendse F; De Caluwe H; Schmits M. (1990). Competition in heathland along an experimental gradient of nutrient availability. Oikos,
57,310-318.
R-1
-------
Aerts R; Toet S. (1997). Nutritional controls on carbon dioxide and methane emission from Carex-dominated peat soils. Soil Biol Biochem, 29,
1683-1690.
Aerts R; de Caluwe H. (1999). Nitrogen deposition effects on carbon dioxide and methane emissions from temperate peatland soils. Oikos, 84,
44-54.
Aerts R; Verhoeven JTA; Whigham DF. (1999). Plant-mediated controls on nutrient cycling in temperate fens and bogs. Ecology, 80,2170-
2181.
Agelakopoulou T; Bassiotis I; Metaxa E; Roubani-Kalantzopoulou F. (2007). Benzene and toluene influence with or without nitrogen dioxide on
inorganic pigments of works of art-Part II. Atmos Environ, 41,2009-2018.
Aherne J; Larssen T; Dillon PJ; Cosby BJ. (2004). Effects of climate events on elemental fluxes from forested catchments in Ontario, Canada:
Modelling drought-induced redox processes. Water Air Soil Pollut, 4, 37-48.
Aherne J; Larssen T; Cosby BJ; Dillon PJ. (2006). Climate variability and forecasting surface water recovery from acidification: Modelling
drought-induced sulphate release from wetlands. Sci Total Environ, 365,186-199.
Ainsworth EA; Long SP. (2005). What have we learned from 15 years of free-air CO2 enrichment (FACE)? A meta-analytic review of the
responses of photosynthesis, canopy properties and plant production to rising CO2. New Phytol, 165,351 -371.
Ainsworth EA; Rogers A. (2007). The response of photosynthesis and stomatal conductance to rising [CO2]: Mechanisms and environmental
interactions. Plant Cell Environ, 30,258-270.
Akiyama H; Yagi K; Yan XY. (2005). Direct N2O emissions from rice paddy fields: Summary of available data. Glob Biogeochem Cycles, 19,
doi;10.1029/2004GB002378.
Alban DH. (1982). Effects of nutrient accumulation by aspen, spruce, and pine on soil properties. Soil Sci Soc Am J, 46,853-861.
Alberti M; Booth D; Hill K; Coburn B; Avolio C; Coe S; Spirandelli D. (2007). The impact of urban patterns on aquatic ecosystems: an empirical
analysis in Puget lowland sub-basins. Landsc Urban Plan, 80, 345-361.
Albinet A; Leoz-Garziandia E; Budzinski H; Viilenave E. (2006). Simultaneous analysis of oxygenated and nitrated polycyclic aromatic
hydrocarbons on standard reference material 1649a (urban dust) .and on natural ambient air samples by gas chromatography-mass
spectrometry with negative ion chemical ionisation. J Chromatogr A, 1121,106-113.
Alcamo J; Shaw R; Hordijk L. (1990). The RAINS model of acidification: Science and stratigies in Europe. Dordrecht, The Netherlands: Kluwer
Academic Publishers.
Alcorn SH; Lurmann FW. (2004). Southern California children's health study exposure database (final report). Petaluma, CA: Sonoma
Technology, Inc., (Report no. STI-95230-2453-FR3). Prepared for Division of Environmental Health; University of Southern California
School of Medicine.
Aldous AR. (2002). Nitrogen retention by Sphagnum mosses: Responses to atmospheric nitrogen deposition and drought. Can J Bot, 80,721 -
731.
Alewell C; Gehre M. (1999). Patterns of stable S isotopes in a forested catchment as indicators for biological S turnover. Biogeochemistry, 47,
317-331.
Alexander RB; Smith RA; Schwarz GE. (2000). Effect of stream channel size on the delivery of nitrogen to the Gulf of Mexico. Nature, 403,
758-761.
Alexander RB; Smith RA; Schwarz GE; Preston SD; Brakebill JW; Srinivasan R; Pacheco PA. (2001). Atmospheric nitrogen flux from the
watersheds of major estuaries of the United States: An application of the SPARROW watershed model. In: Valigura R;, Alexander R;
Castro M; Meyers T; Paerl H; Stacey P; Turner RE (Eds.), Nitrogen loading in coastal water bodies: An atmospheric perspective. (Vol.
monograph 57, pp. 119-170). Washington, DC: American Geophysical Union.
Alexander RB; Johnes RJ; Boyer EW; Smith RA. (2002). A comparison of models for estimating the riverine export of nitrogen from large
watersheds. Biogeochemistry, 57/58, 295-339.
Alexander RB; Boyer EW; Smith RA; Schwarz GE; Moore RB. (2007). The role of headwater streams in downstream water quality. J Am Water
Resources Assoc, 43,41-59.
Alicke B; Hebestreit K; Stutz J; Piatt U. (1999). Iodine oxide in the marine boundary layer. Nature, 397, 572-573.
Allan BJ; McFiggans G; Plane JMC; Coe H. (2000). Observations of iodine monoxide in the remote marine boundary layer. J Geophys Res,
105,14363-14369.
Allegrini I; Febo A; Perrino C. (Eds.) (1992). Field intercomparison exercise on ammonium measurement. Proceedings from Rome, Italy April
29-May 4,1990. Air Pollution Research Report #37. Brussels, Belgium: Commission of European Communities (CEC).
Allen DJ; Pickering KE; Molod A. (1997). An evaluation of deep convective mixing in the Goddard Chemical Transport Model using ISCCP
cloud parameters. J Geophys Res, 102,25467-25476.
Allen DJ; Pickering K; Stenchikov G; Thompson A; Kondo Y. (2000). A three-dimensional total odd nitrogen (NOy).simulation during SONEX
using a stretched-grid chemical transport model. J Geophys Res, 105,3851-3876
Allen EB; Padgett PE; Bytnerowicz A; Minich R. (1998). Nitrogen deposition effects on coastal sage vegetation of southern California (No.
PSW-GTR-166; pp 131-139): U.S. Forest Service; U.S. Department of Agriculture (USDA).
Allen EB; Sirulnik AG; Egerton-Warburton L; Kee SN; Bytnerowicz A; Padgett PE; Temple PJ; Fenn ME; Poth MA; Meixner T. (2005). Air
pollution and vegetation change in southern California coastal sage scrub: A comparison with chaparral and coniferous forest. In: Kus
BE; Beyers JL (Eds.), Planning for biodiversity: Bringing research and management together, (general technical report PSW-GTR-195;
pp. 79-95). Albany, CA; Pacific Southwest Research Station; U.S. Forest Service; U.S. Department of Agriculture (USDA).
Allen EB; Bytnerowicz A; Fenn ME; Minnich RA; Allen MF. (2007). Impact of anthropogenic N deposition on weed invasion, biodiversity, and
the fire cycle at Joshua Tree National Park (No. PMIS 72123). University of California, Riverside. Center for Conserv Biol (CCB); and
Forest Fire Laboratory of the U.S. Forest Service; U.S. Department of Agriculture (USDA).
R-2
-------
Allen EB; Rao LE; Steers RJ. (2008). Impacts of atmospheric nitrogen deposition on vegetation and soils at Joshua Tree National Park. In:
Webb RH; Fenstermaker LF; Heaton JS; Hughson DL; McDonald EV; Miller DM (Eds.), The Mojave Desert: Ecosystem processes and
sustainability. Reno, NV: University of Nevada Press.
Allen NS; Hershey AE. (1996). Seasonal changes in chlorophyll a response to nutrient amendments in a North Shore tributary of Lake
Superior. J North Am Benthol Soc, 15,170-178.
Alley RB; Clark PU; Huybrechts P; Joughin I. (2005). Ice-sheet and sea-level changes. Science, 310,456-460.
Almeida E; Morcillo M; Rosales B. (2000a). Atmospheric corrosion of zinc; Part 2: Marine atmospheres. British Corrosion Journal, 35,289-296.
Almeida E; Morcillo M; Rosales B. (2000b). Atmospheric corrosion of mild steel. Part II - Marine atmospheres. Mater Corros, 51,865-874.
Aimer B; Dickson W; Ekstrom C; Hornstrom E; Miller U. (1974). Effects of acidification on Swedish lakes. Ambio, 3,30-36.
Altshuller AP; Linthurst RA. (Eds.). (1984). The acidic deposition phenomenon and its effects: Critical assessment review papers. (Report no.
EPA-600/8-83-016-BF ). Washington, DC: U.S. Environmental Protection Agency.
Alvarez R; Weilenmann M; Favez J-Y. (2008). Evidence of increased mass fraction of NO2 within real-world NOx emissions of modern light
vehicles—derived from a reliable online measuring method. Atmos Environ, 42,4699-4707.
Alveteg M; Sverdrup H. (2002). Manual for regional assessments using the SAFE model (draft version 8 April 2002). Lund, Sweden:
Department of Chemical Engineering II, Lund University.
Alvo R; Hussell DJT; Berrill M. (1988). The breeding success of common loons (Gaviaimmei) in relation to alkalinity and other lake
characteristics in Ontario. Can J Zool, 66,746-752.
Ambrose RA Jr; Woll TA; Martin JL. (1995). The water quality analysis simulation program, WASP5, Part A: Model documentation. Athens, GA;
Ecosystems Division; National Exposure Research Laboratory; Office of Research and Development; U.S. Environmental Protection
Agency.
Ambus P; Robertson GP. (2006). The effect of increased N deposition on nitrous oxide, methane and carbon dioxide fluxes from unmanaged
forest and grassland communities in Michigan. Biogeochemistry, 79,315-337.
Ambus P; Zechmeister-Boltenstern S; Butterbach-Bahl K. (2006). Sources of nitrous oxide emitted from European forest soils. Biogeosciences,
3,135-145.
Ammann M; von Ballmoos P; Stalder M; Suter M; Brunold C. (1995). Uptake and assimilation of atmospheric N02"N by spruce needles (Picea
abies): A field study. Water Air Soil Pollut, 85,1497-1502.
Ammann M; Kalberer M; Jost DT; Tobler L; Rossler E; Piguet D; Gaggeler HW; Baltensperger U. (1998). Heterogeneous production of nitrous
acid on soot in polluted air masses. Nature, 395,157-160.
Ammann M; Siegwolf R; Pichlmayer F; Suter M; Saurer M; Brunold C. (1999). Estimating the uptake of traffic-derived NO2 from N-15
abundance in Norway spruce needles. Oecologia, 118,124-131.
Amundson RG; McLean DC. (1982). Influence of oxides of nitrogen on crop growth and yield: An overview. In Schneider T; Grant L (Eds.), Air
pollution by nitrogen oxides. (Series: Studies in environmental science Volume 21; Proceedings of the U.S.-Dutch international
symposium held in Maastricht, The Netherlands in May 1982; pp 501-510). Amsterdam, The Netherlands: Elsevier Scientific
Publishing.
AnastasioC; Newberg JT; Williams DK; Chu GB; Matthew BM. (1999). Photoformation of hydroxyl radical in sea salt particles. Presented at
1999 Fall Meeting December 13-17,1999; San Francisco, CA. Sponsored by American Geophysical Union; Trans American Geophys
Union, 80, Abstract F147.
Anderson HR; Spix C; Medina S; Schouten JP; Castellsague J; Rossi G; Zmirou D; Touloumi G; Wojtyniak B; Ponka A; Bacharova L; Schwartz
J; Katsouyanni K. (1997). Air pollution and daily admissions for chronic obstructive pulmonary disease in 6 European cities: Results
from the APHEA project. Eur Respir J, 10,1064-1071.
Anderson NJ; Renberg I; Segerstrom U. (1995). Diatom production responses to the development of early agriculture in a boreal forest lake-
catchment (Kassjon, Northern Sweden). J Ecol, 83,809-822.
Andreae MO. (1991). Biomass burning: Its history, use, and distribution and its impact on environmental quality and global climate. In: Levine,
JS (Ed.), Global biomass burning: Atmospheric, climatic, and biospheric implications (pp. 1-21). Cambridge, MA: MIT Press.
Andreae MO; Crutzen PJ. (1997). Atmospheric aerosols: Biogeochemical sources and role in atmospheric chemistry. Science, 276,1052-1058.
Andrews JA; Schlesinger WH. (2001). Soil CO2 dynamics, acidification, and chemical weathering in a temperate forest with experimental CO2
enrichment. Glob Biogeochem Cycles, 15,149-162.
Aneja VP; Chauhan JP; Walker JT. (2000). Characterization of atmospheric ammonia emissions from swine waste storage and treatment
lagoons. J Geophys Res, 105,11535-11546.
Aneja et al., 1998 ambient NH3 concentration
Ansari AS; Pandis SN. (2000). Water absorption by secondary organic aerosol and its effect on inorganic aerosol behavior. Environ Sci
Technol, 34,71-77.
Antoine ME. (2001). Ecophysiology of the cyanolichen, Lobaria oregana. M.S. Thesis, Corvallis, OR; Oregon State University.
Antoine ME. (2004). An Ecophysiological approach to quantifying nitrogen fixation by Lobaria oregano. Bryologist, 107,82-87.
Appel BR; Tokiwa Y; Haik M. (1981). Sampling of nitrates in ambient air. Atmos Environ, 15,283-289.
Appel BR; Tokiwa Y; Kothny EL; Wu R; Povard V. (1988). Evaluation of procedures for measuring atmospheric nitric acid and ammonia. Atmos
Environ, 22,1565-1573.
Appel KW; Gilliland A; Eder B. (2005). An operational evaluation of the 2005 release of models-3 CMAQ version 4.5. Washington, DC:
Atmospheric Sciences Modeling Division, Air Resources Laboratory, National Oceanic and Atmospheric Administration (NOAA).
Appelo CAJ; Postma D. (1993). Geochemistry, groundwater, and pollution. Rotterdam, The Netherlands: AA Balkema Publishers.
R-3
-------
Appel K; Bhave PV; Gilliland AB; Sarwar G; Roselle SJ. (2008). Evaluation of the community multiscale air quality (CMAQ) model version 4.5:
Sensitivities impacting model performance; Part lla€"particulate matter. Atmospheric Environment 42:6057-6066.ApSimon HM;
Cowell D. (1996). The benefits of reduced damage to buildings from abatement of sulphur dioxide emissions. Energy Policy, 24,651 -
654.
Arbaugh MJ; Miller PR; Carroll JJ; Takemoto BL; Proctor T. (1998). Relationships of ozone exposure to pine injury in the Sierra Nevada and
San Bernardino Mountains of California, USA. Environ. Pollut, 101,291-301.
Arbaugh MJ; Peterson DL; Miller PR. (1999). Air pollution effects on growth of ponderosa pine, Jeffrey pine, andbigcone Douglas fir. In: Miller
PR; McBride JR (Eds.), Oxidant air pollution impacts in the montane forests of southern California: A case study of the San Bernardino
Mountains, (pp. 179-207). New York, NY: Springer-Verlag.
Arbaugh M; Bytnerowicz A; Grulke N; Fenn M; Poth M; Temple T; Miller P. (2003). Photochemical smog effects in mixed conifer forests along a
natural gradient of ozone and nitrogen deposition in the San Bernardino Mountains. Environ Int, 29,401-406.
Arbizzani R; Casellato U; Fiorin E; Nodari L; Russo U; Vigato PA. (2004). Decay markers for the preventative conservation and maintenance of
paintings. Journal of Cultural Heritage, 5,167-182.
Arey J; Zielinska B; Atkinson R; Winer AM; Ramdahl T; Pitts JN Jr. (1986). The formation of nitro-PAH from the gas-phase reactions of
fluoranthene and pyrene with the OH radical in the presence of NOx. Atmos Environ, 20,2339-2345.
Arey J; Atkinson R; Zielinska B; McElroy PA. (1989). Diurnal concentrations of volatile polycydic aromatic hydrocarbons and nitroarenes during
a photochemical air pollution episode in Glendora, California. Environ Sci Technol, 23,321-327.
Arey J. (1998). Atmospheric reactions of PAHs including formation of nitroarenes. In: Neilson AH (Ed.), Anthropogenic compounds. Part I.
PAHs and related compounds - chemistry (Volume 3, Series: Hutzinger O (Ed.): The handbook of environmental chemistry, pp. 347-
385). New York, NY: Springer-Verlag.
Arheimer B; Wittgren HB. (1994). Modelling the effects of wetlands on regional nitrogen transport. Ambio, 23, 378-386.
Arndt U; Billen N; Seufert G; Ludwig W; Borkhart K; Ohnesorge B. (1990). Visible injury responses. Environ Pollut, 68,355-62
ArnebrandtK; Baath E; Soderstrom B. (1990). Changes in microfungal community structure after fertilization of scots pine forest soils with
ammonium nitrate or urea. Soil Biol Biochem, 22,309-312.
Arnold CL; Gibbons CJ. (1996). Impervious surface coverage: Emergence of a key environmental indicator. J Am Planning Assoc, 62,243-258.
Arnold JR; Dennis RL; Tonnesen GS. (1998). Advanced techniques for evaluating Eulerian air quality models: Background and methodology.
Paper presented at the American Meteorological Society 10th Joint Conference on the Applications of Air Pollution Meteorology with
the Air & Waste Management Association January 11-16,1998, Phoenix, AZ. (pp 1-5). Boston, MA: American Meteorological Society.
Arnold JR; Dennis RL. (2001). First results from operational testing of the U.S. EPA models-3/community multiscale model for air quality
(CMAQ). In :Gryning S-E; Schiermeier FA (Eds.), Air pollution modelling and its application XIV. Proceedings of the Millennium
NATO/CCMS international technical meeting on air pollution modeling and its application; May 15-19,2000 in Boulder, Colorado. (Part
6: pp 651-658). New York, NY: Kluwer Academic Publishing.
Arnold JR; Dennis RL; Tonnesen GS. (2003). Diagnostic evaluation of numerical air quality models with specialized ambient observations:
Testing the community multiscale air quality modeling system (CMAQ) at selected SOS 95 ground sites. Atmos Environ, 37,1185-
1198.
Arnold JR; Dennis RL. (2006). Testing CMAQ chemistry sensitivities in base case and emissions control runs at SEARCH and SOS99 surface
sites in the southeastern U.S. Atmos Environ, 40,5027-5040.
Arnold JR; Hartsell BE; Luke WT; Ullah SMR; Dasgupta PK; Huey LG; Tate P. (2007). Field test of four methods for gas-phase ambient nitric
acid. Atmos Environ, 41,4210-4226.
Arnold JR; Luke WT. (2007). Nitric acid and the origin and size segregation of aerosol nitrate aloft during BRACE 2002. Atmos Environ, 41,
4227-4241.
Arp PA; OjaT; Marsh M. (1996). Calculating critical S and N loads and current exceedances for upland forests in southern Ontario, Canada.
Can J For Res, 26,696-709.
Arrigo KR. (2005). Marine microorganisms and global nutrient cycles. Nature, 437,349-355.
Arrington DV; Lindquist RC. (1987). Thickly mantled Karstof the Interlachen, Florida area. In: Beck BF; Wilson WL (Eds.), Karst hydrology:
Engineering and environmental applications. Proceedings of the 2nd multidisciplinary conference on sinkholes and the engineering
and environmental impacts of Karst. February 9-11,1987 - Orlando Florida, (pp. 31-39). Rotterdam, The Netherlands; AA Balkema
Publishers.
Arrow K; SolowR; Portney P; Learner E; Radner R; Schuman H. (1993). Report of the NOAA Panel on Contingent Valuation. Federal Register
58(10), 4601-4614.
Ashby JA; Bowden WB; Murdoch PS. (1998). Controls on denitrification in riparian soils in headwater catchments of a hardwood forest in the
Catskill Mountains, USA. Soil Biol Biochem, 30,853-864.
Ashenden T. (1979). The effects of long-term exposures to SO2 and NO2 pollution on the growth oWactylis glomerata L. and Poapratensis L.
Environ Pollut, 18,249-258.
Ashendorff A; Principe MA; Seeley A; LaDuca J; Beckhardt L; Faber WJ; Mantus J. (1997). Watershed protection for New York City's supply.
Journal of the American Water Works Association, 89,75-88.
Ashworth SH; Allan BJ; Plane JMC. (2002). High resolution spectroscopy of the OIO radical: Implications for the ozone-depleting potential of
iodine. Geophys Res Lett, 29,1456, doi: 10.1029/2001GL013851.
Askey A; Lyon SB; Thompson GE; Johnson JB; Wood GC; Sage PW; Cooke MJ. (1993). The effect of atmospheric fly-ash particulates on the
atmospheric corrosion of zinc and mild steel. Corros Sci, 34,1055-1081.
Asman WAH; Drukker B; Janssen AJ. (1988). Modeled historical concentrations and depositions of NH3 and NH4 in Europe. Atmos Environ, 22,
725-735.
R-4
-------
Asman WAH; van Jaarsveld JA. (1992). A variable-resolution transport model applied for NHx for Europe. Atmos Environ, 26,445-464.
Asman WAH; Sutton MA; Schjorring J. (1998). Ammonia: Emission, atmospheric transport and deposition. New Phytol, 139,27-48.
Asner GP; Townsend AR; Riley WJ; Matson PA; Neff JC; Cleveland CC. (2001). Physical and biogeochemical controls over terrestrial
ecosystem responses to nitrogen deposition, Biogeochemistry, 54,1-39
Asp H; Berggren D. (1990). Phosphate and calcium uptake in beech (Fagus sylvatica) in the presence of aluminium and natural fulvic acids.
Physiol Plant, 80,307-314.
Atkinson R; Winer AM; Pitts JN Jr. (1986). Estimation of night-time N2O5 concentrations from ambient NO2 and NO3 radical concentrations and
the role of N2O5 in night-time chemistry. Atmos Environ, 20, 331-339.
Atkinson R; Arey J; Zielinska B; Aschmann SM. (1990). Kinetics and nitro-products of the gas-phase OH and NO3 radical-initiated reactions of
naphthalene-ds, fluoranthene-dio, and pyrene. Int J Chem Kinetics, 22, 999-1014.
Atkinson R; Baulch DL; Cox RA; Hampson RF Jr; Kerr JA; Troe J. (1992). Evaluated kinetic and photochemical data for atmospheric chemistry:
Supplement IV, IUPAC subcommittee on gas kinetic data evaluation for atmospheric chemistry. J Phys Chem Ref Data, 21,1125-
1568.
Atkinson R; Arey J. (1994). Atmospheric chemistry of gas-phase polycyclic aromatic hydrocarbons: Formation of atmospheric mutagens.
Environ Health Perspect, 4,117-126.
Atkinson R. (2000). Atmospheric chemistry of VOCs and NOx. Atmos Environ, 34,2063-2101.
Audet C; Wood CM. (1988). Do rainbow trout (Salmo gairdnen) acclimate to low pH? Can J Fish Aquat Sci, 45,1399-1405.
Ausset P; Del Monte M; Lefevre RA. (1999). Embryonic sulphated black crusts on carbonate rocks in atmospheric simulation chamber and in
the field: role of carbonaceous fly-ash. Atmos Environ, 33,1525-1534.
Avis PG; McLaughlin DJ; Dentinger BC; Reich PB. (2003). Long-term increases in nitrogen supply alters above and belowground
ectomycorrhizal communities and increases dominance of Russula species. New Phytol, 160,239-253.
Axler RP; Reuter JE. (1996). Nitrate uptake by phytoplankton and periphyton: Whole-lake enrichments and mesocosm-15N experiments in an
oligotrophic lake. Limnol Oceanogr, 41, 659-671.
Ayensu E; Claasen DVR; Collins M; Dearing A; Fresco L; Gadgil M; Gitay H; Glaser G; Juma C; Krebs J; Lenton R; Lubchenco J; McNeely JA;
Mooney HA; Pinstrup-Andersen P; Ramos M; Raven P; Reid WV; Samper C; Sarukhan J, Schei P; Tundisi JG; Watson RT; Guanhua
X; Zakri AH. (1999). Ecology: International ecosystem assessment. Science, 286,685-686.
Babich P; Davey M; Allen G; Koutrakis P. (2000). Method comparisons for particulate nitrate, elemental carbon, and PM2.5 mass in seven U.S.
cities. J Air Waste Manage Assoc, 50,1095-1105.
Baddeley JA; Thompson DBA; Lee JA. (1994). Regional and historical variation in the nitrogen content of Racomitrium lanuginosum in Britain
in relation to atmospheric nitrogen deposition. Environ Pollut, 84,189-196.
Baez S; Fargione J; Moore Dl; Collins SL; Gosz JR. (2007). Atmospheric nitrogen deposition in the northern Chihuahuan desert: Temporal
trends and potential consequences. J Arid Environ, 68, 640-651.
Baggs EM; Richter M: Cadisch G; Hartwig UA. (2003). Denitrification in grass swards is increased under elevated atmospheric CO2. Soil Biol
Biochem, 35. 729-732.
Baggs EM; Blum H. (2004). CH4 oxidation and emissions of CH4 and N2O from Lolium perenne swards under elevated atmospheric CO2. Soil
Biol Biochem, 36, 713-723.
Bailey SW; Hornbeck JW; Driscoll CT; Gaudette HE. (1996). Calcium inputs and transport in a base-poor forest ecosystem as interpreted by Sr
isotopes. Water Resour Res, 32,707-719.
Bailey SW; Horsley SB; Long RP; Hallet RA. (1999). Influence of geologic and pedologic factors on health of sugar maple on the Allegheny
Plateau, U.S. In: Horsley SB; Long RP. (Eds.), Sugar maple ecology and health: Proceedings of an international symposium (General
technical report NE-261; pp. 63-65): Radnor, PA: Northeastern Research Station; U.S. Forest Service; U.S. Department of Agriculture
(US DA).
Bailey SW; Buso DC; Likens GE. (2003). Implications of sodium mass balance for interpreting the calcium cycle of a forested ecosystem.
Ecology, 84, 471-484.
Bailey SW; Horsley SB; Long RP; Hallett RA. (2004). Influence of edaphic factors on sugar maple nutrition and health on the Allegheny
Plateau. Soil Sci Soc Am J, 68,243-252.
Bailey SW; Horsley SB; Long RP. (2005). Thirty years of change in forest soils of the Allegheny Plateau, Pennsylvania. Soil Sci Soc Am J, 69,
681-690.
Baird M; Zabowski D; Everett RL. (1999). Wildfire effects on carbon and nitrogen in inland coniferous forests. Plant Soil, 209,233-243.
Bajwa R; Read DJ. (1985).The biology of the mycorrhizae in the Ericaceae. IX. Peptides as nitrogen sources for the ericoid endophyte and for
mycorrhizal and non-mycorhizzal plants. New Phytol, 101,459-467.
Baker JL; Laflen JM. (1983). Water quality consequences of conservation tilage. J Soil Water Conserv, 38,186-193.
Baker JP; Schofield CL. (1980). Aluminum toxicity to fish as related to acid precipitation and Adirondack surface water quality. In: Drablos, D;
Tollan A (Eds.), Ecological impact of acid precipitation: Proceedings of an international conference; held in Sandefjord, Norway in
March 11-14,1980. (SNSF project; pp. 292-295). Oslo, Norway: Swiss National Science Foundation (SNSF).
Baker JP; Schofield CL. (1982). Aluminum toxicity to fish in acidic waters. Water Air Soil Pollut, 18,289-309.
Baker JP; Schofield CL. (1985). Acidification impacts on fish populations: A review. In: Adams D; Page WP (Eds.), Acid deposition:
environmental, economic, and policy issues (pp. 183-221). New York and London: Plenum Press.
Baker JP; Gherini SA; Christensen SW; Driscoll CT; Gallagher J; Munson RK; Newton RM; Reckhow KH; Schofield CL. (1990a). Adirondack
lakes survey: An interpretive analysis offish communities and water chemistry, 1984-1987. Ray Brook, NY: Adirondack Lakes Survey
Corporation.
R-5
-------
Baker JP; Bernard DP; Christensen SW; Sale MJ. (1990b). Biological effects of changes in surface water acid-base chemistry. (State of
science / technology report #13).Washington DC: National Acid Precipitation Assessment Program (NAPAP).
Baker JP; Christensen SW. (1991). Effects of acidification on biological communities in aquatic ecosystems. In: Charles DF; Christie S (Eds.),
Acidic deposition and aquatic ecosystems: Regional case studies (pp. 83-106). New York, NY: Springer-Verlag New York, Inc
Baker JP; Van Sickle J; Gagen CJ; DeWalle DR; Sharpe WE; Carline RF; Baldigo BP; Murdoch PS; Bath DW; Kretser WA; Simonin HA;
Wigington PJ Jr. (1996a). Episodic acidification of small streams in the northeastern United States: Effects on fish populations. Ecol
Appl, 6,423-437.
Baker LA; Brezonik PL. (1988). Dynamic model of in-lake alkalinity generation. Water Resour Res, 24,65-74.
Baker LA; Pollman CD; Eilers JM. (1988). Alkalinity regulation in softwater Florida lakes. Water Resour Res, 24,1069-1082.
Baker LA; Kaufmann PR; Herlihy AT; Eilers JM. (1990c). Acidic deposition: State of science and technology. In Report 9 - Current status of
surface-water acid-base chemistry. Final report (No. PB-92-100452/XAB). Washington, DC: Prepared in cooperation with Council on
Environmental Quality, and Wisconsin Univ.-Eau Claire. Dept. of Biology. National Acid Precipitation Assessment Program (NAPAP)
Baker LA. (1991a). Regional estimates of atmospheric dry deposition. In: Charles DF; Christie S (Eds.). Acidic deposition and aquatic
ecosystems: Regional case studies, (pp 645-652). New York, NY: Springer-Verlag.
Baker LA; Kaufmann PR; Herlihy AT; Eilers JM. (1991a). Current status of surface water acid-base chemistry. In: Irving PM (Ed.). Acidic
deposition: State of science and technology; Volume II: Aquatic processes and effects. (State of Science/Technology report no 9;
pp. 9-5-9-367; 9-A1; 9-B1; 9-C1; 9-D1; and 9-CP1). Washington, DC: National Acid Precipitation Assessment Program (NAPAP).
Baker LA. (1991b). Appendix A: Ion enrichment analysis for the regional case studies project. In: Charles DF, Christie S (Eds.). Acidic
deposition and aquatic ecosystems: Regional case studies, (pp. 641-644). New York, NY: Springer-Verlag.
Baker LA; Christensen SW. (1991). Effects of acidification on biological communities. In: Charles DF (Ed.).Acidic deposition and aquatic
ecosystems: Regional case studies. New York, NY: Springer-Verlag, pp. 83-106.
Baker LA; Herlihy AT; Kaufmann PR; Eilers JM. (1991b). Acidic lakes and streams in the United States: The role of acidic deposition. Science,
252,1151-1154.
Baker TR; Allen HL; Schoeneberger MM; Kress LW. (1994). Nutritional response of loblolly pine exposed to ozone and simulated acid rain.
Can J For Res, 24,453-461.
Baldigo BP; Murdoch PS. (1997). Effect of stream acidification and inorganic aluminum on mortality of brook trout (Salvelinus fontinalis) in the
Catskill Mountains, New York. Can J Fish Aquat Sci, 54,603-615.
Baldigo BP; Murdoch PS; Burns DA. (2005). Stream acidification and mortality of brook trout (Salvelinus fontinalis) in response to timber
harvest in Catskill Mountain watersheds, New York, USA. Can J Fish Aquat Sci, 62,1168-1183.
Baldigo BP; Lawrence GB; Simonin HA. (2007). Persistent mortality of brook trout in episodically acidified streams of the southwestern
Adirondack Mountains, New York. Trans Am Fish Soc, 136,121-134.
Balkanski YJ; Jacob DJ; Gardener GM; Graustein WC; Turekian KK. (1993). Transport and residence times of tropospheric aerosols inferred
from a global three-dimensional simulation of 210Pb. J Geophys Res, 98,20573-20586.
Bamford HA; Baker JE. (2003). Nitro-polycyclic aromatic hydrocarbon concentrations and sources in urban and suburban atmospheres of the
Mid-Atlantic region. Atmos Environ, 37,2077-2091.
Banzhaf S; Burtraw D; Evans D; Krupnick A. (2006). Valuation of natural resource improvements in the Adirondacks. Land Econ, 82,445-464.
Barbier EB. (1991). An approach to economic valuation of tropical wetlands: With examples from Guatemala and Nicaragua. In: Girvan NP;
Simons D (Eds.). Caribbean ecology and economics, (pp 207-231). St Michael, Barbados: Caribbean Conservation Association.
Barbier EB; Acreman M; Knowler D. (1997). Economic valuation of wetlands: A guide for policy makers and planners. Prepared in collaboration
with the Department of Environmental Economics and Environ Manage of the University of York, UK, the Institute of Hydrology, and
lUCN-The World Conservation Union. Published in Gland, Switzerland by the Ramsar Convention Bureau, with financial support from
the U.K. Department of the Environment and the Swedish International Development Cooperation Agency.
Bardwell CA; Maben JR; Hurt JA; Keene WC; Galloway JN; Boatman JF; Wellman DL. (1990). A technique using high-flow, dichotomous filter
packs for measuring major atmospheric chemical constituents. Glob Biogeochem Cycles, 4,151-163.
Barinaga M. (1990). Where have all the froggies gone? Science, 247,1033-1034.
Barker DH; Vanier C; Naumburg E; Charlet TN; Nielsen KM; Newingham BA; Smith SD. (2006). Enhanced monsoon precipitation and nitrogen
deposition affect leaf traits and photosynthesis differently in spring and summer in the desert shrub Larrea tridentata. New Phytol, 169,
799-808.
Barmuta LA; Cooper SD; Hamilton SK; Kratz KW; Melack JM. (1990). Responses of zooplankton and zoobenthos to experimental acidification
in a high-elevation lake (Sierra Nevada, California, USA). Freshwater Biol, 23,571-586.
Barnard R; Leadley PW; Hungate BA. (2005). Global change, nitrification, and denitrification: A review. Glob Biogeochem Cycles, 19,
doi:10.1029/2004GB002282.
Barnes I; Bastian V; Becker KH; Overath RD. (1991). Kinetic studies of the reactions of IO, BrO and CIO with dimethylsulfide. Int J Chem
Kinetics, 23, 579-591.
Baron J; Norton SA; Beeson DR; Hermann R. (1986). Sediment diatom and metal stratigraphy from Rocky Mountain lakes with special
reference to atmospheric deposition. Can J Fish Aquat Sci, 43,1350-1362.
Baron J. (1992).Surface waters. In: Baron J (Ed.). Biogeochemistry of a subalpine ecosystem: Loch Vale watershed. (Ecological Studies, series
Volume 90, pp 142-186). New York, NY: Springer-Verlag.
Baron JS; Ojima DS; Holland EA; Parton WJ. (1994). Analysis of nitrogen saturation potential in Rocky Mountain tundra and forest: Implications
for aquatic systems. Biogeochemistry, 27,61-82.
R-6
-------
Baron JS; Allstott J; Newkirk BK. (1995). Analysis of long term sulfate and nitrate budgets in a Rocky Mountain basin. In: Tonnessen KA;
Williams MW; Tranter M (Eds.). Biogeochemistry of seasonally snow-covered catchments. Proceedings from an international
symposium held during the XXI Assembly of the International Union of Geodesy and Geophysics in Boulder, CO, in July 1-14,1995.
(IAHS publication no 228, pp 255-261). Wallingford, Oxfordshire, U.K. Centre for Ecology and Hydrology: International Union of
Geodesy and Geophysics (IAHS) Press.
Baron JS; Rueth HM; Wolfe AM; Nydick KR; Allstott EJ; Minear JT; Moraska B. (2000). Ecosystem responses to nitrogen deposition in the
Colorado Front Range. Ecosystems, 3,352-368.
Baron JS. (2006). Hindcasting nitrogen deposition to determine ecological critical load. Ecol Appl, 16,433-439.
Barrie LA; Bottenheim JW; Schnell RC; Crutzen PJ; Rasmussen RA. (1988). Ozone destruction and photochemical reactions at polar sunrise in
the lower Arctic atmosphere. Nature, 334,138-141.
Barton L; Kiese R; Gatter D; Butterbach-Bahl K; Buck R; Hinz C; Murphy DV. (2008). Nitrous oxide emissions from a cropped soil in a semi-arid
climate. Glob Chang Biol, 14,177-192.
Bashkin VN; Kozlov MY; Priputina IV; Avrumuchev AV; Dedkova IS. (1995). Calculation and mapping of critical loads of S, N, and acidity of
ecosystems of northern Asia. Water Air Soil Pollut, 85, 2395-2400.
Basiliko N; Moore TR; Jeannotte R; Bubier JL. (2006). Nutrient input and carbon and microbial dynamics in an ombrotrophic bog. Geomicrobiol
J, 23, 531-543.
Battoe LE; Lowe EF. (1992). Acidification of Lake Annie, Highlands Co., FL. Water Air Soil Pollut, 65,69-80.
Bauer GA; Bazaaz FA; Minocha R; Long S; Magill AH; Aber JD; Berntson GM. (2004). Effects of chronic N additions on tissue chemistry,
photosynthetic capacity, and carbon sequestration potential of a red pine (Pinus resinosa Ait. .stand in the NE United States. For Ecol
Manage, 196,173-186.
Bayley SE; Schindler DW; Parker BR; Stainton MP; Beaty KG. (1992). Effects of forest fire and drought on acidity of a base-poor boreal forest
stream: Similarities between climatic warming and acidic precipitation. Biogeochem, 17,191-204.
Beamish RJ; Harvey HH. (1972). Acidification of the La Cloche Mountain lakes, Ontario, and resulting fish mortalities. J Fish Res Bd Can, 29,
1131-1143.
Beamish RJ. (1976). Acidification of lakes in Canada by acid precipitation and the resulting effects on fishes. Water Air Soil Pollut, 6,501-514.
Beck MB. (1987). Water quality modeling: A review of the analysis of uncertainty. Water Resour Res, 23,1393-1442.
Beck MB; Ravelz JR; Mulkey LA; Barnwell TO. (1997). On the problem of model validation for predictive exposure assessments. Stochastic
Hydrology and Hydraulis, 11,229-254.
Beck MB; Chen J. (2000). Assuring the quality of models designed for predictive tasks. In: Saltelli A; Chan K; Scott EM (Eds.), Handbook on
sensitivity analysis, (pp 401-420). Chichester, UK: Wiley.
Becker DS; Bigham GN. (1995). Distribution of mercury in the aquatic food web of Onondaga Lake, New York. Water Air Soil Pollut, 80,563-
571.
Bedford BL; Walbridge MR; Aldous A. (1999). Patterns in nutrient availability and plant diversity of temperate North American Wetlands:
Within-stand nutrient cycling in Wetland Ecosystems. Ecology, 80,2151-2169.
Bedford BL; Godwin KS. (2003). Fens of the United States: Distribution, characteristics, and scientific connection versus legal isolation.
Wetlands, 23,608-629.
BeedlowPA; Tingey DT; Phillips DL; HogsettWE; Olszyk DM. (2004). Rising atmospheric CO2 and carbon sequestration in forests. Front Ecol
Environ, 2,315-322.
Beggs GL; Gunn JM. (1986). Response of lake trout (Salvelinus namaycush and brook trout (S. fbntinalis) to surface water acidification in
Ontario. Water Air Soil Pollut, 30, 711-717.
Behnke W; George C; Sheer V; Zetzsch C. (1997). Production and decay of CINO2 from the reaction of gaseous N2O5 with NaCI solution: Bulk
and aerosol experiments. J Geophys Res, 102,3795-3804.
Beier C; Hultberg H; Moldan F; Wright RF. (1995). MAGIC applied to roof experiments (Risdalsheia, N; Gardsjon, S; Klosterhede, DK) to
evaluate the rate of reversibility of acidification following experimentally reduced acid deposition. Water Air Soil Pollut, 85,1745-1751.
Bell JNB. (1985). SO2 effects on the productivity of grass species. In: Winner WE; Mooney HA; Goldstein RA (Eds.). Sulfur dioxide and
vegetation: Physiology, ecology, and policy issues, (pp. 250-263). Stanford, CA: Stanford University Press.
Belnap J; Sigal L; Moir W; Eversman S. (1993). Identification of sensitive species, in lichens as bioindicators of air quality. In: Huckaby LS
(Ed.), Lichens as bioindicators of air quality (pp. 67-88). Fort Collins, Colorado: Rocky Mountain Forest and Range Experimental
Station, U.S. Forest Service, U. S. Department of Agriculture.
Benner RL; Stedman DH. (1989). Universal sulfur detection by chemiluminescence. Anal. Chem, 61,1268-1271.
Benner RL; Stedman DH. (1990). Field evaluation of the sulfur chemiluminescence detector. Environ. Sci. Technol, 24,1592-1596.
Bennett D; Goble R; Linthurst R. (1985). The acidic deposition phenomenon and its effects: Critical assessment document (Report no.
EPA/600/8-85/001). Washington, DC: Office of Acid Deposition, Environmental Monitoring, and Quality Assurance; National Exposure
Research Laboratory; Office of Research and Development; U.S. Environmental Protection Agency.
Benoit JM; Fitzgerald WF; Damman AWH. (1998). The biogeochemistry of an ombrotrophic bog: Evaluation and use as an archive of
atmospheric mercury deposition. Environ Res, 78,118-133.
Benoit JM; Gilmour CC; Mason RP; Heyes A. (1999). Sulfide controls on mercury speciation and bioavailability to methylating bacteria in
sediment pore waters. Environ Sci Technol, 33,951-957.
Benoit JM; Mason RP; Gilmour CC; Aiken GR. (2001). Constants for mercury binding by dissolved organic matter isolates from the Florida
Everglades. Geochim Cosmochim Acta, 65,4445-4451.
R-7
-------
Benoit JM; Gilmour C; Heyes A; Mason RP; Miller C. (2003). Geochemical and biological controls over methylmercury production and
degradation in aquatic ecosystems. In: Chai Y, Braids OC (Eds.), Biogeochemistry of environmentally important trace elements, ACS
Symposium Series no. 835 (pp. 262-297). Washington, D.C.: American Chemical Society.
Benstead JP; Deegan LA; Peterson BJ; Huryn AD; Bowden WB; Suberkropp K; Buzby KM; Green AC; Vacca JA. (2005). Responses of a
beaded Arctic stream to short-term N and P fertilisation. Freshw Biol, 50,277-290.
Bentley J; Horst R. (1998). Commercial forestry benefits of alternative emission controls for the NOx SIP Call (Report prepared by Science
Applications International Corporation (SAIC #68-D-98-113) for the Office of Air Quality Planning and Standards). Research Triangle
Park, NC; Office of Air Quality Planning and Standards; U.S. Environmental Protection Agency.
Berendse F. (1994). Competition between plant populations at low and high nutrient supplies. Oikos, 71,253-260.
Berendse F; Breemen NV; Rydin H; Buttler A; Heijmans M; Hoosbeek MR; Lee JA; Mitchell A; Saarinen T; Vassander H; Wallen B. (2001).
Raised atmospheric CO2 levels and increased N deposition cause shifts in plant species composition and production in Sphagnum
bogs. Glob Chang Biol, 7,591-598.
Berg B. (2000). Initial rates and limit values for decomposition of Scots pine and Norway spruce needle litter: A synthesis for N-fertilized forest
stands. Can J For Res, 30,122-135.
Bergametti G; Remoudaki E; Losno R; Steiner E; Chatenet B. (1992). Source, transport and deposition of atmospheric phosphorus over the
northwestern Mediterranean. J Atmos Chem, 14, 501-513.
Bergeron C; Husak J; Jason Unrine J; RomanekC; Hopkins W. (2007). Influence of feeding ecology on blood mercury concentrations in four
species of turtles. Environ Toxicol Chem, 26,1733-1741.
Berggren D; Mulder J. (1995). The role of organic matter in controlling aluminum solubility in acidic mineral soil horizons. Geochim Cosmochim
Acta, 59,4167-4180.
Bergman HL; Mattice JS; Brown DJA. (1988). Lake acidification and fisheries project: adult brook trout (Salvelinus fbntinalis). Can J Fish Aquat
Sci, 45,1561-1562.
Bergstrom A; Blomqvist P; Jansson M. (2005). Effects of atmospheric nitrogen deposition on nutrient limitation and phytoplankton biomass in
unproductive Swedish lakes. Limnol Oceanogr, 50,987-994.
Bergstrom A; Jansson M. (2006). Atmospheric nitrogen deposition has caused nitrogen enrichment and eutrophication of lakes in the northern
hemisphere. Glob Chang Biol, 12,635-643.
Bergstrom JC; Taylor LO. (2006). Using meta-analysis for benefits transfer: Theory and practice. Ecol Econ, 60,351-360.
Bernhardt ES; Likens GE. (2002). Dissolved organic carbon enrichment alters nitrogen dynamics in a forest stream. Ecology, 83,1689-1700.
Bernot MJ; Dodds WK; Gardner WS; McCarthy MJ; Sobolev D; Tank JL. (2003). Comparing denitrification estimates for a Texas estuary by
using acetylene inhibition and membrane inlet mass spectrometry. Appl Environ Microbiol, 69,5950-5956.
Berresheim H; Eisele FL; Tanner DJ; Mclnnes LM; Ramsey-Bell DC; Covert DS. (1993). Atmospheric sulfur chemistry and cloud condensation
nuclei (CCN) concentrations over the northeastern Pacific coast. J Geophys Res, 98,12701-12711
Berresheim H; Wine PH; Davis DD. (1995). Sulfur in the atmosphere. In: Singh HB (Ed.), Composition, chemistry, and climate of the
atmosphere (pp. 251-307). New York, NY: Van Nostrand Reinhold.
Bert D; Leavitt SW; Duponey JL. (1997). Variations in wood 613C and water-use efficiency of Abies alba during the last century. Ecology, 78,
1588-1595.
Bertness MD; Crain C; Holdredge C; Sala N. (2008). Eutrophication and consumer control of New England salt marsh primary productivity.
Conserv Biol, 22,131-9.
Bertram TH; Heckel A; Richter A; Burrows JP; Cohen RC. (2005). Satellite measurements of daily variations in soil NOx emissions. Geophys
Res Lett, 32, doi: 10.1029/2005GL024640
Bey I; Jacob DJ; Yantosca RM; Logan JA; Field B; Fiore AM; Li Q; Liu H; Mickley LJ; Schultz MG. (2001). Global modeling of tropospheric
chemistry with assimilated meteorology: Model description and evaluation. J Geophys Res, 106,23073-23095.
Bezabeh DZ; Bamford HA; Schantz MM; Wise SA. (2003). Determination of nitrated polycyclic aromatic hydrocarbons in diesel particulate-
related standard reference materials by using gas chromatography/mass spectrometry with negative ion chemical ionization. Anal
Bioanal Chem, 375,381-388.
Bhattacharya A; Mudgal R; Taneja A. (2004). Acid deposition and critical load analysis in Agra, India. J Hazard Mater, 106,157-160.
Billen N; Schatzle H; Seufert G; Arndt U. (1990). Performance of some growth variables. Environ Pollut, 68,419-434.
BillettMF; Parker-Jervis F; Fitzpatrick EA; Cresser MS. (1990). Forest soil chemical changes between 1949/50 and 1987. European Journal of
Soil Science, 41,133-145.
Billings WD. (1978). Plants and the ecosystem. (3rd Edition, pp 1-62; 83-108). Belmont, CA: Wadsworth Publishing Co.
Binkley D; Richter D. (1987). Nutrient cycles and H+budgets of forest ecosystems. Adv Ecol Res, 16,2-52
Binkley D. (1995). The influence of tree species on forest soils: Processes and patterns. In: Mead DJ; Cornforth IS (Eds.). Proceedings of the
Trees and Soil Workshop, (pp 1-33). Canterbury, New Zealand; Lincoln University Press.
Binkowski FS; Arunachalam S; Adelman Z; Pinto JP. (2007). Examining photolysis rates with a prototype online photolysis module in CMAQ.
Journal of Applied Metereology and Climatology, 46,1252-1256.
Birks HJB; Berge F; Boyle JF; Cumming BF. (1990). A palaeoecological test of the land-use hypothesis for recent lake acidification in south-
west Norway using hill-top lakes. J Paleolimnol, 4,69-65.
Bishop GA; Stedman DH. (1996). Measuring the emissions of passing cars. Acc Chem Res, 29,489-495.
Blair RB. (1990). Water quality and the summer distribution of common loons in Wisconsin. Passenger Pigeon, 52,119-126.
Blake L; Goulding KWT; MottCJB; Johnston AE. (1999). Changes in soil chemistry accompanying acidification over more than 100 years
under woodland and grass at Rothamsted Experimental Station, UK. Eur J Soil Sci, 50,401-412.
R-8
-------
Blanchard CL; Roth PM; Tanenbaum SJ; Ziman SD; Seinfeld JH. (2000). The use of ambient measurements to identify which precursor
species limit aerosol nitrate formation. J Air Waste Manage Assoc, 50,2073-2084.
Blanchard CL; Hidy GM. (2003). Effects of changes in sulfate, ammonia, and nitric acid on particulate nitrate concentrations in the southeastern
United States. J Air Waste Manage Assoc, 53,283-290.
Blanchard P; Brook JR; Brazil P. (2002). Chemical characterization of the organic fraction of atmospheric aerosol at two sites in Ontario,
Canada. J Geophys Res, 107,8348, doi:10.1029/2001JD000627.
Blancher PJ; McNicol DK. (1988). Breeding biology of tree swallows in relation to wetland acidity. Can J Zool, 66, 842-849.
Blancher PJ; McNicol DK. (1991). Tree swallow diet in relation to wetland activity. Can J Zool, 69, 2629-2637.
Blaschke H. (1990). Mycorrhizal populations and fine root development on Norway spruce exposed to controlled doses of gaseous pollutants
and simulated acidic rain treatments. Environ Pollut, 68,409-418.
Blett T; Geiser L; Porter E. (2003). Air pollution-related lichen monitoring in national parks, forests, and refuges: Guidelines for studies intended
for regulatory and management purposes. (Report #NPS-D2292). http://www2.nature.nps.gov/air/Pubs/pdf/Lichen Studies.pdf.
Corvallis, OR and Denver, CO; Air Resources Division; National Prk Service; U.S. Forest Service ; U.S. Department of the Interior,
U.S. Department of Agriculture (USDA).
Bloom AJ; Chapin FSI; Mooney HA. (1985). Resource limitation in plants - an economic analogy. Annu Rev Ecol Syst, 16,363-392.
Bloom NS. (1992). On the chemical form of mercury in edible fish and marine invertebrate tissue. Can J Fish Aquat Sci, 49,1010-1017.
Bloom SC; Takacs LL; Da Silva AM; Ledvina D. (1996). Data assimilation using incremental analysis updates. Monthly Weather Rev, 124,
1256-1271.
Bloom SC; da Silva A; Dee D; Bosilovich M; Chern J-D; Pawson S; Schubert S; Sienkiewicz M; Stajner I; Tan W-W; Wu M-L. (2005).
Documentation and validation of the Goddard Earth Observing System (GEOS) data assimilation system-version 4. (Technical report
series on global modeling and data assimilation, 104606, v. 26). Greenbelt, MD: Goddard Space Flight Center; National Aeronautics
and Space Administration (NASA)
Bloss WJ; Lee JD; Johnson GP; Sommariva R; Heard DE; Saiz-Lopez A; Plane JMC; McFiggans G; Coe H; Flynn M; Williams P; Rickard AR;
Fleming ZL. (2005). Impact of halogen monoxide chemistry upon boundary layer OH and HO2 concentrations at a coastal site.
Geophys Res Lett, 32, doi:10.1029/2004GL022084
Blucher DB; Svensson JE; Johansson LG. (2005). Influence of ppb levels of SO2 on the atmospheric corrosion of aluminum in the presence of
NaCI. J Electrochem Soc, 152, B397-B404.
Blum JD; Klaue A; NezatCA; Driscoll CT; Johnson CE; Siccama TG; Eagar C; Fahey TJ; Likens GE. (2002). Mycorrhizal weathering of apatite
as an important calcium source in base-poor forest ecosystems. Nature, 417,729-731.
Bobbink R; Heil GW; Raessen MBAG. (1992a). Atmospheric deposition and canopy exchange processes in heathland ecosystems. Environ
Pollut, 75, 29-37.
Bobbink R; Boxman D; Fremstad E; Heil G; Houdijk A; Roelofs J. (1992b). Critical loads for nitrogen eutrophication of terrestrial and wetland
ecosystems based upon changes in vegetation and fauna. In: Grennfelt P, Thornelof E (Eds.), Critical loads for nitrogen. Nord
(Miljorapport Volume 41, pp 111-159). Copenhagen; Nordic Council of Ministers.
Bobbink R; Hornung M; Roelofs JGM. (1996). Empirical nitrogen critical loads for natural and semi-natural ecosystems. Annex III. In: Manual
on methodologies and criteria for mapping critical levels/loads and geographical areas where they are exceeded. (Proceddings from
UNECE Convention on Long-Range Transboundary Air Pollution). Berlin, Germany; Umvelt Bundes Arnt.
Bobbink R. (1998). Impacts of tropospheric ozone and airborne nitrogenous pollutants on natural and semi-natural ecosystems: A commentary.
New Phytol, 139,161-168.
Bobbink R; Hornung M; Roelofs JGM. (1998). The effects of air-borne nitrogen pollutants on species diversity in natural and semi-natural
European vegetation. J Ecol, 86,717-738.
Bobbink R; Ashmore M; Braun S; Fluckiger W; van den Wyngaert IJJ. (2003). Empirical nitrogen critical loads for natural and semi-natural
ecosystems: 2002 update. In: Achermann B; Bobbink R (Eds.), Empirical critical loads for nitrogen (pp. 43-170). Berne, Switzerland:
Swiss Agency for Environment, Forest and Landscape SAEFL.
Bobrowski, N; Honninger, G; Galle, B; Piatt, U. (2003). Detection of bromine monoxide in a volcanic plume. Nature (London, U.K.), 423:273-
276.
Bockstael, NE; McConnell, KE; Strand, IE. (1988). Benefits from improvements in Chesapeake Bay water quality. Volume III of Benefit analysis
using indirect or imputed market methods. Washington, DC: U.S. Environmental Protection Agency.
Bockstael N; McConnell K; Strand I. (1989a). A random utility model for sportfishing: Some preliminary results for Florida. Mar Resource Econ,
6, 245-260.
Bockstael NE; McConnell KE; Strand IE. (1989b). Measuring the benefits of improvements in water quality: The Chesapeake Bay. Mar
Resource Econ, 6,1-18.
Bodelier PLE; Roslev P; Heckel T; Frenze, P. (2000). Stimulation by ammonium-based fertilizers of methane oxidation in soil around rice roots.
Nature, 403, 421-424.
Bodelier PLE; Laanbroek HJ. (2004). Nitrogen as a regulatory factor of methane oxidation in soils and sediments. FEMS Microbiol Ecol, 47,
265-277.
Boening DW. (2000). Ecological effects, transport, and fate of mercury: A general review. Chemosphere40,1335-1351.
Boersma KF; Eskes HJ; Brinksma EJ. (2004). Error analysis for tropospheric NO2 retrieval from space. J Geophys Res, 109, doi:
10.1029/2003JD003962.
Boesch DF; Brinsfield RB; Magnien RE. (2001). Chesapeake Bay eutrophication: Scientific understanding, ecosystem resoration, and
challenges for agriculture. J Environ Qual, 30, 303-320.
R-9
-------
Boggs JL; McNulty SG; Gavazzi MJ; Myers JM. (2005). Tree growth, foliar chemistry, and nitrogen cycling across a nitrogen deposition
gradient in southern Appalachian deciduous forests. Can J For Res, 35,1901-1913.
Bohlke JK; Denver JM. (1995). Combined use of ground-water dating, chemical and isotopic analyses to resolve the history and fate of nitrate
contamincation in 2 agricultural watersheds, Atlantic coastal-plain, Maryland. Water Resour Res, 31, 2319-2339.
Boke H; Gokturk EH; Caner-Saltik EN; Demirci S. (1999). Effect of airborne particle on SO2—calcite reaction. Appl Surf Sci, 140,70-82.
Bondietti EA; McLaughlin SB. (1992). Evidence of historical influences of acidic deposition on wood and soil chemistry. In: Johnson DW;
Lindberg SE, (Eds.).Atmospheric deposition and forest nutrient cycling: a synthesis of the integrated forest study. (Series: Billings WD;
Golley F; Lange OL; Olson JS; Remmert H, [Eds.] Ecological studies analysis and synthesis: Vol.ume 91, pp 358-377). New York, NY:
Springer-Verlag.
Boonpragob K; Nash TH, III; Fox CA. (1989). Seasonal deposition patterns of acidic ions and ammonium to the lichen Ramalina menziesiiJay\.
in southern California. Environ Exp Bot, 29,187-197.
Boonpragob K; Nash THI. (1991). Physiological responses of the lichen Ramalina menziesii Tayl. to the Los Angeles urban environment.
Environ Exp Bot, 31,229-238.
Booth CE; McDonald DG; Simons BP; Wood CM. (1988). Effects of aluminum and lowpH on net ion fluxes and ion balance in the brook trout
(Salvelinus fbntinalis). Can J Fish Aquat Sci, 45,1563-1574.
Boring CB; Al-Horr R; Genfa Z; Dasgupta PK; Martin M. W; Smith, W. F. (2002).Field measurement of acid gases and soluble anions in
atmospheric particulate matter using parallel plate wet denuder and alternating filter-based automated analysis system. Anal. Chem.
74:1256-1268.
Borken W; Beese F; Brumme R; Lamersdorf N. (2002). Long-term reduction in nitrogen and proton inputs did not affect atmospheric methane
uptake and nitrous oxide emission from a German spruce forest soil. Soil Biol Biochem, 34,1815-1819.
Bormann FH; Likens GE. (1967). Nutrient cycling. Science, 155,424-429.
Bormann FH; Likens GE. (1979). Pattern and process in a forested ecosystem. New York, NY: Springer Verlag.
Bormann FH; Likens GE; Fisher DW; Pierce RS. (1968). Nutrient loss accelerated by clear-cutting of a forest ecosystem. Science, 159,822-
884.
Bormann FH; Likens GE; Melillo JM. (1977). Nitrogen budget for an aggrading northern hardwood forest ecosystem. Science, 196,981-982.
Borum J. (1996). Shallow waters and land/sea boundaries. In: Jorgensen BB; Richardson K, (Eds.). Eutrophication in coastal marine
ecosystems, (pp 179-203). Washington DC: American Geophysical Union.
Bouwes NW; Schneider R. (1979). Procedures in estimating benefits of water quality change. Am J Agric Econ, 61, 535-539.
Bouwman AF; Van Drecht G; Knoop JM; Beusen AHW; Meinardi CR. (2005). Exploring changes in river nitrogen export to the world's oceans.
Glob Biogeochem Cycles, 19, doi: 10.1029/2004GB002314.
Bovensmann H; Burrows JP; Buchwitz M; Frerick J; Noel S; Rozanov W; Chance KV; Goede APH. (1999). SCIAMACHY: Mission objectives
and mesurement modes. J Atmos Sci, 56,127-150.
Bowden, RD, Melillo JM; Steudler PA; Aber JD. 1991. Effects of nitrogen additions on annual nitrous-oxide fluxes from temperate forest soils in
the northeastern Unitied States. J Geophys Res, 96,9321-9328.
Bowden RD; Rullo G; Stevens GR; Steudler PA. (2000). Soil fluxes of carbon dioxide, nitrous oxide, and methane at a productive temperate
deciduous forest. J Environ Qual, 29,268-276.
Bowden RD; Davidson E; Savage K; Arabia C; Steudler P. (2004). Chronic nitrogen additions reduce total soil respiration and microbial
respiration in temperate forest soils at Harvard Forest. For Ecol Manage, 196,43-56.
Bowen JL;Valiela I. (2001). Historical changes in atmospheric nitrogen deposition to Cape Cod, Massachusetts, USA. Atmos Environ, 35,
1039-1051.
Bowman WD. (1992). Inputs and storage of nitrogen in winter snowpack in an alpine ecosystem. Arct Alp Res, 24,211-215.
Bowman WD; Theodose TA; Schardt JC; Conant RT. (1993). Constraints of nutrient availability on primary production in two alpine tundra
communities. Ecology, 74,2085-2097.
Bowman WD. (1994). Accumulation and use of nitrogen and phosphorus following fertilization in two alpine tundra communities. Oikos, 70,
261-270.
Bowman WD; Theodose TA; Fisk MC. (1995). Physiological and production responses of plant growth forms to increases in limiting resources
in alpine tundra: Implications for differential community response to environmental change. Oecologia, 101, 217-227.
Bowman WD; Steltzer H. (1998). Positive feedbacks to anthropogenic nitrogen deposition in Rocky Mountain alpine tundra. Ambio, 27,514-
517.
Bowman WD. (2000). Biotic controls over ecosystem response to environmental change in alpine tundra of the Rocky Mountains. Ambio, 29,
396-400.
Bowman WD; Bilbrough CJ. (2001). Influence of a pulsed nitrogen supply on growth and nitrogen uptake in alpine graminoids. Plant Soil, 233,
283-290.
Bowman WD; Fisk MC. (2001). Primary production. In: Bowman WD; Seastedt TR. (Eds.). Structure and function of an alpine ecosystem:
Niwot Ridge, Colorado, (pp 177-197). Oxford, UK: Oxford University Press.
Bowman WD; Gartner JR; Holland K; Wiedermann M. (2006). Nitrogen critical loads for alpine vegetation and terrestrial ecosystem response:
are we there yet? Ecol Appl, 16,1183-1193.
Boxman AW; Blanck K; Brandrud T-E; Emmett BA; Gundersen P; Hogervorst RF; Kjonass OJ; Persson H; Timmermann V. (1998a). Vegetation
and soil biota response to experimentally-changed nitrogen inputs in coniferous forest ecosystems of the NITREX project. For Ecol
Manage, 101,65-79.
Boxman AW; van der Ven PJM; Roelofs JGM. (1998b). Ecosystem recovery after a decrease in nitrogen input to a Scots pine stand at
Ysselsteyn, The Netherlands. For Ecol Manage, 101,155-163.
R-10
-------
Boyd J; Banzhaf S. (2007). What Are Ecosystem Services? The need for standardized environmental accounting units. Ecol Econ, 63,616-
626.
Boyer EW; Howarth RW. (2002). The nitrogen cycle at regional to global scales: report of the International SCOPE Nitrogen Project. Dordrecht,
The Netherlands: Kluwer Academic Publishers.
Boyer EW; Goodale CL; Jaworski NA; Howarth RW. (2002). Anthropogenic nitrogen sources and relationships to riverine nitrogen export in the
northeastern USA. Biogeochemistry, 57/58,137-169.
Boyer EW; Alexander RB; Parton WJ; Li C; Butterbach-Bahl K; Donner SD; Skaggs RW. (2006a). Modeling denitrification in terrestrial and
aquatic ecosystems at regional scales. Ecol Appl, 16,2123-2142.
Boyer E; Howarth RW; Galloway JN; Dentener FJ; Green PA; Vorosmarty CJ. (2006b). Riverine nitrogen export from the continents to the
coasts. Glob Biogeochem Cycles, 20.
Boyle KJ; Poor PJ; Taylor LO. (1999). Estimating the demand for protecting freshwater lakes from eutrophication. Am J Agric Econ, 81,1118-
1122.
Boynton WR; Garber JH; Summers R; Kemp WM. (1995). Inputs, transformations, and transport of nitrogen and phosphorus in Chesapeake
Bay and selected tributaries. Estuaries, 18,285-314.
Bradford DF; Swanson C; Gordon MS. (1992). Effects of low pH and aluminum on two declining species of amphibians in the Sierra Nevada,
California. J Herpetol, 26:369-377.
Bradford DF; Tabatabai F; Graber DM. (1993). Isolation of remaining populations of the native frog, Rana muscosa, by introduced fishes in
Sequoia and Kings Canyon National Parks, California. Conserv Biol, 7,882-888.
Bradford MA; Ineson P; Wookey PA; Lappin-Scott HM. (2001). The effects of acid nitrogen and acid sulphur deposition on CH4 oxidation in a
forest soil: A laboratory study. Soil BiolBiochem, 33,1695-1702.
Bradley K; Drijber RA; Knops J. (2006). Increased N availability in grassland soils modifies their microbial communities and decreases the
abundance of arbuscular mycorrhizal fungi. Soil Biol Biochem, 38,1583-1595.
Bradshaw J; Davis D; Crawford J; Chen G; Shetter R; Muller M; Gregory G; Sachse G; Blake D; Heikes B; Singh H; Mastromarino J; Sandholm
S. (1999). Photofragmentation two-photon laser-induced fluorescence detection of NO2 and NO: Comparison of measurements with
model results based on airborne observations during PEM-Tropics A. Geophys Res Lett, 26,471-474.
Bragazza L; Freeman C; Jones T; Rydin H; Limpens J; Fenner N; Ellis T; Gerdol R; Hajek M; Hajek T. Lacumin P; Kutnar L; Tahvanainen T;
Toberman H. (2006). Atmospheric nitrogen deposition promotes carbon loss from peat bogs. Proc Natl Acad Sci USA, 103,19386-
19389.
Bragazza L; Freeman C. (2007). High nitrogen availability reduces polyphenol content in Sphagnum peat. Sci Total Environ, 377,439-443.
Brand GJ; Nelson MD; Wendt DG; Nimerfro KK. (2000.). The hexagon/panel system for selecting FIA plots under an annual inventory. In:
McRoberts RE; Reams GA; Van Deusen PC (Eds.) Proceedings of the first annual forest inventory and analysis symposium. (General
technical report NC-213; pp. 8-13). St. Paul, MN; North Central Research Station; U.S. Forest Service; U.S. Department of Agriculture
(US DA).
Brander LM; Florax RJGM; Vermaat JE. (2006). The empirics of wetland valuation: A comprehensive summary and a meta-analysis of the
literature. Environmental and Resource Economics, 33,223-250.
Branfireun BA; Heyes A; Roulet NT. (1996). The hydrology and methylmercury dynamics of a Precambrian Shield headwater peatland. Water
Resour Res, 32,1785-1794.
Branfireun BA; Roulet NT; Kelly CA; Rudd WM. (1999). In situ sulphate stimulation of mercury methylation in a boreal peatland: toward a link
between acid rain and methylmercury contamination in remote environments. Glob Biogeochem Cycles, 13,743-750.
Branfireun BA; Bishop K; Roulet, NT; Granberg, G; Nilsson, M. (2001). Mercury cycling in boreal ecosystems: The long-term effect of acid rain
constituents on peatland pore water methylmercury concentrations. Geophys Res Lett, 28,1227-1230.
Brauer M; Koutrakis P; Keeler GJ; Spengler JD. (1991). Indoor and outdoor concentrations of inorganic acidic aerosols and gases. J Air Waste
Manage Assoc, 41,171-181.
Braun S; Schindler C; Volz R; Fluckiger W. (2003). Forest Damages by the Storm 'Lothar' in permanent observation plots in Switzerland: The
significance of soil acidification and nitrogen deposition. Water Air Soil Pollut, 142,327-340.
Bredemeier M; Blanck K; Xu YJ; Tietema A; Boxman AW; EmmettB; Moldan F; Gundersen P; Schleppi P; Wright RF. (1998). Input-output
budgets at the NITREX sites. For Ecol Manage, 101, 57-64.
Breiner J; Gimeno BS; Fenn M. (2007). Calculation of theoretical and empirical nutrient N critical loads in the mixed conifer ecosystems of
southern California. ScientificWorldJournal, 7,198-205.
Brenner M; Binford MW. (1988). Relationships between concentrations of sedimentary variables and trophic state in Florida lakes. Can J Fish
Aquat Sci, 45, 294-300.
Brezonik PL; Baker LA; Perry TE. (1987). Mechanisms of alkalinity generation in acid sensitive softwater lakes. In: Hites R; Eisenreich SJ,
(Eds.). Sources and fates of aquatic pollutants. (Advances in Chemistry Series Volume 216. pp 229-262). Washington, DC: American
Chemical Society.
Brezonik PL; Eaton JG; Frost TM; Garrison PJ; Kratz TK; Mach CE; McCormick JH; Perry JA; Rose WA; Sampson CJ; Shelly BCL; Swenson
WA; Webster KE. (1993). Experimental acidification of Little Rock Lake, Wisconsin: Chemical and biological changes over the pH
range6.1 to 4.7. Can J Fish Aquat Sci, 50,1101-1121.
Bricker OP; Rice KC. (1989). Acidic deposition to streams: a geology-based method predicts their sensitivity. Environ Sci Technol, 23,379-385.
Bricker SB; Clement CG; Pirhalla DE; Orlando SP; Farrow DGG. (1999). National estuarine eutrophication assessment: Effects of nutrient
enrichment in the nation's estuaries. Silver Spring, MD: Special Projects Office and the National Centers for Coastal Ocean Science,
National Ocean Service, National Oceanic and Atmospheric Administration (NOAA).
R-11
-------
Bricker S; Longstaff B; Dennison W; Jones A; Boicourt K; Wicks C; Woerner J. (2007). Effects of nutrient enrichment in the nation's estuaries: A
decade of change, http://ccmaserver.nos.noaa.aov/publications/eutroupdate/. (NOAA Coastal Ocean Program Decision Analysis
Series No. 26). Silver Spring, MD: National Centers for Coastal Ocean Science, National Oceanic and Atmospheric Administration
(NOAA).
Bridgham SD; Pastor J; McClaugherty CA; Richardson CJ. (1995). Nutrient-use efficiency: A litterfall index, a model, and a test along a nutrient
availability gradient in North Carolina peatlands. American Naturalist, 145,1-21.
Bridgham SD; Pastor J; Janssens J; Chapin C; MaltererT. (1996). Multiple limiting gradients in peatlands: A call for a new paradigm. Wetlands,
16,45-65.
Britton KO; Berrang P; Mavity E. (1996). Effects of pretreatment with simulated acid rain on the severity of dogwood anthracnose. Plant Dis, 80,
646-649.
Broderick SJ; Cullen P; Maher W. (1988). Denitrification in a natural wetland receiving secondary treated effluent. Water Res, 4,431-439.
Brook JR; Sirois A; Clarke JF. (1996). Comparison of dry deposition velocities for SO2, HNO3, and SO42" estimated with two inferential models.
Water Air Soil Pollut, 87,205-218.
Brook JR; Dann TF. (1999). Contribution of nitrate and carbonaceous species to PM2.5 observed in Canadian cities. J Air Waste Manage
Assoc, 49,193-199.
Brooks ML. (1999). Alien annual grasses and fire in the Mojave Desert. Madrono, 46,13-19.
Brooks ML; Esque TC. (2002). Alien annual plants and wildfire in desert tortoise habitat: Status, ecological effects, and management.
Chelonian Conserv Biol, 4,330-340.
Brooks ML. (2003). Effects of increased soil nitrogen on the dominance of alien annual plants in the Mojave Desert. J Appl Ecol, 40,344-353.
Brooks ML; D'Antonio CM; Richardson DM; Grace JB; Keeley J; DiTomaso JM; Hobbs RJ; Pellant M; Pyke D. (2004). Effects of invasive alien
plants on fire regimes. Bioscience, 54,677-688.
Brooks PD; Williams MW; Schmidt SK. (1996). Microbial activity under alpine snowpacks, Niwot Ridge, Colorado. Biogeochemistry, 32,93-113.
Broske R; Kleffmann J; Wiesen P. (2003). Heterogeneous conversion of NO2 on secondary organic aerosol surfaces: A possible source of
nitrous acid (HONO).in the atmosphere? Atmos Chem Phys, 3,469-474.
Brouwer R; Langford IH; Bateman I; Turner K. (1999). A meta-analysis of wetland contingent valuation studies. Regional Environ Change, 1,
47-57.
Brown DJA. (1982). The effect of pH and calcium on fish and fisheries. Water Air Soil Pollut, 18,343-351.
Brown DJA. (1983). Effect of calcium and aluminum concentrations on the survival of brown trout (Salmo trutta).at lowpH Bull Environ Contam
Toxicol, 30, 582-587.
Brown SS; Neuman JA; Ryerson TB; Trainer M; Dube WP; Holloway JS; Warneke C; De Gouw JA; Donnelly SG; Atlas E; Matthew B;
Middlebrook AM; Peltier R; Weber RJ; Stohl A; Meagher JF; Fehsenfeld FC; Ravishankara AR. (2006a). Nocturnal odd-oxygen budget
and its implications for ozone loss in the lower troposphere. Geophys Res Lett, 33, doi:10.1029/2006GL025900
Brown SS; Ryerson TB; Wollny AG; Brock CA; Peltier R; Sullivan AP; Weber RJ; Dube WP; Trainer M; Meagher JF; Fehsenfeld FC;
Ravishankara AR. (2006b). Variability in nocturnal nitrogen oxide processing and its role in regional air quality. Science, 311,67-70.
Briihl C; Poschl U; Crutzen PJ; Steil B. (2000). Acetone and PAN in the upper troposphere: Impact on ozone production from aircraft
emissions. Atmos Environ, 34,3931-3938.
Brumme R; Beese F. (1992). Effects of liming and nitrogen fertilization on emission of CO2 and N2O from a temporate forest. J Geophys Res,
97,12851-12858.
Brunet J, Diekmann, M, Falkengren-Grerup, U. (1998). Effects of nitrogen deposition on field layer vegetation in south Swedish oak forests.
Environ Pollut, 102,35-40.
Brunner I; Godbold DL. (2007). Tree roots in a changing world. J For Res, 12, 78-82.
Bubier JL; Moore TR; Bledzki LA. (2007). Effects of nutrient addition on vegetation and carbon cycling in an ombrotrophic bog. Glob Change
Biol, 13,1168-1186.
Buck S; Dodds W; Fisher J; Hart D; Parker A; Stevenson J; Watson V; Welch E. (2000). Nutrient criteria technical guidance manual: Rivers and
streams (Report no EPA-822-B-00-002). Washington, DC; Office of Water; Office of Science and Technology; U.S. Environmental
Protection Agency.
Buckler DR; Mehrle PM; Cleveland L; Dwyer FJ. (1987). Influence of pH on the toxicity of aluminum and other inorganic contaminants to east
coast striped bass. Water Air Soil Pollut, 35,97-106.
Bugas PE Jr; Mohn LO; Kauffman JW. (1999). Impacts of acid deposition on fish populations in the St. Marys River, Augusta County, Virginia.
Banisteria 13,191-200.
Buhr MP; Trainer M; Parrish DD; Sievers RE; Fehsenfeld FC. (1992). Assessment of pollutant emission inventories by principal component
analysis of ambient air measurements. Geophys Res Lett, 19,1009-1012.
Buhyoff GJ; Leuschner WA; Wellman JD. (1979). Aesthetic impacts of southern pine beetle damage. J Environ Manage, 8,261-267.
Buhyoff GJ; Wellman JD. (1980). The specification of a non-linear psychophysical function for visual landscape dimensions. Journal of Leisure
Research, 12,257-272.
Buhyoff GJ; Wellman JD; Daniel, TC. (1982). Predicting scenic quality for mountain pine beetle and western spruce budworm damaged forests.
Forest Science, 28,827-838.
Buijsman E; Aben JMM; Van Elzakker BG; Mennen MG. (1998). An automatic atmospheric ammonia network in the Netherlands set-up and
results. Atmos Environ, 32,317-324.
Bukaveckas P; ShawW. (1998). Phytoplankton responses to nutrient and grazer manipulations among northeastern lakes of varying pH. Can J
Fish Aquat Sci, 55,958-966.
R-12
-------
Bulger AJ. (1986). Coincident peaks in serum osmolality and heat-tolerance rhythms in seawater-acclimated killifish (Fundulus heteroclitus).
Physiol Zool, 59,169-174.
Bulger AJ; Lien L; Cosby BJ; Henriksen A. (1993). Trout status and chemistry from the Norwegian thousand lake survey: Statistical analysis.
Can J Fish Aquat Sci, 50, 575-585.
Bulger AJ; Dolloff CA; Cosby BJ; Eshleman KN; Webb JR; Galloway JN. (1995). The "Shenandoah National Park: fish in sensitive habitats"
(SNP: FISH) Project. An integrated assessment of fish community responses to stream acidification. Water Air Soil Pollut, 85,309-
314.
Bulger AJ; Cosby, BJ; Dolloff, CA; Eshleman, KN; Webb, JR; Galloway, JN. (1999). SNP:FISH, Shenandoah National Park: Fish in sensitive
habitats. Project final report -Volume I: Project description and summary of results; Volume II: Stream water chemistry and discharge,
and synoptic water quality surveys. Volume III: Basin-wide habitat and population inventories, and behavioral responses to acid in a
laboratory stream. Volume IV: Stream bioassays, aluminum toxicity, species richness and stream chemistry, and models of
susceptibility to acidification. (Project Completion Report to the National Park Service). Charlottesville, VA, Department of
Environmental Sciences; University of Virginia.
Bulger AJ; Cosby BJ; Webb JR. (2000). Current, reconstructed past, and projected future status of brook trout (Salvelinus fbntinalis) streams in
Virginia. Can J Fish Aquat Sci, 57,1515-1523.
Bull KR; Roberts RD; Inskip MJ; Goodman GT. (1977) Mercury concentrations in soil, grass, earthworms and small mammals near an industrial
emission source. Environ Pollut, 12,135-140.
Burges SJ; Wigmosta MS; Meena JM. (1998). Hydrological effects of landuse change in a zero-order catchment. J Hydro Engr, 3,86-97.
Burkholder JM; Dickey DA; Kinder CA; Reed RE; Mallin MA; Mclver MR; Cahoon LB; Melia G; Brownie C; Smith J; Deamer N; Springer J;
Glasgow HB; Toms D. (2006). Comprehensive trend analysis of nutrients and related variables in a large eutrophic estuary: a decadal
study of anthropogenic and climatic influences. Limnol Oceanogr, 51,463-487.
Burns DA; Kendall C. (2002). Analysis of 615N and 6180 to differentiate NO3" sources in runoff at two watersheds in the Catskill Mountains of
New York. Water Resour Res, 38,1051. doi: 10.1029/2001WR000292.
Burns DA. (2004). The effects of atmospheric nitrogen deposition in the Rocky Mountains of Colorado and southern Wyoming, USA—A critical
review. Environ Pollut, 127,257-269.
Burrows JP; Weber M; Buchwitz M; Rozanov V; Ladstatter-Weilienmayer A; Richter A; DeBeek R; Hoogen R; Bramstedt K; Eichmann K-U;
Eisinger M. (1999).The global ozone monitoring experiment (GOME): Mission concept and first scientific results. J Atmos Sci, 56,151-
175.
Burton AJ; Pregitzer KS; Hendrick RL. (2000). Relationships between fine root dynamics and nitrogen availability in Michigan northern
hardwood forests. Oecologia, 125,389-399.
Burton TM; Stanford RM; Allan JW. (1985). Acidification effects on stream biota and organic matter processing. Can J Fish, Aquat Sci, 42,669-
675.
Burtraw D; Krupnick A; Austin D; Farrell D; Mansur, E. (1997). The costs and benefits of reducing acid rain. (Discussion Paper 97-31-REV).
Washington, DC: Resources for the Future (RFF).
Bushong SJ; Bachmann RW. (1989). In situ nutrient enrichment experiments with periphyton in agricultural streams. Hydrobiologia, 178,1-10.
Butnor JR; Johnsen KH; Oren R; Katul, GG. (2003). Reduction of forest floor respiration by fertilization on both carbon dioxide-enriched and
reference 17 year-old loblolly pine stands. Glob Change Biol, 9,849-861.
Butterbach-Bahl, K; Gasche, R; Huber, C. H; Kreutzer, K; Papen, H. (1998). Impact of N-input by wet deposition on N-trace gas fluxes and
CH4-oxidation in spruce forest ecosystems of the temperate zone in Europe. Atmos. Environ. 32,559-564.
Butterbach-Bahl, K; Breuer, L; Gasche, R; Willibald, G; Papen, H. (2002a). Exchange of trace gases between soils and the atmosphere in
Scots pine forest ecosystems of the northeastern German lowlands. 1. Fluxes of N2O, NO/NO2 and CH4 at forest sites with different N-
deposition. For. Ecol. Manage. 167:123-134.
Butterbach-Bahl, K; Gasche, R; Willibald, G; Papen, H. (2002b).Exchange of N-gases at the Hoglwald forest: A summary. Plant Soil 240:117-
123.
Bytnerowicz A; Miller PR; Olszyk DM. (1987a). Dry deposition of nitrate, ammonium and sulfate to a Ceanothus crassifolius canopy and
surrogate surfaces. Atmos Environ, 21,1749-1757.
Bytnerowicz A; Miller PR; Olszyk DM; Dawson PJ; Fox CA. (1987b). Gaseous and particulate air pollution in the San Gabriel Mountains of
southern California. Atmospheric Environment 21:1805-1814.
Bytnerowicz A; Fenn ME. (1996). Nitrogen deposition in California forests: A review. Environ Pollut, 92,127-146.
Bytnerowicz A; Percy K; Riechers G; Padgett P; Krywult M. (1998a). Nitric acid vapor effects on forest trees: Deposition and cuticular changes.
Chemosphere, 36,697-702.
Bytnerowicz A; Dueck T; Godzik S. (1998b). Nitrogen oxides, nitric acid vapor, and ammonia. In: Flagler RB (Ed.), Recognition of air pollution
injury to vegetaion: A pictoral atlas (2nd Edition, pp. 5-1-5-6). Pittsburgh, PA: Air & Waste Managment Association.
Bytnerowicz, A; Arbaugh, M. J; Schilling, S. L. (1998c). Proceedings of the international symposium on air pollution and climate change: Effects
on forest ecosystems. Held in Riverside, CA in February 1996. http://www.fs.fed.us/psw/publications/documents/atr-166/ (General
technical report PSW-GTR-166). Albany, CA: Pacific Southwest Research Station; U.S. Forest Service; U.S. Department of
Agriculture (USDA).
Bytnerowicz A; Padgett P; Percy K; Krywult M; Riechers G; Horn J. (1999). Direct effects of nitric acid on forest vegetation. In: Miller PR,
McBride J (Eds.), Oxidant air pollution impacts in the Montane forests of Southern California: The San Bernardino Mountains case
study, Ecological Series 134. (pp. 270-287). New York: Springer.
Bytnerowicz A. (2002). Physiological/ecological interactions between ozone and nitrogen deposition in forest ecosystems. Phyton, 42,13-28.
R-13
-------
Bytnerowicz A; Tausz M; Alonso R; Jones D; Johnson R; Grulke N. (2002). Summer-time distribution of air pollutants in Sequoia National Park,
California. Environ Pollut, 118,187-203.
Byun DW; Ching JKS (Eds.). (1999). Science algorithms of the EPA models-3 community multiscale air quality model (CMAQ) modeling
system. (Report EPA/60O/R-99/030). Research Triangle Park, NC; Atmospheric Modeling Division; National Exposure Research
Laboratory; Office of Research and Development; U.S. Environmental Protection Agency.
Byun D; Schere KL. (2006). Review of the governing equations, computational algorithms, and other components of the models-3 community
multiscale air quality (CMAQ) modeling system. Appl Mech Rev, 59,51-77.
Caddy, JF. (1993). Toward a comparative evaluation of human impacts on fishery ecosystems of enclosed and semi-enclosed seas. Rev. Fish.
Sci. 1:57-95.
Caffrey JM; Murrell M; Wigand C; McKinney R. (2007). Effect of nutrient loading on biogeochemical and microbial processes in a New England
salt marsh. Biogeochemistry, 82,251-264.
Caine, N. (1989).Hydrograph separation in a small alpine basin based on inorganic solute concentrations. J. Hydrol, 1, 89-101.
Cairns J Jr; Pratt JR. (1993). A history of biological monitoring using benthic macroinvertebrates. In: Rosenberg DM; Resh VH (Eds.),
Freshwater biomonitoring and benthic macroinvertebrates (pp. 10-28). New York: Chapman and Hall.
Calanni J; Berg E; Wood M; Mangis D; Boyce; Weathers W; Sievering H. (1999). Atmospheric nitrogen deposition at a conifer forest: Response
of free amino acids in Engelmann spruce needles. Environ Pollut 105:79-89.
Caldeira, K; Wickett, ME. (2003). Anthropogenic carbon and ocean pH. Nature, 425,365.
Caldwell TG; Johnson DW; Miller WW; Quails RG. (2002). Forest floor carbon and nitrogen losses due to prescription fire. Soil Sci Soc Am J,
66, 262-267.
Callaway JM; Darwin, RF; Nesse, RJ. (1986). Economic valuation of acidic deposition damages: preliminary results from the 1985 NAPAP
assessment. Water Air Soil Pollut. 31:1019-1034.
Calvert JG; Su F; Bottenheim JW; Strausz OP. (1978). Mechanism of the homogeneous oxidation of sulfur dioxide in the troposphere. Atmos
Environ, 12,197-226.
Calvert JG; Yarwood G; Dunker AM. (1994). An evaluation of the mechanism of nitrous acid formation in the urban atmosphere. Res Chem
Intermed, 20,463-502.
Camargo JA; Alonso, A; Salamanca, A. (2005). Nitrate toxicity to aquatic animals: A review with new data for freshwater invertebrates.
Chemosphere 58:1255-1267.
Camargo JA; Ward, JV. (1995). Nitrate (N03~N).toxicity to aquatic life: A proposal of safe concentrations for two species of nearctic freshwater
invertebrates. Chemosphere 31: 3211-3216.
Cameron TA; Englin J. (1997). Welfare effects of changes in environmental quality under individual uncertainty about use. Rand J Econ, 28,
s45-s70.
Camp DC; Stevens RK; Cobourn WG; Husar RB; Collins JF; Huntzicker JJ; Husar JD; Jaklevic JM; McKenzie RL; Tanner RL; Tesch JW.
(1982). Intercomparison of concentration results from fine particle sulfur monitors. Atmos Environ, 16,911-916.
Campbell DH; Clow DW; Ingersoll GP; Mast MA; Spahr NE; Turk JT. (1995). Processes controlling the chemistry of two snowmelt-dominated
streams in the Rocky Mountains. Water Resour Res, 31,2811-2821.
Campbell DH; Baron JS; Tonnessen KA; Brooks PD; Schuster PF. (2000a). Controls on nitrogen flux in alpine/subalpine watersheds. Water
Resour Res, 36,37-48.
Campbell, JL; Hornbeck, JW; McDowell, WH; Buso, DC; Shanley, JB; Likens, GE. (2000b). Dissolved organic nitrogen budgets for upland,
forested ecosystems in New England. Biogeochemistry 49:123-142.
Campbell DH; Kendall C; Chang CCY; Silva SR; Tonnessen KA. (2002). Pathways for nitrate release from an alpine watershed: determination
using 615N and6180. Water Resour Res, 31, 2811-2821.
Campbell DH; Muths E; Turk JT; Corn PS. (2004). Sensitivity to acidification of subalpine ponds and lakes in north-western Colorado.
Hydrological Processes, 18,2817-2834.
Campbell JL; Mitchell MJ; Groffman PM; Christenson LM. (2005). Winter in northeastern North America: An often overlooked but critical period
for ecological processes. Front Ecol Environ, 3,314-322.
Canary JD; Harrison RB; Compton JE; Chappell HN. (2000). Additional carbon sequestration following repeated urea fertilization of second-
growth Douglas-fir stands in western Washington. For EcolManage, 138,225-232.
Cape JN; Leith ID; Fowler D; Murray MB; Sheppard LJ; Eamus D; Wilson RHF. (1991). Sulfate and ammonium in mist impair the frost
hardening of red spruce seedlings. NewPhytol, 118,119-126.
Cape JN; Sheppard LJ; Fowler D; Harrison AF; Parkinson JA; Dao P; Paterson IS. (1992). Contribution of canopy leaching to sulphate
deposition in a Scots pine forest. Environ Pollut, 75,229-236.
Cape JN. (1993). Direct damage to vegetation caused by acid rain and polluted cloud: Definition of critical levels for forest trees. Environ Pollut,
82,167-180.
Cape JN. (1997). Photochemical oxidants: What else is in the atmosphere besides ozone? Phyton-Annales Rei Botanicae, 37,45-57.
Cape JN. (2003). Effects of airborne volatile organic compounds on plants. Environ Pollut, 122,145-157.
Cappellato R; Peters NE; Meyers TP. (1998). Above-ground sulfur cycling in adjacent coniferous and deciduous forests and watershed sulfur
retention in the Georgia piedmont, USA. Water Air Soil Pollut, 103,151-171.
Caraco NF; Cole JJ; Likens GE. (1989). Evidence for sulphate-controlled phosphorus release from sediments of aquatic systems. Nature, 341,
316-318.
Cardelino CA; Chameides WL. (2000). The application of data from photochemical assessment monitoring stations to the observation-based
model. Atmos Environ, 34,2325-2332.
R-14
-------
Carlson RW. (1983). Interaction between SO2 and NO2 and their effects on photosynthetic properties of soybean Glycine max. Environ Pollut,
32,11-38.
Carpenter LJ; Sturges WT; Penkett SA; Liss PS; Alicke B; Hebestreit K; Piatt U. (1999). Short-lived alkyl iodides and bromides at Mace Head,
Ireland: Links to biogenic sources and halogen oxide production. J Geophys Res, 104,1679-1689.
Carpenter LJ. (2003). Iodine in the marine boundary layer. Chem Rev, 103,4953-4962.
Carpenter SR; Caraco NF; Correll DL; Howarth RW; Sharpley AN; Smith VH. (1998). Nonpoint pollution of surface waters with phosphorus and
nitrogen. Ecol Appl, 8,559-568.
Carreiro MM; Sinsabaugh RL; Repert DA; Parkhurst DF. (2000). Microbial enzyme shifts explain litter decay responses to simulated nitrogen
deposition. Ecology, 81,2359-2365.
Carroll MA; Bertman, SB; Shepson, PB. (2001). Overview of the program for research on oxidants, photochemistry, emissions and transport
(PROPHET) summer 1998 measurements intensive. J Geophys Res, 106,24275-24288.
Carroll JJ; Miller, PR; Pronos, J. (2003). Historical perspectives on ambient ozone and its effects on the Sierra Nevada. In: Bytnerowicz, A;
Arbaugh, MJ; Alonso, R (Eds.). Ozone air pollution in the Sierra Nevada: Distribution and effects on forests; (Series: Developments in
Environmental Science: Volume 2; pp. 33-54). New York, NY: Elsevier.
Carslaw DC. (2005). Evidence of an increasing NO2/NOX emissions ratio from road traffic emissions. Atmos Environ, 39,4793-4802.
Carslaw DC; Beevers SD. (2005). Estimations of road vehicle primary NO2 exhaust emission fractions using monitoring data in London. Atmos
Environ, 39,167-177.
Carslaw DC; Carslaw N. (2007). Detecting and characterising small changes in urban nitrogen dioxide concentrations. Atmos Environ, 41,
4723-4733.
Carson RT; Mitchell, RC. (1993).The value of clean water: The public's willingness to pay for boaTable, fishable, and swimmable quality water.
Water Resour. Res., 29,2445-2454.
Carson R; Flores NE; Martin KM; Wright. JL. (1996). Contingent valuation and revealed preference methodologies: Comparing the estimates
for quasi-public goods. Land Econ, 72,80-99.
Carson RT; Hanemann MW. (2005). Contingent Valuation. In: Maler K-G; Vincent JR (Eds.), Handbook of environmental economics. Volume 2.
Valuation of environmental changes (pp. 821-936). Amsterdam, The Netherlands: Elsevier.
Carter WPL. (1990). A detailed mechanism for the gas-phase atmospheric reactions of organic compounds. Atmos Environ, 24,481-518.
Carter WPL. (1995). Computer modeling of environmental chamber studies of maximum incremental reactivities of volatile organic compounds.
Atmos Environ, 29,2513-2527.
CASAC. (2005). Recommendations to the Clear Air Act Advisory Committee (CASAC), Air Quality Management Work Group, phase I and next
steps, January 2005. Washington, DC: Clean Air Act Advisory Committee.
Caspersen JP; Pacala SW; Jenkins JC; Hurtt GC; Moorcroft PR; Birdsey RA. (2000). Contributions of land-use history to carbon accumulation
in U.S. forests. Science, 290,1148-1151.
Castaldi S; Smith KA. (1998). The effect of different N substrates on biological N2O production from forest and agricultural light textured soils.
Plant Soil, 199,229-238.
Castro MS; Steudler PA; Melillo JM; Aber JD; Millham S. (1992). Exchange of N2O and CH4 between the atmosphere and soils in spruce fir
forests in the northeastern Unitied States. Biogeochem, 18,119-135.
Castro MS; Morgan IRP. (2000). Input-output budgets of major ions for a forested watershed in western Maryland. Water Air Soil Pollut, 119,
121-137.
Castro MS; Driscoll CT; Jordan TE; Reay WG; Boynton WR; Seitzinger SP; Styles RV; Cable JE. (2001). Contribution of atmospheric
depostition to the total nitrogen loads to thirty-four estuaries on the Atlantic and Gulf Coasts of the United States. In: Valigura RA;
Alexander RB; Castro MS; Meyers TP; Paerl HW; Stacey PE; Turner RE (Eds.), Nitrogen loading in coastal water bodies: An
atmospheric perspective (Monograph Voume 57, pp. 77-106). Washington, DC: American Geophysical Union.
Castro MS; Driscoll CT. (2002). Atmospheric nitrogen deposition to estuaries in the Mid-Atlantic and Northeastern United States. Environ Sci
Technol, 36, 3242-3249.
Castro MS; Driscoll CT; Jordan TE; Reay WG; Boynton WR. (2003). Sources of nitrogen to estuaries in the United States. Estuaries, 26,803-
814.
CDC (2004). The health consequences of smoking: A report of the Surgeon General, http://www.cdc.aov/tobacco/data statistics/sar/sar 2004/.
Atlanta, GA; Office on Smoking and Health; National Center for Chronic Disease Prevention and Health Promotion; Centers for
Disease Control and Prevention (CDC); U.S. Department of Health and Human Services.
Ceccotti SP; Messick DL. (1997). A global review of crop requirements, supply, and environmental impact on nutrient sulphur balance. In: Cram
WJ; De Kok LJ; Stulen I; Brunold C; Rennenberg H (Eds.), Sulphur metabolism in higher plants (pp. 155-163). Leiden, The
Netherlands: Backhuys Publishers.
Committee on Environment and Natural Resources (CENR).of the National Science and Technology Council. (1997+)
Centre for Ecology and Hydrology. (2003).Status of UK critical loads: critical loads methods, data & maps.
http://www.airquality.co.uk/archive/reports/cat11/maintext_7may.pdf. Monks Wood, UK; UK National Focal Centre.
Certini G. (2005). Effects of fire on properties of forest soils: A review. Oecologia, 143,1-10.
CFR. (2006). Lower detectable limit—(1).technical definition. (CFR 40: Section 53.23c.). Washington DC: U.S. Code of Federal Regulations
(CFR).
Chadwick MA; Huryn, AD. (2003). Effect of a whole-catchment N addition on stream detritus processing. J. North Am. Benthol. Soc. 22:194-
206.
Chadwick MA; Huryn, AD. (2005). Response of stream macroinvertebrate production to atmospheric N deposition and channel drying. Limnol
Oceanogr, 50,228-236.
R-15
-------
Chameides WL. (1984). The photochemistry of a remote marine stratiform cloud. J Geophys Res, 89,4739-4755.
Chameides WL; Stelson AW. (1992). Aqueous-phase chemical processes in deliquescent sea-salt aerosols: A mechanism that couples the
atmospheric cycles of S and sea salt. J Geophys Res, 97, 20565-20580.
Chan ASK; Steudler PA; Bowden RD; Gulledge J; Cavanaugh CM. (2005). Consequences of nitrogen fertilization on soil methane consumption
in a productive temperate deciduous forest. Biol Fertil Soils, 41,182-189.
Chanasyk DS; Whitson IR; Mapfumo E; Burke JM; Prepas EE. (2003). The impacts of forest harvest and wildfire on soils and hydrology in
temperate forests: A baseline to develop hypotheses for the Boreal Plain. J Environ Eng Sci, 2, S51-S62.
Chang JS; Brost RA; Isaksen ISA; Madronich S; Middleton P. (1987). A three-dimensional Eulerian acid deposition model - physical concepts
and formulation. J Geophys Res, 92,14681-14700.
Chang MC; Sioutas C; Kim S; Gong H Jr; Linn WS. (2000). Reduction of nitrate losses from filter and impactor samplers by means of
concentration enrichment. Atmos Environ, 34,85-98.
Chang ME; Hartley DE; Cardelino C; Chang W-L. (1996). Inverse modeling of biogenic emissions. Geophys Res Lett, 23,3007-3010.
Chang ME; Hartley DE; Cardelino C; Hass-Laursen D; Chang WL. (1997). On using inverse methods for resolving emissions with large spatial
inhomogeneities. J Geophys Res, 102,16023-16036.
Chapin FS III. (1980). The mineral nutrition of wild plants. Annu Rev Ecol Syst, 11,233-260.
Chapin FS III; Bloom, AJ; Field, CB; Waring, RH. (1987). Plant responses to multiple environmental factors. Bioscience 37:49-57.
Chapin FS III. (1991). Integrated responses of plants to stress: a centralized system of physiological responses. Bioscience 41: 29-36.
Chapin FS III; Moilanen, L; Kielland, K. (1993). Preferential use of organic nitrogen for growth by a non-mycorrhizal arctic sedge. Nature 361:
150-153.
Chapin FS III; Sala OE; Burke IC; Grime JP; Hooper DU; Lauenroth WK; Lombard A; Mooney HA; Mosier AR; Naeem S; Pacala SW; Roy J;
Steffen WL; Tilman D. (1998). Ecosystem consequences of changing biodiversity. Bioscience, 48,45-52.
Chapin FS.III; Ruess, RW. (2001). The roots of the matter. Nature, 411,749,751-752.
Chappelka, AH; Samuelson, LJ. (1998). Ambient ozone effects on forest trees of the eastern United States: A review. NewPhytol. 139:91-108.
Chapuis-Lardy L; Wrage N; Metay A; Chotte JL; Bernoux M. (2007). Soils, a sink for N2O? A review. Glob Chang Biol, 13,1-17.
Charbonneau R; Kondolf GM. (1993). Land use change in California: nonpoint source water quality impacts. Environ Manage, 17,453-460.
Charles, DF; Norton, SA. (1986). Paleolimnological evidence for trends in atmospheric deposition of acids and metals. In: NRC (Ed.),.Acid
deposition: Long-term trends. (NRC Committee on monitoring and assessment of trends in acid deposition; pp 335-435). Washington,
DC: National Academy Press.
Charles DF; Batarbee RW; Renberg I; van Dam H; Smol JP. (1990a). Paleoecological analysis of lake acidification trends in North America and
Europe using diatoms and chrysophytes. In: Acidic Precipitation; Series Volume 4: Soils, aquatic processes, and lake acidification.
Advances in environmental science (pp. 207-276). New York: Springer-Verlag.
Charles DF; Binford MW; Furlong ET; Hites RA; Mitchell MJ; Norton SA; Oldfield F; Paterson MJ; Smol JP; Uutala AJ. (1990b). Paleoecological
investigation of recent lake acidification in the Adirondack Mountains, NY. J Paleolimnol, 3,195-241.
Charles DF. (1991). Christie, S. (Ed.). Acidic deposition and aquatic ecosystems: Regional case studies. New York: Springer-Verlag.
Chatfield RB; Crutzen PJ. (1984). Sulfur dioxide in remote oceanic air: cloud transport of reactive precursors. J Geophys Res, 89,7111-7132.
Chen CW; Gherini SA; Dean JD; Hudson RJM; Goldstein RA. (1984a). Development and calibration of the integrated lake-watershed
acidification model. In: Schnoor JL (Ed.), Modeling of total acid precipitation impacts (pp. 175-203). Stoneham, MA: Butterworth
Publishers.
Chen CW; Gherini, SA. Peters NE; Murdoch, PS; Newton, RM; Goldstein, RA. (1984b). Hydrologic analyses of acidic and alkaline lakes. Water
Resour Res, 20,1875-1882.
Chen CW; Gomez LE. (1989). Comment on surface water chemistry. Environ Sci Technol, 23, 752-754
Chen CY; Stemberger RS; Kamman NC; Mayes BM; Folt CL. (2005). Patterns of Hg bioaccumulation and transfer in aquatic food webs across
multi-lake studies in the Northeast U.S. Ecotoxicology, 14,135-147.
Chen L; Driscoll, CT. (2004a). Modeling the response of soil and surface waters in the Adirondack and Catskill regions of New York to changes
in atmospheric deposition and historical land disturbance. Atmos Environ, 38,4099-4109.
Chen, L; Driscoll, CT. (2004b). A regional assessment of the response of the acid-base status of lake-watersheds in the Adirondack region of
New York to changes in atmospheric deposition using PnET-BGC. (pp. 1 -27). Syracuse, NY: Department of Civil and Environmental
Engineering.
Chen L; Driscoll, CT. (2004c). An evaluation of processes regulating spatial and temporal patterns in lake sulfate in the Adirondack region of
New York. Glob Biogeochem Cycles, 18,10.
Chen L; Driscoll CT. (2005a). Regional application of an integrated biogeochemical model to northern New England and Maine. Ecol Appl, 15,
1783-1797.
Chen L; Driscoll CT. (2005b). Regional assessment of the response of acid-base status of lake-watersheds in the Adirondack region of New
York to changes in atmospheric deposition using PnET-BGC. Environ Sci Technol, 39,787-794.
Chen, L; Driscoll, CT. (2005c). A two-layer model to simulate variations in surface water chemistry draining a northern forest watershed. Water
Resour Res, 41, doi: 10.1029/2004WR003625.
Chock DP; Winkler SL. (1994). A comparison of advection algorithms coupled with chemistry. Atmos Environ, 28,2659-2675.
Choi S-C; Chase T Jr; Bartha R. (1994). Metabolic pathways leading to mercury methylation in Desulfbvibrio desulfiiricans LS. Appl Environ
Microbiol, 60,4072-4077.
Choi W; Leu M-T. (1998). Nitric acid uptake and decomposition on black carbon (soot) surfaces: Its implications for the upper troposphere and
lower stratosphere. J Phys Chem, 102,7618-7630.
R-16
-------
Chorover J; Visousek, PM; Everson, DA; Esperanza, AM; Turner, D. (1994). Solution chemistry profiles of mixed-conifer forests before and
after fire. Biogeochem, 26,115-144.
Chow JC; Zielinska B; Watson JG; Fujita EM; Richards HW; Neff WD; Dietrich D; Hering SV. (1998). Northern Front Range air quality study.
Volume A: Ambient measurements, Final Report. A companion document to the Northern Front Range air quality study final report
(Watson et al., 1998a). http://www.nfraas.colostate.edu/dri.html Fort Collins, CO; Prepared for Cooperative Institute for Research in
the Atmosphere; Colorado State University Foothills Campus.
Chow JC; Watson JG. (1999). Ion chromatography in elemental analysis of airborne particles. In: Landsberger S, Creatchman M. (Eds.).
Elemental analysis of airborne particles, (Series: Vo-Dinh T (Ed.) .Advances in environmental industrial and process control
technologies, Volume 1, pp 97-137). Newark, NJ: Gordon and Breach Science Publishers.
Christensen SW; Beauchamp JJ; Shaakir-Ali JA; Coe J; Baker JP; Smith EP; Gallagher J. (1990). Patterns of fish distribution in relation to
lake/watershed characteristics: Regression analyses and diagnostics. In: Interpretative analysis of the Adirondack Lakes Survey. Final
report to the Adirondack Lakes Survey Corporation (pp. A-1-A-62).
Christensen TR; Michelsen A; Jonasson S; Schmidt IK. (1997). Carbon dioxide and methane exchange of a subarctic heath in response to
climate change related environmental manipulations. Oikos, 79,34-44.
Christie CE; Smol JP. (1993). Diatom assemblages as indicators of lake rophic status in southeastern Ontario lakes. J Phycol, 29,575-586.
Church MR; Shaffer PW; Thornton KW; Cassell DL; Liff CI; Johnson MG; Lammers DA; Lee JJ; Holdren GR; Kern JS; Liegel LH; Pierson SM;
Stevens DL; Rochelle BP; Turner RS. (1992). Direct/delayed response project: Future effects of long-term sulfur deposition on stream
chemistry in the mid-Appalachian region of the eastern United States (No. EPA/600/R-92/186). Corvallis, OR: U.S. Environmental
Protection Agency.
Cione NK; Padgett PE; Allen EB. (2002). Restoration of a native shrubland impacted by exotic grasses, frequent fire and nitrogen deposition in
southern California. Restor Ecol, 10,376-384.
Civerolo KL; Dickerson RR. (1998). Nitric oxide soil emissions from tilled and unfilled cornfields. Ag For Meteor, 90,307-311.
Clair TA; Dennis IF; Amiro PG; Cosby BJ. (2004). Past and future chemistry changes in acidified Nova Scotian Atlantic salmon (Salmo salai)
rivers: A dynamic modeling approach. Can J Fish Aquat Sci, 61,1965-1975.
Clark CM; Cleland EE; Collins SL; Fargione JE; Gough L; Gross KL; Pennings SC; Suding KN; Grace JB. (2007). Environmental and plant
community determinants of species loss following nitrogen enrichment. Ecol Lett, 10,596-607.
Clark CM; Tilman D. (2008). Loss of plant species after chronic low-level nitrogen deposition to prairie grasslands. Nature, 451,712-715.
Clarke JF; Edgerton ES; Martin BE. (1997). Dry deposition calculations for the Clean Air Status and Trends Network. Atmospheric Environment
31:3667-3678.
Cleary PA; Wooldridge PJ; Cohen RC. (2002). Laser-induced fluorescence detection of atmospheric NO2 with a commercial diode laser and a
supersonic expansion. Appl Optics, 41,6950-6956.
Clegg NA; Toumi R. (1998). Non-sea-salt-sulphate formation in sea-salt aerosol. J Geophys Res, 103,31095-31102.
Cleland EE; Chiariello NR; Loarie SR; Mooney HA; Field CB. (2006). Diverse responses of phenology to global changes in a grassland
ecosystem. Proc Natl Acad Sci USA, 103,13740-13744.
Clemitshaw K. (2004). A review of instrumentation and measurement techniques for ground-based and airborne field studies of gas-phase
tropospheric chemistry. Crit Rev Environ Sci Technol, 34,1-108.
Cloern, JE. (2001). Our evolving conceptual model of the coastal eutrophication problem. Mar Ecol Prog Ser, 210,223-253.
Cloern JE; Jassby AD; Schraga TS; Dallas KL. (2006). What is causing the phytoplankton increase in San Francisco Bay? In: The pulse of the
estuary: Monitoring and managing water quality in the San Francisco estuary, (pp 62-70).
http://www.sfei.ora/rmp/pulse/2006/ThePulse2006 LowRes-Screen.pdf Oakland, CA: San Francisco Estuary Institute (SFEI) and The
Regional Monitoring Program for Water Quality in the San Francisco Estuary.
Clough WS. (1975). The deposition of particles on moss and grass surfaces. Atmospheric Environment: 1113-1119.
Clow DW; Mast MA. (1999). Long-term trends in stream water and precipitation chemistry at five headwater basins in the northeastern United
States. Water Resour Res, 35,541-554.
Clow DW; Sueker JK. (2000). Relations between basin characteristics and stream water chemistry in alpine/subalpine basins in Rocky
Mountain National Park, Colorado. Water Resour Res, 36,49-62.
Clow DW; Sickman JO; Striegl RG; Krabbenhoft DP; Elliot JG; Dornblaser M; Roth, DA; Campbell DH. (2003). Changes in the chemistry of
lakes and precipitation in high-elevation national parks in the western United States, 1985-1999. Water Resour Res, 39,1171, doi:
10.1029/2002WR001533.
Cobourn WG; Husar RB; Husar JD. (1978). Continuous in situ monitoring of ambient particulate sulfur using flame photometry and thermal
analysis. (In: Husar RB; Lodge, JP, Jr; Moore, DJ., (Eds.), Sulfur in the atmosphere: Proceedings of the international symposium; held
in Dubrovnik, Yugoslavia; September 1977). Atmos Environ, 12,89-98.
Cobourn, WG; Husar, RB. (1982). Diurnal and seasonal patterns of particulate sulfur and sulfuric acid in St. Louis, July 1977-June 1978. Atmos
Environ, 16,1441-1450.
Code of Federal Regulations. (2006). Lower detectable limit—(1) technical definition. CFR 40:§53.23c.
Coffman DJ; Hegg DA. (1995). A preliminary study of the effect of ammonia on particle nudeation in the marine boundary layer. J Geophys
Res, 100,7147-7160.
Cohen RC. (1999). Laser-induced fluorescence detection of atmospheric NO2 at parts per trillion mixing ratios: Implications for nitrogen oxide
photochemistry in the stratosphere and troposphere. Abstr Pap Am Chem Soc, 218, 262.
Cohn RD; Dennis RL. (1994). The evaluation of acid deposition models using principal component spaces. Atmos Environ, 28,2531-2543.
R-17
-------
Cole D; Rapp M. (1981). Elemental cycling in forest ecosystems. In: Reichle D (Ed.), Dynamic properties of forest ecosystems (Volume 23 from
Series: International biological programme pp 341-409). Cambridge, UK: Cambridge University Press.
Collier TK. (2003). Forensic ecotoxicology: Establishing causality between contaminants and biological effects in field studies. Human and
Ecological Risk Assessment: An International Journal, 9,259-266.
Colombini MP; Modugno F; Fuoco R; Tognazzi A. (2002). A GC-MS study on the deterioration of lipidic paint binders. Microchem J, 73,175-
185.
Compeau GC; Bartha R. (1985). Sulfate-reducing bacteria: Principal methylators of mercury in aNOxic estuarine sediment. Appl Environ
Microbiol, 50,498-502.
Compton JE; Watrud LS; Porteous A; DeGrood S. (2004). Response of soil microbial biomass and community composition to chronic nitrogen
additions at Harvard forest. For Ecol Manage, 196,143-158.
Confer JL; KaaretT; Likens GE. (1983). Zooplankton diversity and biomass in recently acidified lakes. Can J Fish Aquat Sci, 40,36-42.
Conley DJ; SchelskeCL; Stoermer EF. (1993). Modification of the biogeochemical cycle of silica with eutrophication. Mar Ecol Prog Ser, 101,
179-192.
Conley DJ; Markager S; Andersen J; Ellermann T; Svendsen LM. (2002). Coastal eutrophication and the Danish National Aquatic Monitoring
and Assessment Program. Estuaries, 25,848-861.
Conrad R. (1996). Soil microorganisms as controllers of atmospheric trace gases (H2, CO, CH4, OCS, N2O, and NO). Microbiol Rev, 60,609-
640.
Cook, ER, Zedaker, SM. (1992).The dendroecology of red spruce decline. In: Eagar C; Adams MB (Eds.). Ecology and decline of red spruce in
the eastern United States, (pp 192-231). New York, NY: Springer-Verlag.
CookE. R; D'Arrigo, R. D; Mann, M. E. (2002).A well-verified, multiproxy reconstruction of the winter North Atlantic Oscillation Index since A.D.
1400. J Clim, 15,1754-1764.
Cook RB; Kreis, RG Jr; Kingston JC; Camburn KE; Norton SA; Mitchell, MJ; Fry B; Shane LCK (1990). An acidic lake in northern Michigan. J
Paleolimnol, 3,13-34.
Cook RB; Jager HI. (1991). Upper Midwest: The effects of acidic deposition on lakes. In: Charles DF, Christie S. (Eds.). Acidic deposition and
aquatic ecosystems: regional case studies, (pp 421-466). New York: Springer-Verlag.
Cook RB; Elwood JW; Turner RR; Bogle MA; Mulholland PJ; Palumbo AV. (1994). Acid-base chemistry of high-elevation streams in the Great
Smoky Mountains. Water Air Soil Pollut, 72,331-356.
Cooper AB. (1990). Nitrate depletion in the riparian zone and stream channel of a small headwater catchment. Hydrobiologia, 202,13-26.
Cooper OR; Stohl A; Trainer M; Thompson AM; Witte JC; Oltmans SJ; Morris G; Pickering KE; Crawford JH; Chen G; Cohen RC; Bertram TH;
Wooldridge P; Perring A; Brune W; Merrill J; Moody JL; Tarasick D; Nedelec P; Forbes G; Newchurch M; Schmidlin F; Johnson BJ;
Turquety S; Baughcum SL; Ren X; Fehsenfeld FC; Meagher JF; Spichtinger N; Brown CC; McKeen SA; McDermid, IS; Leblanc T.
(2006a).Large upper tropospheric ozone enhancements above mid-latitude North America during summer: In situ evidence from the
IONS and MOZAIC ozone measurement network. J Geophys Res, 111, doi:10.1029/2006JD007306.
Cooper OR; Stohl A; Trainer M; Thompson A; Witte J; Oltmans S; Morris G; Pickering K; Crawford J; Chen G; Cohen R; Bertram T; Wooldridge
P; Perring A; Brune W; Merrill J; Moody J; Tarasisck D; Nedelec P. (2006b). Influence of lightning NOx on the summertime upper
tropospheric ozone enhancement above eastern North America. EOS Transactions, American Geophysical Union, 87.
Cooper SR; Brush GS. (1991). Long-term history of Chesapeake Bay aNOxia. Science, 254,992-996.
Corn PS; Stolzenburg W; Bury RB. (1989). Acid precipitation studies in Colorado and Wyoming: interim report of surveys of mountain
amphibians and water chemistry. (Biological report 80 [40.26]). U.S. Fish and Wildlife Service.
Cosby BJ; Hornberger GM; Galloway JN. (1985a). Modeling the effects of acid deposition: Assessment of a lumped parameter model of soil
water and streamwater chemistry. Water Resour Res, 21,51-63.
Cosby BJ; Wright RF; Hornberger GM; Galloway JN. (1985b). Modelling the effects of acid deposition: Estimation of long-term water quality
responses in a small forested catchment. Water Resour Res, 21,1591-1601.
Cosby BJ; Hornberger GM; Galloway JN; Wright RF. (1985c). Time scales of catchment acidification: a quantitative model for estimating
freshwater acidification. Environ Sci Technol, 19,1144-1149.
Cosby BJ; Jenkins A; Ferrier RC; Miller JD; Walker TAB. (1990). Modelling stream acidification in afforested catchments: Longterm
reconstructions at two sites in central Scotland. J Hydrol, 120,143-162.
Cosby BJ; Ryan PF; Webb JR; Hornberger GM; Galloway JN; Charles DF. (1991). Mountains of western Virginia. In: Charles DF; Christie S
(Ed.), Acidic deposition and aquatic ecosystems: Regional case studies (pp. 297-318). New York, NY: Springer-Verlag.
Cosby BJ; Wright RF; Gjessing E. (1995). An acidification model (MAGIC) with organic acids evaluated using whole-catchment manipulations
in Norway. J Hydrol, 170,101-122.
Cosby BJ; Norton SA; Kahl JS. (1996). Using a paired catchment manipulation experiment to evaluate a catchment-scale biogeochemical
model. Sci Total Environ, 183,49-66.
Cosby BJ; Ferrier RC; Jenkins A; Emmett BA; Wright RF; Tietema A. (1997). Modelling the ecosystem effects of nitrogen deposition: Model of
ecosystem retention and loss of inorganic nitrogen (MERLIN). Hydrol Earth Sys Sci, 1,137-158.
Cosby BJ; Ferrier RC; Jenkins A; Wright RF. (2001). Modeling the effects of acid deposition: Refinements, adjustments and inclusion of
nitrogen dynamics in the MAGIC model. Hydrol Earth Sys Sci, 5,499-517.
Cosby BJ; Webb JR; Galloway JN; Deviney FA. (2006). Acidic deposition impacts on natural resources in Shenandoah National Park.
(Technical Report NPS/NER/NRTR—2006/066) Philadelphia, PA; Northeast Region of the National Park Service; U.S. Department of
the Interior.
Costanza R; d'Arge R; de Groot R; Farber S; Grasso M; Hannon B; Limberg K; Naeem S; O'Neill RV; Paruelo J; Raskin RG; Sutton P; van den
Belt M. (1997). The value of the world's ecosystem services and natural capital. Nature, 387,253-259.
R-18
-------
Courchesne F; Cote B; Fyles JW; HendershotWH; Biron PM; Roy AG; Turmel MC. (2005). Recent changes in soil chemistry in a forested
ecosystem of southern Quebec, Canada. Soil Sci Soc Am J, 69,1298-1313.
Cowardin LM; Golet FC; LaRoe ET. (1979). Classification of wetland and deepwater habitats of the United States; (Report no. FWS/OBS-
79/31). Fish and Wildlife Service, U.S. Department of the Interior.
Cowell D; Apsimon H. (1996). Estimating the cost of damage to buildings by acidifying atmospheric pollution in Europe. Atmos Environ, 30,
2959-2968.
Cowling EB, Dochinger LS (1980). Effects of acidic precipitation on health and the productivity of forests. In: Miller PR (tech coordinator).
Proceedings of symposium on effects of air pollutants on mediterranean and temperate forest ecosystems, June 22- 27,1980,
Riverside, CA. (General technical report PSW-43, pp 165-173). Berkeley, CA; U.S. Forest Service; U.S. Department of Agriculture
(USDA).
Cowling, EB; Chameides WL; Kiang CS; Fehsenfeld FC; Meagher JF. (2000). Introduction to special Section: Southern Oxidant Study
Nashville/Middle Tennessee Ozone Study, part 2. J Geophys Res, 105,9075-9077.
Cowling EB; FurinessC. (Eds.) (2001). The state of the southern oxidants study (SOS): Policy-relevant findings in ozone and PM2.5 pollution
research 1994-2000. (Report SOS-2). Raleigh, NC; SOS-Office of the Director; North Carolina State University.
Craig BW; Friedland AJ. (1991). Spatial patterns in forest composition and standing dead red spruce in montane forests of the Adirondacks and
northern Appalachians. Environ Monit Assess, 18,129-140.
Crain CM. (2007). Shifting nutrient limitation and eutrophication effects in marsh vegetation across estuarine salinity gradients. Estuaries and
Coasts, 30, 26-34.
Crawford J; Davis D; Olson J; Chen G; Liu S; Fuelberg H; Hannan J; Kondo Y; Anderson B; Gregory G. Sachse G; Talbot R; Viggiano A;
Heikes B; Snow J; Singh H; Blake D. (2000). Evolution and chemical consequences of lightning-produced NOx observed in the North
Atlantic upper troposphere. J Geophys Res, 105,19795-19809.
Crill PM; Martikainen PJ; Nykanen H; Silvola J. (1994). Temperature and N-fertilization effects on methane oxidation in a drained peatland soil.
Soil Biol Biochem, 26,1331-1339.
Cristol DA; Brasso RL; Condon AM; Fovargue RE; Friedman SL; Hallinger KK; Monroe AP; White AE. (2008). The movement of aquatic
mercury through terrestrial food webs. Science, 320,335-335.
Crocker TD. (1985). On the value of the condition of a forest stock. Land Econ, 61,244-254.
Crocker TD; Forster BA. (1986). Atmospheric deposition and forest decline. Water Air Soil Pollut, 31,1007-1017.
Cronan CS; Reiners WA; Reynolds RC Jr; Lang GE. (1978). Forest floor leaching: Contributions from mineral, organic, and carbonic acids in
New Hampshire subalpine forests. Science, 200,309-311
Cronan CS; Aiken GR. (1985). Chemistry and transport of soluble humic substances in forested watersheds of the Adirondack Park, New York.
Geochim Cosmochim Acta, 49,1697-1705.
Cronan CS; Schofield CL. (1979). Aluminum leaching response to acid precipitation: Effects on high elevation watersheds in the Northeast.
Science, 204, 304-306.
Cronan CS; Walker W; Bloom P. (1986). Predicting aqueous aluminum concentrations in natural waters. Nature, 140-143.
Cronan CS; Goldstein RA. (1989a). ALBIOS: A Comparison of aluminum biogeochemistry in forested watersheds exposed to acidic deposition.
In: Adriano DC; Havas M. (Eds.). Acidic precipitation: Case studies. Volume 1 of series: Advances in environmental science, (pp 113-
135). New York, NY: Springer-Verlag.
Cronan CS; April R; Bartlett RJ; Bloom PR; Driscoll CT; Gherini SA; Henderson GS; Joslin JD; Kelly JM; Newton RM; Parnell RA; Patterson
HH; Raynal DJ; Schaedle M; Schofield CL; Sucoff El; Tepper HB; Thornton FC. (1989b). Aluminum toxicity in forests exposed to
acidic deposition: The ALBIOS results. Water Air Soil Pollut, 48,181-192.
Cronan CS; Schofield CL. (1990). Relationships between aqueous aluminum and acidic deposition in forested watersheds of North America
and northern Europe. Environ Sci Technol, 24,1100-1105.
Cronan CS; Perdue EM; Gjessing ET. (1990a). Patterns of organic acid transport from forested watersheds to aquatic ecosystems. In: Perdue
M; Glaze W; Gjessing ET (Eds.). Organic acids in aquatic ecosystems, (Report of the Dahlem workshop on organic acids in aquatic
ecosystems; Berlin, Germany May 7-12,1989). (pp. 245-260). New York, NY: John Wiley and Sons, Ltd.
Cronan, CS; Driscoll, CT; Newton, RM; Kelly JM; Schofield CL; Bartlett RJ; April R. (1990b). A comparative analysis of aluminum
biogeochemistry in a northeastern and a southeastern forested watershed. Water Resour Res, 26,1413-1430.
Cronan CS; Grigal DF. (1995). Use of calcium/aluminum ratios as indicators of stress in forest ecosystems. J Environ Qual, 24,209-226.
Crosley DR. (1996). NOy blue ribbon panel. J Geophys Res, 101,2049-2052.
Crutzen PJ. (1976). Upper limits on atmospheric ozone reductions following increased application of fixed nitrogen to the soil. Geophys Res
Lett, 3,169-172.
Crutzen PJ; Gidel LT. (1983). A two-dimensional photochemical model of the atmosphere. 2. The tropospheric budgets of the anthropogenic
chlorocarbons, CO, CH4, CH3CI and the effect of various NOx sources on tropospheric ozone. J Geophys Res, 88,6641-6661.
Cumming BF; Smol JP; Kingston JC; Charles DF; Birks HJB; Camburn KE; Dixit SS; Uutala AJ; Selle AR. (1992). How much acidification has
occurred in Adirondack region lakes (New York, USA) since preindustrial times? Can J Fish Aquat Sci, 49,128-141.
Cumming BF; Davey KA; Smol JP; Birks HJB. (1994). When did acid-sensitive Adirondack lakes (New York, USA).begin to acidify and are they
still acidifying? Can J Fish Aquat Sci, 51,1550-1568.
Cumming JR; Weinstein LH. (1990). Aluminum-mycorrhizal interactions in the physiology of pitch pine seedlings. Plant Soil, 125,7-18.
Curtis CJ; Emmett BA; Reynolds B; Shilland J. (2006). How important is N2O production in removing atmospherically deposited nitrogen from
UK moorland catchments? . Soil Biol Biochem, 38, 2081-2091.
Dagnall RM; Thompson KC; West TS. (1967).Molecular-emission in cool flames. Part I. The behaviour of sulphur species in a hydrogen-
nitrogen diffusion flame and in a shielded air-hydrogen flame. Analyst, 92, 506-512.
R-19
-------
Dahlgren RA; Walker WJ. (1993). Aluminum release rates from selected spodosol Bs horizons: Effect of pH and solid-phase aluminum pools.
Geochim Cosmochim Acta, 57,57-66.
Dahlgren RA. (1994). Soil acidification and nitrogen saturation from weathering of ammonium-bearing rock. Nature, 368,838-841.
Dahlgren RA; Driscoll CT. (1994). The effects of whole-tree clear cutting on soil processes at Hubbard Brook Experimental Forest, New
Hampshire, USA. Plant Soil, 58,239-262.
Dahlman L; Nasholm T; Palmqvist K. (2002). Growth, nitrogen uptake, and resource allocation in the two tripartite lichens Nephroma arcticum
and Peltigera aphthosa during nitrogen stress. New Phytol, 153,307-315.
Dahlman L; Persson J; Nasholm T; Palmqvist K. (2003). Carbon and nitrogen distribution in the green algal lichens Hypogymniaphysodes and
Platismatia glauca in relation to nutriend supply. Planta, 217,41-48.
Dahlman L; Persson J; Palmqvist K; Nasholm T. (2004). Organic and inorganic nitrogen uptake in lichens. Planta, 219,459-467.
Dail DB; Davidson EA; Chorover J. (2001). Rapid abiotic transformation of nitrate in an acid forest soil. Biogeochemistry, 54,131-146.
Daily GC. (1997). Introduction: What are ecosystem services. In: Daily GC (Ed.). Nature's services: Societal dependence on natural
ecosystems (pp. 1-10). Washington, DC: Island Press.
Daily GC; Ehrlich PR. (1999). Managing earth's ecosystems: An interdisciplinary challenge. Ecosystems, 2,277-280.
Dalai RC; Wang WJ; Robertson GP; Parton WJ. (2003). Nitrous oxide emission from Australian agricultural lands and mitigation options: A
review. Australian J Soil Res, 41,165-195.
Dalsgaard T; Tharndrup B; Canfield DE. (2005). Anaerobic ammonium oxidation (anammox) in the marine environment. Res Microbiol, 156,
457-464.
Damian L; Fako R. (2000). Weathering structural steels corrosion in atmospheres of various degrees of pollution in Romania. Materials and
Corrosion, 51,574-578.
Dangles O; Malmqvist B; Laudon H. (2004). Naturally acid freshwater ecosystems are diverse and functional: evidence from boreal streams.
Oikos, 104,149-155.
Dante JF; Kelly RG. (1993). The evolution of the adsorbed solution layer during atmospheric corrosion and its effects on the corrosion rate of
copper. J Electrochem Soc, 140,1890-1897.
D'Antonio CM; Vitousek PM. (1992). Biological invasions by exotic grasses, the grass/fire cycle, and global change. Ann Rev Ecol Syst, 23,63-
87.
Darby FA; Turner RE. (2008). Below-and aboveground Spartina alterniflora production in a Louisiana salt marsh. Estuaries and Coasts, 31,
223-231.
Das B; Vinebrooke RD; Sanchez-Azofeifa A; Rivard B; Wolfe AP. (2005). Inferring sedimentary chlorophyll concentrations with reflectance
spectroscopy: A novel approach to reconstructing historical changes in the trophic status of mountain lakes. Can J Fish Aquat Sci, 62,
1067-1078.
Dasch JM. (1987). Measurement of dry deposition to surfaces in deciduous and pine canopies. Environmental Pollution 44:261-277.
Daum PH; Kleinman LI; Hills AJ; Lazrus AL; Leslie ACD; Busness K; Boatman J. (1990). Measurement and interpretation of concentrations of
H2O2 and related species in the upper midwest during summer. J Geophys Res, 95,9857-9871.
Daum PH; Kleinman LI; Newman L; Luke WT; Weinstein-Lloyd J; Berkowitz CM; Busness KM. (1996). Chemical and physical properties of
plumes of anthropogenic pollutants transported over the North Atlantic during the North Atlantic Regional Experiment. J Geophys Res,
101,29029-29042.
David MB; Driscoll CT. (1984). Aluminium speciation and equilibria in soil solutions of a haplorthod in the Adirondack Mountains (New York,
USA). Geoderma, 33,297-318.
David MB; Mitchell MJ; Scott TJ. (1987). Importance of biological processes in the sulfur budget of a northern hardwood ecosystem. Biol Fertil
Soils, 5, 258-264.
David MB; Lawrence GB. (1996). Soil and soil solution chemistry under red spruce stands across the northeastern United States. Soil Sci,
161(3), 314-328.
David M; Vance G; Kahl J. (1999). Chemistry of dissolved organic carbon at Bear Brook Watershed, Maine: Stream water response to
(NH4)2S04 additions. Environ Monit Assess, 55,149-163.
Davidson CI; Wu Y-L. (1990). Dry deposition of particles and vapors. In Lindberg SE, Page AL, Norton SA (Eds.), Acidic precipitation: vol. 3,
sources, deposition, and canopy interactions (pp. 103-216). New York: Springer-Verlag, Inc.
Davidson EA; Schimel JP. (1995). Microbial processes of production and consumption of nitric oxide, nitrous oxide and methane. In: Matson
PA, Harriss RC (Eds.), Biogenic trace gases: Measuring emissions from soil and water (pp. 327-357). New York, NY: Blackwell
Science.
Davidson EA; Kingerlee W. (1997). A global inventory of nitric oxide emissions from soils. Nutr Cycling Agroecosyst, 48,37-50.
Davidson EA; Keller M; Erickson HE; Verchot LV; Veldkamp E. (2000). Testing a conceptual model of soil emissions of nitrous and nitric
oxides. Bioscience, 50,667-680.
Davidson EA; Chorover J; Dail DB. (2003). A mechanism of abiotic immobilization of nitrate in forest ecosystems: the ferrous wheel hypothesis.
Glob Chang Biol, 9,228-236.
Davidson EA; Seitzinger S. (2006). The enigma of progress in denitrification research. Ecol Appl, 16,2057-2063.
Davies L; Bates JW; Bell JNB; James PW; Purvis OW. (2007). Diversity and sensitivity of epiphytes to oxides of nitrogen in London. Environ
Pollut, 146,299-310.
Davis RB; Anderson DS; Berge F. (1985). Paleolimnological evidence that lake acidification is accompanied by loss of organic matter. Nature,
316,436-438.
Davis RB; Anderson DS; Norton SA; Whiting MC. (1994). Acidity of twelve northern New England (USA) lakes in recent centuries. J
Paleolimnol, 12,103-154.
R-20
-------
Day DA; Wooldridge PJ; Dillon MB; Thornton JA; Cohen RC. (2002). A thermal dissociation laser-induced fluorescence instrument for in situ
detection of NO2, peroxy nitrates, alkyl nitrates, and HNO3. J Geophys Res, 107, doi: 10.1029/2001JD000779.
Day DA; Dillon MB; Woolridge PJ; Thornton JA; Rosen RS; Wood EC; Cohen RC. (2003). On alkyl nitrates, O3, and the missing NOy. J
Geophys Res, 108, doi 10.1029/2003JD003685.
De Bakker AJ. (1989). Effects of ammonia emission on epiphytic lichen vegetation. Acta Bot Neerl, 38, 337-342.
De Coninck F. (1980). Major mechanisms in formation of spodic horizons. Geoderma, 24,101-128.
De Haan DO; Brauers T; Oum K; Stutz J; Nordmeyer T; Finlayson-Pitts BJ. (1999). Heterogeneous chemistry in the troposphere: Experimental
approaches and applications to the chemistry of sea salt particles. Int Rev Phys Chem, 18,343-385.
de Laat ATJ; Zachariasse M; Roelofs GJ; Van Velthoven P; Dickerson RR; Rhoads KP; Oltmans SJ; Lelieveld J. (1999). Tropospheric O3
distribution over the Indian Ocean during spring 1995 evaluated with a chemistry-climate model. J Geophys Res, 104,13881 -13894.
De Schrijver A; Verheyen K; Mertens J; Staelens J; Wuyts K; Muys B. (2008). Nitrogen saturation and net ecosystem production. (Brief
communicatin arising from: F. Magnani et al. (2007). Nature, 447,849-850; Magnani et al. reply). Nature, 451, E1.
de Steiguer JE; Pye JM. (1988). Using scientific opinion to conduct forestry air pollution economic analyses. In: Jobstl A (Ed.). Proceedings of
symposium on the economic assessment of damage caused to forests by air pollutants; held in Gmunden, Austria September 13-17,
1988. IUFRO Working Party S4.04-02 (pp. 133-142).
de Steiguer JE; Pye JM; LoveCS. (1990). Air pollution damage to U.S. forests: A survey of perceptions and estimates of scientists. J For, 88,
17-22.
de Vries W. (1988). Critical deposition levels for nitrogen and sulphur on Dutch forest ecosystems. Water Air Soil Pollut, 42,221-239.
de Vries W; Posch M; Kamari J. (1989). Simulation of the long-term soil response to acid deposition in various buffer ranges. Water Air Soil
Pollut, 48, 349-390.
de Vries W; Reinds GJ; Posch M; Kamari J. (1994). Simulation of soil response to acidic deposition scenarios in Europe. Water Air Soil Pollut,
78,215-246.
de Vries W; Reinds GJ; Vel E. (2003). Intensive monitoring of forest ecosystems in Europe: 2. Atmospheric deposition and its impacts on soil
solution chemistry. For Ecol Manage, 174,97-115.
de Vries W; Reinds GJ; Gundersen P; Sterba H. (2006). The impact of nitrogen deposition on carbon sequestration in European forests and
forest soils. Glob Chang Biol, 12,1151-1173.
de Vries W; Kros H; Reinds GJ; Wamelink W; Mol J; van Dobben H; Bobbink R; EmmettB; SmartS; Evans C; SchlutowA; Kraft P; Belyazid S;
Sverdrup H; van Hinsberg A; Posch M; Hettelingh J-P. (2007). Developments in deriving critical limits and modelling critical loads of
nitrogen for terrestrial ecosystems in Europe. (Alterra report no 1382). Wageningen, The Netherlands: Wageningen University and
Research Centre - Alterra.
de Vries W; Solberg S; Dobbertin M; Sterba H; Laubhahn D; Reinds GJ; Nabuurs GJ; Gundersen P; Sutton MA. (2008). Ecologically
implausible carbon response? (Brief communicatin arising from: F. Magnani et al. (2007). Nature, 447,849-850; Magnani et al. reply).
Nature, 451, E1-E3.
De Witt HA; Mulder J; Nygaard PH; Aamlid D. (2001). Testing the aluminium toxicity hypothesis: A field manipulation experiment in mature
spruce forest in Norway. Water Air Soil Pollut, 130,995-1000.
DeBano LF; Conrad CE. (1978). The effect of fire on nutrients in a chaparral ecosystem. Ecology, 59,489-497.
DeBell LJ. (2006). Spatial and seasonal patterns and temporal variability of haze and its constituents in the United States: (CIRA IMPROVE
network report IV). Fort Collins, CO; Cooperative Institute for Research in the Atmosphere (CIRA); Colorado State University.
DeCaria AJ; Pickering KE; Stenchikov GL; Scala JR; Stith JL; Dye JE; Ridley BA; Laroche P. (2000). A cloud-scale model study of lightning-
generated NOx in an individual thunderstorm during STERAO-A. J Geophys Res, 105,11601-11616.
DeCaria AJ; Pickering KE; Stenchikov GL; Ott LE. (2005). Lightning-generated NOx and its impact on tropospheric ozone production: a three-
dimensional modeling study of a stratosphere-troposphere experiment: Radiation, aerosols and ozone (STERAO-A).thunderstorm. J
Geophys Res, 110, doi: 10.1029/2004JD005556.
DeConinck F. (1980). Major mechanisms in formation of spodic horizons. Geoderma, 24,101-128.
Deevey ES; Binford MW; Brenner M; Whitmore TJ. (1986). Sedimentary records of accelerated nutrient loading in Florida lakes. Hydrobiologia,
143,49-53.
DeForest JL; Zak DR; Pregitzer KS; Burton AJ. (2005). Atmospheric nitrate deposition and enhanced dissolved organic carbon leaching: Test
of a potential mechanism. Soil Sci Soc Am J, 69,1233-1237.
DeHayes DH; Schaberg PG; Hawley GJ; Strimbeck GR. (1999). Acid rain impacts on calcium nutrition and forest health. Bioscience, 49,789-
800.
Dehri I; Yazici B; Erbil M; Galip H. (1994). The effects of SO2 and NH3 on the atmospheric corrosion of galvanized iron sheet. Corros Sci, 36,
2181-2191.
Del Giorgio PA; Cole JJ; Cimberlic A. (1997). Respiration rates in bacteria exceed phytoplankton production in unproductive aquatic system.
Nature, 385,148-151.
Del Grosso SJ; Parton WJ; Mosier AR; Hartman MD; Brenner J; Ojima DS; Schimel DS. (2001). Simulated interaction of carbon dynamics and
nitrogen trace gas fluxes using the DAYCENT model. In: Schaffer M; Ma L; Hansen S (Eds.), Modeling carbon and nitrogen dynamics
for soil management, (pp 303-332). Boca Raton, Florida: CRC Press,
del Monte M; Rossi P. (1997). Fog and gypsum crystals on building materials. Atmos Environ, 31,1637-1646.
Delaune RD; Lindau CW; Sulaeman E; Jugsujinda A. (1998). Nitrification and denitrification estimates in a Louisiana swamp forest soil as
assessed by N-15 isotope dilution and direct gaseous measurements. Water Air Soil Pollut, 106,149-161.
Delgado JA; Mosier AR; Follett RH; Follett RF; Westfall DG; Klemedtsson LK; Vermeulen J. (1996). Effects of N management on N2O and CH4
fluxes and N-15: Recovery in an irrigated mountain meadow. Nutr Cycling Agroecosys, 46,127-134.
R-21
-------
D'Elia CJ; Sanders JG; Boynton WR. (1986). Nutrient enrichment studies in a coastal plain estuary: phytoplankton growth in large-scale,
continuous cultures. Can J Fish Aquat Sci, 43,397-406.
Delmas R. (1982). On the emission of carbon, nitrogen and sulfur in the atmosphere during bushfires in intertropical savannah zones. Geophys
Res Lett, 9, 761-764.
DeLucia EH; Hamilton JG; Naidu SL; Thomas RB; Andrews JA; Finzi A; Lavine M; Matamala R; Mohan JE; Hendrey GR; Schlesinger WH.
(1999). Net primary production of a forest ecosystem with experimental CO2 enrichment. Science, 284,1177-1179.
Denning AS; Baron J; Mast MA; Arthur M. (1991). Hydrologic pathways and chemical composition of runoff during snowmelt in Loch Vale
Watershed, Rocky Mountain National Park, Colorado, USA. Water Air Soil Pollut, 59,107-123.
Dennis A; Fraser M; Anderson S; Allen D. (2002). Air pollutant emissions associated with forest, grassland, and agricultural burning in Texas.
Atmos Environ, 36,3779-3792.
Dennis RL; Binkowski FS; Clark TL; McHenry JN; McHenry SJ; Raynolds SK. (1990). Selected applications of the regional acid deposition
model and engineering model, appendix 5F (part 2).of NAPAP SOS/T report 5. In: Irving PM, ed. National Acid Precipitation
Assessment Program: State of Science and Technology, Volume 1. Washington, DC: National Acid Precipitation Assessment
Program.
Dennis, R. (1997). Using the regional acid deposition model to determine the nitrogen deposition airshed of the Chesapeake Bay watershed.
In: Baker JE (Ed.), Atmospheric deposition of contaminants to the Great Lakes and coastal waters, (pp 393-413). Pensacola, FL:
Society of Environmental Toxicology and Chemistry Press.
Dennis RL; Mathur R. (2001). Airshed domains for modeling atmospheric deposition of oxidized and reduced nitrogen to the Neuse/Pamlico
system of North Carolina. Hydrol Sci Technol, 17,107-118.
Dennis RL; Tonnesen GS; Mathur R. (2001). Nonlinearities in the sulfate secondary fine particulate response to NOx emissions reductions as
modeled by the regional acid deposition model. In: Gryning S-E;. Schiermeier FA (Eds.), Air pollution modeling and its application, XIV
(Proceedings of the millennium NATO/CCMS international technical meeting on air pollution modeling and its application, held May
15-19,2000 in Boulder, Colorado pp 193-202). New York, NY: Kluwer Academic/Plenum Publishers.
Dennis TD; MacAvoy SE; Steg MB; Bulger AJ. (1995). The association of water chemistry variables and fish condition in streams of
Shenandoah National Park (USA). Water Air Soil Pollut, 85,365-370.
Dennis TE; Bulger AJ. (1995). Condition factor and whole-body sodium concentrations in a freshwater fish: Evidence for acidification stress and
possible ionoregulatory over-compensation. Water Air Soil Pollut, 85,377-382.
Dennis TE; Bulger AJ. (1999). The susceptibility of blacknose dace (Rhinichthys atratulus).to acidification in Shenandoah National Park. In:
Bulger AJ; Cosby BJ; Dolloff CA; Eshleman KN; Galloway JN; Webb. JR (Eds.), Shenandoah National Park: Fish in sensitive habitats.
Project final report, Volume IV: Stream bioassays, aluminum toxicity, species richness and stream chemistry, and models of
susceptibility to acidification. Chapter 6B. Project completion report to the National Park Service. Charlottesville, VA; Department of
Environmental Sciences, University of Virginia.
Dentener FJ; Crutzen PJ. (1993). Reaction of N2O5 on tropospheric aerosols: impact on the global distributions of NOx, O3, and OH. J Geophys
Res, 98,7149-7163.
Dentener F; Stevenson D; Cofala J; Mechler R; Amann M; Bergamaschi P; Raes F; Derwent R. (2005). The impact of air pollutant and
methane emission controls on tropospheric ozone and radiative forcing: CTM calculations for the period 1990-2030. Atmos Chem
Phys, 5,1731-1755.
Dentener F; Stevenson D; Ellingsen K; Van Noije T; Schultz M; Amann M; Atherton C; Bell N; Bergmann D; Bey I; Bouwman L; Butler T; Cofala
J; Collins B; Drevet J; Doherty R; Eickhout B; Eskes H; Fiore A; Gauss M; Hauglustaine D; Horowitz L; Isaksen IS; Josse B; Lawrence
M; Krol M; Lamarque JF; Montanaro V; Muller JF; Peuch VH; Pitari G; Pyle J; Rast S; Rodriguez I; Sanderson M; Savage NH; Shindell
D; Strahan S; Szopa S; Sudo K; Van Dingenen R; Wild O; Zeng G. (2006a). The global atmospheric environment for the next
generation. Environ Sci Technol, 40,3586-3594.
Dentener F; Drevet J; Lamarque JF; Bey I; Eickhout B; Fiore AM; Hauglustaine D; Horowitz LW; Krol M; Kulshrestha UC; Lawrence M; Galy-
Lacaux C; Rast S; Shindell D; Stevenson D; Van Noije T; Atherton C; Bell N; Bergman D; Butler T; Cofala J; Collins B; Doherty R;
Ellingsen K; Galloway J; Gauss M; Montanaro V; Muller JF; Pitari G; Rodriguez J; Sanderson M; Solman F; Strahan S; Schultz M;
Sudo K; Szopa S; Wild O. (2006b). Nitrogen and sulfur deposition on regional and global scales: a multi-model evaluation. Glob
Biogeochem Cycle, 20, doi: 10.1029/2005GB002672.
Derwent RG; Collins WJ; Johnson CE; Stevenson DS. (2001). Transient behaviour of tropospheric ozone precursors in a global 3-D CTM and
their indirect greenhouse effects. Climatic Change, 49,463-487.
DesGranges J-L; Darveau M. (1985). Effect of lake acidity and morphometry on the distribution of aquatic birds in southern Quebec. Holarctic
Ecology, 8,181-190.
DesGranges J-L; Houde B. (1989). Effects of acidity and other environmental parameters on the distribution of lacustrine birds in Quebec. In:
DeGranges, JL (Ed.). Studies on the effects of acidification on aquatic wildlife in Canada: Lacustrine birds and their habitats in
Quebec. (Canadian Wildlife Service occasional paper No. 67 pp 7-41). Ottawa, Ontario, Canada; Canadian Wildlife Service;
Environment Canada.
Deviney FA; Rice KC; Hornberger GM. (2006). Time series and recurrence interval models to predict the vulnerability of streams to episodic
acidification in Shenandoah National Park, Virginia. Water Resour Res, 42, doi: 10.1029/2005WR004740.
Devito KJ; Dillon PJ; LaZerte BD. (1989). Phosphorus and nitrogen retention in five Precambrian Shield wetlands. Biogeochemistry, 8,185-
204.
Devito KJ; Hill AR. (1997). Sulphate dynamics in relation to groundwater-surface water interactions in headwater wetlands of the southern
Canadian Shield. Hydrol Process, 11,485-500.
R-22
-------
Devito KJ; Hill AR. (1999). Sulphate mobilization and pore water chemistry in relation to groundwater hydrology and summer drought in two
conifer swamps on the Canadian Shield. Water Air Soil Pollut, 113,97-114.
Devol A. (2003). Solution to a marine mystery. Nature, 422,575-576.
DeWald LE; Sucoff El; Ohno T; Buschena CA. (1990). Response of northern red oak (Quercus rubral.) seedlings to soil solution aluminum.
Can J For Res, 20,331-336.
DeWalle DR; Dinicola RS; Sharpe WE. (1987). Predicting baseflow alkalinity as an index to episodic stream acidification and fish presence.
Water Resour Bull, 23, 29-35.
DeWalle DR; Swistock BR. (1994a). Causes of episodic acidification in five Pennsylvania streams on the northern Appalachian Plateau. Water
Resour Res, 30,1955-1963.
DeWalle DR; Swistock BR. (1994b). Differences in oxygen-18 content of throughfall and rainfall in hardwood and coniferous forests. Hydrol
Processes, 8,75-82.
DeWalle DR; Kochenderfer JN; Adams MB; Miller GW; Gilliam FS; Wood F; Odenwald-Clemens SS; Sharpe WE. (2006). Vegetation and
acidification. In: Adams MB; DeWalle DR; Horn JL (Eds.), The Fernow watershed acidification study. (Volume 11 of Environmental
Pollution Series, pp 137-188). Dordrecht, The Netherlands: Springer.
Diaz RJ; Solow A. (1999). Ecological and economic consequences of hypoxia: Topic 2 Report for the integrated assessment on hypoxia in the
Gulf of Mexico. (NOAA Coastal Ocean Program Decision Analysis Series No
16).http://www8.nos.noaa.aov/nccos/pubinfo.aspx?id=514 . Silver Spring, MD: National cean Service Coastal Ocean Program;
National Oceanic and Atmospheric Administration (NOAA). U.S. Department of Commerce.
Dibb JE; Arsenault M; Peterson MC. (2002). Fast nitrogen oxide photochemistry in Summit, Greenland snow. Atmos Environ, 36,2501-2511.
Dickerson RR; Huffman GJ; Luke WT; Nunneracker LJ; Pickering KE; Leslie ACD; Lindsey CG; Slinn WGN; Kelly TJ; Daum PH; Delany AC;
Greenberg JP; Zimmerman PR; Boatman JF; Ray JD; Stedman DH. (1987). Thunderstorms: An important mechanism in the transport
of air pollutants. Science, 235,460-464.
Dickerson RR; Rhoads KP; Carsey TP; Oltmans SJ; Burrows JP; Crutzen PJ. (1999). Ozone in the remote marine boundary layer: a possible
role for halogens. J Geophys Res, 104, 21385-21395.
Dickson W. (1978). Some effects of the acidification of Swedish lakes. (Proceedings from the 20th Congress, internationale vereinigung fur
theoretische und angewandte limnologie; held in Copenhagen, Denmark; August 8,1977). Verh InternatVerein Limnol, 20,851-856.
Diemer M. (1997). Effects of elevated CO2 on gas exchange characteristics of alpine grassland. Acta Oecologica, 18,177-182.
Dierberg FE; Brezonik PL. (1983). Nitrogen and phosphorus mass balances in natural and sewage-enriched cypress domes. J Appl Ecol, 20,
323-337.
Dighton J; Tuininga AR; Gray DM; Huskins RE; Belton T. (2004). Impacts of atmospheric deposition on New Jersey pine barrens forest soils
and communities of ectomycorrhizae. For Ecol Manage, 201,131-144.
Dijkstra FA; Smite MM. (2002). Tree species effects on calcium cycling: The role of calcium uptake in deep soils. Ecosystems, 5; 385-398.
Dillon PJ; Reid RA; Girard R. (1986). Changes in the chemistry of lakes near Sudbury, Ontario, following reduction of SO2 emissions. Water Air
Soil Pollut, 31,59-65.
Dillon PJ; LaZerte BD. (1992). Response of Plastic Lake catchment, Ontario, to reduced sulphur deposition. Environ. Pollut, 77, 211-217.
Dillon PJ; Molot LA; Futter M. (1997). The effect of El Nino-related drought on the recovery of acidified lakes. Environ Monit Assess, 46,105-
111.
Dise NB; Wright RF. (1995). Nitrogen leaching from European forests in relation to nitrogen deposition. For Ecol Manage, 71,153-161.
Dise NB; Matzner E; Gundersen P. (1998). Synthesis of nitrogen pools and fluxes from European forest ecosystems. Water Air Soil Pollut, 105,
143-154.
Dise NB; Verry ES. (2001). Suppression of peatland methane emission by cumulative sulfate deposition in simulated acid rain.
Biogeochemistry, 53,143-160.
Dixit SS; Keller W; Dixit AS; Smol JP. (2001). Diatom-inferred dissolved organic carbon reconstructions provide assessments of past UV-B
penetration in Canadian Shield lakes. Can J Fish Aquat Sci, 58,543-550.
Dobson JE; Rush RM; Peplies RW. (1990). Forest blowdown and lake acidification. Ann Assoc Amer Geogr, 80,343-361.
Dodds WK. (2006). Eutrophication and trophic state in rivers and streams. Limnol Oceanogr, 51,671-680.
Dodge MC. (1977). Combined use of modeling techniques and smog chamber data to derive ozone-precursor relationships. In: Dimitriades B
(Ed.). Proceedings from the international conference on photochemical oxidant and its control, Volume II (EPA/600/3/77-001 b pp 881 -
889). Research Triangle Park, NC; Environmental Sciences Research Laboratory, U.S. Environmental Protection Agency.
Domine F; Shepson PB. (2002). Air-snow interactions and atmospheric chemistry. Science, 297,1506-1510.
Dommen J; Prevot ASH; Hering AM; Staffelbach T; Kok GL; Schillawski RD. (1999). Photochemical production and aging of an urban air mass.
J Geophys Res, 104, 5493-5506.
Donagi AE; Goren Al. (1979). Use of indicator plants to evaluate atmospheric levels of nitrogen dioxide in the vicinity of a chemical plant.
Environ Sci Technol, 13, 986-989.
Dortch Q; Whitledge TE. (1992). Does nitrogen or silicon limit phytoplankton production in the Mississippi River plume and nearby regions?
Continental Shelf Research, 12,1293-1309.
Doskey PV; Kotamarthi VR; Fukui Y; Cook DR; Breitbeil FW, III; Wesely ML. (2004). Air-surface exchange of peroxyacetyl nitrate at a
grassland site. J Geophys Res, 109, doi 10.1029/2004JD004533
Downing JA; McCauley E. (1992). The nitrogen-phosphorus relationship in lakes. Limnol Oceanogr, 37, 936-945.
Downing JA. (1997). Marine nitrogen: phosphorus stoichiometry and the global N:P cycle. Biogeochemistry, 37,237-252.
Drake BG; Gonzalez-Meier MA; Long SP. (1997). More efficient plants: a consequence of rising atmospheric CO2? Ann Rev Plant Physiol Mol
Biol, 48, 609-639.
R-23
-------
Drenovsky RE; Richards JH. (2005). Nitrogen addition increases fecundity in the desert shrub Sarcobatus vermiculatus. Oecologia, 143,349-
356.
Drever Jl; Hurcomb DR. (1986). Neutralization of atmospheric acidity by chemical weathering in an alpine drainage basin in the North Cascade
Mountains. Geology, 14, 221-224.
Drevnick PE; Canfield DE; Gorski PR; Shinneman ALC; Engstrom DR; Muir DCG; Smith GR; Garrison PJ; Cleckner LB; Hurley JP; Noble RB;
Otter RR; Oris JT. (2007). Deposition and cycling of sulfur controls mercury accumulation in Isle Royale fish. Environ Sci Technol, 41,
7266-7272.
Driscoll CT Jr; Baker JP; Bisogni JJ Jr; SchofieldCL. (1980). Effect of aluminum speciation on fish in dilute acidified waters. Nature, 284,161-
164.
Driscoll CT. (1984). A procedure for the fractionation of aqueous aluminum in dilute acidic waters. Int J Environ Anal Chem, 16,267-283.
Driscoll CT; Bisogni JJ. (1984). Weak acid/base systems in dilute acidified lakes and streams of the Adirondack region of New York State. In:
Schnoor J.L., ed. Modeling of total acid precipitation, (pp 53-72). Boston, MA: Butterworth Publishers.
Driscoll CT; Newton RM. (1985). Chemical characteristics of Adirondack lakes. Environ Sci Technol, 19,1018-1024.
Driscoll CT; Van Breemen N; Mulder J. (1985). Aluminum chemistry in a forested spodosol. Soil Sci Soc Am J, 49,437-444.
Driscoll CT; Yatski CP; Unangst FJ. (1987a). Longitudinal and temporal trends in the water chemistry of the North Branch of the Moose River.
Biogeochemistry, 3,37-61.
Driscoll CT; Wyskowski BJ; Cosentini CC; Smith ME. (1987b). Processes regulating temporal and longitudinal variations in the chemistry of a
low-order woodland stream in the Adirondack region of New York. Biogeochemistry, 3,225-241.
Driscoll CT; Johnson NM; Likens GE; Feller MC. (1988). Effects of acidic deposition on the chemistry of headwater streams: A comparison
between Hubbard Brook, New Hampshire, and Jamieson Creek, British Columbia. Water Resour Res, 24,195-200.
Driscoll CT; Likens GE; Hedin LO; Eaton JS; Bormann FH. (1989). Changes in the chemistry of surface waters. Environ Sci Technol, 23,137-
143.
Driscoll CT; Newton RM; Gubala CP; Baker JP; Christensen SW. (1991). Adirondack Mountains. In: Charles DF; Christie S (Eds.), Acidic
deposition and aquatic ecosystems: Regional case studies (pp. 133-202). New York, NY: Springer-Verlag.
Driscoll CT; Van Dreason R. (1993). Seasonal and long-term temporal patterns in the chemistry of Adirondack lakes. Water Air Soil Pollut, 67,
319-344.
Driscoll CT; Yan C; Schofield CL; Munson R; Holsapple J. (1994a). The mercury cycle and fish in the Adirondack lakes. Environ Sci Technol,
28,136A-143A.
Driscoll CT; Lehtinen MD; Sullivan TJ. (1994b). Modeling the acid-base chemistry of organic solutes in Adirondack, New York, lakes. Water
Resour Res, 30,297-306.
Driscoll CT; PostekKM; KretserW; Raynal DJ. (1995). Long-term trends in the chemistry of precipitation and lake water in the Adirondack
region of New York, USA. Water Air Soil Pollut, 85,583-588.
Driscoll CT; Postek KM. (1995). The chemistry of aluminum in surface waters. In: Sposito G (Ed.), The environmental chemistry of aluminum
(pp. 363-418). Chelsea, Ml: Lewis Publishers.
Driscoll CT; Likens GE; Church MR. (1998). Recovery of surface waters in the northeastern U.S. from decreases in atmospheric deposition of
sulfur. Water Air Soil Pollut, 105,319-329.
Driscoll CT; Lawrence GB; Bulger AJ; Butler TJ; Cronan CS; Eagar C; Lambert KF; Likens GE; Stoddard JL; Weather KC. (2001a). Acid rain
revisited: Advances in scientific understanding since the passage of the 1970 and 1990 Clean Air Act amendments. (Vol. 1, No. 1).
Hanover, NH: Hubbard Brook Research Foundation; Science Links Publication.
Driscoll CT; Lawrence GB; Bulger AJ; Butler TJ; Cronan CS; Eagar C; Lambert KF; Likens GE; Stoddard JL; Weather KC. (2001b). Acidic
deposition in the northeastern United States: Sources and inputs, ecosystem effects, and management strategies. Bioscience, 51,
180-198.
Driscoll C; Whitall J; Aber E; Boyer M; Castro C; Cronan C; Goodale P; Groffman C; Hopkinson K; Lambert G; Lawrence G; Ollinger S.
(2003a). Nitrogen pollution in the northeastern United States: Sources, effects and management options. Bioscience, 53, 357-374.
Driscoll CT; Driscoll KM; Mitchell M J; Raynal DJ. (2003b). Effects of acidic deposition on forest and aquatic ecosystems in New York State.
Environ Pollut, 123,327-336.
Driscoll CT; Driscoll KM; Roy KM; Mitchell MJ. (2003c). Chemical response of lakes in the Adirondack Region of New York to declines in acidic
deposition. Environ Sci Technol, 37,2036-2042.
Driscoll C; Whitall J; Aber E; Boyer M; Castro C; Cronan C; Goodale P; Groffman C; Hopkinson K; Lambert G; Lawrence G; Ollinger S.
(2003d). Nitrogen pollution: Sources and
consequences in the U.S. Northeast. Environ, 45,8-22.
Driscoll CT; Driscoll KM; Roy KM; Dukett J. (2007a). Changes in the chemistry of lakes in the Adirondack Region of New York following
declines in acidic deposition. Appl Geochem, 22,1181 -1188
Driscoll CT; Han Y-J; Chen CY; Evers DC; Lambert KF; Holsen TM; Kamman NC; Munson RK. (2007b). Mercury contamination in forest and
freshwater ecosystems in the northeastern United States. Bioscience, 57,17-28.
Driscoll CT; Fallon-Lambert K; Chen L. (2007c). Acidic deposition: sources and ecological effects. In: Visgilio GR; Whitelaw DM, (Eds.). Acid in
the environment: Lessons learned and future prospects. New York, NY: Springer.
Drohan JR; Sharpe WE. (1997). Long-term changes in forest soil acidity in Pennsylvania, USA. Water Air Soil Pollut, 95,299-311.
Drohan P; Stout S; Petersen G. (1999). Spatial relationships between sugar maple (Acersacharum Marsh.), sugar maple decline, slope,
aspect, and atmospheric deposition in northern Pennsylvania. In: Horsley SB; Long RP, (Eds.). Sugar Maple Ecology and Health:
Proceedings of an International Symposium held in Warren, PA in June, 1998. (General technical report NE-261). Radnor, PA; U.S.
Forest Service; U.S. Department of Agriculture (USDA).
R-24
-------
Drohan PJ; DeWalle DR. (2002). Defoliation and atmospheric deposition influences on spring baseflow chemistry in 56 Pennsylvania mixed
land-use watersheds. Water Air Soil Pollut, 133,31-48.
Drohan PJ; Stout SL; Petersen GW. (2002). Sugar maple (Acersaccharum Marsh.) decline during 1979-1989 in northern Pennsylvania. For
Ecol Manage, 170,1-17.
Duan L; Hao JM; Xie SD; Du K. (2000a). Critical loads of acidity for surface waters in China. Sci. Total Environ, 246,1-10.
Duan L; Xie SD; Zhou ZP; Hao JM. (2000b). Critical loads of acid deposition on soil in China. Water Air Soil Pollut, 118,35-51.
Duan L; Xie SD; Zhou ZP; YeXM; Hao JM. (2001). Calculation and mapping of critical loads for S, N and acidity in China. Water Air Soil Pollut,
130,1199-1204.
Dubowski Y; Sumner AL; Menke EJ; Gaspar DJ; Newberg JT; Hoffman RC; Penner RM; Hemminger JC; Finlayson-Pitts BJ. (2004).
Interactions of gaseous nitric acid with surfaces of environmental interest. Phys Chem Chem Phys, 6,3879-3888.
Duce RA; Liss PS; Merrill JT; Atlans EL; Buat-Menard P; Hicks BB; Miller JM; Prospero JM; Atimoto R; Church TM; Ellis W; Galloway JN;
Hansen L; Jickells TD; Knap AH; Reinhardt KH; Schneider B; Soudine A; Tokos JJ; Tsunogai S; Wollast R; Ahou M (1991) The
atmospheric input of trace species to the world ocean. Global Biogeochem. Cycles, 5,193-259
Dudhia J. (1993). A nonhydrostatic version of the Penn State-NCAR mesoscale model: Validation tests and simulation of an Atlantic cyclone
and cold front. Monthly Weather Rev, 121,1493-1513.
DueckTA; Dorel FG; Ter Horst R; Van de Eerden LJM. (1990). Effects of ammonia, ammonium sulphate andsulpur dioxide on the frost
sensitivity of Scots pine (Pinus sylvestris L.).Water Air Soil Pollut, 54,35-49.
Dueck TA; Van der Eerden LJM; Breemsterboer B; Elderson J. (1991). Nitrogen uptake and allocation by Calluna vulgaris (L.). Hull and
Deschampsia flexuosa (L.).Trin. exposed to 15NH3. Acta Bot Neerl, 40,257-267.
Dumont E; Harrison JA; Kroeze C; Bakker EJ; Seitzinger SP. (2005). Global distribution and sources of dissolved inorganic nitrogen export to
the coastal zone: Results from a spatially explicit, global model. Glob Biogeochem Cycles, 19, GB4S02, doi: 10.1029/2005GB002488.
Dunlea EJ; Herndon SC; Nelson DD; Volkamer RM; San Martini F; Sheehy PM; Zahniser MS; Shorter JH; Wormhoudt JC; Lamb BK; Allwine
EJ; Gaffney JS; Marley NA; Grutter M; Marquez, C; Blanco S; Cardenas B; Retama.A; Ramon Villegas CR; Kolb CE; Molina LT;
Molina MJ. (2007). Evaluation of nitrogen dioxide chemiluminescence monitors in a polluted urban environment. Atmos Chem Phys, 7,
2691-2704.
Dupont J; Clair TA; Gagnon C; Jeffries DS; Kahl JS; Nelson S; Peckenham J. (2005). Estimation of critical loads of acidity and water sensitivity
in New England states and eastern Canadian provinces. Environ Monit Assess, 109,275-291.
Durant JL; Busby WF Jr; Lafleur AL; Penman BW; Crespi CL. (1996). Human cell mutagenicity of oxygenated, nitrated and unsubstituted
polycyclic aromatic hydrocarbons associated with urban aerosols. Mutat Res, 371,123-157.
Durham JL; Wilson WE; Bailey EB. (1978). Application of an S02-denuder for continuous measurement of sulfur in submicrometric aerosols.
Atmos Environ, 12,883-886.
Duriscoe DM; Stolte KW. (1989). Photochemical oxidant injury to ponderosa pine (Pinus ponderosa Laws.).and Jeffrey pine (Pinus jeffreyi
Grev. and Balf.).in the national parks of the Sierra Nevada of California. In: Olson RK; Lefohn AS, eds. Effects of air pollution on
western forests [an A&WMA symposium held in Anaheim, CA; in June 1989; (APCA transactions series: no. 16 pp. 261-278)].
Pittsburgh, PA: Air and Waste Management Asssociation.
Dyrness CT; Van Cleve K; Levison JD. (1989). The effect of wildfire on soil chemisty in four forest types in interior Alaska. Can J For Res, 19,
1389-1396.
Eadie BJ; McKee BA; Lansing MB; Robbins JA; Metz S; Trefry JH. (1994). Records of nutrient-enhanced coastal productivity in sediments from
the Louisiana continental shelf. Estuaries, 17,754-765.
EagarC; Adams MB. eds. (1992). Ecology and decline of red spruce in the eastern United States. (Ecological studies: Volume 96). New York,
NY: Springer-Verlag.
EagarC; Van Miegroet H; McLaughline SB; Nicholas NS. (1996). Evaluation of effects of acidic deposition to terrestrial ecosystems in class I
areas of the Southern Appalachians (No. A technical report to the Southern Appalachian Mountains Initiative (SAMI)). Tennessee
Valley Authority (TVA), Asheville, NC; U.S. Forest Service, U.S. Department of Agriculture (USDA).
Eder B; Yu S. (2006). A performance evaluation of the 2004 release of models-3 CMAQ. Atmos Environ, 40,4811 -4824.
Edgerton ES; Hartsell BE; Saylor RD; Jansen JJ; Hansen DA; Hidy GM. (2006). The southeastern aerosol research and characterization study.
Part 3: Continuous measurements of fine particulate matter mass and composition. J Air Waste Manage Assoc, 56,1325-1341.
Edgerton ES; Saylor RD; Hartsell BE; Jansen JJ; Alan Hansen D. (2007). Ammonia and ammonium measurements from the southeastern
United States. Atmos Environ, 41,3339-3351.
Edmondson WT. (1969). Eutrophication in North America. Washington, D.C.: National Academy of Sciences, Printing and Publishing Office.
Edmondson WT. (1991). The uses of ecology: Lake Washington and beyond. Seattle, WA: University of Washington Press.
Edwards D; Brown JA; Whitehead C. (1987). Endocrine and other physiological indicators of acid stress in the brown trout. Ann Soc R Zool
Belg, 117,331-342.
Edwards JH; Horton BD. (1977). Aluminum-induced calcium deficiency in peach seedlings. J Am Soc Hortic Sci, 102,459-461.
Edwards PJ; Gregory JD; Allen HI. (1999). Seasonal sulfate deposition and export patterns for a small Appalachian watershed. Water Air Soil
Pollut, 110,137-155.
Edwards PJ; Wood F; Kochenderfer JN. (2002a). Baseflow and peakflow chemical responses to experimental applications of ammonium
sulphate to forested watersheds in north-central West Virginia, USA. Hydrological Processes, 16,2287-2310.
Edwards PJ; Kochenderfer JN; Coble DW; Adams MB. (2002b). Soil leachate responses during 10 years of induced whole-watershed
acidification. Water Air Soil Pollut, 140,99-118.
Edwards PM; Helvey JD. (1991). Long-term ionic increases from a central Appalachian forested watershed. J Environ Qual, 20,250-255.
R-25
-------
Egerton-Warburton L; Allen EB. (2000). Shifts in Arbuscular mycorrhizal communities along an anthropogenic nitrogen deposition gradient.
EcolAppI, 10,484-496.
Egerton-Warburton LM; Graham RC; Allen EB; Allen MF. (2001). Reconstruction of the historical changes in mycorrhizal fungal communities
under anthropogenic nitrogen deposition. Proc R Soc B, 268,2479-2484.
Eide I; Neverdal G; Thorvaldsen B; Grung B; Kvalheim OM. (2002). Toxicological evaluation of complex mixtures by pattern recognition:
correlating chemical fingerprints to mutagenicity. Environ Health Perspect, 1106, 985-988.
Eilers JM; Glass GE; Webster, KE; Rogalla, JA. (1983). Hydrologic control of lake susceptibility to acidification. Can J Fish Aquat Sci, 40,1896-
1904.
Eilers JM; Kanciruck P; McCord RA; Overton WS; Hook L; Blick DJ; Brakke DF; Kellar PE; DeHaan MS; Silverstein ME; Landers DH. (1987).
Characteristics of lakes in the western United States. Volume II. Data compendium for selected physical and chemical variables (No.
EPA-600/3-86/054b). Washington, DC: U.S. Environmental Protection Agency.
Eilers JM; Brakke DF; Landers DH. (1988). Chemical and physical characteristics of lakes in the Upper Midwest, United States. Environ Sci
Technol, 22,164-172.
Eimers MC; Dillon PJ. (2002). Climate effects on sulphate flux from forested catchments in south-central Ontario. Biogeochemistry, 61,337-
355.
Eimers MC; Buttle JM; Watmough SA. (2007). The contribution of rain-on-snow events to annual N03~N export from a forested catchment in
south-central Ontario, Canada. Appl Geochem, 22,1105-1110.
Eisele FL; Mauldin L; Cantrell C; Zondlo M; Apel E; Fried A; Walega J; Shetter R; Lefer B; Flocke F; Weinheimer A; Avery M; Vay S; Sachse G;
Podolske J; Diskin G; Barrick JD; Singh HB; Brune W; Harder H; Martinez M; Bandy A; Thornton D; Heikes B; Kondo Y; Riemer D;
Sandholm S; Tan D; Talbot R; Dibb J. (2003). Summary of measurement intercomparisons during TRACE-P. J Geophys Res, 108,
doi: 10.1029/2002JD003167.
Ekstrom EB; Morel FMM; Benoit JM. (2003). Mercury methylation independent of the acetyl-coenzyme a pathway in sulfate-reducing bacteria.
Appl Environ Microbiol, 69,5414-5422.
Ekstrom EB; Morel FMM. (2004). Mercury methylation by sulfate-reducing bacteria independent of vitamin B12. Biogeochemistry, 51,968-970.
Eldred R. (2001). Sulfur-Sulfate Trends for IMPROVE. University of California-Davis.
http://vista.cira.colostate.edu/improve/publications/GravLit/005 Sulfur-Sulfate Trends/005 sulfur-sulfate trends.htm.
Eliason SA; Allen EB. (1997). Exotic grass competition in suppressing native shrubland re-establishment. Restor Ecol, 5,245-255.
Ellenberg H. (1985). Veranderungen der floa mitteleuropas unter dem einfluss von diingung und immissionen. Schweiz Z Forstwesten, 136,
19-39.
Ellenberg H. (1987). Floristic changes due to eutrophication. In Asman WAH; Diederen HSMA (Eds.). Proceedings of the ammonia and
acidification symposium of the European association for the science of air pollution (EURASAP), held in Bilthoven, The Netherlands.
April 13 -15,1987; (pp 301-308). European Association for the Science of Air Pollution (EURASAP).
Elliott EM; Kendall C; Wankel S; Burns DA; Boyer EW; Harlin K; Bain DJ; Butler TJ. (2007). Nitrogen isotopes as indicators of NOx source
contributions to atmospheric nitrate deposition across the midwestern and northeastern United States. Environ Sci Technol, 41, 7661-
7667.
Ellis H; Bowman M. (1994). Critical loads and development of acid rain control options. J Environ Eng, 120,273-290.
Ellis H; Ringold PL; Holdren GR Jr. (1996). Emissions reductions and ecological response: management models for acid rain control.
Socioecon Plann Sci, 30,15-26.
Ellison AM; Gotelli NJ. (2002). Nitrogen availability alters the expression of carnivory in the northern pitcher plant, Sarracenia purpurea. Proc
Natl Acad Sci USA, 99,4409-4412.
Elser JJ; Marzolf ER; Goldman CR. (1990). Phosphorus and nitrogen limitation of phytoplankton growth in the freshwaters of North America: A
review and critique of experimental enrichments. Can J Fish Aquat Sci, 47,1468-1477.
Elser JJ; Hayakawa K; Urabe J. (2001). Nutrient limitation reduces food quality forzooplankton: Daphnia response to seston phosphorus
enrichment. Ecology, 82,898-903.
Elser JJ; Bracken MES; Cleland EE; GrunerDS; Harpole WS; Hillebrand II H; Ngai JT; Seabloom EW; Shurin JB; Smith JE. (2007). Global
analysis of nitrogen and phosphorus limitation of primary producers in freshwater, marine, and terrestrial ecosystems. Ecol Lett, 10,
1135-1142.
Elsom DM. (2002). Atmospheric pollution: a global problem, 2nd Edition. Oxford,UK: Blackwell Publishers.
Elvir JA; Wiersma GB; White AS; Fernandez IJ. (2003). Effects of chronic ammonium sulfate treatment on basal area increment in red spruce
and sugar maple at the Bear Brook Watershed in Maine. Can J For Res, 33,862-869.
Elvir JA; Wiersma GB; Day ME; Greenwood MS; Fernandez I J. (2006). Effects of enhanced nitrogen deposition on foliar chemistry and
physiological processes of forest trees at the Bear Brook Watershed in Maine. For Ecol Manage, 221,207-214.
Elwood JW; Sale MJ; Kaufmann PR; Cada GF. (1991). Southern Blue Ridge Province. In: Charles DF, Christie S (Eds.). Acidic deposition and
aquatic ecosystems: Regional case studies. (Chapter 11, pp 319-364). New York, NY: Springer-Verlag.
Emmett BA; Boxman D; Bredemeier M; Gundersen P; Kjonaas OJ; Moldan F; Schleppi P; Tietema A; Wright RF. (1998). Predicting the effects
of atmospheric nitrogen deposition in conifer stands: Evidence from the NITREX ecosystem-scale experiments. Ecosystems, 1,352-
360.
Emmett BA. (1999). The impact of nitrogen on forest soils and feedbacks on tree growth. Water Air Soil Pollut, 116,65-74.
Emmons LK; Hauglustaine DA; Muller J-F; Carroll MA; Brasseur GP; Brunner D; Staehelin J; Thouret V; Marenco A. (2000). Data composites
of airborne observations of tropospheric ozone and its precursors. J Geophys Res, 105,20497-20538.
Engel-Cox JA; Holloman CH; Coutant BW; Hoff RM. (2004). Qualitative and quantitative evaluation of MODIS satellite sensor data for regional
and urban scale air quality. Atmos Environ, 38,2495-2509.
R-26
-------
Englin J; Cameron T; Mendelsohn R; Parsons G; Shankle S. (1991). Valuation of damages to recreational trout fishing in the Upper Northeast
due to acidic deposition. (Report no PNL-7683). Richland, WA: Pacific Northwest Laboratory, (prepared for National Acidic
Precipitation Assessment Program (NAPAP)).
Engstrom DR; Swain EB; Henning TA; Brigham MA; Brezonik PL. (1994). Atmospheric mercury deposition to lakes and watersheds: A
quantitative reconstruction from multiple sediment cores. In: Baker LA (Ed.), Environmental chemistry of lakes and reservoirs, Volume
237 in Series - Advances in chemistry (pp. 33-66). Washington, DC: American Chemical Society.
Engstrom P; Dalsgaard T; Hulth S; Aller RC. (2005). Anaerobic ammonium oxidation by nitrate (anammox): Implications for N2 production in
coastal marine sediments. Geochim Cosmochim Acta, 69,2057-2065.
Epp DJ; Al-Ani KS. (1979). The effect of water quality on rural nonfarm residential property values. Am J Agric Econ, 61,529-534.
Epstein HE; Burke IC; Lauenroth WK. (2002). Regional patterns of decomposition and primary production rates in the U.S. Great Plains.
Ecology, 83, 320-327.
Erickson DJ III; Seuzaret C; Keene WC; Gong SL. (1999). A general circulation model based calculation of HCI and CINO2 production from
sea-salt dechlorination: reactive chlorine emissions inventory. J Geophys Res, 104,8347-8372.
Erickson H; Keller M; Davidson EA. (2001). Nitrogen oxide fluxes and nitrogen cycling during postagricultural succession and forest fertilization
in the humid tropics. Ecosystems, 4, 67-84.
Eriksson E. (1981). Aluminium in groundwater, possible solution equilibria. Nord Hydrol, 12,43-50.
Eriksson MOG. (1983). The role offish in the selection of lakes by non-piscivorous ducks: mallard, teal and goldeneye. Wildfowl, 34,27-32.
Eriksson MOG. (1986). Fish delivery, production of young, and nest density of Osprey (Pandion haliaetus).in southwest Sweden. Can J Zool,
64,1961-1965.
Eriksson MOG. (1987). Some effects of freshwater acidification on birds in Sweden. In: Diamond AW, Filion, FL (Eds.). The value of birds.
(ICBP Birdlife International Technical Publication Series No. 6, pp 183-190). Cambridge, UK: International Council for Bird
Preservation (ICBP).
Eshleman KN. (1988). Predicting regional episodic acidification of surface waters using empirical techniques. Water Resour Res, 24,1118-
1126.
Eshleman KN; Davies TD; Tranter M; Wigington PJ Jr. (1995). A two-component mixing model for predicting regional episodic acidification of
surface waters during spring snowmelt periods. Water Resour Res, 31 (4), 1011 -1021.
Eshleman KN; Morgan II RP; Webb JR; Deviney FA; Galloway JN. (1998). Temporal patterns of nitrogen leakage from mid-Appalachian
forested watersheds: role of insect defoliation. Water Resour Res, 34,2005-2116.
Eshleman KN; Hyer KE. (2000). Discharge and water chemistry at the three intensive sites. In: Bulger AJ; Cosby BJ; Dolloff CA; Eshelman KN;
Webb JR; Galloway JH (Eds.). Shenandoah National Park: Fish in sensitive habitats (SNP: FISH) project: An integrated assessment
offish community responses to stream acidification. Final report (Volumes 1-4); Volume II (Chapter 4, pp 51-92). Charlottesville, VA;
Department of Environmental Sciences; University of Virginia.
Eshleman KN; Gardner RH; Seagle SW; Castro NM; Fiscus DA; Webb JR; Galloway JN; Deviney FA; Herlihy AT. (2000). Effects of
disturbance on nitrogen export from forested lands of the Chesapeake Bay watershed. Environ Monit Assess, 63,187-197.
Eshleman KN; Fiscus DA; Castro NM; Webb JR; Herlihy AT. (2004). Regionalization of disturbance-induced nitrogen leakage from mid-
Appalachian forests using a linear systems model. Hydrol Process, 18,2713-2725.
Ettershank G; Ettershank J; Bryant M; Whitford WG. (1978). Effects of nitrogen fertilization on primary production in a Chihuahuan desert
ecosystem. J Arid Environ, 1,135-139.
Evans CD; Monteith DT. (2001). Chemical trends at lakes and streams in the UK acid waters monitoring network, 1988-2000: Evidence for
recent recovery at a national scale. Hydrol Earth Sys Sci, 5, 351-366.
Evans CD; Chapman PJ; Clark JM; Monteith D; Cresser MS. (2006). Alternative explanations for rising dissolved organic carbon export from
organic soils. Glob Change Biol, 12,2044-2053.
Everett KR; Brown J. (1982). Some recent trends in the physical and chemical characterization and mapping of tundra soils, arctic slope of
Alaska. Soil Sci., 133,264-280.
EverhartWH; Youngs WD. (1981). Principles of fishery science (2nd edition.). Ithaca, NY: Comstock/Cornell University Press.
Evers DC; Kaplan JD; Meyer MW; Reaman PS; Braselton WE; Major A; Burgess N; Scheuhammer AM. (1998). A geographic trend in mercury
measured in common loon feathers and blood. Environ Toxicol Chem, 17,173-183.
Evers DC; Taylor KM; Major A; Taylor RJ; Poppenga RH; Scheuhammer AM. (2003). Common loon eggs as bioindicators of methylmercury
availability in North America. Ecotoxicology, 12,69-81.
Evers DC; Clair TA. (2005). Mercury in northeastern North America: A synthesis of existing databases Ecotoxicology, 14,7-14.
Evers DC; Burgess NM; Champoux L; Hoskins B; Major A; Goodale WM; Taylor RJ; Poppenga R. (2005). Patterns and interpretation of
mercury exposure in freshwater avian communities in northeastern North America. Ecotoxicology, 14,193-222.
Evers DC; Han Y; Driscoll CT; Kamman NC; Goodale MW; Lambert KF; Holsen TM; Chen CY; Clair TA; Butler T. (2007). Biological mercury
hotspots in the northeastern United States and southeastern Canada. Bioscience, 57,29-43.
Exley C; Phillips MJ. (1988). Acid rain: Implications for the farming of salmonids. In: Muir JF, Roberts RJ (Eds.). Recent advances in
aquaculture (Volume 3, pp. 225-342). London, UK: Croom Helm.
Falkengren-Grerup U. (1986). Soil acidification and vegetation changes in deciduous forest in southern Sweden. Oecologia, 70,339-347.
Falkengren-Grerup U. (1989). Soil acidification and its impact on ground vegetation. Ambio, 18,179-183.
Falkengren-Grerup U; Eriksson H. (1990). Changes in soil, vegetation and forest yield between 1947 and 1988 in beech and oak sites of
southern Sweden. For Ecol Manage, 38,37-53.
Falkengren-Grerup U. (1998). Nitrogen response of herbs and graminoids in experiments with simulated acid soil solution. Environ Pollut, 102,
93-99.
R-27
-------
Fan Z; Chen D; Birla P; Kamens RM. (1995). Modeling of nitro-polycyclic aromatic hydrocarbon formation and decay in the atmosphere. Atmos
Environ, 29,1171-1181.
Fangmeier A; Hadwiger-Fangmeier A; Van der Eerden L; Jager HJ. (1994a). Effects of atmospheric ammonia on vegetation - A review. Environ
Pollut, 86,43-82.
Fangmeier A; Brockerhoff U; Gruters U; Jager H-J. (1994b). Growth and yield responses of spring wheat (Triticum aestivum L. cm. Turbo)
grown in open-top chambers to ozone and water stress. Environ Pollut, 83,317-325.
FAO/IFA. (2001). Global estimates of gaseous emissions of NH3, NO and N2O from agricultural land. Rome, Italy:
fftp://fto.fao.ora/agl/aaIl/docs/globest.pdf).Food and Agriculture Organization of the United Nations (FAO).and International Fertilizer
Industry Association (IFA).
Farag AM; Woodward DF; Little EE; Steadman B; Vertucci FA. (1993). The effects of low pH and elevated aluminum on Yellowstone cutthroat
trout (Oncorhynchus clarkibouvieri). Environ Toxicol Chem, 12,719-731.
Faraji M; Kimura Y; Allen D. (2005). Comparison of the carbon bond and SAPRC photochemical mechanisms under conditions relevent to
southeast Texas. (Project report H12.8HRB). http://files.harc.edu/Projecte/AirQuality/Projects/H012.2004/8HRB/H12-
8HRBMechanismReport.pdf. Houston, TX: Texas Environmental Research Consortium (TERC).
Farmer AM; Bates JW; Bell JNB. (1992). Ecophysiological effects of acid rain on bryophytes and lichens. In: Bates JW; Farmer AM (Eds.),
Bryophytes and lichens in a changing environment. Oxford, UK: Claredon Press.
Farmer DK; Wooldridge PJ; Cohen RC. (2006). Application of thermal-dissociation laser induced fluorescence (TD-LIF).to measurement of
HNO3, Zalkyl nitrates, Zperoxy nitrates, and NO2 fluxes using eddy covariance. Atmos Chem Phys, 6,3471-3486.
Farmer GJ; Beamish FWH. (1969). Oxygen consumption of Tilapia nilotica in relation to swimming speed and salinity. J Fish Res Bd Can, 26,
2807-2821.
Farwell SO; Rasmussen RA. (1976). Limitations of the FPD and ECD in atmospheric analysis: A review. J Chromatogr Sci, 14,224-234.
Fast JD; Zaveri RA; Bian RX; Chapman EG; Easter RC. (2002). Effect of regional-scale transport on oxidants in the vicinity of Philadelphia
during the 1999 NE-OPS Field Campaign. J Geophys Res (Atmos) 107(D16),m 10.1029/2001 JD000980.
Federal Register. (1993). Report of the NOAA Panel on contingent valuation. FR58:4601 -4614.
Federal Register. (1995). National ambient air quality standards for nitrogen dioxide: proposed decision. FR60:52874-52889.
Federal Register. (1996). National ambient air quality standards for nitrogen dioxide: final decision. FR61:52852-52856.
Federal Register. (2006). National ambient air quality standards for particulate matter. Final rule. FR71:61144-61233.
Federal Register. (2007). Endangered and threatened wildlife and plants; designation of critical habitat for the bay checkerspot butterfly
(,Euphydryas editha bayensis). FR50:48177-48218.
Federer CA; Hornbeck JW; Tritton LM; Martin CW; Pierce RS. (1989). Long-term depletion of calcium and other nutrients in eastern U.S.
forests. Environ Manage, 13,593-601.
Fehsenfeld FC; Dickerson RR; Hubler G; Luke WT; Nunnermacker LJ; Williams EJ; Roberts JM; Calvert JG; Curran CM; Delany AC; Eubank
CS; Fahey DW; Fried A; Gandrud BW; Langford AO; Murphy PC; Norton RB; Pickering KE; Ridley BA. (1987). A ground-based
intercomparison of NO, NOx, and NOy measurement techniques. J Geophys Res, 92,14710-14722.
Fehsenfeld FC; Trainer M; Parrish DD; Volz-Thomas A; Penkett S. (1996a). North Atlantic regional experiment (NARE) 1993 summer intensive:
Foreword. J Geophys Res, 101,28869-28875.
Fehsenfeld FC; Daum P; Leaitch WR; Trainer M; Parrish DD; Hubler, G. (1996b). Transport and processing of O3 and O3 precursors over the
North Atlantic: An overview of the 1993 North Atlantic Regional Experiment (NARE).summer intensive. J Geophys Res, 101,28877-
28891.
Fehsenfeld FC; Huey LG; Sueper DT; Norton RB; Williams EJ; Eisele FL; Mauldin RL, III; Tanner DJ. (1998). Ground-based intercomparison of
nitric acid measurement techniques. J Geophys Res, 103,3343-3353.
Fehsenfeld FC; Ancellet G; Bates TS; Goldstein AH; Hardesty RM; Honrath R; Law KS; Lewis AC; Leaitch R; McKeen S; Meagher J; Parrish
DD; Pszenny AAP; Russell PB; Schlager H; Seinfeld J; Talbot R; Zbinden R. (2006). International consortium for atmospheric
research on transport and transformation (ICARTT): North America to Europe -Overview of the 2004 summer field study. J Geophys
Res, 111, D23S01, doi:10.1029/2006JD007829
Feilberg A; Kamens RM; Strommen MR; Nielsen T. (1999). Modeling the formation, decay, and partitioning of semivolatile nitro-polycyclic
aromatic hydrocarbons (nitronaphthalenes).in the atmosphere. Atmos Environ, 33,1231-1243.
Fenn ME; Dunn PH. (1989). Litter decomposition across an air-pollution gradient in the San Bernardino Mountains. Soil Sci Soc Am J, 53,
1560-1567.
Fenn ME; Poth MA; Johnson DW. (1996). Evidence for nitrogen saturation in the San Bernardino Mountains in southern California. For Ecol
Manage, 82,211-230.
Fenn ME; Bytnerowicz A. (1997). Summer throughfall and winter deposition in the San Bernardino Mountains in southern California. Atmos
Environ, 31,673-683.
Fenn ME; Poth MA. (1998). Indicators of nitrogen status in California forests. (General technical report PSW-GTR-166). Washington, DC; U.S.
Forest Service; U.S. Department of Agriculture (USDA).
Fenn ME; Poth MA; Aber JD; Baron JS; Bormann BT; Johnson DW; Lemly AD; McNulty SG; Ryan DF; Stottlemyer R. (1998). Nitrogen excess
in North American ecosystems: Predisposing factors, ecosystem responses, and management strategies. Ecol Appl, 8,706-733.
Fenn ME; Poth MA. (1999). Temporal and spatial trends in streamwater nitrate concentrations in the San Bernardino Mountains, southern
California. J Environ Qual, 28,822-836.
Fenn ME; Poth MA; Schilling S; Grainger D. (2000). Throughfall and fog deposition of nitrogen and sulfur at an N-limited and N-saturated site in
the San Bernardino Mountains, southern California. Can J For Res, 30,1476-1488.
Fenn ME; Poth MA. (2001). A case study of nitrogen saturation in western U.S. forests. Scientific WorldJournal, 1 433-439.
R-28
-------
Fenn ME; De Bauer, LI; Hernandez-Tejeda, T. (2002a).Urban air pollution and forests: resources at risk in the Mexico City air basin. (Volume
156, Ecological studies). New York, NY; Springer-Verlag.
Fenn ME., Poth MA; Arbaugh MJ. (2002b). A throughfall collection method using mixed bed ion exchange resin columns. In: Proceedings of
the International Symposium on Passive Sampling of Gaseous Air Pollutants in Ecological Effects Research held in Riverside,
California; April 9,2001. Scientific World Journal, 2,122-130.
Fenn ME; Baron; JS, Allen, EB, Rueth, HM, Nydick KR, Geiser, L, Bowman, WD, Sickman, JO, Meixner, T, Johnson, DW, Neitlich, P. (2003a).
Ecological effects of nitrogen deposition in the western United States. Bioscience, 53,404-420.
Fenn ME; Haeuber R; Tonnesen GS; Baron JS; Grossman-Clark S; Hope D; Jaffe DA; Copeland S; Geiser L; Rueth HM; Sickman JO. (2003b).
Nitrogen emissions, deposition, and monitoring in the western United States. Bioscience, 53,391-403.
Fenn ME; Poth MA; Bytnerowicz A; Sickman JO; Takemoto BK. (2003c). Effects of ozone, nitrogen deposition, and other stressors on montane
ecosystems in the Sierra Nevada. In: Bytnerowicz A; Arbaugh MJ; Alonso R (Eds.). Ozone air pollution in the Sierra Nevada:
distribution and effects on forests. (Developments in Environmental Science: Volume 2; pp. 111-155). New York, NY: Elsevier.
Fenn ME; Poth MA. (2004). Monitoring nitrogen deposition in throughfall using ion exchange resin columns: a field test in the San Bernarndino
Mountains. J Environ Qual, 33,2007-2014.
Fenn ME; Huntington TG; McLaughlin SB; Eagar C; Gomez A; Cook RB. (2006). Status of soil acidification in North America. J For Sci, 52,3-
13.
Fenn ME; Geiser L; Bachman R; Blubaugh TJ; Bytnerowicz A. (2007). Atmospheric deposition inputs and effects on lichen chemistry and
indicator species in the Columbia River Gorge, USA. Environ Pollut, 146, 77-91.
Fenn ME; Jovan S; Yuan F; Geiser L; Meixner T; Gimeno BS. (2008). Empirical and simulated critical loads for nitrogen deposition in California
mixed conifer forests. Environ Pollut, 155,492-511.
Fernandez IJ; Rustad LE; Norton SA; Kahl JS; Cosby BJ. (2003). Experimental acidification causes soil base-cation depletion at the Bear
Brook watershed in Maine. Soil Sci Soc Am J, 67,1909-1919.
Ferrari JB. (1999). Fine-scale patterns of leaf litterfall and nitrogen cycling in an old-growth forest. Can. J Forest Res, 29, 291-302.
Ferrier RC; Helliwell RC; Cosby BJ; Jenkins A; Wright RF. (2001). Recovery from acidification of lochs in Galloway, south-west Scotland, UK:
1979-1998. Hydrol Earth Sys Sci, 5,421-431.
Field CB; Mooney HA. (1986). The photosynthesis-nitrogen relationship in wild plants. In: Givinish TJ (Ed.), The economy of plant form and
function, (pp 25-55). Cambridge, UK: Cambridge University Press.
Field C. B; Mortsch, L. D; Brklacich, M; Forbes, D. L; Kovacs, P; Patz, J. A; Running, S. W; Scott, M. J. (2007).North America. In: Climate
change 2007: impacts, adaptation and vulnerability. (Contribution of Working Group 11 to the fourth assessment report of the
Intergovernmental Panel on Climate Change; pp. 617-852).http://www.ipcc.ch/pdf/assessment-report/ar4/wa2/ar4-wa2-chapter14.pdf.
Cambridge, UK: Cambridge University Press.
Fields RF. (1988). Physiological responses of lichens to air pollutant fumigations. In: Nash TH III; Wirth V (Eds.), Lichens, bryophytes and air
quality. Volume 30; Bibl. Lichenol. (pp 175-200). Berlin/ Stuttgart, Germany: Cramer Publisher.
Findlay DL; Kasian SEM. (1996). The effect of incremental pH recovery on the Lake 223 phytoplankton community. Can J Fish Aquat Sci, 53,
856-864.
Findlay DL; Hecky RE; Kasian SEM; Stainton MP; Hendzel LL; Schindler EU. (1999). Effects on phytoplankton of nutrients added in
conjunction with acidification. FreshwBiol, 41,131-145.
Findlay DL. (2003). Response of phytoplankton communities to acidification and recovery in Killarney Park and the Experimental Lakes Area,
Ontario. Ambio, 32,190-195.
Findlay SEG. (2005). Increased carbon transport in the Hudson River: Unexpected consequence of nitrogen deposition? Front Ecol Environ, 3,
133-137.
Finlayson-Pitts BJ; Ezell MJ; Pitts JN Jr. (1989). Formation of chemically active chlorine compounds by reactions of atmospheric NaCI particles
with gaseous N2O5 and CIONO2. Nature, 337,241-244.
Finlayson-Pitts BJ; Pitts JN Jr. (2000). Chemistry of the upper and lower atmosphere: theory, experiments and applications. San Diego, CA:
Academic Press.
Finlayson-Pitts BJ; Wingen LM; Sumner AL; Syomin D; Ramazan KA. (2003). The heterogeneous hydrolysis of NO2 in laboratory system and in
outdoor and indoor atmospheres: An integrated mechanism. Phys Chem Chem Phys, 5,223-242.
Finley BD; Saltzman ES. (2006). Measurement of Cb in coastal urban air. Geophys Res Lett, 33, L11809, doi:10.1029/2006GL025799.
Finley MT; Stendell RC. (1978). Survival and reproductive success of black ducks fed methyl mercury. Environ Pollut, 16,51-64.
Finley MT; Stickel WH; Christensen RE. (1979). Mercury residues in tissues of dead and surviving birds fed methylmercury. Bull Environ
Contam Toxicol, 21,105-110.
Finzi AC; Van Breemen N; Canham CD. (1998). Canopy tree-soil interactions within temperate forests: Species effects on soil carbon and
nitrogen. Ecol Appl, 8,440-446.
Fiore AM; Jacob DJ; Field BD; Streets DG; Fernandes SD; Jang C. (2002). Linking ozone pollution and climate change: the case for controlling
methane. Geophys Res Lett, 29(19), 10.1029/2002GL015601.
Fiore A; Jacob DJ; Liu H; Yantosca RM; Fairlie TD; Li Q. (2003). Variability in surface ozone background over the United States: implications
for air quality policy. J Geophys Res, 108,4787, doi: 10.1029/2003JD003855 (Correction Published 2004).
Firestone MK; Davidson EA. (1989). Microbiological basis of NO and N2O production and consumption in soil. In: Andreae MO; Schimel DS
(Eds.), Exchange of trace gases between terrestrial ecosystems and the atmosphere: Report of the Dahlem workshop on exchange of
frace gases between terrestrial ecosystems and the atmosphere (pp. 7-21). Berlin, New York, NY: Wiley.
Fischer E; Pszenny A; Keene W; Maben J; Smith A; Stohl A; Talbot R. (2006). Nitric acid phase partitioning and cycling in the New England
coastal atmosphere. J Geophys Res, 111, D23S09, doi:10.1029/2006JD007328.
R-29
-------
Fisher DC; Oppenheimer M. (1991). Atmospheric nitrogen deposition and the Chesapeake Bay estuary. Ambio, 20,102-108.
Fisher TR; Harding LW Jr; Stanley DW; Ward LG. (1988). Phytoplankton, nutrients, and turbidity in the Chesapeake, Delaware, and Hudson
estuaries. Estuarine Coastal Shelf Sci, 27,61-93.
Fisher TR; Lee K-Y; Berndy H; Benitez JA; Norton MM. (1998). Hydrology and chemistry of the Choptank River basin in the Chesapeake Bay
drainage. Water Air Soil Pollut, 105, 387-397.
Fisher TR; Hagy III JD; Boynton WR; Williams MR. (2006). Cultural eutrophication in the Choptank and Patuxent estuaries of Chesapeake Bay.
Limnol Oceanogr, 51,435-447.
Fishman J; Larsen JC. (1987). Distribution of total ozone and stratospheric ozone in the tropics-Implications for the distribution of tropospheric
ozone. J Geophys Res, 92,6627-6634.
Fishman J; Creilson JK; Wozniak AE; Crutzen PJ. (2005).The interannual variability of stratospheric and tropospheric ozone determined from
satellite measurements. J Geophys Res, 110, doi: 10.1029/2005JD005868.
Fisk MC; Schmidt SK. (1996). Microbial responses to nitrogen additions in alpine tundra soil. Soil Biol Biogeochem, 28,751-755.
Fisk MC; Schmidt SK; SeastedtTR. (1998). Topographic patterns of above-and belowground production and nitrogen cycling in alpine tundra.
Ecology, 79, 2253-2266.
Fisk MC; Zak DR; Crow TR. (2002). Nitrogen storage and cycling in old- and second-growth northern hardwood forests. Ecology, 83,73-87.
Fitzhugh RD; Lovett GM; Venterea RT. (2003). Biotic and abiotic immobilization of ammonium, nitrite, and nitrate in soils developed under
different tree species in the Catskill Mountains, New York, USA. Glob Chang Biol, 9,1591-1601.
Fjeld E; Haugen TO; V0llestad LA. (1998). Permanent impairment in the feeding behavior of grayling (Thymallus thymallus).exposed to
methylmercury during embryogenesis. Sci Total Environ, 213,247-254.
Fjellheim A; Raddum GG; Sagen T. (1985). Effect of aluminum at low pH on the mortality of elvers (Anguilla anguilla L). J Exp Zool, 262,
247-254.
Flower RJ; Battarbee RW. (1983). Diatom evidence for recent acidification of two Scottish Lochs. Nature, 305,130-133.
Flowers PJ; Vaux HJ; Gardner PD; Mills TJ. (1985). Changes in recreation values after fire in the northern Rocky Mountains (Research note
PSW-373). Berkeley, CA; Pacific Southwest Forest and Range Experiment Station; U.S. Forest Service; U.S. Department of
Agriculture (USDA).
Flum T; Nodvin SC. (1995). Factors affecting streamwater chemistry in the Great Smoky Mountains, USA. Water Air Soil Pollut, 85,1707-1712.
Ford J; Stoddard JL; Powers CF. (1993). Perspectives in environmental monitoring: an introduction to the U.S. EPA long-term monitonring
(LTM)project. Water Air Soil Pollut, 67,247-255.
Fore LS; Karr JR; Wisseman RW. (1996). Assessing invertebrate responses to human activities: evaluating alternative approaches. J North Am
Benthol Soc, 15,212-231.
Forkner RE; Hunter MD. (2000). What goes up must come down? Nutrient addition and predation pressure on oak herbivores. Ecology, 81,
1588-1600.
Forster P; Ramaswamy V; Artaxo P; Betss R; Fahey DW; Haywood J; Lean J; Lowe DC; Myhre G; Nganga J; Prinn R; Raga G; Schultz M; Van
Dorland R. (2007). Changes in Atmospheric constituents and in radiative Forcing. In: Solomon S, Qin D, Manning M, Chen Z, Marquis
M, Averyt KB, Tignor M & Miller HL (Eds.). Climate Change 2007: The Physical Science Basis. (Contribution of Working Group I to the
Fourth Assessment Report of the Intergovernmetnal Panel on Climate Change (IPCC).
Foster KL; Plastridge RA; Bottenheim JW; Shepson PB; Finlayson-Pitts BJ; Spicer CW. (2001). The role of Br2 and BrCI in surface ozone
destruction at polar sunrise. Science, 291,471-474.
Foster NW; Nicolson JA; Hazlett PW. (1989). Temporal variation in nitrate and nutrient cations in drainage waters from a deciduous forest. J
Environ Qual, 18,238-244.
Foster NW; Mitchell M J; Morrison IK; Shepard JP. (1992). Cycling of acid and base cations in deciduous stands of Huntington Forest, New
York, and Turkey Lakes, Ontario. Can J For Res, 22,167-174.
Fowler D; Cape JN; Unsworth MH. (1989). Deposition of atmospheric pollutants on forests. Philosophical Transactions of the Royal Society of
London B 324: 247-265.
Fowler D; Duyzer JH; Baldocchi DD. (1991). Inputs of trace gases, particles and cloud droplets to terrestrial surfaces. Proceedings of the Royal
Society of Edinburgh Sect B: Biol Sci 97:35-59.
Fowler D; MacDonald J; Leith ID; Hargreaves J; Martynoga R. (1995). The response of peat wetland methane emissions to temperature, water
table and sulphate deposition. In: Heij GJ; Erisman JW (Eds.), Acid rain research: Do we have enough answers? (pp. 485-487). New
York: Elsevier Science.
Fox GA. (1991). Practical causal inference for ecoepidemiologists. J Toxicol Environ Health, 33,359-373.
Fox-Rabinovitz MS; Takacs LL; Govindaraju RC. (2002). A variable-resolution stretched-grid general circulation model and data assimilation
system with multiple areas of interest: Studying the anomalous regional climate events of 1998. J Geophys Res, 107,4768,
doi:10.1029/2002JD002177
Fraser MP; Cass GR; Simoneit BRT; Rasmussen RA. (1998). Air quality model evaluation data for organics. 5: C6-C22 nonpolar and semipolar
aromatic compounds. Environ Sci Technol, 32,1760-1770.
Fraterrigo JM; Turner MG; Pearson SM; Dixon P. (2005). Effects of past land use on spatial heterogeneity of soil nutrients in southern
Appalachian forests. Ecol Monogr, 75,215-230.
Freda J; McDonald DG. (1988). Physiologic correlates of interspecific variation in acid tolerance in fish. J Exp Biol, 136,243-258.
Frederick PC; Spalding MG; Dusek R. (2002). Wading birds as bioindicators of mercury contamination in Florida, USA: Annual and geographic
variation. Environ Toxicol Chem, 21,163-167.
Freeman MA. Ill (2003). The measurement of environmental and resource values: Theory and methods. 2nd Edition. Washington, D.C: RFF
Press.
R-30
-------
Fremstad E; Paal J; Mols T. (2005). Impacts of increased nitrogen supply on Norwegian lichen-rich alpine communities: A10 year experiment.
J Ecol, 93,471-481.
Frenette JJ; Richard Y; Moreau G. (1986). Fish responses to acidity in Quebec lakes: A review. Water Air Soil Pollut, 30,461-475.
Frey SD; Knorr M; Parrent JL; Simpson RT. (2004). Chronic nitrogen enrichment affects the structure and function of the soil microbial
community in temperate hardwood and pine forests. For Ecol Manage, 196,159-171.
Friedmann AS; Watzin MC; Brinck-Johnsen T; Leiter JC. (1996). Low levels of dietary methylmercury inhibit growth and gonadal development
in juvenile walleye. Aquat Toxicol, 35,265-278.
Friend A; Stevens A; Knox R; Cannell M. (1997). A process-based, terrestrial biosphere model of ecosystem dynamics (Hybrid v3.0). Ecol
Model, 95, 249-287.
Frost GJ; McKeen SA; Trainer M; Ryerson TB; Neuman JA; Roberts JM; Swanson A; Holloway JS; Sueper DT; Fortin T; Parrish DD;
Fehsenfeld FC; Flocke F; Peckham SE; Grell GA; Kowal D; Cartwright J; Auerbach N; Habermann T. (2006a). Effects of changing
power plant NOx emissions on ozone in the eastern United States: Proof of concept. J Geophys Res, 111.
Frost PC; Stelzer RS; Lamberti GA; Elser JJ. (2002). Ecological stoichiometry of trophic interactions in the benthos: Understanding the role of
C:N:P ratios in lentic and lotic habitats. J N Am Benthol Soc, 21,515-528.
Frost TM; Montz PK; Kratz TK; Badillo T; Brezonik PL; Gonzalez MJ; Rada RG; WatrasCJ; Webster KE; Wiener JG; Williamson CE; Morris
DP. (1999). Multiple stresses from a single agent: Diverse responses to the experimental acidification of Little Rock Lake, Wisconsin.
Limnol Oceanogr, 44,784-794.
Frost TM; Fischer JM; Klug JL; Arnott SE; Montz PK. (2006b). Trajectories of zooplankton recovery in the Little Rock Lake whole-lake
acidification experiment. Ecol Appl, 16,353-367.
Fu T-M; Jacob DJ; Palmer PI; Chance K; Wang YX; Barletta B; Blake DR; Stanton JC; Pilling MJ. (2006). Space-based formaldehyde
measurements as constraints on volatile organic compound emissions in east and south Asia, and implications for ozone. J Geophys
Res, 112, D06312, doi:10.1029/2006JD007853
Fuentes M; Raftery AE. (2005). Model evaluation and spatial interpolation by Bayesian combination of observations with outputs from
numerical models. Biometrics, 61,36-45.
Fuller RD; Driscoll CT; Lawrence GB; Nodvin SC. (1987). Processes regulating sulphate flux after whole-tree harvesting. Nature, 325,707-710.
Funk JL; Giardina CP; Knohl A; Lerdau MT. (2006). Influence of nutrient availability, stand age, and canopy structure on isoprene flux in a
Eucalyptus saligna experimental forest. J Geophys Res, 111, doi:10.1029/2005JG000085.
Furutani H; Akimoto H. (2002). Development and characterization of a fast measurement system for gas-phase nitric acid with a chemical
ionization mass spectrometer in the marine boundary layer. J Geophys Res, 107, doi 10.1029/2000JD000269.
Gaffney JS; Bornick RM; Chen Y-H; Marley NA. (1998). Capillary gas chromatographic analysis of nitrogen dioxide and pans with luminol
chemiluminescent detection. Atmos Environ, 32,1445-1454.
Gagen CJ; Sharpe WE; Carline RF. (1993). Mortality of brook trout, mottled sculpins, and slimy sculpins during acidic episodes. Trans Am Fish
Soc, 122,616-628.
Gaio-Oliveira G; Dahlman L; Palmqvist K; Martins-Lougao MA; Maguas C. (2005). Nitrogen uptake in relation to excess supply and its effects
on the lichens Evernia prunastri (L.)Ach and Xanthoria parietina (L ). Th. Fr. Planta, 220,794-803.
Galbally IE; Bentley ST; Meyer CP. (2000). Mid-latitude marine boundary-layer ozone destruction at visible sunrise observed at Cape Grim,
Tasmania, 41degrees S. Geophys Res Lett, 27,3841-3844.
Gallagher J; Baker JL. (1990). Current status offish communities in Adirondack lakes. In: Adirondack Lakes Survey: An interpretative analysis
offish communities and water chemistry, 1984-87. (pp. 3-48). Ray Brook, NY: Adirondack Lake Survey Corporation.
Galloway JN; Norton SA; Church MR. (1983). Freshwater acidification from atmospheric deposition of sulfuric acid: A conceptual model.
Environ Sci Technol, 17, 541A-545A.
Galloway JN; Schlesinger WH; Levy H; Michaels AF; Schnoor JL. (1995). Nitrogen-fixation: anthropogenic enhancement, environmental
response. Glob Biogeochem Cycles, 9,235-252.
Galloway J. (1995). Acid deposition: Perspectives in time and space. Water Air Soil Pollut, 85,15-24.
Galloway JN. (1996). Anthropogenic mobilization of sulphur and nitrogen: Immediate and delayed consequences. Ann Rev Energy Environ, 21,
261-292.
Galloway JN. (1998). The global nitrogen cycle: Changes and consequences. Environ Pollut, 102,15-24.
Galloway JN; Cowling EB. (2002). Reactive nitrogen and the world: 200 years of change. Ambio, 31, 64-71.
Galloway JN; Aber JD; Erisman JW; Seitzinger SP; Howarth RW; Cowling EB; Cosby J. (2003). The nitrogen cascade. Bioscience, 53,341 -
356.
Gao W; Wesely ML; Lee IY. (1991). A numerical study of the effects of air chemistry on fluxes of NO, NO2, and O3 near the surface. J Geophys
Res, 96,18761-18769.
Gao W; Wesely ML; Doskey PV. (1993). Numerical modeling of the turbulent diffusion and chemistry of NOx, O3, isoprene, and other reactive
trace gases in and above a forest canopy. J Geophys Res, 98,18339-18353.
Garcia E; Carignan R. (2000).Mercury concentrations in northern pike (ESOx /uc/'us).from boreal lakes with logged, burned, or undisturbed
catchments. Can J Fish Aquat Sci, 57,129-135.
Garner JHB; Pagano T; Cowling EB. (1989). An evaluation of the role of ozone, acid deposition, and other airborne pollutants in the forests of
eastern North America (No. General Technical Report SE-59). Asheville, NC: U.S. Department of Agriculture, Forest Service,
Southeastern Forest Experiment Station.
R-31
-------
Garner JHB. (1994). Nitrogen oxides, plant metabolism, and forest ecosystem response. In: Alscher RG; Wellburn A (Eds.). Plant responses to
the gaseous environment: Molecular, metabolic, and physiological aspects, (Proceedings from the 3rd International symposium on air
pollutants and plant metabolism, held in Blacksburg.VA at Virginia Polytechnic Institute inJune 1992; pp 301-314). Norwell, MA:
Kluwer Academic Publishers.
Garten CT, Jr.; Hanson PJ. (1990). Foliar retention of 15N-nitrate and 15N-ammonium by red maple (Acer rubrum) and white oak (Quercus
alba) leaves from simulated rain. Environmental and Experimental Botany 30:333-342.
Gates DM. (1968). Transpiration and leaf temperature. Annu Rev Plant Biol, 19,211-238.
Gauci V; Dise N; Fowler D. (2002). Controls on suppression of methance flux from a peat bog subjected to simulated acid rain sulfate
deposition. Glob Biogeochem Cycles, 16,4-1 -4-12. doi: 10.1029/2000GB001370
Gauci V; Fowler D; Chapman SJ; Dise NB. (2004). Sulfate deposition and temperature controls on methane emission and sulfur forms in peat.
Biogeochemistry, 71,141-162.
Gauslaa Y. (1995). Lobarion, an epiphytic community of ancient forests, threatened by acid rain. Lichenologist, 27,59-76.
Gbondo-Tugbawa SS; Driscoll CT; Aber JD; Likens GE. (2001). Evaluation of an integrated biogeochemical model (PnET-BGC) at a northern
hardwood forest ecosystem. Water Resour Res, 37,1057-1070.
Gbondo-Tugbawa SS; Driscoll CT. (2002a). Evaluation of the effects of future controls on sulfur dioxide and nitrogen oxide emissions on the
acid-base status of a northern forest ecosystem. Atmos Environ, 36,1631-1643.
Gbondo-Tugbawa SS; Driscoll CT; Mitchell MJ; Aber JD; Likens GE. (2002a). A model to simulate the response of a northern hardwood forest
ecosystem to changes in S deposition. Ecol Appl, 12, 8-23.
Gbondo-Tugbawa SS; Driscoll CT. (2003). Factors controlling long-term changes in soil pools of exchangeable basic cations and stream acid
neutralizing capacity on a northern hardwood forest ecosystem. Biogeochem, 63,161-185.
Gebauer RLE; Reynolds JF; Tenhunen JD. (1995). Growth and allocation of the arctic sedges Eriophorum angustifblium and E. vaginatum'.
effects of variable soil oxygen and nutrient availability. Oecologia, 104,330-339.
Geiser LH; Neitlich PN. (2007). Air pollution and climate gradients in western Oregon and Washington indicated by epiphytic macrolichens.
Environ Pollut, 145,203-218.
GenfaZ; Dasgupta PK. (1989). Fluorometric measurement of aqueous ammonium ion in a flow injection system. Anal Chem, 61,408-412.
Gensemer RW; Playle RC. (1999). The bioavailability and toxicity of aluminum in aquatic environments. Crit Rev Environ Sci Technol, 29, 315-
450.
George C; Strekowski RS; Kleffmann J; Stemmler K; Ammann M. (2005). Photoenhanced uptake of gaseous NO2 on solid organic compounds:
A photochemical source of HONO? Faraday Discuss, 130,195-210.
Gerlach TM. (2004).Volcanic sources of tropospheric ozone-depleting trace gases. Geochem Geophys Geosys, 5, doi:
10.1029/2004GC000747.
Gerritsen J; Carlson RE; Dycus DL; Faulkner C; Gibson GR. Harcum J; Markowitz SA. (1998). Lake and reservoir bioassessment and
biocriteria: Technical guidance document. (Report no EPA 841-B-98-007). http://www.epa.aov/owow/monitorina/tech/lakes.html.
Washington, DC; Office of Water; U.S. Environmental Protection Agency.
Gery MW; Whitten GZ; Killus JP; Dodge MC. (1989). A photochemical kinetics mechanism for urban and regional scale computer modeling. J
Geophys Res, 94,12925-12956.
Geyer A; Alicke B; Konrad S; Schmitz T; Stutz J; Piatt U. (2001). Chemistry and oxidation capacity of the nitrate radical in the continental
boundary layer near Berlin. J Geophys Res, 106,8013-8025.
Geyer A; Piatt U. (2002). Temperature dependence of the NO3 loss frequency: A new indicator for the contribution of NO3 to the oxidation of
monoterpenes and NOx removal in the atmosphere. J Geophys Res, 107,4431, doi:10.1029/2001JD001215.
Gherini SA; Mok L; Hudson RJ; Davis GF; Chen CW; Goldstein RA. (1985). The ILWAS model: formulation and application. Water Air Soil
Pollut, 26,425-459.
Giardina CP; Binkley D; Ryan MG; Fownes JH; Senock RS. (2004). Belowground carbon cycling in a humid tropical forest decreases with
fertilization. Oecologia, 139,545-550.
Gibson TL. (1983). Sources of direct-acting nitroarene mutagens in airborne particulate matter. Mutat Res, 122,115-121.
Gidel LT. (1983). Cumulus cloud transport of transient tracers. J Geophys Res, 88,6587-6599.
Gilliam FS; Adams MB; Yurish BM. (1996). Ecosystem nutrient responses to chronic nutrient inputs at Fernow Experimental Forest, West
Virginia. Can J For Res, 26,196-205.
Gilliam FS; Yurish BM; Adams MB. (2001). Temporal and spatial variation of nitrogen transformations in nitrogen-saturated soils of a central
Appalachian hardwood forest. Can J For Res, 31,1768-1785.
Gilliam FS. (2006). Response of the herbaceous layer of forest ecosystems to excess nitrogen deposition. J Ecol, 94,1176-1191.
Gilliam FS; Hockenberry AW; Adams MB. (2006). Effects of atmospheric deposition on the herbaceous layer of a central Appalachian
hardwood forest. J Torrey Bot Soc, 133,240-254.
Gilliland A. (2001). A sensitivity study of the discrete Kalman filter (DKF).to initial condition discrepancies. J Geophys Res, 106,17939-17952.
Gilliland AB; Dennis RL; Roselle SJ; Pierce TE; Bender LE. (2001). Developing seasonal ammonia emission estimates with an inverse
modeling technique. Scientific WorldJournal, 1,356-362.
Gilliland AB; Dennis RL; Roselle SJ; Pierce TE. (2003). Seasonal NH3 emission estimates for the eastern United States based on ammonium
wet concentrations and an inverse modeling method. J Geophys Res, 108,4477, doi:10.1029/2002JD003063.
Gilmour CC; Henry EA. (1991). Mercury methylation in aquatic systems affected by acid deposition. Environ Pollut, 71,131-169.
Gilmour CC; Henry EA; Mitchell R. (1992). Sulfate stimulation of mercury methylation in freshwater sediments. Environ Sci Technol, 26,2281-
2287.
R-32
-------
Gilmour CC; Riedel GS; Ederington MC; Bell JT; Gill GA; Stordal MC. (1998). Methylmercury concentrations and production rates across a
trophic gradient in the northern Everglades. Biogeochemistry, 40,327-345.
Gipson GL. (1999). Process analysis (Chapter 16). In: Byun DW; Ching JKS. (Eds.). Science algorithms in models-3 community multiscale air
quality modeling system (CMAQ). (EPA-600/R-99/030, pp 16-1 -16-20), Research Triangle Park, NC; U.S. Environmental Protection
Agency.
Gipson GL; Young JO. (1999). Gas phase chemistry (Chapter 8). In: Byun DW; Ching JKS. (Eds.). Science algorithms in models-3 community
multiscale air quality modeling system (CMAQ). (EPA-600/R-99/030, pp 8-1 - 8-99), Research Triangle Park, NC; U.S. Environmental
Protection Agency.
Gitay H; Brown S; Easterling W; Jallow B (2001). Ecosystems and their goods and services.
http://www.arida.no/publications/other/ipcc tar/?src=/climate/ipcc tar/wa2/196.htm In: McCarthy JJ; Osvaldo FC; Leary NA; Dokken
DJ; White KS (Eds.). Climate change 2001: Impacts, adaptation, and vulnerability: Contribution of working group II to the third
assessment report of the intergovernmental panel on climate change (IPCC). (pp 237-342). New York, NY: Cambridge University
Press
Glooschenko V; Blancher P; Herskowitz J; Fulthorpe R; Rang S. (1986). Association of wetland acidity with reproductive parameters and insect
prey of the eastern kingbird (Tyrannus tyrannus) near Sudbury, Ontario. Water Air Soil Pollut, 30, 553-567.
Gobler CJ; Davis TW; Coyne KJ; Boyer GL. (2007). Interactive influences of nutrient loading, zooplankton grazing, and microcystin synthetase
gene expression on cyanobacterial bloom dynamics in a eutrophic New York lake. Harmful Algae, 6,119-133.
Gochfeld M. (1980). Tissue distribution of mercury in normal and abnormal young common terns. Mar Pollut Bull, 11,362-366.
Godbold DL; Fritz E; Huttermann A. (1988). Aluminum toxicity and forest decline. Proc Natl Acad Sci USA, 85, 3888-3892.
Goebes MD; Strader R; Davidson C. (2006). An ammonia emission inventory for fertilizer application in the United States. Atmos Environ, 37.
2539-2550.
Goede RW; Barton BA. (1990). Organismic indices and an autopsy-based assessment as indicators of health and condition offish. Am Fish
Soc Symp, 8,80-93.
Goldan PD; Kuster WC; Albritton DL; Fehsenfeld FC; Connell PS; Norton RB; Huebert BJ. (1983). Calibration and tests of the filter-collection
method for measuring clean-air, ambient levels of nitric acid. Atmos Environ, 17,1355-1364.
Goldan PD; Trainer M; Kuster WC; Parrish DD; Carpenter J; Roberts JM; Yee JE; Fehsenfeld FC. (1995). Measurements of hydrocarbons,
oxygenated hydrocarbons, carbon monoxide, and nitrogen oxides in an urban basin in Colorado: Implications for emission inventories.
J Geophys Res, 100,22771-22783
Goldan PD; Kuster WC; Fehsenfeld FC. (1997). Nonmethane hydrocarbon measurements during the tropospheric OH photochemistry
experiment. J Geophys Res, 102,6315-6324.
Goldan PD; Parrish DD; Kuster WC; Trainer M; McKeen SA; Holloway J; Jobson BT; Sueper DT; Fehsenfeld FC. (2000). Airborne
measurements of isoprene, CO, and anthropogenic hydrocarbons and their implications. J Geophys Res, 105,9091 -9105.
Golden DM; Smith GP. (2000). Reaction of OH + NO2 + M: A New View. J PhysChem A, 104, 3991-3997
Goldman CR. (1988). Primary productivity, nutrients, and transparency during the early onset of eutrophication in ultra-oligotrophic Lake Tahoe,
California-Nevada. Limnol Oceanogr, 33,1321-1333.
Goldstein RA; Gherini SA; Chen CW; Mok L; Hudson RJM; Kramer JR; Brown DJA; Chester PF. (1984). Integrated acidification study (ILWAS):
A mechanistic ecosystem analysis. Philos Trans R Soc Lond B Biol Sci, 305,409-425.
Goldstein AH; Schade GW. (2000). Quantifying biogenic and anthropogenic contributions to acetone mixing ratios in a rural environment.
Atmos Environ, 34,4997-5006.
Gondal MA. (1997). Laser photoacoustic spectrometer for remote monitoring of atmospheric pollutants. Laser photoacoustic spectrometer for
remote monitoring of atmospheric pollutants. Appl Opt, 36,3195-3201.
Gondal MA; Mastromarino J. (2001). Pulsed laser photoacoustic detection of SO2 near 225.7 nm. Appl Opt, 40,2010-2016.
Goodale CL; Aber JD; McDowell WH. (2000). The long-term effects of disturbance on organic and inorganic nitrogen export in the White
Mountains, New Hampshire. Ecosystems, 3,433-450.
Goodale CL; Aber JD. (2001). The long-term effects of land-use history on nitrogen cycling in northern hardwood forests. Ecol Appl, 11,253-
267.
Goodale C; Lajtha K; Nadelhoffer KJ; Boyer EW; Jaworski NA. (2002). Forest nitrogen sinks in large eastern U.S. watersheds: estimates from
forest inventory and an ecosystem model. Biogeochemistry, 57/58,239-266.
Goodale CL; Aber JD; Vitousek PM. (2003). An unexpected nitrate decline in New Hampshire streams. Ecosystems, 6, 75-86.
Goodale CL; Aber JD; Vitousek PM; McDowell WH. (2005). Long-term decreases in stream nitrate: successional causes unlikely; possible links
to DOC? Ecosystems, 8, 334-337.
Goransson A; EldhusetTD. (1987). Effects of aluminum on growth and nutrient uptake of Betula pendula seedlings. Physiol Plant, 69,193-199.
Gordon C; Wynn JM; Woodin SJ. (2001). Impacts of increased nitrogen supply on high Arctic heath: The importance of bryophytes and
phosphorus availability. New Phytol, 149,461-471.
Gorham E; Janssens JA; Wheeler GA; Glaser PH. (1987). Natural and anthropogenic acidification of peatlands. In: Hutchinson TC; Meema KM
(Eds.). Effects of atmospheric pollutants on forests, wetlands, and agricultural ecosystems. (Proceedings from the NATO advanced
research workshop on effects of acid deposition of forests, wetlands and agricultural ecosystems (1985, Toronto). (NATO ASI series
G: Ecololgical sciences pp. 493-512). New York: Springer-Verlag
Goriup PD. (1989). Acidic air pollution and birds in Europe. Oryx, 23,82-86.
Gotelli NJ; Ellison AM. (2002). Nitrogen deposition and extinction risk in the northern pitcher plant, Sarraceniapurpurea. Ecology, 83,2758-
2765.
R-33
-------
Gotelli NJ; Ellison AM. (2006). Food-web models predict species abundances in response to habitat change. PLoS Biol, 4, e324,
doi:10.1371/journal.pbio.0040324.
Goulder LH; Kennedy D. (1997). Valuing ecosystem services: Philosophical bases and empirical methods. In: Daily GC (Ed.). Nature's
services: Societal Ddpendence on natural ecosystems, (pp 23-47). Washington, DC: Island Press.
Graetz DA; Pollman CD; Roof B; Will E. (1985). Effects of acidic treatments on soil chemistry and microbiology. In: Florida acid deposition
study. Phase IV report. (Vol. I. ESE No. 83-152-0106/0207/0307; pp. 4-5 - 4-123). Gainesville, FL: Environmental Science and
Engineering, Inc
Gramlich FW. (1977). The demand for clean water: The case of the Charles River. Natl Tax J, 30,183-194.
Gran G. (1952). Determination of the equivalence point in potentiometric titrations. Analyst, 77, 661-671.
Granberg G; Sundh I; Svensson BH; Nilsson M. (2001). Effects of temperature and nitrogen and sulfur deposition, on methane emission from a
boreal mire. Ecology, 82,1982-1998.
Grant EHC; Jung RE; Rice KC. (2005). Stream salamander species richness and abundance in relation to environmental factors in
Shenandoah National Park, Virginia. Am Midi Nat, 153,348-356.
Green PA; Vorosmarty CJ; Meybeck M; Galloway JN; Peterson BJ; Boyer EW. (2004). Pre-industrial and contemporary fluxes of nitrogen
through rivers: A global assessment based on typology. Biogeochemistry, 68, 71-105.
Greenhut GK; Ching JKS; Pearson R Jr; Repoff TP. (1984). Transport of ozone by turbulence and clouds in an urban boundary layer. J
Geophys Res, 89,4757-4766.
Greenhut GK. (1986). Transport of ozone between boundary layer and cloud layer by cumulus clouds. J Geophys Res, 91,8613-8622.
Gregory GL; Hoell JM Jr; Torres AL; Carroll MA; Ridley BA; Rodgers MO; Bradshaw J; Sandholm S; Davis DD. (1990). An intercomparison of
airborne nitric oxide measurements: A second opportunity. J Geophys Res, 95,10129-10138.
Grell GA; Dudhia J; Stauffer DR. (1995). A description of the fifth-generation Penn State/NCAR mesoscale model (MM5). [NCAR technical note
NCAR/TN-398 + STR], Boulder, CO; Mesoscale and Microscale meteorology Division, National Center for Atmospheric Research; and
Forcast Systems Laboratory, National Oceanic and Atmospheric Administration (NOAA).
Grell GA; Emeis S; Stockwell WR; Schoenemeyer T; Forkel R; Michalakes J; Knoche R; Seidl W. (2000). Application of a multiscale, coupled
MM5/chemistry model to the complex terrain of the VOTALP valley campaign. Atmos Environ, 34,1435-1453.
Grenon F; Bradley RL; Joanisse G; Titus BD; Prescott CE. (2004). Mineral N availability for conifer growth following clearcutting: responsive
versus non-responsive ecosystems. For Ecol Manage, 188,305-316.
Grieb TM; Driscoll CT; Gloss SP; Schofield CL; Bowie GL; Porcella DB. (1990). Factors affecting mercury accumulation in fish in the Upper
Michigan Peninsula. Environ Toxicol Chem, 9, 919-930.
Grier CC (1975). Wildfire effects on nutrient distribution and leaching in a coniferous ecosystem. Can J For Res, 5,599-607.
Griffin K; Bashkin M; Thomas R; Strain B. (1997). Interactive effects of soil nitrogen and atmospheric carbon dioxide on root/rhizosphere
carbon dioxide efflux from loblolly and Ponderosa pine seedlings. Plant Soil, 190,11 -18.
Griffiths RW; Keller W. (1992). Benthic macroinvertebrate changes in lakes near Sudbury, Ontario following a reduction in acid emissions. Can
J Fish Aquat Sci, 49, 63-75.
Grill, D. (1971). Pufferkapazitat gesunder und rauchgeschadigter Fichtennadeln. Z. Pflanzenkr Pflanzenschutz, 78,612-622.
Grimm NB; Fisher SG. (1986). Nitrogen limitation in a Sonoran Desert stream. J North Am Benthol Soc, 5,2-15.
Grippo MA; Heath AG. (2003). The effect of mercury on the feeding behavior of fathead minnows (Pimephales promelas). Ecotoxicol Environ
Saf, 55,187-198.
Groffman PM. (1994). Denitrification in freshwater wetlands. Curr Topics Wetland Biogeochem, 1,15-35.
Groffman PM; Altabet AM; Bohlke JK; Butterbach-Bahl K; David MB; Firestone MK; Giblin AE; Kana TM; Nielsen LP; Voytek MA. (2006).
Methods for measuring denitrification: diverse approaches to a difficult problem. Ecol Appl, 16, 2091-2122.
Grosjean D; Fung K; Harrison J. (1983). Interactions of polycyclic aromatic hydrocarbons with atmospheric pollutants. Environ Sci Technol, 17,
673-679.
Grosjean D; Grosjean E; Williams EL. (1993). Fading of artists' colorants by a mixture of photochemical oxidants. Atmos Environ, 27,765-772.
Grosjean D; Grosjean E; Williams EL II. (1994). Fading of colorants by atmospheric pollutants: reflectance spectroscopy studies. Sci Total
Environ, 151,213-226.
Grosjean D. (2003). Ambient PAN and PPN in southern California from 1960 to the SCOS97-NARSTO. Proceedings from the SCO-97-
NARSTO Data Analysis Conference, February 13-15,2001; Diamond Bar, CA. Atmos Environ, 37, S221-S238.
Grosovsky AJ; Sasaki JC; Arey J; Eastmond DA; Parks KK; Atkinson R. (1999). Evaluation of the potential health effects of the atmospheric
reaction products of polycyclic aromatic hydrocarbons. HEI Research Report #084. Boston, MA: Health Effects Institute (HEI).
Gross A; Stockwell WR. (2003). Comparison of the EMEP, RADM2 and RACM mechanisms. J Atmos Chem , 44,151-170.
Grossenbacher JW; Couch T; Shepson PB; Thornberry T; Witmer-Rich M; Carroll MA; Faloona I; Tan D; Brune W; Ostling K; Bertman S.
(2001). Measurements of isoprene nitrates above a forest canopy. J Geophys Res, 106,24429-24438.
Grossman GD; Ratajczak RE Jr; Crawford MK; Freeman MC. (1998). Assemblage of organization in stream fishes: effects of environmental
variation and interspecific interactions. Ecol Monogr, 68,395-420.
Grulke NE; Andersen CP; Fenn ME; Miller PR. (1998). Ozone exposure and nitrogen deposition lowers root biomass of ponderosa pine in the
San Bernardino Mountains, California. Environ Pollut, 103,63-73.
Grulke NE; Balduman L. (1999). Deciduous conifers: high N deposition and O3 exposure effects on growth and biomass allocation in
ponderosa pine. Water Air Soil Pollut, 116, 235-248.
Grulke N; Schilling S. (2008). Air Pollution and Climate Change. Washington, DC; Climate Change Resource Center; U.S. Forest Service; U.S.
Department of Agriculture (USDA). http://www.fs.fed.us/ccrc/topics/air-pollution.shtml.
R-34
-------
Grulke NE; Minnich RA; Paine TD; Seybold SJ; Chavez D; Fenn ME; Riggan PJ; Dunn A. (2008). Air pollution increases forest susceptibility to
wildfires: A case study in the San Bernardino Mountains in southern California. (An air pollution and climate change case study). In:
Bytnerowicz A; Arbaugh MJ; Riebau AR; Anderson, C (Eds.) Wildland fires and air pollution; Section III: Ecological Impacts of Forest
Fires and Air Pollution (Volume 8 in Developments in Environmental Sciences Series, pp 365-403). Amsterdam, The Netherlands:
Elsevier Ltd.
Grtinhage L; Dammgen U; Haenel HD; Jager HJ. (1992). Vertikale fliisse von spurengasen in der bodennahen atmosphare. Landbauforsch
Voelkenrode, 128,201-245.
Gubala CP; Driscoll CT; Newton RM; Schofield CL. (1991). Chemistry of a near-shore lake region during spring snowmelt. Environ Sci Technol,
25, 2024-2030.
Guderian R; Tingey DT; Rabe R. (1985). Effects of photochemical oxidants on plants. In Guderian R. (Ed.).Air pollution by photochemical
oxidants: Formation, transport, control, and effects on plants. (Volume 52 in Series: Billings WD; Golley F; Lange OL; Olson JS;
Remmert H (Eds.). Ecological studies: Analysis and synthesis, pp 129-333). Berlin, Germany: Springer-Verlag.
Guenther FR; Dorko WD; Miller WR; Rhoderick GC. (1996). The NIST traceable reference material program for gas standards. (NIST special
publication 260-126). Washington, DC; National Institute of Standards and Technology; U.S. Department of Commerce.
Guenther A; Geron C; Pierce T; Lamb B; Harley P; Fall R. (2000). Natural emissions of non-methane volatile organic compounds, carbon
monoxide, and oxides of nitrogen from North America. Atmos Environ, 34,2205-2230.
Guerold F; Boudot JP; Jacquemin G; Vein D; Merlet D; Rouiller J. (2000). Macroinvertebrate community loss as a result of headwater stream
acidification in the Vosges Mountains (N-E France). Biodivers Conserv, 9,767-783.
Gulledge J; Doyle AP; Schimel JP. (1997). Different NH^-inhibition patterns of soil Cm -oxidizer populations across sites. Soil Biol Biochem,
29,13-21.
Gulledge J; Schimel JP. (2000). Controls on soil carbon dioxide and methane fluxes in a variety of taiga forest stands in interior Alaska.
Ecosystems, 3,269-282.
Gundersen P; Rasmussen L. (1995). Nitrogen mobility in a nitrogen limited forest at Klosterhede, Denmark, examined by NH4NO3 addition. For
Ecol Manage, 71,75-88.
Gundersen P; Callesen I; de Vries W. (1998). Nitrate leaching in forest ecosystems is controlled by forested floor C/N ratio. Environ Pollut, 102,
403-407.
Gundersen P; Schmidt IK; Raulund-Rasmussen K. (2006). Leaching of nitrate from temperate forests - Effects of air pollution and forest
management. Environ Rev, 14,1-57.
Gunn JM; Keller W. (1984a). In situ manipulation of water chemistry using crushed limestone and observed effects on fish. Fisheries, 9,19-24.
Gunn JM; Keller W. (1984b). Spawning site water chemistry and lake trout (Salvelinus namaycush).sac fry survival during spring snowmelt.
Can J Fish Aquat Sci, 41, 319-329.
Gunn JM; Noakes DLG. (1987). Latent effects of pulse exposure to aluminum and lowpH on size, ionic composition and feeding efficiency of
lake trout (Salvelinus fontinalis).a\ev\ns. Can J Fish Aquat Sci, 44,1418-1424.
Gunn JM; McMurtry MJ; Casselman JM; Keller W; Powell MJ. (1988). Changes in the fish community of a limed lake near sudbury, Ontario:
Effects of chemical neutralization or reduced atmospheric deposition of acids? Water Air Soil Pollut, 41,113-136.
Gunn JM. (1989). Survival of lake charr (Salvelinus namaycush) embryos under pules exposure to acidic runoff water. In: Nriagu JO;
Lakshminarayana JSS, (Eds.). Aquatic toxicology and water quality management. (Advances in environmental science and
technology: Volume 22, pp 23-45).New York, NY: John Wiley & Sons.
Gunn JM; Keller W. (1990). Biological recovery of an acid lake after reductions in industrial emissions of sulphur. Nature, 355,431 -433.
Gunn J; Keller W; Negusanti J; Potvin R; Beckett P; Winterhalder K. (1995). Ecosystem recovery after emission reductions: Sudbury, Canada.
Water Air Soil Pollut, 85,1783-1788.
Gunn JM; Mills KH. (1998).The potential for restoration of acid-damaged lake trout lakes. Restor Ecol, 6,390-397.
Gunthardt-Goerg MS; Schmutz P; Matyssek R; Bucher JB. (1996). Leaf and stem structure of poplar (Populus x euramericana) as influenced
by O3, NO2, their combination, and different soil N supplies. Can J For Res, 26, 649-657.
Gunthardt-Goerg MS; McQuattie CJ; Maurer S; Frey B. (2000). Visible and microscopic injury in leaves of five deciduous tree species related to
current critical ozone levels. Environ Pollut, 109,489-500.
Gupta UC; Munro DC. (1969). Influence of sulfur, molybdenum and phosphorus on the chemical composition and yield of brussels sprouts and
of molybdenum on sulfur contents of several plant species grown in the greenhouse. Soil Sci, 107,114-118.
Gupta VK; Mehla DS. (1980). Influence of sulphur on the yield and concentration of copper, manganese, iron and molybdenum in Berseem
('Trifolium alexandrinum) grown on two different soils. Plant Soil, 56,229-234.
Gysels K; Delalieux F; Deutsch F; Van Grieken R; Camuffo D; Bernardi A; Sturaro G; Busse HJ; Wieser M. (2004). Indoor environment and
conservation in the Royal Museum of Fine Arts, Antwerp, Belgium. Journal of Cultural Heritage, 5,221-230.
Hachmoller B; Matthews RA; Brakke DF. (1991). Effects of riparian community structure, sediment size, and water-quality on the
macroinvertebrate communities in a small, suburban stream. Northwest Science, 65,125-132.
Haefele M; Kramer RA; Holmes TP. (1992). Estimating the total value of forest quality in high elevation spruce-fir forests. In: Payne C; Bowker
JM; Reed PC (Eds.). The economic value of wilderness. Proceedings of the conference held in Jackson, WY May 8-11,1991; General
technical report #SE-78 (pp 91-96). Asheville, NC; Southeastern Forest Experimental Station, U.S. Forest Service, U.S. Department of
Agriculture (USDA).
Hagy JD; Boynton WR; Keefe CW; Wood KV. (2004). Hypoxia in Chesapeake Bay, 1950-2001: Long-term change in relation to nutrient loading
and river flow. Estuaries, 27, 634-658.
Haines TA; Baker JP. (1986). Evidence of fish population responses to acidification in the eastern United States. Water Air Soil Pollut, 31, 605-
629.
R-35
-------
Hains JC; Chen LWA; Taubman BF; Doddridge BG; Dickerson RR. (2007). A side-by-side comparison of filter-based PM2.5 measurements at a
suburban site: A closure study. Atmos Environ, 41,6167-6184.
Hall SJ; Matson PA. (1999). Nitrogen oxide emissions after nitrogen additions in tropical forests. Nature, 400,152-155.
Hall SJ; Matson PA. (2003). Nutrient status of tropical rain forests influences soil N dynamics after N additions. Ecol Monogr, 73,107-129.
Hall LW. (1987). Acidification effects on larval striped bass, Morone saxatilis, in Chesapeake Bay tributaries: A review. Water Air Soil Pollut, 35,
87-96.
Hall Rl; Leavit PR; Quinlan R; Dixit AS; Smol JP. (1999). Effects of agriculture, urbanization, and climate on water quality in the northern Great
Plains. Limnol Oceanogr, 44,739-756.
Hall RJ; Likens GE; Fiance SB; Hendrey GR. (1980). Experimental acidification of a stream in the Hubbard Brook Experimental Forest, New
Hampshire. Ecology, 61,976-989.
Hall RJ; Ide FP. (1987). Evidence of acidification effects on stream insect communities in central Ontario between 1937 and 1985. Can J Fish
Aquat Sci, 44,1652-1657.
Hallett RA; Bailey SW; Horsley SB; Long RP. (2006). Influence of nutrition and stress on sugar maple at a regional scale. Can J For Res, 36,
2235-2246.
Hallingback T. (1991). Blue-green algae and cyanophilic lichens threatened by air pollution and fertilization. Sven Bot Tidskr, 85,87-104.
Hallingback T; Kellner O. (1992). Effects of simulated nitrogen rich and acid rain on the nitrogen-fixing lichen Peltigera aphthosa (L.)Willd. New
Phytol, 120,99-103.
Hallock-Waters KA; Doddridge BG; Dickerson RR; Spitzer S; Ray JD. (1999). Carbon monoxide in the U.S. mid-Atlantic troposphere: Evidence
for a decreasing trend. Geophys Res Lett, 26,2861-2864.
Hallquist M; Stewart DJ; Stephenson SK; Cox RA. (2003). Hydrolysis of N2O5 on sub-micron sulfate aerosols. Phys Chem Chem Phys, 5,
3453-3463.
Hamburg SP; Yanai RD; Arthur MA; Blum JD; Siccama TG. (2003). Biotic control of calcium cycling in northern hardwood forests: Acid rain and
aging forests. Ecosystems, 6,399-406.
Hameed S; Pinto JP; Stewart RW. (1979). Sensitivity of the predicted CO-OH-CH4 perturbation to tropospheric NOx concentrations. J Geophys
Res, 84, 763-768.
Hames RS; Rosenberg KV; Lowe JD; Barker SE; Dhondt AA. (2002). Adverse effects of acid rain on the distribution of the wood thrush
Hylocichla mustelina in North America. Proc Natl Acad Sci USA, 99,11235-11240.
HammittWE; Patterson ME; Noe FP. (1994). Identifying and predicting visual preferences of southern Appalachian forest recreation vistas.
Landsc Urban Plan, 29 171-183.
Hanemann M. (1991). Willingness to pay and willingness to accept: How much can they differ? Am Econ Rev, 81, 635-647.
Hanke M; Umann B; Uecker J; Arnold F; Bunz H. (2003). Atmospheric measurements of gas-phase HNO3 and SO2 using chemical ionization
mass spectrometry during the MINATROC field campaign 2000 on Monte Cimone. Atmos Chem Phys, 3,417-436.
Hanley N; Shogren J; White B. (1997). Environmental economics in theory and practice. New York, NY: Oxford University Press.
Hansen DA; Edgerton ES; Hartsell BE; Jansen JJ; Kandasamy N; Hidy GM; Blanchard CL. (2003). The southeastern aerosol research and
characterization study: Part 1 - Overview. J Air Waste Manage Assoc, 53,1460-1471.
Hansen J; Sato M; Ruedy R; Lo K; Lea DW; Medina-Elizade M. (2006). Global temperature change. Proc Natl Acad Sci, USA, 103,14288-
14293.
Hansen K; Draaijers GPJ; Ivens WPMF; Gundersen P; Van Leeuwen NFM. (1994). Concentration variations in rain and canopy throughfall
collected sequentially during individual rain events. Atmos Environ, 28,3195-3205.
Hanson PJ; Rott K; Taylor GE; Gunderson CA; Lindberg SE; Ross-Todd BM. (1989). NO2 deposition to elements representative of a forest
landscape. Atmos Environ, 23,1783-1794.
Hanson PJ; Lindberg SE. (1991). Dry deposition of reactive nitrogen compounds: A review of leaf, canopy, and non-foliar measurements.
Atmos Environ, 25,1615-1634.
Hanson PJ; Garten CT Jr. (1992). Deposition of H15N03 vapour to white oak, red maple, and loblolly pine foliage: Experimental observations
and a generalized model. New Phytol, 122, 329-337.
Hao JM; YeXM; Duan L; Zhou ZP. (2001). Calculating critical loads of sulfur deposition for 100 surface waters in China using the MAGIC
model. Water Air Soil Pollut, 130,1157-1162.
Harasmis GM; Chu DS. (1987). Acid rain effects on waterfowl: Use of black duck broods to assess food resources of experimentally acidified
wetlands. In: Diamond AW; Filion FL (Eds.), The value of birds. (ICBP technical publication no 6; pp 173-181). Cambridge, United
Kingdom: International Council for Bird Preservation (ICBP).
Harder JW; Williams EJ; Baumann K; Fehsenfeld FC. (1997). Ground-based comparison of NO2, H2O, and O3 measured by long-path and in
situ techniques during the 1993 tropospheric OH photochemistry experiment. J Geophys Res, 102,6227-6243.
Harding LW; Perry ES. (1997). Long-term increase of phytoplankton biomass in Chesapeake Bay, 1950-1994. Mar Ecol Prog Ser, 157,39-52.
Harding, LWJ; Mallonee ME; Perry ES. (2002). Toward a predictive understanding of primary productivity in a temperate, partially stratified
estuary. Estuar Coast Shelf Sci, 55,437-463.
Hari P; Raivonen M; Vesala T; Munger JW; Pilegaard K; Kulmala M. (2003). Atmospheric science: Ultraviolet light and leaf emission of NOx.
Nature, 422,134.
Harley PC; Litvak ME; Sharkey TD; Monson RK. (1994). Isoprene emission from velvet bean leaves: Interactions among nitrogen availability,
growth photon flux density, and leaf development. Plant Physiol, 105,279-285.
Harmon SM; King JK; Gladden JB; Chandler GT; Newman LA. (2004). Methylmercury formation in a wetland mesocosm amended with sulfate.
Environ Sci Technol, 38, 650-656.
R-36
-------
Harpole WS; Potts DL; Suding KN. (2007). Ecosystem responses to water and nitrogen amendment in a California grassland. Glob Change
Biol, 13, 2341-2348.
Harriman R; Morrison BRS. (1982). Ecology of streams draining forested and non-forested catchments in an area of central Scotland subject to
acid precipitation. Hydrobiologia, 88,251-263.
Harris GW; Carter WPL; Winer AM; Pitts JN Jr; Piatt U; Perner D. (1982). Observations of nitrous acid in the Los Angeles atmosphere and
implications for predictions of ozone-precursor relationships. Environ Sci Technol, 16,414-419.
Harris HH; Pickering I J; George GN. (2003). The chemical form of mercury in fish. Science, 301,1203.
Harris RC; Rudd JWM. (2007). Whole-ecosystem study shows rapid fish-mercury response to changes in mercury deposition. Proc Natl Acad
Sci USA, 104,16586-16591.
Harrison RM; Kitto A-MN. (1994). Evidence for a surface source of atmospheric nitrous acid. Atmos Environ, 28,1089-1094.
Harrison RM; Peak JD; Collins GM. (1996). Tropospheric cycle of nitrous acid. J Geophys Res, 101,14429-14439.
Hartman MD; Baron JS; Ojima DS. (2005a). Calculating pre-measurement atmospheric deposition, stream chemistry, and soil chemistry for an
alpine watershed in Rocky Mountain National Park, Colorado. In: Abstracts of the Ecological Society of America 90th annual meeting;
held in Montreal, Canada; Aug 7-12,2005.
Hartman MD; Baron JS; Ojima DS; Parton WJ. (2005b). Modeling the timeline for surface water acidification from excess nitrogen deposition
for Rocky Mountain National Park. In: Abstracts of the George Wright Society biennial conference; Philadelphia, PA.
Hartman MD; Baron JS; Ojima DS. (2007). Application of a coupled ecosystem-chemical equilibrium model, DayCent-Chem, to stream and soil
chemistry in a Rocky Mountain watershed. Ecol Modell, 200,493-510.
Harvey H; Lee C. (1982). Historical fisheries changes related to surface water pH changes in Canada. In: Johnson RE (Ed.), Acid rain/ fisheries
(pp. 44-55). Bethesda, MD: American Fisheries Society.
Harwell MA; Myers V; Young T; Bartuska A; Gassman N; Gentile JH; Harwell CC; Appelbaum S; Barko J; Causey B; Johnson C; McLean A;
Smola R; Templet P; Tosini S. (1999). A framework for an ecosystem integrity report card. Bioscience, 49,543-556.
HaslerAD. (1947). Eutrophication of lakes by domestic drainage. Ecology, 28,383-395.
Hassan R; Scholes R; Ash N. (2005). Ecosystem change and human well-being. In: Hassan R; Scholes R. (Eds). Current state and trends
assessment: Volume 1: Findings of the condition and trends working group (Chapter 5;
http://www.millenniumassessment.org/en/Condition.aspx). Nairobi, Kenya; Millennium Ecosystem Assessment; United Nations
Environment Programme (UNEP): Island Press.
Hauer FR; Spencer CN. (1998). Phosphorous and nitrogen dynamics in streams associated with wildfire: A study of immediate and longterm
effects. Int J Wildland Fire, 8,183-198.
Havas M; Hutchinson TC; Likens GE. (1984). Red herrings in acid rain research. Environ Sci Technol, 18,176A-186A.
Havas, M. (1985). Aluminum bioaccumulation and toxicity to Daphnia magna in soft water at lowpH. Can J Fish Aquat Sci, 42,1741-1748.
Havas M. (1986). Effects of acid deposition on aquatic ecosystems. In Stern A (Ed.), Air pollution (Vol. VI, pp. 351-389). New York, NY:
Academic Press.
Havas M; Woodfine DG; Lutz P; Yung K; Maclsaac HJ; Hutchinson TC. (1995). Biological recovery of two previously acidified, metal-
contaminated lakes near Sudbury, Ontario, Canada. Water Air Soil Pollut, 85,791-796.
Havens KE; Yan ND; Keller W. (1993). Lake acidification: effects on crustacean zooplankton populations. Environ Sci Technol, 27,1621-1624.
Havens KE; Carlson RE. (1998). Functional complementarity in plankton communities along a gradient of acid stress. Environ Pollut, 101,427-
436.
Hawksworth DL. (2002). Bioindication calibrated scales and their utility. In: Nimis PL; Scheidegger C; Wolseley PA (Eds.), Monitoring with
lichens-monitoring lichens: Proceedings of the NATO Advanced Research Workshop on Lichen Monitoring (pp. 11-20). Netherlands:
Kluwer Academic Publishers.
Hawley GJ; Schaberg PG; Eagar C; Borer CH. (2006). Calcium addition at the Hubbard Brook Experimental Forest reduced winter injury to red
spruce in a high-injury year. Can J Forest Res, 36,2544-2549
Haya K; Waiwood BA. (1981). Acid pH and chorionase activity of Atlantic salmon (Salmo salai) eggs. Bull Environ Contam Toxicol, 27, 7-12.
Hayden KL; Anlauf KG; Hastie DR; Bottenheim JW. (2003). Partitioning of reactive atmospheric nitrogen oxides at an elevated site in southern
Quebec, Canada. J Geophys Res, 108, doi: 10.1029/2002JD003188.
Hayden MJ; Ross DS. (2005). Denitrification as a nitrogen removal mechanism in a Vermont peatland. J Environ Qual, 34,2052-2061.
Haynes RW; Adams DM. (1992). Assessing economic impacts of air pollution damage to U.S. forests. In: de Steiguer JE (Ed.),The economic
impact of air pollution on timber markets: Studies from North America and Europe. (General technical report no. SE-75). Asheville, NC;
Southeastern Forest Experiment Station ; U.S. Forest Service; U.S. Department of Agriculture (USDA).
Hazlett PW; Gordon AM; Voroney RP; Sibley PK. (2007). Impact of harvesting and logging slash on nitrogen and carbon dynamics in soils from
upland spruce forests in northeastern Ontario. Soil Biol Biogeochem, 39,43-57.
He Y; Zhou X; Hou J; Gao H; Bertman SB. (2006). Importance of dew in controlling the air-surface exchange of HONO in rural forested
environments. Geophys Res Lett, 33, L02813, doi: 10.1029/2005GL024348.
Heal GM; Barbier EB; Boyle KJ. (2005). Valuing Ecosystem Services: Toward Better Environmental Decision Making.
http://www.nap.edu/openbook.php?isbn=030909318X. Washington, DC: The National Academies Press.
Heber AJ; Ni JQ; Haymore BL; Duggirala RK; Keener KM. (2001). Air quality and emission measurement methodology at swine finishing
buildings. Trans of the ASAE, 44,1765-1778.
Hebestreit K; Stutz J; Rosen D; Matveiv V; Peleg M; Luria M; Piatt U. (1999). DOAS measurements of tropospheric bromine oxide in mid-
latitudes. Science, 283,55-57.
Heck WW; Adams RM; Cure WW; Heagle AS; Heggestad HE; Kohut RJ; Kress LW; Rawlings JO; Taylor OC. (1983). A reassessment of crop
loss from ozone. Environ Sci Technol, 17,573A-581A.
R-37
-------
Hecnar SJ. (1995). Acute and chronic toxicity of ammonium-nitrate fertilizer to amphibians from southern Ontario. Environ Toxicol Chem, 14,
2131-2137.
Hedin LO; Likens GE; Postek KM; Driscoll CT. (1990). A field experiment to test whether organic acids buffer acid deposition. Nature, 345,798-
800.
Hedin LO; Granat L; Likens GE; Buishand TA; Galloway JN; Butler TJ; Rodhe H. (1994). Steep declines in atmospheric base cations in regions
of Europe and North America. Nature, 367,351-354.
Hedin LO; Armesto JJ; Johnson AH. (1995). Patterns of nutrient loss from unpolluted, old-growth temperate forests: evaluation of
biogeochemical theory. Ecology, 76,493-509.
Hedin LO; von Fischer JC; Ostrom NE; Kennedy BP; Brown MG; Robertson GP. (1998). Thermodynamic constraints on nitrogen
transformations and other biogeochemical processes at soil-stream interfaces. Ecology, 79,684-703.
Heeb NV; Schmid P; Kohler M; Gujer E; Zennegg M; Wenger D; Wichser A; Ulrich A; Gfeller U; Honegger P; Zeyer K; Emmenegger L;
Petermann JL; Czerwinski J; Mosimann T; Kasper M; Mayer A. (2008). Secondary effects of catalytic diesel particulate filters:
conversion of PAHs versus formation of nitro-PAHs. Environ Sci Technol, 42,3773-3779.
Hefting MM; Bobbink R; De Caluwe H. (2003). Nitrous oxide emission and denitrification in chronically nitrate-loaded riparian buffer zones. J
Environ Qual, 32,1194-1203.
Heil GW; DiemontWH. (1983). Raised nutrient levels change heathland in grassland. Vegetation, 53,113-120.
Heil GW; Bruggink M. (1987). Competition for nutrients between Calluna vulgaris (L.) Hull and Molinia caerulea (L.) Moench. Oecologia, 73,
105-107.
Heinz G. (1974). Effects of low dietary levels of methyl mercury on mallard reproduction. Bull Environ Contam Toxicol, 11,386-392.
Heinz GH. (1979). Methylmercury: reproductive and behavioral effects on three generations of mallard ducks. J Wildl Manage, 43,394-401.
Hemond HF. (1980). Biogeochemistry of Thoreau's Bog, Concord, Massachusetts. Ecol Monogr, 50,507-526.
Hemond HF. (1990). Wetlands as the source of dissolved organic carbon to surface waters. In: Perdue M; Gjessing ET (Eds.), Organic acids in
aquatic ecosystems (pp. 301-313). New York, NY: Wiley & Sons.
Hemond HF. (1994). Role of organic acids in acidification of fresh waters. In Steinberg C; Wright R (Eds.), Acidification of freshwaters:
Implications for the future. (Report of the Dahlem workshop on acidification of freshwater ecosystems; held in Berlin, Germany;
September-October 1992 (pp. 103-116). New York, NY: John Wiley & Sons.
Hendrickson OQ. (1990). Asymbiotic nitrogen fixation and soil metabolism in three Ontario forests. Soil Biol Biochem, 22,967-971.
Hendry CD; Brezonik PL. (1984). Chemical composition of softwater Florida lakes and their sensitivity to acid precipitation. JAWRA, 20, 75-86.
Hennigan CJ; Sandholm S; Kim S; Stickel RE; Huey LG; Weber RJ. (2006). Influence of Ohio River valley emissions on fine particle sulfate
measured from aircraft over large regions of the eastern United States and Canada during INTEX-NA. J Geophys Res, 111,
doi:10.1029/2006JD007282.
Henriksen A. (1980).Acidification of freshwaters - a large scale titration. In: Drablos D; Tollan A, (Eds.), Ecological impact of acid precipitation:
Proceedings of an international conference; held in Sandefjord, Norway in March 1980. (SNSF Project; pp 68-74). Oslo, Norway; The
Norwegian Interdisciplinary Research Programme.
Henriksen A. (1984). Changes in base cation concentrations due to freshwater acidification. Verh Internat Verein Limnol, 22,692-698.
Henriksen A; Skogheim OK; Rosseland BO. (1984). Episodic changes in pH and aluminium-speciation kill fish in a Norwegian salmon river.
Vatten, 40,255-260.
Henriksen A; Kamari J; Posch M; Wilander A. (1992). Critical loads of acidity: Nordic surface waters. Ambio, 356-363.
Henriksen, A; Dillon PJ. (2001). Critical load of acidity to surface waters in south-central Ontario, Canada. I: Application of the Steady-State
Water Chemistry SSWC model. (Acid Rain Research Report no 2001:52). Oslo, Norway; Norwegian Institute for Water Research
(NIVA).
Henriksen A; Posch M. (2001). Steady-state models for calculating critical loads of acidity for surface waters. Water Air Soil Pollut: Focus, 1,
375-398.
Henriksen A; Dillon PJ; Aherne J. (2002). Critical loads of acidity to surface waters in south-central Ontario, Canada: Regional application of
the steady-state water chemistry model. Can J Fish Aquat Sci, 59,1287-1295.
Hering SV; Lawson DR; Allegrini I; Febo A; Perrino C; Possanzini M; Sickles JE, II; Anlauf KG; Wiebe A; Appel BR; John W; Ondo J; Wall S;
Braman RS; Sutton R; Cass GR; Solomon PA; Eatough DJ; Eatough NL; Ellis EC; Grosjean D; Hicks BB; Womack JD; Horrocks J;
Knapp KT; Ellestad TG; Paur RJ; Mitchell WJ; Pleasant M; Peake E; MacLean A; Pierson WR; Brachaczek W; Schiff HI; Mackay Gl;
SpicerCW. (1988). The nitric acid shootout: Field comparison of measurement methods. Atmos Environ, 22,1519-1539.
Hering S; Cass G. (1999). The magnitude of bias in the measurement of PM2.5 arising from volatilization of particulate nitrate from Teflon filters.
J Air Waste Manage Assoc, 49,725-733.
Herlihy AT; Kaufman PR; Mitch ME. (1991). Stream chemistry in the eastern United States 2. Current sources of acidity in acidic and low acid-
neutralizing capacity streams. Water Resour Res, 27,629-642.
Herlihy AT; Kaufmann PR; Church MR; Wigington PJ Jr; Webb JR; Sale MJ. (1993). The effects of acidic deposition on streams in the
Appalachian Mountain and Piedmont region of the mid-Atlantic United States. Water Resour Res, 29, 2687-2703.
Herlihy AT; Larsen DP; Paulsen SG; Urquhart NS; Rosenbaum BJ. (2000). Designing a spatially balanced, randomized site selection process
for regional stream surveys: The EMAP mid-Atlantic pilot study. Environ Monit Assess, 63,95-113.
Hess PG. (2001). Model and measurement analysis of springtime transport and chemistry of the Pacific Basin. J Geophys Res, 106,12689-
12717.
Hicke JA; Asner GA; Randerson JT; Tucker CJ; Los S; Birdsey RA; Jenkins JC; Field CB. (2002). Trends in North American net primary
productivity derived from satellite observations, 1982-1998. Glob Biogeochem Cycles, 16,2-1 - 2-15, doi: 10.1029/2001 GB001550.
R-38
-------
Hicks BB; Baldocchi DD; Meyers TP; Hosker RP Jr; Matt DR. (1987). A preliminary multiple resistance routine for deriving dry deposition
velocities from measured quantities. Water Air Soil Pollut, 36, 311 -330.
Hicks BB; McMillen RT; Turner RS; Holdren J; Strickland TC. (1993). A national critical loads framework for atmospheric deposition effects
assessment: III: Deposition characterization. Environ, Manage, 17, 343-353.
Higley B; Carrick HJ; Brett MT; Luecke C; Goldman, CR. (2001). The effects of ultraviolet radiation and nutrient additions on periphyton
biomass and composition in a sub-alpine lake (Castle Lake, USA). Int Rev Hydrobiol, 86,147-163.
Hill AB. (1965). The Environment and Disease: Association or Causation? Proc R Soc Med, 58,295-300.
Hill AR. (1996). Nitrate removal in stream riparian zones. J Environ Qual, 25,743-755.
Hill BH; Elonen CM; Jicha TM; Cotter AM; Trebitz AS; Danz NP. (2006). Sediment microbial enzyme activity as an indicator of nutrient limitation
in Great Lakes coastal wetlands. Freshw Biol, 51,1670-1683.
Hill WR; Knight AW. (1988). Nutrient and light limitation of algae in 2 northern California streams. J Phycol, 24,125-132.
Hilliard JC; Wheeler RW. (1979). Nitrogen dioxide in engine exhaust. Paper presented at the SAE Passenger Car Meeting; Dearborn, Ml, June
11-15,1979. (Paper #790691). SAE Trans, 88, 790691.
Hindar A; Posch M; Henriksen A; Gunn J; Snucins E. (2000). Development and application of the FAB model to calculate critical loads of S and
N for lakes in the Killarney Provincial Park (Ontario, Canada). (NIVA report SNO 4202-2000). Oslo, Norway: Norwegian Institute for
Water Research (NIVA).
Hindar A; Posch M; Henriksen A. (2001). Effects of in-lake retention of nitrogen on critical load calculations. Water Air Soil Pollut, 130,1403-
1408.
Hinga KR; Keller AA; Oviatt CA. (1991). Atmospheric deposition and nitrogen inputs to coastal waters. Ambio, 20,256-260.
Hintelmann H; Evans RD. (1997). Application of stable isotopes in environmental tracer studies - measurement of monomethylmercury
(CH3Hg+) by isotope dilution ICP-MS and detection of species transformation. Fresenius J Anal Chem, 358,378-385.
Hobbie SE. (1996). Temperature and plant species control over litter decomposition in Alaskan tundra. Ecol Monogr, 66,503-522.
Hoell JM; Davis DD; Liu SC; Newell RE; Akimoto H; McNeal RJ; Bendura RJ. (1997). The Pacific exploratory mission-west phase B: February-
March, 1994. J Geophys Res, 102, 28223-28239.
Hoell JM; Davis DD; Jacob DJ; Rodgers MO; Newell RE; Fuelberg HE; McNeal RJ; Raper JL; Bendura RJ. (1999). Pacific exploratory mission
in the tropical Pacific: PEM-Tropics A, August-September 1996. J Geophys Res, 104, 5567-5583.
Hoffman DJ; Moore JM. (1979). Teratogenic effects of external egg applications of methyl mercury in the mallard, Anas platyrynchos.
Teratology, 20,453-461.
Hofmann D; Butler J; Dutton E; Elkins J; Ogren J; Oltmans S; Tans P; Schnell R. (2004). Variation in the direct radiative climate forcing by well-
mixed greenhouse gases over the past 25 years. Geophys Res Abstr, 6,00545, doi: 1607-7962/gra/EGU04-A-00545.
Hogan GD; Rennenberg H; Fink S. (1998a). Role and effect of sulfur in tree biology. In Maynard DG (Ed.), Sulfur in the environment (pp. 173-
217). New York, NY: Marcel Dekker.
Hogan K; Marcus A; Smith R; White P. (1998b). Integrated exposure uptake biokinetic model for lead in children: Empirical comparisons with
epidemiologic data. Environ Health Perspect, 106,1557-1567.
Hogberg P; Fan H; Quist M; Binkleys D; Oloftamm C. (2006). Tree growth and soil acidification in response to 30 years of experimental
nitrogen loading on boreal forest. Glob Chang Biol, 12,489-499.
Hogbom L; Hogberg P. (1991). Dynamics of soil nitrate after forest fertilization as monitored by the plant nitrate reductase assay. Forest Ecol
Manage, 44, 223-238.
Holdren GR Jr; Strickland TC; Cosby BJ; Marmorek D; Bernard D; Santore R; Driscoll CT; Pardo L; Hunsaker C; Turner RS; Aber J. (1993a). A
national critical loads framework for atmospheric deposition effects assessment: IV. Model selection, applications and critical loads
mapping. Environ Manage, 17,355-363.
Holdren GR; Strickland TC; Rosenbaum BJ; Turner RE; Ryan PF; McDowell MK; Bishop GD. (1993b). Comparison of selected critical loads
estimation approaches for assessing the effects of sulfate deposition on lakes in the northeastern United States. (Report no
EPA/600/R93/074). Washington, D.C; U.S. Environmental Protection Agency.
Holdren GR; Strickland TC; Shaffer PW; Ryan PF; Ringold PL; Turner RS. (1993c). Sensitivity of critical load estimates for surface waters to
model selection and regionalization schemes. J Environ Qual, 22,279-289.
Holland EA; Lamarque JF. (1997). Modeling bio-atmospheric coupling of the nitrogen cycle through NOx emissions and NOy deposition. Nut-
Cycling Agroecosyst, 48,7-24.
Holland EA; Braswell BH; Lamarque JF; Townsend A; Sulzman JM; Muller JF; Dentener F; Brasseur G; Levy H; Penner JE; Roelofs GJ.
(1997). Variations in the predicted spatial distribution of atmospheric nitrogen deposition and their impact on carbon uptake by
terrestrial ecosystems. J Geophys Res, 102,15849-15866.
Holland EA; Dentener FJ; Braswell BH; Sulzman JM. (1999). Contemporary and pre-industrial global reactive nitrogen budgets.
Biogeochemistry, 46,7-43.
Holland EA; Braswell BH; Sulzman J; Lamarque JF. (2005). Nitrogen deposition onto the United States and western Europe: Synthesis of
observations and models. Ecol Appl, 15,38-57.
Holland HD. (1978). The chemistry of the atmosphere and oceans. New York, NY: John Wiley & Sons.
Hollenhorst SJ; Brock SM; Freimund WA; Twery M J. (1993). Predicting the effects of gypsy moth on near-view aesthetic preferences and
recreation appeal. For Sci, 39, 28-40.
Hollwedel J; Wenig M; Beirle S; Kraus S; Kiihl S; Wilms-Grabe W; Piatt U; Wagner T. (2004). Year-to-year variations of spring time polar
tropospheric BrO as seen by GOME. Adv Space Res, 34,804-808.
Holmes RW; Whiting MC; Stoddard JL. (1989). Changes in diatom-inferred pH and add neutralizing capacity in a dilute, high elevation, Sierra
Nevada lake since AD 1825. Freshw Biol, 21,295-310.
R-39
-------
Holmes TP. (1992). Economic welfare impacts of air pollution damage to forests in the southern United States. In: de Steiguer JE (Ed.), The
economic impact of air pollution on timber markets: Studies from North America and Europe. (General Technical Report SE-75;
pp. 19-26). Asheville, NC; Southeastern Forest Experiment Station; U.S. Forest Service; U.S. Department, of Agriculture (USDA).
Holmes TP; Kramer RA. (1996). Contingent valuation of ecosystem health. Ecosystem Health, 2,56-60.
Holmes TP; Murphy EA; Bell KP. (2006). Exotic forest insects and residential property values. Agricultural and Resource Economics Review,
35,155-166.
Holt CA; Yan N D; Somers KM. (2003). pH 6 as the threshold to use in critical load modeling for zooplankton community change with
acidification in lakes of south-central Ontario: accounting for morphometry and geography. Can J Fish Aquat Sci, 60,151-158.
HoltC; Yan ND. (2003). Recovery of crustacean zooplankton communities from acidification in Killarney Park, Ontario, 1971-2000: pH 6 as a
recovery goal. R Swed Acad Sci, 32, 203-207.
Holtze KE; Hutchinson NJ. (1989). Lethality of lowpH and Al to early life stages of six fish species inhabiting precambrian shield waters in
Ontario. Can J Fish Aquat Sci, 46,1188-1202.
Holzmueller E; Jose S; Jenkins M; Camp A; Long A. (2006). Dogwood anthracnose in eastern hardwood forests: What is known and what can
be done? Journal of Forestry, 104,21-26.
Holzmueller EJ; Jose S; Jenkins MA. (2007). Influence of calcium, potassium, and magnesium on Cornus florida L. density and resistance to
dogwood anthracnose. Plant Soil, 290,189-199.
Honrath RE; Guo S; Peterson MC; Dziobak MP; Dibb JE; Arsenault MA. (2000). Photochemical production of gas phase NOx from ice crystal
N03". J Geophys Res, 105,24183-24190.
Honrath RE; Lu Y; Peterson MC; Dibb JE; Arsenault MA; Cullen NJ; Steffen K. (2002). Vertical fluxes of NOx, HONO, and HNO3 above the
snowpack at Summit, Greenland. Atmos Environ, 36,2629-2640.
Hooper DU; Vitousek PM. (1997). The effects of plant composition and diversity on ecosystem processes. Science, 277,1302-1305.
Hooper DU, Johnson L. (1999). Nitrogen limitation in dryland ecosystems: responses to geographical and temporal variation in precipitation.
Biogeochemistry, 46, 247-293.
Hoppel WA; Caffrey PF. (2005). Oxidation of S(IV).in sea-salt aerosol at high pH: ozone versus aerobic reaction. J Geophys Res, 110, D23202,
doi:10.1029/2005JD006239.
Horii CV. (2002). Tropospheric reactive nitrogen speciation, deposition, and chemistry at Harvard Forest [dissertation], Cambridge, MA:
Harvard University.
Horii CV; Munger JW; Wofsy SC; Zahniser M; Nelson D; McManus JB. (2004). Fluxes of nitrogen oxides over a temperate deciduous forest. J
Geophys Res, 109, D08305: doi: 10.1029/2003JD004326.
Horii C; Munger JW; Wofsy S; Zahniser M; Nelson D; McManus JB. (2006). Atmospheric reactive nitrogen concentrations and flux budgets at a
northeastern U.S forest site. Ag For Meteor, 136,159-174.
Hornbeck JW; Bailey SW; Buso DC; Shanley JB. (1997). Streamwater chemistry and nutrient budgets for forested watersheds in New England"
variability and management implications. For Ecol Manage, 93, 73-89.
Hornberger GM; Cosby BJ; Wright RF. (1989). Historical reconstructions and future forecasts of regional surface water acidification in
southernmost Norway. Water Resource Research, 25,2009-2018.
Hornung M; Bull KR; Cresser M; Hall J; Langan SJ; Loveland P; Smith C. (1995). An empirical map of critical loads of acidity for soils in Great
Britain. Environ Pollut, 90,301-310.
Horowitz LW; Liang J; Gardner GM; Jacob DJ (1998). Export of reactive nitrogen from North America during summertime: Sensitivity to
hydrocarbon chemistry. J Geophys Res, 103,13451-13476
Horowitz LW; Walters S; Mauzerall DL; Emmons LK; Rasch PJ; Granier C; Tie X; Lamarque J-F; Schultz MG; Tyndall GS; Orlando JJ;
Brasseur GP. (2003). A global simulation of tropospheric ozone and related tracers: description and evaluation of MOZART, version 2.
J Geophys Res, 108,4784, doi:10.1029/2002JD002853.
Horowitz LW; Fiore AM; Milly GP; Cohen RC; Perring A; Wooldridge PJ; Hess PG; Emmons LK; Lamarque JF. (2007). Observational
constraints on the chemistry of isoprene nitrates over the eastern United States. J Geophys Res, 112,10.1029/2006JD007747.
Horsley SB; Long RP; Bailey SW; Hallet RA; Hall TJ. (1999). Factors contributing to sugar maple decline along topographic gradients on the
glaciated and unglaciated Allegheny Plateau. In: Horsely SB; Long RP, (Eds.), Sugar maple ecology and health: proceedings of an
international symposium; June 1998; (General technical report NE-261, pp 60-62). Warren, PA. Radnor, PA; U.S. Forest Service; U.S.
Department of Agriculture (USDA).
Horsley SB; Long RP; Bailey SW; Hallett RA; Hall TJ. (2000). Factors associated with the decline disease of sugar maple on the Allegheny
Plateau. Can J For Res, 30,1365-1378.
Hosker RP; Lindberg SE. (1982). Review Article: Atmospheric deposition and plant assimilation of airborne gases and particles. Atmos Environ,
16,889-910.
Houdijk A; Verbeek PJM; Van Dijk HFG; Roelofs JGM. (1993). Distribution and decline of endangered herbaceous heathland species in
relation to the chemical composition of the soil. Plant Soil, 148,137-143.
Houghton RA. (2003). Why are estimates of the terrestrial carbon balance so different? Glob Chang Biol, 9,500-509.
Houle D; Carignan R; Ouimet R. (2001). Soil organic sulfur dynamics in a coniferous forest. Biogeochemistry, 53,105-124.
Houlton BZ; Driscoll CT; Fahey TJ; Likens GE; Groffman PM; Bernhardt ES; Buso DC. (2003). Nitrogen dynamics in ice storm-damaged forest
ecosystems: Implications for nitrogen limitation theory. Ecosystems, 6,431-443.
Houston DR. (1994). Major new tree epidemics: beech bark disease. Annu Rev Phytopathol, 32,75-87.
Howarth RW; Billen G; Swaney D; Townsend A; Jaworski N; Lajtha K; Downing JA; Elmgren R; Caraco N; Jordan T; Berendse F; Freney J;
Kudeyarov V; Murdoch PS; Zhao-Liang Z. (1996). Regional nitrogen budgets and riverine N & P fluxes for the drainages to the North
Atlantic Ocean: natural and human influences. Biogeochemistry, 35, 75-139.
R-40
-------
Howarth RW; Anderson D; Cloern J; Elfring C; Hopkinson C; Lapointe B; Malone T; Marcus N; McGlathery K; Sharpley A; Walker D. (2000).
Nutrient pollution of coastal rivers, bays, and seas. Iss Ecol, 1-15.
Howarth RW, Sharpley A; Walker D. (2002a). Sources of Nutrient Pollution to Coastal Waters in the United States: Implications for Achieving
Coastal Water Quality Goals. Estuaries, 25:656-676.
Howarth RW; Boyer EW; Pabich WJ; Galloway JN. (2002b). Nitrogen use in the United States from 1961-2000 and potential future trends.
Ambio, 31, 88-96.
Howarth RW; Marino R. (2006). Nitrogen as the limiting nutrient for eutrophication in coastal marine ecosystems: evolving views over three
decades. Limnol Oceanogr, 51,364-376.
Howarth RW; Swaney DP; Boyer EW; Marino R; Jaworski N; Goodale C. (2006). The influence of climate on average nitrogen export from
large watersheds in the Northeastern United States. Biogeochemistry, 79,163-186.
Howarth RW. (2007). Atmospheric deposition and nitrogen pollution in coastal marine ecosystems. In: Visgilio GR, Whitelaw DM (Eds.), Acid in
the Environment. New York, NY: Springer.
Howes BL; Dacey JWH; Goehringer DD. (1986). Factors controlling the growth form of Spartina alterniflora. feedbacks between above-ground
production, sediment oxidation, nitrogen and salinity. J Ecol, 74,881-898.
Hrabik TR; Watras CJ. (2002). Recent declines in mercury concentration in a freshwater fishery: isolating the effects of de-acidification and
decreased atmospheric mercury deposition in Little Rock Lake. Sci Total Environ, 297,229-237.
Huberty LE; Gross KL; Miller CJ. (1998). Effects of nitrogen addition on successional dynamics and species diversity in Michigan old-fields. J
Ecol, 86, 794-803.
Hiibler G; Alvarez RII; Daum P; Dennis R; Gillani N; Kleinman L; Luke W; Meagher J; Rider D; Trainer M; Valente R. (1998). An overview of
the airborne activities during the Southern Oxidants Study (SOS) 1995 Nashville/Middle Tennessee ozone study. J Geophys Res, 103,
22245-22259.
Hudman RC; Jacob DJ; Cooper OR; Evans MJ; Heald CL; Park RJ; Fehsenfeld F; Flocke F; Holloway J; Hubler G; Kita K; Koike M; Kondo Y;
Neuman A; Nowak J; Oltmans S; Parrish D; Roberts JM; Ryerson T. (2004). Ozone production in transpacific Asian pollution plumes
and implications for ozone air quality in California. J Geophys Res, 109, doi:10.1029/2004JD004974.
Hudman RC; Jacob DJ; Turquety S; Leibensperger EM; Murray LT; Wu S; Gilliland AB; Avery M; Bertram TH; Brune W; Cohen RC; Dibb JE;
Flocke FM; Fried A; Holloway J; Neuman JA; Orville R; Perring A; Ren X; Sachse GW; Singh HB; Swanson A; Wooldridge PJ. (2007).
Surface and lightning sources of nitrogen oxides over the United States: magnitudes, chemical evolution, and outflow. J Geophys Res,
112, doi:10.1029/2006JD007912.
Huebert DB; L'Hirondelle SJ; Addison PA. (1985). The effects of sulfur dioxide on netC02 assimilation in the lichen Evernia mesomorpha Nyl.
New Phytol, 100,643-651.
Huebert BJ; Zhuang L; Howell S; Noone K; Noone B. (1996). Sulfate, nitrate, methanesulfonate, chloride, ammonium, and sodium
measurements from ship, island, and aircraft during the Atlantic Stratocumulus Transition Experiment/Marine Aerosol Gas Exchange.
J Geophys Res, 101,4413-4423.
Huey LG; Dunlea EJ; Lovejoy ER; Hanson DR; Norton RB; Fehsenfeld FC; Howard CJ. (1998). Fast time response measurements of HNO3 in
air with a chemical ionization mass spectrometer. J Geophys Res, 103,3355-3360.
Huey LG; Tanner DJ; Slusher DL; Dibb JE; Arimoto R; Chen G; Davis D; Buhr MP; Nowak JB; Mauldin RL. (2004). CIMS measurements of
HNO3 and S02 at the South Pole during ISCAT 2000. Atmos Environ, 38, 5411-5421.
Hulsman PF; Powles PM; Gunn JM. (1983). Mortality of walleye eggs and rainbow trout yolk-sac larvae in low-pH waters of the LaCloche
Mountain area, Ontario. Trans Am Fish Soc, 112,680-683.
Hultberg H; Andersson I. (1982). Liming of acidified lakes: induced long-term changes. Water Air Soil Pollut, 18,311-331.
Humphreys FR; Lambert MJ; Kelly J. (1975). The occurrence of sulphur deficiency in forests. In: McLachlan, K. D., ed. Sulphur in Australasian
agriculture, (pp. 154-162.). Sydney, Australia: Sydney University Press.
Hungate BA; Lund CP; Pearson HL; Chapin FS. (1997). Elevated CO2 and nutrient addition alter soil N cycling and N trace gas fluxes with early
season wet-up in a California annual grassland. Biogeochem, 37,89-109.
Hunsaker C; Graham R; Turner RS; Ringold PL; Holdren GR Jr; Strickland TC. (1993). A national critical loads framework for atmospheric
deposition effects assessment: II. Defining assessment end points, indicators, and functional subregions. Environ Manage, 17,335-
341.
Hunter MD; Schultz JC. (1995). Fertilization mitigates chemical induction and herbivore responses within damaged oak trees. Ecology, 76,
1226-1232.
Hunter ML; Jones JJ; Gibbs KE; Moring JR. (1986). Duckling responses to lake acidification: do black ducks and fish compete? Oikos, 47,26-
32.
Huntington TG. (2000). The potential for calcium depletion in forest ecosystems of southeastern United States: review and analysis. Glob
Biogeochem Cycles, 14,623-638.
Huntington TG; Hooper RP; Johnson CE; Aulenbach BT; Cappellato R; Blum AE. (2000). Calcium Depletion in a Southeastern United States
Forest Ecosystem. Soil Sci Soc Am J, 64,1845-1858.
Huntrieser H; Feigl C; Schlager H; Schroder F; Gerbig C; Van Velthoven P; Flatoy F; Thery C; Petzold A; Holler H; Schumann U. (2002).
Airborne measurements of NOx, tracer species, and small particles during the European Lightning Nitrogen Oxides Experiment. J
Geophys Res, 107,4113, doi:10.1029/2000JD000209.
Huntrieser H; Schlager H; Roiger A; Lichtenstern M; Schumann U; KurzC; Brunner D; SchwierzC; Richter A; Stohl A. (2007). Lightning-
produced NOx over Brazil during TROCCINOx: airborne measurements in tropical and subtropical thunderstorms and the importance
of mesoscale convective systems. Atmos Chem Phys, 7,2987-3013.
R-41
-------
Huntzicker JJ; Hoffman RS; Ling C-S. (1978). Continuous measurement and speciation of sulfur-containing aerosols by flame photometry.
Atmos Environ, 12,83-88.
Hurd TM; Brach AR; Raynal DJ. (1998). Response of understory vegetation of Adirondack forests to nitrogen additions. Can J For Res, 28,
799-807.
Hurd TM; Raynal DJ; Schwintzer CR. (2001). Symbiotic N2 fixation of Alnus incana ssp. rugosa in shrub wetlands of the Adirondack
Mountains, New York, USA. Oecologia, 126,94-103.
Hurrell JW. (1995). Decadal trends in the North Atlantic Oscillation: regional temperatures and precipitation. Science, 269, 676-679.
Husar RB; Sullivan TJ. (1991). Historical trends in atmospheric sulfur deposition and methods for assessing long-term trends in surface water
chemistry. In: Charles DF; Christie S (Eds.), Acidic deposition and aquatic ecosystems: regional case studies (pp. 65-82). New York:
Springer-Verlag.
Husar RB; Poirot RL. (2005). DataFed and Fastnet: Tools for agile air quality analysis.
http://datafedwiki.wustl.edU/imaaes/c/c4/EM DataFed FASTNET 050720.pdf. Environ Manager, 39-41.
Hutchinson J; Maynard D; Geiser L. (1996). Air quality and lichens - a literature review emphasizing the Pacific Northwest, USA. Pacific
Northwest Region Air Resource Management Program; U.S. Forest Service; U.S. Department of Agriculture (USDA).
Hutchinson NJ; Holtze KE; Munro JR; Pawson TW. (1989). Modifying effects of life stage, ionic strength and post-exposure mortality n lethality
of H+ and Al to lake trout and brook trout. Aquat Toxicol, 15,1-26.
Hutchinson TC; Bozic L; Munoz-Vega G. (1986). Responses of five species of conifer seedlings to aluminum stress. Water Air Soil Pollut, 31,
283-294.
Hyer KE; Webb JR; Eshleman KN. (1995). Episodic acidification of three streams in Shenandoah National Park, Virginia, USA. Water Air Soil
Pollut, 85, 523-528.
Hyvarinen M; Crittenden PD. (1998). Relationships between atmospheric nitrogen inputs and the vertical nitrogen and phosphorus
concentration gradients in the lichen Cladonia portentosa. New Phytol, 140,519-530.
Hyvonen R; Agren Gl; Linder S; Persson T; Cotrufo MF; Ekblad A; Freeman M; Grelle A; Janssens IA; Jarvis PG. (2007). The likely impact of
elevated CO2, nitrogen deposition, increased temperature and management on carbon sequestration in temperate and boreal forest
ecosystems: a literature review. New Phytol, 173,463-480.
IARC. (1989). Diesel and gasoline engine exhausts and some nitroarenes: Summary of data reported and evaluation. Volume 46 of IARC
Monographs on the evaluation of carcinogenic risks to humans. Lyons, France: International Agency for Research on Cancer (IARC);
World Health Organization (WHO).
IARC. (2006). Preamble to the IARC monographs on the evaluation of carcinogenic risks to humans:
http://monoqraphs.iarc.fr/ENG/Preamble/CurrentPreamble.pdf. Lyon, France: International Agency for Research on Cancer (IARC),
World Health Organization (WHO).
ICP. (2007a). Developments in deriving critical limits and modeling critical loads of nitrogen for terrestrial ecosystems in Europe. (Alterra report
no 1382). The International Cooperative Programme (ICP) on Modelling and Mapping of Critical Loads & Levels and Air Pollution
Effects, Risks and Trends works under the Convention on Long-range Transboundary Air Pollution of the United Nations Economic
Commission for Europe (UNECE).
ICP. (2007b). CCE Progress Report 2007: Critical loads of nitrogen and dynamic modeling - Status of European critical loads with focus on
nitrogen. The International Cooperative Programme (ICP) on Modelling and Mapping of Critical Loads & Levels and Air Pollution
Effects, Risks and Trends works under the Convention on Long-range Transboundary Air Pollution of the United Nations Economic
Commission for Europe (UNECE).
lEc. (1999a). Benefits Assessment of Decreased Nitrogen Deposition to Estuaries in the United States Attributable to the 1990 Clean Air Act
Amendments, 1990-2010. Referenced in: Economic benefits of decreasing atmospheric deposition of nitrogen, in Appendix E
(Ecological Effects of Criteria Pollutants). In: The benefits and costs of the Clean Air Act 1990 to 2010: A report to Congress. (Report
# EPA-410-R-99-001). Prepared by Industrial Economics, Inc (lEc) for Office of Air and Radiation; Office of Policy; U.S. Environmental
Protection Agency.
lEc. (1999b). Characterizing the Commercial Timber Benefits from Tropospheric Ozone Reduction Attributable to the 1990 Clean Air Act
Amendments, 1990-2010: Referenced in: Timber production impacts from tropospheric ozone, in Appendix E (Ecological effects of
criteria pollutants). In: The benefits and costs of the Clean Air Act 1990 to 2010: A report to Congress. (Report# EPA-410-R-99-001).
Prepared by Industrial Economics (lEc) Inc, for Office of Air and Radiation; Office of Policy; U.S. Environmental Protection Agency.
lEc. (1999c).Characterizing the forest aesthetics benefits attributable to the 1990 Clean Air Act Amendments, 1990-2010. Referenced in:
Timber production impacts from tropospheric ozone, in Appendix E (Ecological effects of criteria pollutants). In: The benefits and
costs of the Clean Air Act 1990 to 2010: A report to Congress. (Report# EPA-410-R-99-001). Prepared by Industrial Economics (lEc),
Inc, for Office of Air and Radiation; Office of Policy; U.S. Environmental Protection Agency.
Ilvesniemi H. (1992). The combined effect of mineral nutrition and soluble aluminum on Pinus sylvestris and Picea abies seedlings. For Ecol
Manage, 51, 227-238.
Ineson P; Coward P; Benham DG; Robertson SMC. (1998). Coniferous forests as "secondary agricultural" sources of nitrous oxide. Atmos
Environ, 32, 3321-3330.
Ingersoll CG. (1986). The effects of pH, aluminum, and calcium on survival and growth of brook trout (Salvelinus fbntinalis).early life stages
[dissertation], Laramie, WY; University of Wyoming; Department of Zoology and Physiology. Ann Arbor, Ml: ProQuest; publication no.
AAD86-23090.
Ingersoll CG; Gulley DD; Mount DR; Mueller ME; Fernandez JD; Hockett JR; Bergman HL. (1990a). Aluminum and acid toxicity to two strains
of brook trout (Salvelinus fontinalis). Can J Fish Aquat Sci, 47,1641-1648.
R-42
-------
Ingersoll CG; Mount DR; Gulley DD; La Point TW; Bergman HL. (1990b). Effects of pH, aluminum, and calcium on survival and growth of eggs
and fry of brook trout (Salvelinus fontinalis). Can J Fish Aquat Sci, 47,1580-1592.
Inouye RS. (2006). Effects of shrub removal and nitrogen addition on soil moisture in sagebrush steppe. J Arid Environ, 65,604-618.
Interlandi SJ; Kilham SS. (1998). Assessing the effects of nitrogen deposition on mountain waters: a study of phytoplankton community
dynamics. Water Science and Technology, 38,139-146.
Interlandi SJ; Kilham SS; Theriot EC. (1999). Responses of phytoplankton to varied resource availability in large lakes of the Greater
Yellowstone ecosystem. Limnol Oceanogr; 44; 668-682.
Interlandi SJ; Kilham SS. (2001). Limiting resources and the regulation of diversity in phytoplankton communities. Ecology, 82,1270-1282.
IPCC. (1996). Climate Change 1995: The Science of Climate Change. In: Houghton JT (Ed.), Working Group I of the Intergovernmental Panel
on Climate Change (IPCC): Climate Change 1995: The Science of Climate Change. New York, NY: Cambridge University Press.
IPCC. (1999).Aviation and the global atmosphere. Cambridge, United Kingdom: Cambridge University Press. Available:
http://www.ipcc.ch/ipccreports/sres/aviation/index.htm.
IPCC. (2000). Good Practice Guidance and Uncertainty Management in National Greenhouse Gas Inventories: http://www.ipcc-
naaip.iaes.or.ip/public/gp/enalish/. Intergovernmental Panel on Climate Change (IPCC), in response to a request from the United
Nations Framework Convention on Climate Change (UNFCCC). New York, NY: Cambridge University Press.
IPCC. (2001a). Climate change 2001: The scientific basis. In: Houghton JT; Ding Y; Griggs DJ; Noguer M; van der Linden PJ; DaiX; Maskell K;
Johnson CA (Eds.), Working Group I of the intergovernmental panel on climate change (IPCC): Climate change 2001: The scientific
basis. New York, NY: Cambridge University Press.
IPCC. (2001b). Climate change 2001: Impacts, adaptation, and vulnerability. Contribution of working group II to the third assessment report of
the Intergovernmental Panel on Climate Change. Cambridge, United Kingdom: Cambridge University Press.
IPCC. (2007a).Summary for policymakers. In: Solomon S; Qin D; Manning M; Solomon S; Chen Z; Marquis M; Averyt KB; Tignor M; Miller HL
Eds. Climate Change 2007: The physical science basis. Contribution of Working Group I to the fourth assessment report of the
Intergovernmental Panel on Climate Change (IPCC).. http://www.ipcc.ch/ipccreports/ar4-wq 1.htm New York, NY: Cambridge
Unviersity Press.
IPCC. (2007b).CIimate change 2007: Impacts, adaptation and vulnerability. Contribution of Working Group II to the fourth assessment report of
the Intergovernmental Panel on Climate Change, http://www.ipcc.ch/ipccreports/assessments-reports.htm Cambridge, United
Kingdom: Cambridge University Press.lPCC (2007). Climate change impacts, adaptation and vulnerability. Working group II:
Contribution to the intergovernmental panel on climate change (IPCC): Fourth assessment Report. New York, NY: Cambridge
University Press.
IPCS. (2003). Nitrogenated polycyclic aromatic hydrocarbons. (Environmenal Health Criteria no 229). Geneva, Switzerland; International
Programme on Chemical Safety (IPCS); World Health Organization (WHO).
Irving PM; Miller JE; Xerikos PB. (1982). The effect of NO2 and SO2 alone and in combination on the productivity of field-grown soybeans. In:
Schneider T; Grant L (Eds.), Air Pollution by Nitrogen Oxides (pp. 521-531). Amsterdam, The Netherlands: Elsevier Scientific
Publishing Company.
Isopp H; Frehner M; Almeida JPF; Blum H; Daepp M; Hartwig UA; Luscher A; Suter S; Nosberger J. (2000). Nitrogen plays a major role in
leaves when source-sink relations change: C and N metabolism in Lolium perenne growing under free air CO2 enrichment. Aust J
Plant Physiol; 27; 851-858.
Ito M; Mitchell MJ; Driscoll CT. (2002). Spatial patterns of precipitation quantity and chemistry and air temperature in the Adirondack region of
New York. Atmos Environ, 36,1051-1062.
Ito M; Mitchell MJ; Driscoll CT; Roy KM. (2005a). Factors affecting acid neutralizing capacity in the Adirondack region of New York: a solute
mass balance approach. Environ Sci Technol; 39; 4076-4081.
Ito M; Mitchell MJ; Driscoll CT; Roy KM. (2005b). Nitrogen input-output budgets for lake-containing watersheds in the Adirondack region of New
York. Biogeochemistry, 72,283-314.
Ito M; Mitchell MJ; Driscoll CT; Newton RM; Johnson CE; Roy KM. (2007). Controls on surface water chemistry in two lake-watersheds in the
Adirondack region of New York: differences in nitrogen solute sources and sinks. Hydrological Processes, 21,1249-1264.
Ittekot V. (2003). Carbon-silicon interactions. In: Melillo JM; Field CB; Moldan B. Eds. Interactions of the major biogeochemical cycles: global
change and human impacts. Washington, D.C.: Island Press; pp. 311-322.
Izuta TM, H; Ohashi, T; Miyake, H; Totsuka, T. (1993). Effects of peroxyacetyl nitrate on the growth of petunia, kidney bean and radish plants.
Environmental Sciences, 2,63-75.
Izuta, T. (1998). Ecopysiological responses of Japanese forest free species to ozone, simulated acid rain and soil acidification. J Plant Res,
111,471-480.
Jackson DA; Harvey HH. (1995). Gradual reduction and extinction of fish populations in acid lakes. Water Air Soil Pollut, 85,389-394.
Jackson JBC; Kirby MX; Berger WH; Bjorndal KA; Botsford LW; Bourque BJ; Bradbury RH; Cooke R; Erlandson J; Estes JA; Hughes TP;
Kidwell S; Lange CB; Lenihan HS; Pandolfi JM; Peterson CH; Steneck RS; Tegner, MJ; Warner RR. (2001). Historical overfishing and
recent collapse of coastal ecosystems. Science, 293,629-637.
Jacob DJ; Bakwin PS. (1991). Cycling of NOx in tropical forest canopies. In: Rogers JE; Whitman WB. Eds. Microbial production and
consumption of greenhouse gases. Washington, DC: American Society of Microbiology; pp. 237-253.
Jacob DJ; Horowitz LW; Munger JW; Heikes BG; Dickerson RR; Artz RS; Keene WC. (1995). Seasonal transition from NOx- to hydrocarbon-
limited conditions for ozone production over the eastern United States in September. J Geophys Res, 100,9315-9324.
Jacob DJ. (1999). Introduction to atmospheric chemistry. Princeton, NJ: Princeton University Press.
Jacob DJ. (2000). Heterogeneous chemistry and tropospheric ozone. Atmos Environ, 34,2131-2159.
Jacobs TC; Gilliam JW. (1985). Riparian losses of nitrate from agricultural drainage waters. J Environ Qual, 14,472-478.
R-43
-------
Jacobson MZ; Lu R; Turco RP; Toon OP. (1996). Development and application of a new air pollution modeling system-Part I: gas-phase
simulations. Atmos Environ, 30,1939-1963.
Jacobson MZ. (1999). Isolating nitrated and aromatic aerosols and nitrated aromatic gases as sources of ultraviolet light absorption. J Geophys
Res, 104, 3527-3542.
Jacobson MZ. (2002). Atmospheric pollution: history, science, and regulation. Cambridge, UK: Cambridge University Press.
Jaegle L; Jacob DJ; Wennberg PO; Spivakovsky CM; Hanisco TF; Lanzendorf EJ; Hintsa EJ; Fahey DW; Keim ER; Proffitt MH; Atlas EL;
Flocke F; Schauffler S; McElroy CT; Midwinter C; Pfister L; Wilson JC. (1997). Observed OH and HO2 in the upper troposphere
suggests a major source from convective injection of peroxides. Geophys Res Lett, 24, 3181-3184.
Jaegle L; Jacob DJ; Brune WH; Wennberg PO. (2001). Chemistry of HOx radicals in the upper troposphere. Atmos Environ, 35,469-489.
Jaegle L; Martin RV; Chance K; Steinberger L; Kurosu TP; Jacob DJ; Modi Al; Yoboue V; Sigha-Nkamdjou L; Galy-Lacaux C. (2004). Satellite
mapping of rain-induced nitric oxide emissions from soils. J Geophys Res, 109, doi:10.1029/2004JD004787.
Jaegle L; Steinberger L; Martin RV; Chance K. (2005). Global partitioning of NOx sources using satellite observations: Relative roles of fossil
fuel combustion, biomass burning and soil emissions. Faraday Discuss,130,407-423.
Jager HJ; Klein H. (1977). Biochemical and physiological detection of sulfur dioxide injury to pea plants (Pisum sativum). Journal of Air and
Waste Management Association, 27,464-466.
Jager H-J; Klein H. (1980). Biochemical and physiological effects of SO2 on plants. Angewandte Botanik, 54,337-348.
Jagoe CH; Haines TA; Kircheis FW. (1984). Effects of reduced pH on three life stages of Sunapee char Salvelinus alpinus. Bull Environ
Contam Toxicol, 33,430-438.
Jaklevic JM; Loo BW; Fujita TY. (1981). Automatic particulate sulfur measurements with a dichotomous sampler and on-line x-ray fluorescence
analysis. Environ Sci Technol, 15,687-690.
Jakus PM; Smith VK. (1991). Measuring use and nonuse values for landscape amenities: A contingent behavior analysis of gypsy moth control
(RFF Discussion Paper #QE92-07). Washington DC: Resources for the Future (RFF).
Jansson M; Bergstrom AK; Drakare S; Blomquist P. (2001). Nutrient limitation of bacterioplankton and phytoplankton in humic lakes in northern
Sweden. Freshw Biol, 46,653-666.
Jassby AD; Reuter JE; Axler RP; Goldman CR; Hackley SH. (1994). Atmospheric deposition of N and phosphorus in the annual nutrient load of
Lake Tahoe (California-Nevada). Water Resour Res, 30,2207-2216.
Jaworski NA; Howarth RS; Hetling LJ. (1997). Atmospheric deposition of nitrogen oxides onto the landscape contributes to coastal
eutrophication in the northeast United States. Environ Sci Technol, 31,1995-2004.
Jayanty RKM. (2003). Overview of PM2.5 chemical speciation nationwide network program. Presented at the american association for aerosol
research (AAAR) conference: Particulate matter: Atmospheric sciences, exposure and the fourth colloquium on PM and human health,
held in Pittsburgh, PA; March 31 - April 4,2003 Mt Laurel, NJ; American Association for Aerosol Research (AAAR).
http://www.rti.ora/pubs/PM25 Overview Jav Apr2003.pdf
DS; Lam CCL. (1993). Assessment of the effect of acidic deposition on Canadian lakes: determination of critical loads for sulphate
deposition. Water Science and Technology, 28,183-187.
DS; Clair TA; Dillon PJ; Papineau M; Stainton MP. (1995). Trends in surface water acidification at ecological monitoring sites in
southeastern Canada (1981-1993). Water Air Soil Pollut, 85,577-582.
DS. (1997). Canadian acid rain assessment: The effects on Canada's lakes, rivers and wetlands (Volume 3, pp. 224). Burlington,
Ontario: Environment Canada.
DS; Semkin RG; Beall FD; Franklyn J. (2002).Temporal trends in water chemistry in the Turkey Lakes watershed, Ontario, Canada,
1982-1999. Water Air Soil Pollut, 2, 5-22.
DS; Clair TA; Couture S; Dillon PJ; Dupont J; Keller W; McNicol DK; Turner MA; Vet R; Weeber R. (2003). Assessing the recovery of
lakes in southeastern Canada from the effects of acidic deposition. Royal Swedish Academy of Sciences, 32,176-182.
DS; Ouimet R; Aherne J; Arp PA; Balland V; Demerchant I; Dupont J; Franklyn J; Lam CCL; Norouzian F; Watmough SA; Wong I.
(2005). Critical load values and exceedances (Chapter 8.4). In: 2004 Canadian acid deposition science assessment. Ottawa, Ontario;
Meteorological Service of Canada; Environment Canada.
Jenkin ME; Cox RA; Williams DJ. (1988). Laboratory studies of the kinetics of formation of nitrous acid from the thermal reaction of nitrogen
dioxide and water vapour. Atmos Environ, 22,487-498.
Jenkins A; Cosby BJ; Ferrier RC; Walker TAB; Miller JD. (1990). Modelling stream acidification in afforested catchments: an assessment of the
relative effects of acid deposition and afforestation. J Hydrol, 120,163-181.
Jenkins A; Cosby BJ; Ferrier RC; Larssen T; Posch M. (2003). Assessing emission reduction targets with dynamic models: deriving target load
functions for use in integrated assessment. Hydrol Earth Syst Sci, 7,609-617.
Jenkins DH; Sullivan J; Amacher GS. (2002). Valuing high altitude spruce-fir forest improvements: Importance of forest condition and
recreation activity. J For Econ, 8,77-99.
Jenkins DH; Sullivan J; Amacher G S; Nicholas NS; Reaves DW. (2002). Valuing high altitude spruce-fir forest improvements: importance of
forest condition and recreation activity. J For Econ, 8,77-99.
Jenkinson DS; Goulding K; Powlson DS. (1999). Nitrogen deposition and carbon sequestration. Nature, 400, 629-629.
Jeremiason JD; Engstrom DR; Swain EB; Nater EA; Johnson BM; Almendinger JE; Monson BA; Kolka RK. (2006). Sulfate Addition Increases
Methylmercury Production in an Experimental Wetland. Environ Sci Technol, 40,3800-3806.
Jetten MSM; Sliekers O; Kuypers MMM; Dalsgaard T; Van Niftrik L; Cirpus I; Van de Pas-Schoonen KT; Lavik G; Thamdrup B; Le Paslier D;
Op Den Camp HJM; Hulth S; Nielsen LP; Abma W; Third K; Engstrom P; Kuenen JG; Jorgensen BB; Canfield DE; Damste JSS;
Revsbech NP; Fuerst J; Weissenbach J; Wagner M; Schmidt I; Schmid M; Strous M. (2003). Anaerobic ammonium oxidation by
marine and freshwater planctomycete-like bacteria. Applied Microbiology and Biotechnology, 63,107-114.
R-44
Jeffries
Jeffries
Jeffries
Jeffries
Jeffries
Jeffries
-------
Jetten MSM; Cirpus I; Kartal B; van Niftrik L; van de Pas-Schoonen KT; Sliekers 0; Haaijer SCM; Van der Star WRL; Schmid M; Van de
Vossenberg J; Schmidt I; Harhangi HR; van Loosdrecht M; Gijs Kuenen J; Op den Camp H; Strous M. (2005). 1994-2004:10 years of
research on the anaerobic oxidation of ammonium. Biochemical Society Transactions, 33,119-123.
Jimenez P; Baldasano JM; Dabdub D. (2003). Comparison of photochemical mechanisms for air quality modeling. Atmos Environ, 37,4179-
4194.
Jobson BT; Niki H; Yokouchi Y; Bottenheim J; Hopper F; Leaitch R. (1994). Measurements of C2-C6 hydrocarbons during the Polar Sunrise
1992 Experiment: evidence for CI atom and Br atom chemistry. J Geophys Res, 99,25355-25368.
Johannessen M; Henriksen A. (1978). Chemistry of snow meltwater: Changes in concentrations during melting. Water Resour Res, 14, 615-
619.
Johansson A; Lennholm H. (2000). Influences of SO2 and O3 on the ageing of paper investigated by in situ diffuse reflectance FUR and time-
resolved trace gas analysis. Appl Surf Sci, 161,163-169.
Johansson M; Rasanen K; Merila J. (2001). Comparison of nitrate tolerance between different populations of the common frog, Rana
temporaria. Aquat Toxicol, 54,1-14.
Johansson N; Kihlstrom JE. (1975). Pikes (ESOxlucius L.) shown to be affected by lowpH values during first weeks after hatching. Environ
Res, 9,12-17.
Johansson N; Milbrink G. (1976). Some effects of acidified water on the early development of roach (Rutilus rutilus L.) and perch (Perca
fiuviatilis L.). Water Resour Bull, 12,39-48.
Johansson N; Runn P; Milbrink G. (1977). Early development of three salmonid species in acidified water. Zoon, 5,127-132.
John W; Wall SM; Ondo JL. (1988). A new method for nitric acid and nitrate aerosol measurement using the dichotomous sampler. Atmos
Environ, 22,1627-1635.
Johnson AH; Andersen SB; Siccama TG. (1994a). Acid rain and soils of the Adirondacks. I. Changes in pH and available calcium, 1930-1984.
Can J For Res, 24,39-45.
Johnson AH; Friedland AJ; Miller EK; Siccama TG. (1994b). Acid rain and soils of the Adirondacks. III. Rates of soil acidification in a montane
spruce-fir forest atWhiteface Mountain, New York. Can J For Res, 24,663-669.
Johnson CE; Johnson AH; Siccama TG. (1991a). Whole-tree clear-cutting effects on exchangeable cations and soil acidity. Soil Sci Soc Am J,
55, 502-508.
Johnson CE. (1995). Soil nitrogen status 8 years after whole-tree dear-cutting. Can J Forest. Res, 25,1346-1355.
Johnson DW; Cole DW. (1980). Mobility in soils: Relevance to nutrient transport from forest ecosystems. Environ Int, 3,79-90.
Johnson, DW; Cole DW; Bledsoe CS; Cromack K; Edmonds RL; Gessel SP; Grier CC; Richards BN; Vogt KA. (1982a). Nutrient cycling in
forests of the Pacific Northwest. In: Edmonds RL (Ed.), Analysis of coniferous forest ecosystems in the Western United States.
(U.S./IBP synthesis series: Volume 14, pp 186-232). Stroudsburg, PA: Hutchinson Ross Publishing Company.
Johnson DW; Henderson GS; Huff DD; Lindberg SE; Richter DD; Shriner DS; Todd DE; Turner J. (1982b). Cycling of organic and inorganic
sulphur in a chestnut oak forest. Oecologia, 54,141-148.
Johnson DW; Turner J; Kelly JM. (1982c). The effects of acid rain on forest nutrient status. Water Resour Res, 18,449-461.
Johnson DW. (1984). Sulfur cycling in forests. Biogeochem, 1, 29-43.
Johnson DW; Simonin HA; Colquhoun JR; Flack FM. (1987). In situ toxicity tests of fishes in acid waters. Biogeochem, 3,181-208.
Johnson DW; Kelly JM; Swank WT; Cole DW; Van Miegroet H; Hornbeck JW; Pierce RS; Van Lear D. (1988). The effects of leaching and
whole-tree harvesting on cation budgets of several forests. J Environ Qual, 17,418-424.
Johnson DW; Todd DE. (1990). Nutrient cycling in forests of Walker Branch Watershed Tennessee: roles of uptake and leaching in causing soil
changes. J Environ Qual, 19,97-104.
Johnson DW; Cresser MS; Milsson SI; Turner J; Ulrich B; Binkley D; Cole DW. (1991b). Soil changes in forest ecosystems: evidence for and
probable causes. Proceedings of the Royal Society of Edinburg Section B: Biol Sci, 97B, 81-116.
Johnson DW; Van Miegroet H; Lindberg SE; Todd DE; Harrison RB. (1991c). Nutrient cycling in red spruce forests of the Great Smoky
Mountains. Can J For Res, 21,769-787.
Johnson DW; Fernandez IJ. (1992). Soil-mediated effects of atmospheric deposition on eastern U.S. spruce-fir forests. In: Eagar C; Adams MB
(Eds.), Ecology and decline of red spruce in the eastern United States. (Series Volume 96 of: Ecology studies; pp 235-270). New York,
NY: Springer-Verlag.
Johnson DW; Lindberg SE. (1992a). Atmospheric deposition and forest nutrient cycling: A synthesis of the integrated forest study. (Series:
Billings WD; Golley F; Lange OL; Olson JS; Remmert H (Eds.), Ecological Studies: Analysis and Synthesis, Volume 91). New York,
NY: Springer-Verlag.
Johnson DW; Lindberg SE. (1992b). Nitrogen chemistry, deposition, and cycling in forests. In: Johnson DW; Lindberg SE (Eds.), Atmospheric
deposition and forest nutrient cycling: A synthesis of the integrated forest study. (Series: Billings WD; Golley F; Lange OL; Olson JS;
Remmert H (Eds.), Ecological Studies: Analysis and Synthesis, Volume 91; pp. 150-213). New York, NY: Springer-Verlag, Inc
Johnson DW; Swank WT; Vose JM. (1993). Simulated effects of atmospheric sulfur deposition on nutrient cycling in a mixed deciduous forest.
Biogeochemistry, 23,169-196
Johnson DW; Binkley D; Conklin P. (1995). Simulated effects of atmospheric deposition, harvesting, and species change on nutrient cycling in
a loblolly pine forest. For Ecol Manage, 76,29-45.
Johnson DW; Susfalk RB; Brewer PF. (1996). Simulated responses of red spruce forest soils to reduced sulfur and nitrogen deposition. J
Environ Qual, 25,1300-1309.
Johnson DW; Mitchell MJ. (1998). Responses of forest ecosystems to changing sulfur inputs. In: Maynard DG (Ed.), Sulfur in the environment
(pp. 219-262). New York, NY: Marcel Dekker, Inc.
Johnson DW; Curtis PS. (2001). Effects of forest management on soil C and N storage: Metaanalysis. For Ecol Manage, 140, 227-238.
R-45
-------
Johnson DW; Susfalk RB; Caldwell TG; Murphy JD; Miller WW; Walker RF. (2004). Fire effects on carbon and nitrogen budgets in forests.
Water Air Soil Pollut, 4,263-275.
Johnson DW. (2005). Comments on "Thirty years of change in forest soils of the Allegheny Plateau, Pennsylvania." Soil Sci. Soc. Amer. J. 69:
2077, doi: 10.2136/sssaj2005.01971.
Johnson DW; Murphy JF; Susfalk RB; Caldwell TG; Miller WW; Walker RF; Powers RF. (2005a). The effects of wildfire, salvage logging, and
post-fire N-fixation on the nutrient budgets of a Sierran forest. For Ecol Manage, 220,155-165.
Johnson DW; Hoylman AM; Ball JT; Walker RF. (2006). Ponderosa pine responses to elevated CO2 and nitrogen fertilization. Biogeochem, 77,
157-175.
Johnson DW; Todd DE; Trettin CF; Mulholland PJ. (2008). Changes in soil exchangeable K+, Ca2+, and Mg2+ in forests of Walker Branch
Watershed, 1972-2004. Soil Sci Soc Am J, (In press).
Johnson GV. (1991). General Model for Predicting Crop Response to Fertilizer. Agron J, 83, 367-373.
Johnson LB; Richards C; Host GE; Arthur JW. (1997). Landscape influences on water chemistry in Midwestern stream ecosystems. Freshw
Biol, 37,193-208.
Johnson MG; Phillips DL; Tingey DT; Storm MJ. (2000). Effects of elevated C02, N-fertilization, and season on survival of ponderosa pine fine
roots. Can J For Res, 30,220-228.
Johnson NM; Driscoll CT; Eaton JS; Likens GE; McDowell WH. (1981). "Acid Rain", dissolved aluminum and chemical weathering at the
Hubbard Brook Experimental Forest, New Hampshire. Geochim Cosmochim Acta, 45,1421-1437.
Johnston RJ; Besedin EY; Ranson MH; Helm EC. (2005b). What determines willingness to pay per fish? A meta-analysis of recreational fishing
values. Mar Resource Econ, 21,1-32.
Johnson T. (2002). A guide to selected algorithms, distributions, and databases used in exposure models developed by the Office of Air Quality
Planning and Standards [revised draft], http://www.epa.aov/ttn/fera/data/human/report052202.pdf. Chapel Hill, NC Piscataway, New
Jersey: TRJ Environmental, Inc & Environmental and Occupational Health Sciences Institute, prepared for Office of Air Quality
Planning and Standards (OAQPS); U.S. Environmental Protection Agency.
Jones JA; Swanson FJ; Wemple BC; Snyder KU. (2000). Effects of roads on hydrology, geomorphology, and disturbances patches in stream
networks. Conserv Biol, 14,76-85.
Jones ME; Paine TD; Fenn ME; Poth MA. (2004). Influence of ozone and nitrogen deposition on bark beetle activity under drought conditions.
For Ecol Manage, 200,67-76.
Joos F; Prentice IC; House Jl. (2002). Growth enhancement due to global atmospheric change as predicted by terrestrial ecosystem models:
consistent with U.S. forest inventory data. Glob Change Biol, 8,299-303.
Jordan TE; Correll DL; Weller DE. (1993). Nutrient interception by a riparian forest receiving inputs from adjacent cropland. J Environ Qual, 22,
467-473.
Jordan TE; Weller DE. (1996). Human contributions to terrestrial nitrogen flux. Bioscience, 46,655-664.
Joseph DW; Spicer CW. (1978). Chemiluminescence method for atmospheric monitoring of nitric acid and nitrogen oxides. Anal Chem, 50,
1400-1403.
Joslin JD; Mays PA; Wolfe MH; Kelly JM; Garber RW; Brewer PF. (1987). Chemistry of tension lysimeter water and lateral flow in spruce and
hardwood stands. J Environ Qual, 16,152-160.
Joslin JD; Wolfe MH. (1988). Responses of red spruce seedlings to changes in soil aluminum in six amended forest soil horizons. Can J For
Res, 18,1614-1623.
Joslin JD; Wolfe MH. (1989). Aluminum effects on northern red oak seedling growth in six forest soil horizons. Soil Sci Soc Am J, 53,274-281.
Joslin JD; Wolfe MH. (1992a). Red spruce soil solution chemistry and root distribution across a cloud water deposition gradient. Can J For Res,
22, 893-904.
Joslin JD; Wolfe MH. (1992b). Tests of the use of net throughfall sulfate to estimate dry and occult sulfur deposition. Atmos Environ, 63-72.
Joslin JD; Kelly JM; van Miegroet H. (1992). Soil chemistry and nutrition of North American spruce-fir stands: evidence for recent change. J
Environ Qual, 21,12-30.
Joslin JD; Wolfe MH. (1994). Foliar deficiencies of mature Southern Appalachian red spruce determined from fertilizer trials. Soil Sci Soc Am J,
58,1572-1579.
Jouen S; Jean M; HanNOyer B. (2004). Atmospheric corrosion of nickel in various outdoor environments. Corros Sci, 46,499-514.
Jovan S; McCune B. (2005). Air-quality bioindication in the greater central valley of California, with epiphytic macrolichen communities. Ecol
Appl, 15,1712-1726.
Jovan S; McCune B. (2006). Using epiphytic macrolichen communities for biomonitoring ammonia in forests of the greater Sierra Nevada,
California. Water Air Soil Pollut, 170,69-93.
JPL. (2003). Chemical kinetics and photochemical data for use in atmospheric studies. (No. Evaluation number 15. JPL publication 06-2).
Pasadena, CA; Jet Propulsion Laboratory (JPL); National Aeronautics and Space Administration (NASA).
Juggins S; Omerod SJ; Harriman R. (1995). Relating critical loads to aquatic biota. In: Critical loads of acid deposition for UK freshwaters (pp 9-
13). London UK: Department of the Environment.
Juice SM; Fahey TJ; Siccama TG; Driscoll CT; Denny EG; EagarC; Cleavitt NL; Minocha R; Richardson AD. (2006). Response of sugar maple
to calcium addition to Northern Hardwood Forest. Ecology, 87,1267-1280.
JungeCE; Chagnon CW; Manson JE. (1961). Stratospheric aerosols. Journal of Meteorol, 18, 81-108.
Justic D; Rabalais NN; Turner RE; Wiseman WJ Jr. (1993). Seasonal coupling between riverborne nutrients, net productivity and hypoxia. Mar
Pollut Bull, 26,184-189.
Justic D; Rabalais NN; Turner RE; Dortch Q. (1995a). Changes in nutrient structure of river-dominated coastal waters: Stoichiometric nutrient
balance and its conseqeuences. Estuarine Coastal & Shelf Science, 40, 339-356.
R-46
-------
Justic D; Rabalais NN; Turner RE. (1995b). Stoichiometric nutrient balance and origin of coastal eutrophication. Marine Pollution Bulletin, 30,
41-46.
Justic D; Rabalais NN; Turner RE. (1997). Impacts of climate change on net productivity of coastal waters: implications for carbon budget and
hypoxia. Climate Research, 8,225-237.
Kahl JS; Norton SA; Cronan CS; Fernandez IJ; Bacon LC; Haines TA. (1991). Maine. In: Charles DF, Christie S (Eds.), Acidic deposition and
aquatic ecosystems: regional case studies (pp. 203-235). New York, NY: Springer-Verlag.
Kahl JS; Norton SA; Haines TA; Rochette EA; Heath RH; Nodvin SC. (1992). Mechanisms of episodic acidification in low-order streams in
Maine, USA. Environ Pollut (1987), 78,37-44.
Kahl JS; Norton SA; Fernandez IJ; Nadelhoffer KJ; Driscoll CT; Aber JD. (1993a). Experimental inducement of nitrogen saturation at the
watershed scale. Environ Sci Technol, 27,565-568.
Kahl JS; Haines TA; Norton SA; Davis RB. (1993b). Recent trends in the acid-base status of surface waters in Maine, USA. Water Air Soil
Pollut, 67,281-300.
Kahl JS; Scott M. (1994). High elevation lake monitoring in Maine: 1986-1989. August, ME: Maine Department of Environmental Protection.
Kahl J; Norton S; Fernandez I; Rustad L; Handley M. (1999). Nitrogen and sulfur input-output budgets in the experimental and reference
watersheds, Bear Brook Watershed in Maine (BBWM). Environ Monit Assess, 55,113-131
Kahl S. (1999). Responses of Maine surface waters to the Clean Air Act Amendments of 1990. Final report; EPA project CX826563-01-0.
Orono, ME; Gulf of Maine Council on the Marine Environment; Water Research Institute; University of Maine-Orono.
Kain JS; Fritsch JM. (1993). Convective parameterization in mesoscale models: The Kain-Fritsch scheme. In: Emanuel KA; Raymond DJ
(Eds.),The representation of cumulus convection in numerical models. Meteorological Monographs, (Volume 24, pp 165-170). Boston,
MA: American Meteorological Society.
Kainulainen P; Holopainen JK; Oksanen J. (1995). Effects of S02 on the concentrations of carbohydrates and secondary compounds in Scots
Pine (pinus-sylvestris L) and Norway Spruce (picea-abies(L) karst) seedlings. NewPhytol, 130,231-238.
Kaiser EW; Wu CH. (1977). A kinetic study of the gas phase formation and decomposition reactions of nitrous acid. J PhysChem, 81,1701-
1706.
Kalbitz K; Solinger S; Park JH; Michalzik B; Matzner E. (2000). Controls on the dynamics of dissolved organic matter in soils: A review. Soil Sci,
165, 277-304.
Kameda T; Inazu K; Bandow H; Sanukida S; Maeda Y. (2004). Diurnal change of direct-acting mutagenicity of soluble organic fraction of
airborne particles collected at Southern Osaka: Correlation between the mutagenicity, particles-associated nitroarenes, and gaseous
emission. Atmos Environ, 38,1903-1912.
Kamman NC; Lorey PM; Driscoll CT; Estabrook R; Major A; Pientka B; Glassford E. (2004). Assessment of mercury in waters, sediments and
biota of New Hampshire and Vermont lakes sampled using a geographically randomized design. Environ Toxicol Chem, 23(5), 1172-
1186.
Kana TM; Sullivan MB; Cornwell JC; Groszkowski. (1998). Denitrification in estuarine sediments determined by membrane inlet mass
spectrometry. Limnol Oceanogr, 43,334-339.
Kane DA; Rabeni CF. (1987). Effects of aluminum and pH on the early life stages of smallmouth bass (Micropterus dolemieul). Water Res, 21,
633-639.
Kane SM; Caloz F; Leu M-T. (2001). Heterogeneous uptake of gaseous N2O5 by (NH4)2S04, NH4HSO4, and H2SO4 aerosols. J Phys Chem,
105, 6465-6470.
Kaoru Y. (1995). Measuring marine recreation benefits of water quality improvements by the nested random utility model. Resource and
Energy Economics, 17,119-136.
Kaoru Y; Smith VK; Liu JL. (1995). Using random utility models to estimate the recreational value of estuarine resources. Am J Agric Econ, 77,
141-151.
Karnosky DF. (2003). Impacts of elevated atmospheric CO2 on forest trees and forest ecosystems: Knowledge gaps. Environ Int, 29,161-169.
Karr JR; Chu LW. (1999). Restoring life in running rivers: better biological monitoring. Washington, DC: Island Press.
Kasibhatla P; Chameides WL. (2000).Seasonal modeling of regional ozone pollution in the eastern United States. Geophys Res Lett, 27,1415-
1418.
Kaufmann PR; Herlihy AT; Elwood JW; Mitch ME; Overton WS; Sale MJ; Messer JJ; Cougan KA; Peck DV; Reckhow KH; Kinney AJ; Christie
SJ; Brown DD; Hagley CA; Jager HI. (1988). Chemical characteristics of streams in the Mid-Atlantic and Southeastern United States.
Volume I: Population descriptions and physico-chemical relationships (No. EPA/600/3-88/021a). Washington, DC: U.S. Environmental
Protection Agency.
Kaufmann PR; Herlihy AT; Mitch ME; Messer JJ; Overton WS. (1991). Stream chemistry in the eastern United States 1. synoptic survey
design, acid-base status, and regional patterns. Water Resour Res, 27,611-627.
Kauppi PE; Mielikainen K; Kuusela K. (1992). Biomass and carbon budget of European forests, 1971 to 1990. Science, 256,70-74.
Kebabian PL; Robinson WA; Freedman A. (2007). Optical extinction monitor using cw cavity enhanced detection. Rev Sci Instrum, 78,063102.
Keck L; Wittmaack K. (2005). Effect of filter type and temperature on volatilisation losses from ammonium salts in aerosol matter. Atmos
Environ, 39,4093-4100.
Keck L; Wittmaack K. (2006). Simplified approach to measuring semivolatile inorganic particulate matter using a denuded cellulose filter without
backup filters. Atmos Environ, 40,7106-7114.
Keene WC; Jacob DJ; Fan S-M. (1996). Reactive chlorine: a potential sink for dimethylsulfide and hydrocarbons in the marine boundary layer.
Atmos Environ, 306, i-iii.
Keene WC; Savoie DL. (1998). The pH of deliquesced sea-salt aerosol in polluted marine air. Geophys Res Lett, 25,2181-2194.
R-47
-------
Keene WC; Sander R; Pszenny MP; Vogt R; Crutzen PJ; Galloway JN. (1998). Aerosol pH in the marine boundary layer: a review and model
evaluation. Journal of Aerosol Science, 29,339-356.
Keene WC; Khalil MAK; Erickson DJ III; McCulloch A; Graedel TE; Lobert JM; Aucott ML; Gong SL; Harper DB; Kleiman G; Midgley P; Moore
RM; SeuzaretC; Sturges WT; BenkovitzCM; Koropalov V; Barrie LA; Li YF. (1999). Composite global emissions of reactive chlorine
from anthropogenic and natural sources: reactive chlorine emissions inventory. J Geophys Res, 104.8429-8440.
Keene WC; Pszenny AAP; Maben JR; Sander R. (2002). Variation of marine aerosol acidity with particle size. Geophys Res Lett, 29,1101,
doi: 10.1029/2001GL013881.
Keene WC; Pszenny AAP; Maben JR; Stevenson E; Wall A. (2004). Closure evaluation of size-resolved aerosol pH in the New England coastal
atmosphere during summer. J Geophys Res, 109, doi:10.1029/2004JD004801.
Keene WC; Lobert RM; Crutzen PJ; Maben JR; Scharffe DH; Landmann T; Hely C; Brain C. (2006). Emissions of major gaseous and
particulate species during experimental burns of southern African biomass. J Geophys Res, 111, doi: 10.1029/2005JD006319.
Keene WC; Stutz J; Pszenny AAP; Maben JR; Fischer EV; Smith AM; Von Glasow R; Pechtl S; Sive BC; Varner RK. (2007). Inorganic chlorine
and bromine in coastal New England air during summer. J Geophys Res, 112, doi:10.1029/2006JD007689.
Keller JK; Bridgham SD; Chapin CT; Iversen CM. (2005). Limited effects of six years of fertilization on carbon mineralization dynamics in a
Minnesota fen. Soil Biol Biochem, 37,1197-1204.
Keller M; Kaplan WA; Wofsy SC; da Costa JM. (1988). Emissions of N2O from tropical forest soils: Response to fertilization with NH4, NO3",
and PO43". J Geophys Res, 93,1600-1604.
Keller W; Pitblado JR. (1986). Water quality changes in Sudbury area lakes: a comparison of synoptic surveys in 1974-1976 and 1981 -1983.
Water Air Soil Pollut, 29,285-296.
Keller W; Pitblado JR; Conroy Nl. (1986). Water quality improvements in the Sudbury, Ontario, Canada area related to reduced smelter
emissions. Water Air Soil Pollut, 31, 765-774.
Keller W; Yan D. (1991). Recovery of crustacean zooplankton species richness in Sudbury area lakes following water quality improvements.
Can J Fish Aquat Sci, 48,1635-1644.
Keller W. (1992). Introduction and overview to aquatic acidificaiton studies in the Sudbury, Ontario, Canada, area. Can J Fish Aquat Sci, 49, 3-
7.
Keller W; Gunn JM. (1995). Lake water quality improvements and recovering aquatic communities. In: Gunn JM (Ed.), Restoration and
recovery of an industrial region: progress in restoring the smelter-damaged landscape near Sudbury, Canada (pp. 67-80). New York,
NY: Springer-Verlag.
Kellman M; Carty A. (1986). Magnitude of nutrient influxes from atmospheric sources to a central american Pinus caribaea woodland. J Appl
Ecol, 23, 211-226.
Kelly CA; Rudd JWM; Hesslein RH; Schindler DW; Dillon PJ; Driscoll CT; Gherini SA; Hecky RE. (1987). Prediction of biological acid
neutralization in acid sensitive lakes. Biogeochem, 3,129-140.
Kelly CA; Rudd JW; Holoka MH. (2003). Effect of pH on mercury uptake by an aquatic bacterium: implications for Hg cycling. Environ Sci
Technol, 37,2941-2946.
Kelly J; Lambert M. (1972). The relationship between sulphur and nitrogen in the foliage of Pinus radiata. Plant Soil, 37,395-407.
Kelly RH; Parton WJ; Hartman MD; Stretch LK; Ojima DS; Schimel DS. (2000). Intra- and interannual variability of ecosystem processes in
shortgrass steppe. J Geophys Res, 105,20093-20100.
Kelly TJ; Spicer CW; Ward GF. (1990). An assessment of the luminol chemiluminescence technique for measurement of NO2 in ambient air.
Atmos Environ, 24,2397-2403.
Kelso JRM; Jeffries DS. (1988). Response of headwater lakes to varying atmospheric deposition in north-central Ontario. Can J Fish Aquat Sci,
45,1905-1911.
Keltjens WG; Van Loenen E. (1989). Effects of aluminum and mineral nutrition on growth and chemical composition of hydroponically grown
seedlings of five different forest tree species. Plant Soil, 119,39-40.
Kemp WM; Twilley RR; Stevenson JC; Boynton WR; Means JC. (1983). The decline of submerged vascular plants in upper Chesapeake Bay:
summary of results concerning possible causes. Marine Technology Society Journal, 17, 78-89.
Kemp WM; Sampou P; Caffrey J; Mayer M; Henriksen K; Boynton WR. (1990). Ammonium recycling versus denitrification in Cheasapeake Bay
sediments. Limnol Oceanogr, 35,1545-1563.
Kendall C; Silva SR; Chang CC; Campbell DH; Burns DA; Shanley JB. (1995). Use of oxygen and nitrogen isotopes to trace sources of nitrate
during snowmelt in forested catchments. In: Tonnessen KA, Williams MW, Tranter M (Eds.), Biogeochemistry of Seasonally Snow-
Covered Catchments (Vol. Publication no. 228, pp. 339-347): International Association of Hydrologic Sciences.
Kenk G; Fischer H. (1988). Evidence from nitrogen fertilisation in the forests of Germany. Environ Pollut, 54,199-218.
Kesik M; Ambus P; Baritz R. (2005). Inventory of N2O and NO emissions from European forest soils. Biogeosciences, 2,353-375.
Kesik M; Briiggemann N; Forkel R; Kiese R; Knoche R; Li C; Seufert G; Simpson D; Butterbach-Bahl K. (2006). Future scenarios of N2O and
NO emissions from European forest soils. J Geophys Res, 111, doi:10.1029/2005JG000115.
Kessler C; Niederau A; Scholz W. (2006). Estimation of NO2/NOX relations of traffic emissions in Baden-Wurttemberg from 1995 to 2005. Paper
presented at the 2nd conference of Environment & Transport, (Joint with 15th conference of Transport and air pollution). Held in
Reims, France; June 12-14,2006. Atmospheric Compositon Change: The European Network of Excellence (ACCENT).
Khan MA; Hany O; Azeem M; Hasan SA; Jabeen S; Khan MA. (2007). Pollution Profile of Malir River. Pak J Sci Ind Res, 50,96-100.
Khanna PK; Raison RJ. (1986). Effect of fire intensity on solution chemistry of surface soil under a Eucalyptus pauciflora forest. Aust J Soil
Res, 24,423-434.
Khanna PK; Raison RJ; Falkiner RA (1994). Chemical properties of ash derived from Eucalyptus litter and its effects on forest soils. For Ecol
Manage, 66,107-125.
R-48
-------
KieftT; Soroker LE; Firestone MK. (1987). Microbial biomass response to a rapid increase in water potential when dry soil is wetted. Soil Biol
Biochem, 19,119-126.
Kielland, K. (1994). Amino acid absorption by arctic plants: Implications for plant nutrition and nitrogen cycling. Ecology, 75,2373-2383.
Kielland, K. (1995). Landscape patterns of free amino acids in arctic tundra soils. Biogeochem, 31,85-98.
Killingbeck KT. (1996). Nutrients in senesced leaves: keys to the search for potential resorption and resorption proficiency. Ecology, 77,1716-
1727.
Kim BM; Lester J; Tisopulos L; Zeldin MD. (1999). Nitrate artifacts during PM2.5 sampling in the South Coast Air Basin of California. J Air Waste
Manage Assoc, 49(special issue), PM142-153.
Kim EY; Murakami T; Saeki K; Tatsukawa R. (1996). Mercury levels and its chemical form in tissues and organs of seabirds. Arch Environ
Contam Toxicol, 30,259-266.
Kim HJ; Helfand GE; Howitt RE. (1998). An economic analysis of ozone control in California's San Joaquin Valley. J Agric Resource Econ, 23,
55-70.
Kim H; Lee J-T; Hong Y-C; Yi S-M; Kim Y. (2004). Evaluating the effect of daily PM10 variation on mortality. Inhalation Toxicology, 1,55-58.
Kim K-H; Kim M-Y. (2001). Comparison of an open path differential optical absorption spectroscopy system and a conventional in situ
monitoring system on the basis of long-term measurements of SO2, NO2, and O3. Atmos Environ, 35,4059-4072.
Kim S-T; Maedab Y; Tsujino Y. (2004). Assessment of the effect of air pollution on material damages in northeast Asia. Atmos Environ, 38,37-
48.
Kim S-W; Heckel A; McKeen SA; Frost GJ; Hsie EY; Trainer MK; Richter A; Burrows JP; Peckham SE; Grell GA. (2006). Satellite-observed
U.S. power plant NOx emission reductions and their impact on air quality. Geophys Res Lett, 33, doi:10.1029/2006GL027749.
Kimmel WG; Murphey DJ; Sharpe WE; DeWalle DR. (1985). Macroinvetebrate community structure and detritus processing rates in two
southwestern Pennsylvania streams acidified by atmospheric deposition. Hydrobiologia, 124,97-102.
Kincheloe JW; Wedemeyer GA; Koch DL. (1979). Tolerance of developing salmonid eggs and fry to nitrate exposure. Bull Environ Contam
Toxicol, 23, 575-578.
King GM; Schnell S. (1994). Ammonium and nitraite inhibition of methane oxidation by Methylobacter albus BG8 and Methylosinus
trichosporium OB3b at Low Methane Concentrations. Appl Environ Microbiol, 60,3508-3513.
King GM; Schnell S. (1998). Effects of ammonium and non-ammonium salt additions on methane oxidation by Methylosinus trichosporium
OB3b and Maine forest soils. Appl Environ Microbiol, 64,253-257.
King JK; Kostka JE; Frischer ME; Saunders FM. (2000). Sulfate-Reducing Bacteria Methylate Mercury at Variable Rates in Pure Culture and in
Marine Sediments. Appl Environ Microbiol, 66,2430-2437.
King JK; Kostka JE; Frischer ME; Saunders FM; Jahnke RA. (2001a). A Quantitative Relationship that Demonstrates Mercury Methylation
Rates in Marine Sediments Are Based on the Community Composition and Activity of Sulfate-Reducing Bacteria. Environ Sci Technol,
35, 2491-2496.
King JS; Pregitzer KS; Zak DR; Kubiske ME; Holmes WE. (2001b). Correlation of foliage and litter chemistry of sugar maple, Acer saccharum,
as affected by elevated CO2 and varying N availability, and effects on decomposition. Oikos, 94,403-416.
Kingston JC; Birks HJB. (1990). Dissolved organic carbon reconstruction from diatom assemblages in PIRLA project lakes, North America.
Philosophical Transactions of the Royal Society of London, 327, 279-288.
Kingston JC; Cook RB; Kreis RG Jr; Camburn KE; Norton SA; Sweets PR; Binford MW; Mitchell MJ; Schindler SC; Shane LCK; King GA.
(1990). Paleoecological investigation of recent lake acidification in the northern Great Lakes states. J Paleolimnol, 4,153-201.
Kinney KK; Lindroth RL; Jung SM; Nordheim EV. (1997). Effects of CO2 and NO3" availability on deciduous trees: Phytochemistry and insect
performance. Ecology, 78,215-230.
Kirchner JW; Lydersen E. (1995). Base Cation Depletion and Potential Long-Term Acidification of Norwegian Catchments. Environ Sci
Technol, 29,1953-1960.
Kirchstetter TW; Harley RA. (1996). Measurement of nitrous acid in motor vehicle exhaust. Environ Sci Technol, 30,2843-2849.
Kireev SV; Shnyrev SL; Zhiganov AA. (1999). A laser fluorimeter for NO and NO2 in atmosphere. Instruments and Experimental Techniques,
42,701-703.
Kirschbaum MUF. (1994). THe temperature dependence of soil organic matter decomposition, and the effect of global warming on soil organic
C storage. Soil Biol Biochem, 27,753-760.
Kittelson DB; McKenzie R; Vermeersch M; Dorman F; Pui D; Linne M; Liu B; Whitby K. (1978). Total sulfur aerosol concentration with an
electrostatically pulsed flame photometric detector system. Atmos Environ, 12,105-111.
Kittelson DB; Watts WF; Johnson JP; Rowntree C; Payne M; Goodier S; Warrens C; Preston H; Zink U; Ortiz M. (2006). On-road evaluation of
two Diesel exhaust aftertreatment devices. Journal of Aerosol Science, 37,1140-1151.
Kitzler B; Zechmeister-Boltenstern S; Holtermann C; Skiba U; Butterbach-Bahl K. (2006). Nitrogen oxides emission from two beech forests
subjected to different nitrogen loads. Biogeosciences, 3,293-310.
Kjonaas OJ; Wright RF. (1998). Nitrogen leaching from N limited forest ecosystems: The MERLIN model applied to Gardsjon, Sweden. Hydrol
Earth Syst Sci, 2,415-429.
Klauda RJ; Palmer RE. (1987). Responses of blueback herring eggs and larvae to pulses of acid and aluminum. Trans Am Fish Soc, 116,561-
569.
Klauda RJ; Palmer RE; Lenkevich MJ. (1987). Sensitivity of early life stages of blueback herring to moderate acidity and aluminum in soft
freshwater. Estuaries, 10,44-53.
Kleb HR; Wilson SD. (1997). Vegetation effects on soil resource heterogeneity in prairie and forest. American Naturalist, 150, 283-298.
Kleeman MJ; Hughes LS; Allen JO; Cass GR. (1999). Source contributions to the size and composition distribution of atmospheric particles:
Southern California in September 1996. Environ Sci Tech, 33,4331-4341.
R-49
-------
Kleffmann J; Becker KH; Wiesen P. (1998). Heterogeneous NO2 conversion processes on acid surfaces: possible atmospheric implications.
Atmos Environ, 32,2721-2729.
Kleindienst TE. (1994). Recent developments in the chemistry and biology of peroxyacetyl nitrate. Res Chem Intermed, 20,335-384.
Kleinman LI. (1991). Seasonal dependence of boundary-layer peroxide concentration: the low and high NOx regimes. J Geophys Res, 96,
20721-20733.
Kleinman LI; Daum PH; Imre DG; Lee JH; Lee Y-N; Nunnermacker LJ; Springston SR; Weinstein-Lloyd J; Newman L. (2000). Ozone
production in the New York City urban plume. J Geophys Res, 105,14495-14511.
Kleinman LI; Daum PH; Lee Y-N; Nunnermacker LJ; Springston SR; Weinstein-Lloyd J; Rudolph J. (2001). Sensitivity of ozone production rate
to ozone precursors. Geophys Res Lett, 28,2903-2906.
Klemedtsson L; Klemedtsson AK; Moldan F; Weslien P. (1997). Nitrous oxide emission from Swedish forest soils in relation to liming and
simulated increased N-deposition. Biol Fertil Soils, 25,290-295.
Kley D; Crutzen PJ; Smit HGJ; Vomel H; Oltmans SJ; Grassl H; Ramanathan V. (1996). Observations of near-zero ozone concentrations over
the convective Pacific: effects on air chemistry. Science, 274,230-233.
Klironomos JN; Rillig MC; Allen MF; Zak DR; Kubiske M; Pregitzer KS. (1997). Soil fungal-arthropod responses to Populus tremuloides grown
under enriched atmospheric CO2 under field conditions. Glob Change Biol, 3,473-478.
Klopatek JM; Barry MJ; Johnson DW. (2006). Potential canopy interception of nitrogen in the Pacific Northwest, USA. For Ecol Manage, 234,
344-354.
Knapp AK; Smith MD (2001). Variation among biomes in temporal dynamics of aboveground primary production. Science, 291,481-484.
Knight H. (1966). Loss of nitrogen from the forest floor by burning. For Chron, 42.149-152.
Knights JS; Zhao FJ; Spiro B; McGrath SP. (2000). Long-term effects of land use and fertilizer treatments on sulfur cycling. J Environ Qual, 29,
1867-1874.
Knoepp JD; Swank WT. (1994). Long-term soil chemistry changes in aggrading forest ecosystems. Soil Sci Soc Am J, 58,325-331.
Knorr M; Frey SD; Curtis PS. (2005). Nitrogen additions and litter decomposition: A meta-analysis. Ecology, 86,3252-3257.
Kobe RK; Likens GE; Eagar C. (2002). Tree seedling growth and mortality responses to manipulations of calcium and aluminum in a northern
hardwood forest. Can J Forest Res, 32,954-966.
Koch MS; Mendelssohn IA; McKee KL. (1990). Mechanism for the hydrogen sulfide-induced growth limitation in wetland macrophytes. Limnol
Oceanogr, 35,399-408.
Kochy M; Wilson SD. (2001). Nitrogen deposition and forest expansion in the northern Great Plains. J Ecol, 89,807-817.
Koerselman W. (1989). Groundwater and surface water hydrology of a small groundwater-fed fen. Wetlands Ecol Manag, 1,31-43
Koerselman W; Van Kerkhoven MB; Verhoeven JTA. (1993). Release of inorganic N, P, and K in peat soils; effect of temperature, water
chemistry and water level. Biogeochemistry, 20,63-81.
Kooijman AM; Bakker C. (1994). The acidification capacity of wetland bryophytes as influenced by simulated clean and polluted rain. Aquatic
Botany, 48,133-144.
Kopp RJ; Vaughn WJ; Hazilla M; Carson R. (1985). Implications of environmental policy for U.S. agriculture: The case of ambient ozone
standards. J Environ Manage, 20,321-331.
Korb JE; Ranker TA. (2001). Changes in stand composition and structure between 1981 and 1996 in four Front Range plant communities in
Colorado. Plant Ecology, 157,1-11.
Korhonen P; Kulmala M; Laaksonen A; Viisanen Y; McGrawR; Seinfeld JH. (1999). Ternary nucleation of H2SO4, NH3, and H20 in the
atmosphere. J Geophys Res, 104,26349-26353.
Kotchenruther RA; Jaffe DA; Jaegle L. (2001). Ozone photochemistry and the role of peroxyacetyl nitrate in the springtime northeastern Pacific
troposphere: results from the Photochemical Ozone Budget of the Eastern North Pacific Atmosphere (PHOBEA).campaign. J Geophys
Res, 106,28731-28742.
Koutrakis P; Wolfson JM; Slater JL; Brauer M; Spengler JD; Stevens RK; Stone CL. (1988a). Evaluation of an annular denuder/filter pack
system to collect acidic aerosols and gases. Environ Sci Technol, 22,1463-1468.
Koutrakis P; Wolfson JM; Spengler JD. (1988b). An improved method for measuring aerosol strong acidity: results from a nine-month study in
St. Louis, Missouri and Kingston, Tennessee. Atmos Environ, 22,157-162.
Koutrakis P; Thompson KM; Wolfson JM; Spengler JD; Keeler GJ; Slater JL. (1992). Determination of aerosol strong acidity losses due to
interactions of collected particles: results from laboratory and field studies. Atmos Environ, 26,987-995.
Kozlowski TT. (1985). SO2 effects on plant community structure. In: Winner WE; Mooney HA; Goldstein RA (Eds.), Sulfur dioxide and
vegetation: physiology, ecology, and policy issues, (pp 431-453). Stanford, CA: Stanford University Press.
Krabbenhoft DP; Wiener JG; Brumbaugh WG; Olson ML; DeWild JF; Sabin TJ. (1999). A National Plot Study of Mercury Contamination of
Aquatic Ecosystems along Multiple Gradients In Morganwalp DW, Buxton HT (Eds.), U.S. Geological Survey Toxic Substances
Hydrology Program: Proceedings of the technical meeting; held in Charleston, SC March 8-12,1999. (Volume 2 of 3: Contamination of
hydrologic systems and related ecosystems. (U.S. Geological Survey Water resources investigations report #99-4018-B, pp. 147-160).
West Trenton, NJ; U.S. Geological Survey.
Krabbenhoft DP; Olson ML; Dewild JF; Clow DW; Striegel RG; Dornblaser MM; VanMetre P. (2002). Mercury loading and methylmercury
production and cycling in high-altitude lakes from the western United States. Water Air Soil Pollut Focus, 2,233-249.
Kramar D; Goodale WM; Kennedy LM; Carstensen LW; Kaur T. (2005). Relating land cover characteristics and common loon mercury levels
using geographic information systems. Ecotoxicology, 14,253-262.
Kramer A; Holmes TP; Haefel M. (2003). Contingent valuation of forest ecosystem protection. In: Sills EO; Abt KL (Eds.), Forests in a Market
Economy (pp. 303-320). Dordrecht, The Netherlands: Kluwer Academic Publishers.
R-50
-------
Kramer R; Mercer E. (1997). Valuing a global environmental good: U.S. residents willingness to pay to protect tropical rain forests. Land Econ,
73,196-210.
Kramer RA; Eisen-Hecht Jl. (2002). Estimating the economic value of water quality protection in the Catawba River basin. Water Resour Res,
38,1183-1191.
Kratz KW; Cooper SD; Melack JM. (1994). Effects of single and repeated experimental acid pulses on invertebrates in a high altitude Sierra
Nevada stream. Freshw Biol, 32,161-183.
Kress LW; Skelly JM. (1982). Response of several eastern forest tree species to chronic doses of ozone and nitrogen dioxide. Plant Dis, 66,
1149-1152.
Kretser W; Gallagher J; Nicolette J. (1989). Adirondack Lakes Study, 1984-1987: An Evaluation of Fish Communities and Water Chemistry:
Ray Brook, NY; prepared for: Adirondack Lakes Survey (ALS) Corporation.
Kristensen HL; Gundersen P; Callesen I; Reinds G-J. (2004). Throughfall nitrogen deposition has different impacts on soil solution nitrate
concentration in European coniferous and deciduous forests. Ecosystems, 7,180-192.
Kristensen HL; Thorup-Kristensen K. (2006). Roots Below One Meters Depth Are Important for Nitrate Uptake by Annual Crops. Paper
presented at the American society of agronomy - Crop science society of America - Soil science society of America international
annual meetings, held in Indianapolis, IN, November 12-16, 2006. (CD-rom no. 205-9). Published in CD-rom by The American Society
of Agronomy; Crop Science Society of America; Soil Science Society of America.
Kropff MJ. (1990). The effects of long-term open-air fumigation with S02on a field crop of braod bean (Viciafaba L.); III. Quantitative analysis
of damage components. NewPhytol, 115,357-365.
Kropff MJ. (1991). Long-term effects of SO2 on plants, SO2 metabolism and regulation of intracellular pH. Plant Soil, 131,235-245.
Kros J; Reinds GJ; de Vries W; Latour JB; Bollen MJS. (1995). Modelling abiotic site factors in response to atmospheric deposition and upward
seepage. In: Schoute JFT; Finke PA; Veeneklaas FR; Wolfert HP (Eds.), Scenario studies for the rural environment (pp. 445-448):
Kluwer, Dordrecht.
Krug EC; Frink CR. (1983). Acid rain on acid soil: a new perspective. Science, 221,520-525.
Krug EC; Isaacson PJ; Frink CR. (1985). Appraisal of some current hypotheses describing acidification of watersheds. Journal of the Air
Pollution Control Association, 35,109-114.
Krug EC. (1989). Assessment of the theory and hypotheses of the acidification of watersheds. Champaign, IL: Illinois State water Survey.
Krug EC. (1991). Review of acid-deposition-catchment interaction and comments on future research needs. J Hydrol, 128,1-27.
Krupa SV; LeggeAH. (1986). Single or joint effects of coarse and fine particulate sulfur aerosols and ozone on vegetation. In: LeeSD;
Schneider T; Grant LD; Verkerk PJ (Eds.), Aerosols: research, risk assessment and control strategies: Proceedings of the second
U.S.-Dutch international symposium, held in Williamsburg, VA, May 19-25,1985. Organized by the U.S. Environmental Protection
Agency and the Ministry of Housing, Physical Planning, and the Environment of The Netherlands, (pp 879-887). Chelsea, Ml: Lewis
Publishers.
Krupa SV; Legge AH. (1998). Sulphur dioxide, particulate sulphur and its impacts on a boreal forest ecosystem. In: Ambasht RD (Ed.), Modern
trends in ecology and environment (pp. 285-306). Leiden, The Netherlands: Backhuys Publishers.
Krupa SV; Legge AH. (1999). Foliar injury symptoms of Saskatoon serviceberry (Amelanchier alnifolia Nutt.) as a biological indicator of ambient
sulfur dioxide exposures. Environ Pollut, 106,449-454.
Krupa SV. (2003). Effects of atmospheric ammonia (NH3).on terrestrial vegetation: A review. Environ Pollut, 124,179-221.
Krywult M; Bytnerowicz A. (1997). Induction of nitrate reductase activity by nitric acid vapor in California black oak (Quercus kelloggii), canyon
live oak (Quercus chrysolepis), and ponderosa pine (Pinus ponderosa).seedlings. Can J For Res, 27,2101-2104.
Kubiske M; Godbold DL. (2001). Influence of carbon dioxide on the growth and function of roots and roots systems. In: Karnosky DF,
Scarascia-Mungozza GE, Ceulemans R, Innes JL (Eds.), The impacts of carbon dioxide and other greenhouse gases on forest
ecosystems (Research Series Report No. 3 of the IUFRO Task Force on Environmental Change, IUFRO Series #08, pp. 147-192).
Wallingford, Oxfordshire, UK: CABI Publishing.
Kuik OJ; Helming JFM; Dorland C; Spaninks FA. (2000). The economic benefits to agriculture of a reduction of low-level ozone pollution in The
Netherlands. Eur Rev Agric Econ, 27,75-90.
Kulmala M; Pirjola L; Makela JM. (2000). Stable sulphate clusters as a source of new atmospheric particles. Nature, 404,66-69.
Kulmala M; Riipinen I; Sipila M; Manninen HE; PetajaT; Junninen H; Maso MD; Mordas G; Mirme A; Vana M; Hirsikko A; Laakso L; Harrison
RM; Hanson I; Leung C; Lehtinen KE; Kerminen VM. (2007). Toward direct measurement of atmospheric nudeation. Science, 318,
89-92.
Kumar N; Russell AG. (1996). Development of a computationally efficient, reactive subgrid-scale plume model and the impact in the
northeastern United States using increasing levels of chemical detail. J Geophys Res, 101,16737-16744.
Kurz D; Alveteg M; Sverdrup H. (1998). Integrated assessment of soil chemical status. 2. Application of a regionalized model to 622 forested
sites in Switzerland. Water Air Soil Pollut, 105,11-20.
Kuypers MMM; Lavik G; Woebken D; Schmid M; Fuchs B; Amann R; Jorgensen BB; Jetten MSM. (2005). Massive nitrogen loss from the
Benguela upwelling system through anaerobic ammonium oxidation. Proc Natl Acad Sci USA, 102,6478-3483.
Lacroix GL. (1985a). Plasma ionic composition of the Atlantic salmon (Salmo salai), white sucker (Catostomus commersoni), and alewife
(Alosa pseudoharengus) in some acidic rivers of Nova Scotia. Can J Zool, 63, 2254-2261.
Lacroix GL. (1985b). Survival of eggs and alevins of Atlantic salmon (Salmo salai) in relation to the chemistry of interstitial water in redds in
some acidic streams of Atlantic Canada. Can J Fish Aquat Sci, 42,292-299.
Lacroix GL; Townsend DR. (1987). Responses of juvenile Atlantic salmon (Salmo salai) to episodic increases in acidity of Nova Scotia rivers.
Can J Fish Aquat Sci, 44,1475-1484.
R-51
-------
Lafrancois BM; Carlisle DM; Nydick KR; Johnson BM; Baron JS. (2003a). Environmental characteristics and benthic invertebrate assemblages
in Colorado mountain lakes. Western North American Naturalist, 63,137-154.
Lafrancois BM; Nydick KR; Caruso B. (2003b). Influence of nitrogen on phytoplankton biomass and community composition in fifteen Snowy
Range lakes (Wyoming, USA). Arct Antarct Alp Res, 35,499-508.
Lafrancois BM; Nydick KR; Johnson BM; Baron JS. (2004). Cumulative effects of nutrients and pH on the plankton of two mountain lakes. Can
J Fish Aquat Sci, 61,1153-1165.
Lamers LPM; Tomassen HBM; Roelofs JGM. (1998). Sulfate-induced eutrophication and phytotoxicity in freshwater wetlands. Environ Sci
Technol, 32,199-205.
Lamers LPM; Falla SJ; Samborska EM; van Dulken IAR; van Hengstum G; Roelofs JGM. (2002). Factors controlling the extent of
eutrophication and toxicity in sulfate-polluted freshwater wetlands. Limnol Oceanogr, 47,585-593.
Lammel G; Perner D. (1988). The atmospheric aerosol as a source of nitrous acid in the polluted atmosphere. Journal of Aerosol Science, 19,
1199-1202.
Lammel G; Cape JN. (1996). Nitrous acid and nitrite in the atmosphere. Chem Soc Rev, 25,361-369.
Lamsal LN; Martin RV; van Donkelaar A; Steinbacher M. (2008). Ground-level nitrogen dioxide concentrations inferred. J Geophys Res, 113,
doi:10.1029/2007JD009235.
Lan T; Wu S-P; Xu F-L; Wang X-J; Tao S. (2005). DDTs in Airborne Particles with Different Sizes in Tianjin During Winter. Journal of Agro-
Environment Science, 24,1182-1185.
Lan TTN; Thoa NTP; Nishimura R; Tsujino Y; Yokoi M; Maeda Y. (2005). New model for the sulfation of marble by dry deposition sheltered
marble: The indicator of air pollution by sulfur dioxide. Atmos Environ, 39,913-920.
Lancaster J; Real M; Juggins S; Monteith DT; Flower RJ; Beaumont WRC. (1996). Monitoring temporal changes in the biology of acid waters.
Freshwater Biol, 36,179-201.
Landers DH; Eilers JM; Brakke DF; Overton WS; Kellar PE; Silverstein WE; Schonbrod RD; Crowe RE; Linthurst RA; Omernik JM; Teague SA;
Meier EP. (1987). Western lake survey phase I: Characteristics of lakes in the western United States. Volume I: Population
descriptions and physico-chemical relationships. (Report # EPA/600/3-86/054a). Washington, D.C.: U.S. Environmental Protection
Agency.
Landers DH; Overton WS; Linthurst RA; Brakke DF. (1988). Eastern lake survey: regional estimates of lake chemistry. Environ Sci Technol, 22,
128-135.
Lane DR; BassiriRad H. (2002). Differential responses of tallgrass prairie species to nitrogen loading and varying ratios of N03~to NH4+.
Functional Plant Biology, 29,1227-1235.
Langford AO; Fehsenfeld FC. (1992). Natural Vegetation as a Source or Sink for Atmospheric Ammonia: A Case Study. Science, 255,581-583.
Langford AO; Fehsenfeld FC; Zachariassen J; Schimel DS. (1992). Gaseous Ammonia Fluxes and Background Concentrations in Terrestrial
Ecosystems of the United States. Glob Biogeochem Cycles, 6,459-459.
Lapenis AG; Lawrence GB; Andreev AA; Bobroc AA; Torn MS; Harden JW. (2004). Acidification of forest soil in Russia: from 1893 to present.
Glob Biogeochem Cycles, 18,1-13.
Laposata MM; Dunson WA. (1998). Effects of boron and nitrate on hatching success of amphibian eggs. Arch Environ Contam Toxicol, 35,
615-619.
Larsen DP; Urquhart NS. (1993). A framework for assessing the sensitivity of the EMAP design. In: Larsen DP, Christie SJ; Paulson, SG;
Hughes RM; Johnson CB (Eds.), EMAP-Surface Waters 1991 Pilot Report (EPA—620/R-93/003, pp. 4.1-4.37). Corvallis, OR;
Environmental Research Laboratory; U.S. Environmental Protection Agency.
Larsen DP; Thornton KW; Urquhart NS; Paulsen SG. (1994). The role of sample surveys for monitoring the condition of the nation's lakes.
Environ Monit Assess, 32,101-134.
Larson GL; Moore SE. (1985). Encroachment of exotic rainbow trout into stream populations of native brook trout in the Southern Appalachian
Mountains. Trans Am Fish Soc, 114,195-203.
Latty EF; Canham CD; Marks PL. (2004). The Effects of Land-use History on Soil Properties and Nutrient Dynamics in Northern Hardwood
Forests of the Adirondack Mountains. Ecosystems, 7,193-207.
Laudon H; Clair TA; Hemond HF. (2002). Long-term response in episodic acidification to declining SO42" deposition in two streams in Nova
Scotia. Hydrol Earth Sys Sci, 6, 773-782.
Laudon H; Dillon PJ; Eimers MC; Semkin RG; Jeffries DS. (2004). Climate-Induced Episodic Acidification of Streams in Central Ontario.
Environ Sci Technol, 38, 6009-6015.
Laurence JA; Amundson RG; Friend AL; Pell EJ; Temple PJ. (1994). Allocation of carbon in plants under stress: an analysis of the ROPIS
experiments. J Environ Qual, 23,412-417.
Laurenroth WK; Milchunas DG. (1985). SO2 effects on plant community function. In: Winter WE; Mooney HA; Goldstein RA (Eds.), Sulfur
dioxide and vegetation: Physiology, ecology, and policy issues, (pp. 454-477). Sanford, CA: Sanford University Press.
Lawrence GB; David MB. (1977). Response of aluminum solubility to elevated nitrification in soil of a red spruce stand in eastern Maine.
Environ Sci Technol, 31, 825-830.
Lawrence GB; Fuller RD; Driscoll CT. (1987). Release of aluminum following whole-tree harvesting at the Hubbard Brook Experimental Forest,
New Hampshire. J Environ Qual, 16,383-390.
Lawrence GB; David MB; Shortle WC. (1995). A New Mechanism for Calcium Loss in Forest-Floor Soils. Nature, 378,162-165.
Lawrence GB; David MB. (1997). Response of aluminum solubility to elevated nitrification in soil of a red spruce stand in eastern Maine.
Environ Sci Technol, 31, 825-830.
Lawrence GB; David MB; Bailey SW; Shortle WC. (1997). Assessment of soil calcium status in red spruce forests in the northeastern United
States. Biogeochemistry, 38,19-39.
R-52
-------
Lawrence GB; David MB; Shortle WC; Bailey SW; Lovett GM. (1998). Mechanisms of base-cation depletion by acid deposition in forest soils of
the northeastern U.S. In: Sugar Maple Ecology and Health: Proceedings of an International Symposium held in Warren, PA; June, 2-
4,1998. (Pages 75-87).
Lawrence GB; Huntington TG. (1999). Soil-Calcium Depletion Linked to Acid Rain and Forest Growth in the Eastern United States. (Report no
WRIR 98-4267). Reston, VA: U.S. Geological Survey.
Lawrence GB; David MB; Shortel WS; Bailey SW; Lovett GM. (1999a). Mechanisms of base-cation depletion by acid deposition in forest soils
of the northeastern U.S. ln:Horsely, SB; Long RP (Eds.) Sugar maple ecology and health: Proceedings of an international symposium
held in Radner, PA June 2-4,1998. (pp 75-87). U.S. Forest Service; U.S. Department of Agriculture (USDA).
Lawrence GB; David MB; Lovett GM; Murdoch PS; Burns DA; Stoddard JL; Baldigo B; Porter JH; Thompson AW. (1999b). Soil calcium status
and the response of stream chemistry to changing acidic deposition rates. Ecol Appl, 9,1059-1072.
Lawrence GB. (2002). Persistent episodic acidification of streams linked to acid rain effects on soil. Atmos Environ, 36,1589-1598.
Lawrence GB; Momen B; Roy KM. (2004). Use of stream chemistry for monitoring acidic deposition effects in the Adirondack Region of New
York. J Environ Qual, 33,1002-1009.
Lawrence GB; Sutherland JW; Boylen CW; Nierzwicki-Bauer SA; Momen B; Baldigo BP; Simonin HA. (2007). Acid rain effects on aluminum
mobilization clarified by inclusion of strong organic acids. Environ Sci Technol, 41,93-98.
Lawrence GB; Lapenis AG; Berggren D; Aparin BF; Smith KT; Shortle WC; Bailey SW; Varlyguin DL; Babikov B. (2005). Climate dependency
of tree growth suppressed by acid deposition effects on soils in northwest Russia. Environ Sci Technol, 39,2004-2010.
Lawrence GB; Roy KM; Baldigo BP; Simonin HA; Capone SB; Sutherland JW; Nierzwicki-Bauer SA; Boylen CW. (2008). Chronic and episodic
acidification of adirondack streams from Acid rain in 2003-2005 . J Environ Qual, 37, 2264-2274.
Lawson DR. (1990). The Southern California Air Quality Study. J Air Waste Manage Assoc, 40,156-165.
Le Mer J; Roger P. (2001). Production, oxidation, emission and consumption of methane by soils: A review. European Journal of Soil Biology,
37,25-50.
Leaderer BP; Naeher L; Jankun T; Balenger K; Holford TR; Toth C; Sullivan J; Wolfson JM; Koutrakis P. (1999). Indoor, outdoor, and regional
summer and winter concentrations of PM10, PM2.5, SO42-, H+, NH4"*, NO3", NH3, and nitrous acid in homes with and without kerosene
space heaters. Environ Health Perspect, 107,223-231.
Leavitt PR; Findlay DL; Hall Rl; Smol JP. (1999). Algal responses to dissolved organic carbon loss and pH decline during whole-lake
acidification: evidence from paleolimnology. Limnol Oceanogr, 44,757-773.
LeBauer DS; Treseder KK. (2008). Nitrogen limitation of net primary productivity in terrestrial ecosystems is globally distributed. Ecology, 89,
371-379.
Lee EH; Tingey DT. (2003). History of tropospheric ozone for the San Bernardino Mountains of Southern California, 1963-1999. Atmos Environ,
37,2705-2717.
Lefer BL; Shetter RE; Hall SR; Crawford JH; Olson JR. (2003). Impact of clouds and aerosols on photolysis frequencies and photochemistry
during TRACE-P: 1. Analysis using radiative transfer and photochemical box models. J Geophys Res, 108,, 8821,
doi: 10.1029/2002JD003171.
Lefohn A; Husar J; Husar R. (1999). Estimating historical anthropogenic global sulfur emissions patterns for the period 1850—1990. Atmos
Environ, 33, 3435-3444.
Lefohn AS; Oltmans SJ; Dann T; Singh HB. (2001). Present-day variability of background ozone in the lower troposphere. J Geophys Res, 106,
9945-9958.
Leggett ZH; Kelting DL. (2006). Fertilization effects on carbon pools in loblolly pine plantations on two upland sites. Soil Sci Soc Am J, 70,279-
286.
Leigh RJ; Corlett GK; Friess U; Monks PS. (2006). Concurrent multiaxis differential optical absorption spectroscopy system for the
measurement of tropospheric nitrogen dioxide. Appl Opt, 45,7504-7518.
Leino RL; Wilkinson P; Anderson JG. (1987). Histopathlogical changes in the gills of pearl dace, Semotilus margarita, and fathead minnows,
Pimephales promleas, from experimentally acidified Canadian lakes. Can J Fish Aquat Sci, 44,126-134.
Leivestad H. (1982). Physiological effects of acid stress on fish. In Haines TA, Johnson E (Eds.), Acid rain/fisheries (pp. 157-164). Bethesda,
MD: American Fisheries Society.
Lendzian KJ; Kerstiens G. (1988). Interactions between plant cuticles and gaseous air pollutants. Aspects of Applied Biology, 17,97-104.
Lenner M. (1987). Nitrogen dioxide in exhaust emissions from motor vehicles. Atmos Environ, 21,37-43.
Lerdau MT; Munger JW; Jacob DJ. (2000). The NO2 flux conundrum. Science, 289,2293.
Lesack LFW; Melack JM. (1996). Mass balance of major solutes in a rainforest catchment in the central Amazon: Implications for nutrient
budgets in tropical rainforests. Biogeochemistry, 32,115-142.
Leue C; Wenig M; Wagner T; Klimm O; Piatt U; Jahne B. (2001). Quantitative analysis of NOx emissions from Global Ozone Monitoring
Experiment satellite image sequences. J Geophys Res, 106,5493-5505.
Leuenberger-Minger AU; Faller M; Richner P. (2002). Runoff of copper and zinc caused by atmospheric corrosion. Mater Corros, 53,157-164.
Leung F-YT; Colussi AJ; Hoffmann MR; Toon GC. (2002). Isotopic fractionation of carbonyl sulfide in the atmosphere: implications for the
source of background stratospheric sulfate aerosol. Geophys Res Lett, 29, doi: 10.1029/2001GL013955.
Leuschner WA; Young RL. (1978). Estimating the southern pine beetle's impact on reservoir campsites. For Sci, 24, 527-537.
Levelt PF; Ven Den Oord GHJ; Dobber MR; Malkki A; Visser H; de Vries J; Stammes P; Lundell JOV; Saari H. (2006). The ozone monitoring
instrument. IEEE Trans Geosci Remote Sens, 44,1093-1101.
Levin SA. (1998). Ecosystems and the Biosphere as Complex Adaptive Systems. Ecosystems, 1,431 -436.
R-53
-------
Levine, J. S; Bobbe, T; Ray, N; Witt, R. G; Singh, A. (1999).Wildland fires and the environment: A global synthesis. (Report no.
UNEP/DEIAEW/TR.99.1). http://www.na.unep.net/publications/wildfire.pdf. Nairobi, Kenya: Division of Environmental Information;
Assessment an Early Warning (DEIA&EW); United Nations Environment Programme (UNEP).
Levine MA; Whalen SC. (2001). Nutrient limitation of phytoplankton production in Alaskan arctic foothill lakes. Hydrobiologia, 455,189-201.
Levine SN; Shambaugh AD; Pomeroy SE; Braner M. (1997). Phosphorus, nitrogen, and silica as controls on phytoplankton biomass and
species composition in Lake Champlain (USA-Canada). Intern Assoc for Great Lakes Res, 23,131-148.
Levlin SA. (1998). Ecosystems and the biosphere as complex adaptive systems. Ecosystems, 1,431-436.
Lewis GP; Likens GE. (2007). Changes in stream chemistry associated with insect defoliation in a Pennsylvania hemlock-hardwoods forest.
For Ecol Manage, 238,199-211.
Ley RE; Williams MW; Schmidt SK. (2004). Microbial population dynamics in an extreme environment: Controlling factors in talus soils at 3750
m in the Colorado Rocky Mountains. Biogeochemistry, 68,297-311.
Li C; Frolking S; Frolking TA. (1992). A model of nitrous-oxide evolution from soil driven by rainfall events. 1: Model structure and sensitivity. J
Geophys Res, 97, 9759-9776.
Li C; Narayanan V; Harriss RC. (1996). Model estimates of nitrous oxide emissions from agricultural lands in the United States. Glob
Biogeochem Cycles, 10,297-306.
Li CS. (2000). Modeling trace gas emissions from agricultural ecosystems. Nutr Cycling Agroecosyst, 58, 259-276.
Li C; Aber J; Stange F; Butterbach-Bahl K; Papen H. (2000). A process-oriented model of N2O and NO emissions from forest soils: 1. Model
development. J Geophys Res [Atmos.], 105,4369-4384.
Li JH; Tang HX; Bai QZ; Nie YF; Luan ZK. (2000). Aquatic acidification sensitivity for regional environment: a multi-indicator evaluation
approach. Water Air Soil Pollut, 11.251-261.
Liang J; Horowitz LW; Jacob DJ; Wang Y; Fiore AM; Logan JA; Gardner GM; Munger JW. (1998). Seasonal budgets of reactive nitrogen
species and ozone over the United States, and export fluxes to the global atmosphere. J Geophys Res, 103,13435-13450.
Lien L; Raddum GG; Fjellheim A. (1992). Critical loads for surface waters: invertebrates and fish. (Acid rain research report no 21). Oslo,
Norway: Norwegian Institute for Water Research.
Lien L; Raddum GG; Fjellheim A; Henriksen A. (1996). A critical limit for acid neutralizing capacity in Norwegian surface waters, based on new
analyses offish and invertebrate responses. Sci Total Environ, 177,173-193.
Likens GE; Bormann FH; Johnson NM; Fisher DW; Pierce RS. (1970). Effects of forest cutting and herbicide treatment on nutrient budgets in
the Hubbard Brook watershed-ecosystem. Ecol Monogr, 40,23-47.
Likens GE; Bormann FH; Pierce RS; Eaton JS; Johnson NM. (1977). Biogeochemistry of a Forested Ecosystem. New York, NY: Springer
Verlag.
Likens GE; Bormann FH; Pierce RS; Reiners WA. (1978). Recovery of a deforested ecosystem. Science, 199,492-496.
Likens GE. (1985). An ecosystem approach to aquatic ecology: Mirror Lake and its environment: New York, NY: Springer-Verlag.
Likens GE; Bormann FH; Hedin LO; Driscoll CT; Eaton JS. (1990). Dry deposition of sulfur: A 23 year record for the Hubbard Brook Forest
ecosystem. Tellus B, 42,319-329.
Likens GE. (1992). The ecosystem approach: Its use and abuse. Volume 3 of Series: Excellence in ecology. Oldendorf/Luhe, Germany:
Ecology Institute.
Likens GE; Bormann FH. (1995). Biogeochemistry of a forested ecosystem. 2nd Edition. New York, NY: Springer-Verlag.
Likens GE; Driscoll CT; Buso DC. (1996). Long-term effects of acid rain: Response and recovery of a forest ecosystem. Science, 272,244-246.
Likens GE; Driscoll CT; Buso DC; Siccama TG; Johnson CE; Lovett GM; Fahey TJ; Reiners WA; Ryan DF; Martin CW; Bailey SW. (1998). The
biogeochemistry of calcium at Hubbard Brook. Biogeochemistry, 41,89-173.
Likens GE; Butler TJ; Buso DC. (2001). Long- and short-term changes in sulfate deposition: Effects of the 1990 Clean Air Act Amendments.
Biogeochemistry, 52,1-11.
Likens GE; Driscoll CT; Buso DC; Mitchell MJ; Lovett GM; Bailey SW; Siccama TG; Reiners WA; Alewell C. (2002). The biogeochemistry of
sulfur at Hubbard Brook. Biogeochemistry, 60,235-316.
Lilleskov EA; Fahey TJ; Lovett GM. (2001). Ectomycorrhizal Fungal Aboveground Community Change over an Atmospheric Nitrogen
Deposition Gradient. Ecol Appl, 11,397-410.
Lilleskov EA; Fahey TJ; Horton TR; Lovett GM. (2002). Belowground ectomycorrhizal fungal community change over a nitrogen deposition
gradient in Alaska. Ecology, 83,104-115.
Lind CJ; Hem JD. (1975). Effects of organic solutes on chemical reactions of aluminum. Chemistry of aluminum in natural water. (Geological
Survey Water Supply Paper No. 1827). Reston, VA; U.S. Geological Survey.
Lindau CW; Delaune RD; Pardue JH. (1994). Inorganic nitrogen processing and assimilation in a forested wetland. Hydrobiologia, 277,171-
178.
Lindberg SE; Lovett GM. (1985). Field measurements of particle dry deposition rates to foliage and inert surfaces in a forest canopy.
Environmental Science and Technology 19:238-244.
Lindberg SE; Bredemeir M; Schaefer DA; Qi L. (1990). Atmospheric concentrations and deposition during the growing season in conifer forests
in the United States and West Germany. Atmos Environ, 24A, 2207-2220.
Lindberg SE. (1992). Atmospheric deposition and canopy interactions of sulfur. In: Johnson DW; Lindberg SE (Eds.), Atmospheric deposition
and forest nutrient cycling: A synthesis of the integrated forest study. (Series: Billings WD; Golley F; Lange OL; Olson JS; Remmert H
(Eds.), Ecological studies: Analysis and synthesis; Volume. 91, pp. 74-90). New York, NY: Springer-Verlag.
Lindroth, RL; Roth S; Kruger EL; Volin JC; Koss PA. (1997). C02-mediated changes in aspen chemistry: Effects on gypsy moth performance
and susceptibility to virus. Glob Change Biol, 3,279-289.
R-54
-------
Linthurst RA; Landers DH; Eilers JM; Brakke DF; Overton WS; Meier EP; Crowe RE. (1986a). Characteristics of Lakes in the easten United
States. Volume I. Population descriptions and physico-chemical relationships. (EPA-600/4-86-007A). Washington, DC; Office of Acid
Deposition; Environmental Monitoring, and Quality Assurance; U.S. Environmental Protection Agency.
Linthurst RA; Landers DH; Eilers JM; Kellar PE; Brakke DF; Overton WS; Crowe R; Meier EP; Kanciruk P; Jeffries DS. (1986b). Regional
chemical characteristics of lakes in North America Part II: eastern United States. Water Air Soil Pollut, 31, 577-591.
Lipton D. (2003). The value of improved water quality to Chesapeake Bay boaters. (Working Paper #03-16). College Park, MD; Department of
Agricultural and Resource Economics, University of Maryland-College Park.
Lipton D. (2004). The value of improved water quality to Chesapeake Bay boaters. Mar Resour Econ, 19, 265-270.
Little EL Jr. (1971). Atlas of United States trees. Vol. 1: Conifers and important hardwoods (USDA. Misc. Publ., no. 1146).Washington, DC;
U.S. Forest Service; U.S. Department of Agriculture (USDA).
Little P; Wiffen RD. (1977). Emission and deposition of petrol engine exhaust Pb—Deposition of exhaust Pb to plant and soil surfaces, Atmos
Environ, 11,437-447
Litvak ME; Loreto F; Harley PC; Sharkey TD; Monson RK. (1996). The response of isoprene emission rate and photosynthetic rate to photon
flux and nitrogen supply in aspen and white oak trees. Plant Cell and Environment, 19,549-559.
Liu H; Jacob DJ; Bey I; Yantosca RM. (2001a). Constraints from 210Pb and 7Be on wet deposition and transport in a global three-dimensional
chemical tracer model driven by assimilated meteorological fields. J Geophys Res, 106,12109-12128.
Liu S; Munson R; Johnson D; Gherini S; Summers K; Hudson R; Wilkinson K; Pitelka L. (1991). Application of a nutrient cycling model
(NuCM).to a northern mixed hardwood and a southern coniferous forest. Tree Physiology, 9,173-184.
LiuX; Hegg DA; Stoelinga MT. (2001b). Numerical simulation of new particle formation the Northwest Atlantic using the M5 mesoscale model
coupled with sulfur chemistry. J Geophys Res, 106,9697-9715.
Liu X; Chance K; Sioris CE; Kurosu TP; Spurr RJD; Martin RV; Fu T-M; Logan JA; Jacob DJ; Palmer PI; Newchurch MJ; Megretskaia IA;
Chatfield RB. (2006). First directly retrieved global distribution of tropospheric column ozone from GOME: Comparison with the GEOS-
CHEM model. J Geophys Res, 111, doi: 10.1029/2005JD006564.
Lobert JM; Scharffe DH; Hao W-M; Kuhlbusch TA; Seuwen R; Warneck P; Crutzen PJ. (1991). Experimental evaluation of biomass burning
emissions: nitrogen and carbon containing compounds. In: Levine, JS (Ed.), Global biomass burning, (pp 289-304). Cambridge, MA:
MIT Press.
Lobert JM; Keene WC; Logan JA; Yevich R. (1999). Global chlorine emissions from biomass burning: Reactive Chlorine Emissions Inventory. J
Geophys Res, 104,8373-8389.
Locke A; Sprules WG. (1994). Effects of lake acidification and recovery on the stability of zooplankton food webs. Ecology, 75,: 498-506.
Lofts S; Woof C; Tipping E; Clarke N; Mulder J. (2001). Modelling pH buffering and aluminium solubility in European forest soils. European
Journal of Soil Sci, 52,189-204.
Lohman K; Jones JR; Baysinger-Daniel C. (1991). Experimental evidence for nitrogen limitation in a northern Ozark stream. J North Am
Benthol Soc, 10,14-23.
Lohman K; Priscu JC. (1992). Physiological indicators of nutrient deficiency in cladophora (chlorophyta) in the Clark Fork of the Columbia
River, Montana. J Phycol, 28,443-448.
Lohrenz SE; Fahnenstiel GL; Redakje DG; Lang GA; Chen X; Dagg MJ. (1997). Variations in primary production of northern Gulf of Mexico
continental shelf waters linked to nutrient inputs from the Mississippi River. Marine Ecology Progress Series, 155,445-454.
Lokke H; Bak J; Falkengren-Grerup U; Finlay RD; llvesniemi H; Nygaard PH; Starr M. (1996). Critical loads of acidic deposition for forest soils:
Is the current approach adequate. Ambio, 25,510-516.
Longcore JR; Gill JD (Eds.), (1993). Acidic depositions: effects on wildlife and habitats. (Wildlife Society technical review no 93-1). Bethesda,
MD: The Wildlife Society.
Longfellow CA; Ravishankara AR; Hanson DR. (1999). Reactive uptake on hydrocarbon soot: focus on NO2. J Geophys Res, 104,13833-
13840.
Loomis JB; Gonzalez-Caban A; Gregory R. (1996). A contingent valuation study of the value of reducing fire hazards to old-growth forests in
the Pacific Northwest. (Research paper no. PSW-RP-229). Portland, OR; Pacific Southwest Research Station; U.S. Forest Service;
U.S. Department of Agriculture (USDA).
Lotter AF. (1998). The recent eutrophication of Baldeggersee (Switzerland).as assessed by fossil diatom assemblages. Holocene, 8,395-405.
Lovett GM. (1992). Atmospheric deposition and canopy interactions of nitrogen. In: Johnson DW, Lindberg SE (Eds.), Atmospheric deposition
and forest nutrient cycling: A systhesis of the integrated forest study. (Series: Billings WD; Golley F; Lange OL; Olson JS; Remmert H
(Eds.), Ecological studies: Analysis and synthesis; Volume 91; pp. 152-166). New York, NY: Springer-Verlag.
Lovett GM. (1994). Atmospheric deposition of nutrients and pollutants in North America: an ecological perspective. Ecological Applications 4(4):
629-650.
Lovett GM; Ruesink AE. (1995). Carbon and nitrogen assimilation in red oaks (Quercus rubra L.) subject to defoliation and nitrogen stress.
Tree Physiol, 12,259-269.
Lovett GM; Rueth H. (1999). Soil nitrogen transformations in beech and maple stands along a nitrogen deposition gradient. Ecol Appl, 9,1330-
1344.
Lovett GM; Traynor MM; Pouyat RV; Carreiro MM; Zhu W-X; Baxter JW. (2000a). Atmospheric deposition to oak forest along an urban-rural
gradient. Environ Sci Technol, 34,4294-4300.
Lovett GM; Weathers KC; Sobczak WV. (2000b). Nitrogen saturation and retention in forested watersheds of the Catskill Mountains, New York.
Ecol Appl, 10,73-84.
Lovett GM; Christenson LM; Groffman PM; Jones CG; Hart JE; Mitchell MJ. (2002a). Insect defoliation and nitrogen cycling in forests.
Bioscience, 52,335-341.
R-55
-------
Lovett GM; Weathers KC; Arthur MA. (2002b). Control of nitrogen loss from forested watersheds by soil C:N ratio and tree species
composition. Ecosystems, 5,712-718.
Lovett GM; Burns DA; Driscoll CT; Jenkins JC; Mitchell MJ; Rustad L; Shanley JB; Likens GE; Haeuber R. (2007). Who needs environmental
monitoring? Front Ecol Environ, 5,253-260.
Lowe PN; Lauenroth WK; Burke IC. (2002). Effects of nitrogen availability on the growth of native grasses exotic weeds. Journal of Range
Management, 55, 94-98.
Lowrance RR; Todd RL; Asmussen LE. (1983). Waterborne nutrient budgets for the riparian zone of an agricultural watershed. Agric Ecosyst
Environ, 10,371-384.
Lowrance R. (1992). Groundwater nitrate and denitrification in a coastal plain riparian forest. J Environ Qual, 21,401-405.
Lu R; Turco RP; Jacobson MZ. (1997). An integrated air pollution modeling system for urban and regional scales: 1. Structure and
performance. J Geophys Res, 102,6063-6079.
Lu S; Mattson KG; Zaerr JB; Marshall JD. (1998). Root respiration of Douglas-fir seedlings: effects of N concentration. Soil Biol Biochem, 30,
331-336.
Lucassen E; Smolders AJP; van der Salm AL; Roelofs JGM. (2004). High groundwater nitrate concentrations inhibit eutrophication of sulphate-
rich freshwater wetlands. Biogeochemistry, 67,249-267.
Lukac M; Calfapietra C; Godbold DL. (2003). Production, turnover and mycorrhizal colonization of root systems of three Populus species grown
under elevated CO2 (POPFACE). Glob Change Biol, 9,838-848.
Luke WT; Dickerson RR; Ryan WF; Pickering KE; Nunnermacker LJ. (1992). Tropospheric chemistry over the lower Great Plains of the United
States. 2. Trace gas profiles and distributions. J Geophys Res, 97,20647-20670.
Luke WT. (1997). Evaluation of a commercial pulsed fluorescence detector for the measurement of low-level S02 concentrations during the
Gas-Phase Sulfur Intercomparison Experiment. J Geophys Res, 102,16255-16265.
Luke WT; Watson TB; Olszyna KJ; Gunter RL; McMillen RT; Wellman DL; Wilkison SW. (1998). A comparison of airborne and surface trace
gas measurements during the Southern Oxidants Study (SOS). J Geophys Res, 103,22317-22337.
Luo C; Zender CS; Bian H; Metzger S. (2007). Role of ammonia chemistry and coarse mode aerosols in global dimatological inorganic aerosol
distributions. Atmospheric Environment 41:2510-2533.
Lussenhop J; Treonis A; Curtis PS; Teeri JA; Vogel CS. (1998). Response of soil biota to elevated atmospheric CO2 in poplar model systems.
Oecologia, 113,247-251.
Luterbacher J; Dietrich D; Xoplaki E; Grosjean M; Wanner H. (2004). European seasonal and annual temperature variability, trends, and
extremes since 1500. Science, 303,1499-1503.
Lydersen E; Fjeld E; Gjessing ET. (1996). The Humic Lake Acidification Experiment (HUMEX): Main physico-chemical results after five years of
artificial acidification. Environ Int, 22,591-604.
Lynch JA; Corbett ES. (1989). Hydrologic control of sulfate mobility in a forested watershed. Water Resour Res, 25(7), 1695-1703.
Ma J; Richter A; Burrows JP; Null H; Van Aardenne JA. (2006). Comparison of model-simulated tropospheric NO2 over China with GOME-
satellite data. Atmos Environ, 40,593-604.
Maag M; Malinovsky M; Nielson SM. (1997). Kinetics and temperature dependence of potential denitrification in riparian soils. Journal of
Environmental Sciences, 26,215-223.
MacAvoy SW; Bulger AJ. (1995). Survival of brook trout (Salvelinus fontinalis) embryos and fry in streams of different acid sensitivity in
Shenandoah National Park, USA. Water Air Soil Pollut, 85,445-450.
MacDonald JA; Skiba U; Sheppard LJ; Ball B; Roberts JD; Smith KA; Fowler D. (1997). The effect of nitrogen deposition and seasonal
variability on methane oxidation and nitrous oxide emission rates in an upland spruce plantation and moorland. Atmos Environ, 31,
3693-3706.
MacDonald JA; Dise NB; Matzner E; Arbruster M; Gundersen P; Forsius M. (2002). Nitrogen input together with ecosystem nitrogen
enrichment predict nitrate leaching from European forests. Glob Chang Biol, 8,1028-1033.
Mack MC; Schuur EAG; Bret-Harte MS; Shaver GR; Chapin FS. (2004). Ecosystem carbon storage in arctic tundra reduced by long-term
nutrient fertilization. Nature, 431,440-443.
MacLean DC; McCune DC; Weinstein LH; Mandl RH; Woodruff GN. (1968). Effects of acute hydrogen fluoride and nitrogen dioxide exposures
on citrus and ornamental plants of central Florida. Environ Sci Technol, 2,444-449.
MacMillan D; Ferrier B; Hanley N. (2001). Valuation of air pollution effects on ecosystems: A scoping study. (A report to the Department for
Environment, Food and Rural Affairs). Aberdeen, Scotland: University of Aberdeen.
Madronich S. (1987). Photodissociation in the atmosphere. 1. Actinic flux and the effects of ground reflections and clouds. J Geophys Res, 92,
9740-9752.
Magill AH; Aber JD; Berntson GM; McDowell WH; Nadelhoffer KJ; Melillo JM; Steudler P. (2000). Long-term nitrogen additions and nitrogen
saturation in two temperate forests. Ecosystems, 3,238-253.
Magill AH; Aber JD; Currie WS; Nadelhoffer KJ; Martin ME; McDowell WH; Melillo JM; Steudler P. (2004). Ecosystem response to 15 years of
chronic nitrogen additions at the Harvard Forest LTER, Massachusetts, USA. For Ecol Manage, 196,7-28.
Magill AH; Aber JD; Hendricks JJ; Bowden RD; Melillo JM; Steudler PA. (1997). Biogeochemical response of forest ecosystms to simulated
chronic nitrogen dispositon. Ecol Appl, 7,402-415.
Magnani F; Mencuccini M; Borghetti M; Berbigier P; Berninger F; Delzon S; Grelle A; Hari P; Jarvis PG; Kolari P; Kowalski AS; Lankreijer H;
Law BE; Lindroth A; Loustau D; Manca G; Moncrieff JB; Rayment M; Tedeschi V; Valentini R; Grace J. (2007). The human footprint in
the carbon cycle of temperate and boreal forests. Nature, 447,848-850.
R-56
-------
Magnani F; Mencuccini M; Borghetti M; Berninger F; Delzon S; Grelle A; Hari P; Jarvis PG; Kolari P; Kowalski AS. (2008). Magnani et al. reply.
Nature, 451, 3.Mahrt L. (1998). Stratified atmospheric boundary layers and breakdown of models. Theor Comput Fluid Dyn, 11,263-
279.
Makipaa R; Karjalainen T; Pussinen A; Kellomaki S. (1999). Effects of climate change and nitrogen deposition on the carbon sequestration of a
forest ecosystem in the boreal zone. Can J For Res-Revue Canadienne De Recherche Forestiere, 29,1490-1501.
Maler K-G; Vincent JR. (Eds.). (2005). Handbook of Environmental Economics. Vol. 2. Valuation of Environmental Changes. Amsterdam, The
Netherlands: Elsevier.
Malhi Y; Baldocchi DD; Jarvis PG. (1999). The carbon balance of tropical, temperate and boreal forests. Plant Cell and Environment, 22,715-
740.
Maljanen M; Jokinen H; Saari A; Strommer R; Martikainen PJ. (2006). Methane and nitrous oxide fluxes, and carbon dioxide production in
boreal forest soil fertilized with wood ash and nitrogen. Soil Use Manage, 22,151-157.
Malley DF; Chang PS. (1995). Assessing the health of a zooplankton community in a small Precambrian Shield lake during recovery from
experimental acidification. Journal of Aquatic Ecosystem Health, 3, 273-286.
Malm WC; Hand JL. (2007). An examination of the physical and optical properties of aerosols collected in the IMPROVE program. Atmos
Environ, 41:3,407-3,427, 2007.
Malone D J; Conley TR; Fisher PM; Gilbert LW; Harding J; Dellner KG. (1996). Scales of nutrient-limited phytoplankton productivity in
Chesapeake Bay. Estuaries, 19, 371-385.
Malone TC; Crocker LH; Pike SE; Wendler BW. (1988). Influence of river flow on the dynamics of phytoplankton in a partially stratified estuary.
Mar Ecol Progr Ser, 48, 235-249.
Malone TC. (1991). River flow, phytoplankton production and oxygen depletion in Chesapeake Bay. In: Tyson RV, Peason TH (Eds.), Modern
and ancient continentintal shelf aNOxia. Geological Society special publication no. 58 (pp. 83-93). London: The Geological Society.
Malone TC. (1992). Effects of water column processes on dissolved oxygen, nutrients, phytoplankton and zooplankton. In: Smith DE, Leffler M,
Makiernan G (Eds.), Oxygen dynamics in Chesapeake Bay: a synthesis of recent research (pp. 61-112). College Park, MD: Maryland
Sea Grant College Program.
Mann LK; Johnson DW; West DC; Cole DW; Hornbeck JW; Martin CW; Riekerk H; Smith CT; Swank WT; Tritton LM; Van Lear DH. (1988).
Effects on whole-tree and stem-only clearcutting on post harvest hydrologic losses, nutrient capital, and regrowth. For Sci, 34,412-
428.
Manninen S; Huttunen S. (2000). Response of needle sulphur and nitrogen concentrations of Scots pine versus Norway spruce to S02 and
N02. Environ Pollut, 107,421-436.
Mansfield TA; Jones T. (1985). Growth environment interactions in SO2 responses of grasses. In: Winter WE, Mooney HA, Goldstein RA
(Eds.), Sulfur dioxide and vegetation: physiology, ecology, and policy issues, (pp. 332-345). Stanford, CA: Stanford University Press.
Marco A; Blaustein AR. (1999). The effects of nitrite on behavior and metamorphosis in cascades frogs (Rana cascadae). Environ Toxicol
Chem, 18, 949-949.
Margitan JJ; Watson RT. (1982). Kinetics of the reaction of hydroxyl radicals with nitric acid. The J PhysChem, 86, 3819-3824.
Marinoni N; Birelli MP; Rostagno C; Pavese A. (2003). The effects of atmospheric multipollutants on modern concrete. Atmos Environ, 37,
4701-4712.
Markewitz D; Richter DD; Allen HL; Urrego J. (1998). Three decades of observed soil acidification in the Calhoun experimental forest: Has acid
rain made a difference?. So/7. Science Society of America Journal, 62,1428-1439.
Marler RJ; Stromberg JC; Pattern DT. (2001). Growth response of Populus fremontii, Salix gooddingii, and Tamarix ramosissimiseedlings
under different nitrogen and phosphorus concentrations. J Arid Environ, 49,133-146.
Marley NA; Gaffney JS; White RV; Rodriguez-Cuadra L; Herndon SE; Dunlea E; Volkamer RM; Molina LT; Molina MJ. (2004). Fast gas
chromatography with luminol chemiluminescence detection for the simultaneous determination of nitrogen dioxide and peroxyacetyl
nitrate in the atmosphere. Rev Sci Instrum, 75,4595-4605.
Marschner H. (1995). Sulfur. In Mineral nutrition of higher plants (pp. 255-265). London, UK: Academic Press.
Marshall HG; Lacouture R. (1986). Seasonal patterns of growth and composition of phytoplankton in the lower Chesapeake Bay and vicinity.
Estuar Coast Shelf Sci, 23,115-130.
Martilli A; Neftel A; Favaro G; Kirchner F; Sillman S; Clappier A. (2002). Simulation of the ozone formation in the northern part of the Po Valley.
J Geophys Res, 107, doi: 10.1029/2001 JD000534.
Martin CW; Noel DS; Federer CA. (1984). Effects of forest clearcutting in New England on stream chemistry. J Environ Qual, 13, 204-210.
Martin D; Tsivou M; Bonsang B; Abonnel C; Carsey T; Springer-Young M; Pszenny A; Suhre K. (1997). Hydrogen peroxide in the marine
atmospheric boundary layer during the Atlantic Stratocumulus Transition Experiment/Marine Aerosol and Gas Exchange experiment in
the eastern subtropical North Atlantic. J Geophys Res, 102,6003-6015.
Martin RV; Chance K; Jacob DJ; Kurosu TP; Spurr RJD; Bucsela E; Gleason JF; Palmer PI; Bey I; Fiore AM; Li Q; Yantosca RM; Koelemeijer
RBA. (2002). An improved retrieval of tropospheric nitrogen dioxide from GOME. J Geophys Res, 107, doi: 10.1029/2001 JD001027.
Martin RV; Jacob DJ; Chance KV; Kurosu TP; Palmer PI; Evans MJ. (2003). Global inventory of nitrogen oxide emissions constrained by
space-based observations of NO2 columns. J Geophys Res, 108,10.1029/2003JD003453.
Martin RV; Fiore AM; Van Donkelaar A. (2004a). Space-based diagnosis of surface ozone sensitivity to anthropogenic emissions. Geophys
Res Lett, 31, doi: 10.1029/2004GL019416.
Martin RV; Parrish DD; Ryerson TB; Nicks DK Jr; Chance K; Kurosu TP; Jacob DJ; Sturges ED; Fried A; Wert BP. (2004b). Evaluation of
GOME satellite measurements of trophospheric NO2 and HCHO using regional data from aircraft campaigns in the southeastern
United States. J Geophys Res, 109, doi: 10.1029/2004JD004869.
R-57
-------
Martin RV; Sioris CE; Chance K; Ryerson TB; Bertram TH; Wooldridge PJ; Cohen RC; Neuman JA; Swanson A; Flocke FM. (2006). Evaluation
of space-based constraints on nitrogen oxide emissions with regional aircraft measurements over and downwind of eastern North
America. J Geophys Res, 111, doi: 10.1029/2005JD006680.
Martinez M; Arnold T; Perner D. (1999). The role of bromine and chlorine chemistry for arctic ozone depletion events in Ny-Alesund and
comparison with model calculations. Annates Geophysicae, 17,941-956.
Martinez-Arkarazo I; Angulo M; Bartolome L; Etxebarria N; Olazabal MA; Madariaga JM. (2007). An integrated analytical approach to diagnose
the conservation state of building materials of a palace house in the metropolitan Bilbao (Basque Country, North of Spain). Analytica
Chimica Acta, 584,350-359.
Martinson L; Alveteg M; Warfvinge P. (2003). Parameterization and evaluation of sulfate adsorption in a dynamic soil chemistry model. Environ
Pollut, 124,119-125.
Martinson L; Lamersdorf N; Warfvinge P. (2005). The Soiling roof revisited - Slow recovery from acidification observed and modeled despite a
decade of "dean-rain" treatment. Environ Pollut, 135,293-302.
Mason BJ. (1992). Acid rain: its causes and its effects on inland waters. Science Technology and Society Series No 8 1992:i-x, 1-126.
Mason J; Seip HM. (1985). The current state of knowledge on acidification of surface waters and guidelines for further research. Ambio, 14,45-
51.
Mason RR; Wickman BE; Beckwith RC; Paul HG. (1992). Thinning and nitrogen fertilization in a grand fir stand infested with spruce budworm.
part I: insect response. For Sci, 38,235-251.
Massman WJ; Pederson J; Delany A; Grantz D; Denhartog G; Neumann HH; Ondey SP; Pearson R; ShawRH. (1994). An evaluation of the
regional acid deposition model surface module for ozone uptake at 3 sites in the San-Joaquin valley of California. J Geophys Res, 99,
8281-8294.
Matamala R; Gonzalez-Meier MA; Jastrow JD; Norby RJ; Schlesinger WH. (2003). Impacts of fine root turnover on forest NPP and soil C
sequestration potential. Science, 302,1385-1387.
Matamala R; Schlesinger WH. (2000). Effects of elevated atmospheric CO2 on fine root production and activity in an intact temperate forest
ecosystem. Glob Change Biol, 6,967-979.
MathTech. (1994). The regional model farm (RMF): An agricultural sector benefits assessment model: Version 3.0 for personal computers.
(Mathech, Inc; Princeton, NJ; Report to the Office of Air Quality Planning and Standards pp 1-83). Research Triangle Park, NC; U.S.
Environmental Protection Agency.
Mathur R; Dennis RL. (2000). A regional modeling analysis of reduced nitrogen cycling in the eastern United States. In: Preprints of the
symposium on atmospheric chemistry: Issues in the 21st century - held in Long Beach, CA. in January 2000 (p. 85-88). Boston, MA:
American Meteorological Society.
Mathur R; Dennis RL. (2003). Seasonal and annual modeling of reduced nitrogen compounds over the eastern United States: Emissions,
ambient levels, and deposition amounts. J Geophys Res-Atmos, 108.
Matson PA; Volkmann C; Grier CC; Gower ST; Billow C. (1992). Soil nitrogen cycling and nitrous oxide flux in a Rocky Mountain Douglas-fir
forest: effects of fertilization, irrigation and carbon addition. Biogeochemistry, 18,101-117.
Matsumi Y; Murakami S-l; Kono M; Takahashi K; Koike M; Kondo Y. (2001). High-sensitivity instrument for measuring atmospheric NO2. Anal
Chem, 73, 5485-5493.
Matsumi Y; Shigemori H; Takahashi K. (2005). Laser-induced fluorescence instrument for measuring atmospheric SO 2. Atmos Environ, 39,
3177-3185.
Mattson WJ Jr. (1980). Herbivory in relation to plant nitrogen content. Annu Rev Ecol Syst, 11,119-161.
Matuszek JE; Beggs GL. (1988). Fish species richness in relation to lake area, pH, and other abiotic factors in Ontario lakes. Can J Fish Aquat
Sci, 45,1931-1941.
Matzner E; Blanck K; Hartmann G; Stock R. (1989). Needle chlorosis pattern in relation to soil chemical properties in two Norway spruce (P/'cea
abies, Karst) forests of the German Harz Mountains. In: Bucher JB; Bucher-Wallin I, (Eds.), Air pollution and forest decline:
Proceedings of the 14th international meeting for specialists in air pollution effects of forest ecosystems; held in Interlaken, Switzerland
in October 1988. (pp. 195-199). Birmensdorf, Switzerland: International Union of Forestry Research Organizations.
Mauldin RL III; Tanner D J; Eisele FL. (1998). A new chemical ionization mass spectrometer technique for the fast measurement of gas phase
nitric acid in the atmosphere. J Geophys Res, 103, 3361-3367.
May JD; Burdette SB; Gilliam FS; Adams MBeth. (2005). Interspecific divergence in foliar nutrient dynamics and stem growth in a temperate
forest in response to chronic nitrogen inputs. Can J For Res, 35,1023-1030.
May PF; Till AR; Cumming MJ. (1972). Systems analysis of 35sulphur kinetics in pastures grazed by sheep. J Appl Ecol, 9,25-49.
McAuley DG; Longcore JR. (1988a). Foods of juvenile ring-necked ducks: Relationship to wetland pH. J Wildlife Management, 52,177-185.
McAuley DG; Longcore JR. (1988b). Survivial of juvenile ring-necked ducks on wetlands of different pH. J Wildlife Management, 52,169-176.
McClain ME; Boyer EW; Dent CL; Gergel SE; Grimm NB; Goffman PM; Hart SC; Harvey JW; Johnston CA; Mayorga E; McDowell WH; Pinay
G. (2003). Biogeochemical hot spots and hot moments at the interface of aquatic and terrestrial ecosystems. Ecosystems, 6, 301-312.
McClenahen JR. (1987). Effects of simulated throughfall pH and sulfate concentration on a deciduous forest soil. Water Air Soil Pollut, 35,319-
333.
McClenny WA. (Ed.), (2000). Recommended methods for ambient air monitoring of NO, NO2, NOy, and individual NOz species. (Report no.
EPA/600/R-01/005). Research Triangle Park, NC; National Exposure Research Laboratory; U.S. Environmental Protection Agency.
McClenny WA; Williams EJ; Cohen RC; Stutz J. (2002). Preparing to measure the effects of the NOx SIP Call-methods for ambient air
monitoring of NO, NO2, NOy, and individual NOz species. J Air Waste Manage Assoc, 52, 542-562.
McConnell V; Walls M. (2005). The value of open space: Evidence from studies of nonmarket benefits. Washington, DC; Resources for the
Future (RFF).
R-58
-------
McCormick JH; Jensen KM; Anderson LE. (1989). Chronic effects of low pH and elevated aluminum on survival, maturation, spawning, and
embryo-larval development of the fathead minnow in soft water. Water Air Soil Pollut, 43,293-307.
McCormick LH; Steiner KC. (1978). Variation in aluminum tolerance among six genera of trees. For Sci, 24,565-568.
McCulloch RB; FewGS; Murray GC Jr; Aneja VP. (1998). Analysis of ammonia, ammonium aerosols and acid gases in the atmosphere at a
commercial hog farm in eastern North Carolina, USA. Environ Pollut,102,263-268
McCune B. (1988). Lichen communities along O3 and SO2 gradients in Indianapolis. Bryologist, 91,223-228.
McCune B; Geiser L. (1997). Macrolichens of the Pacific Northwest. Corvallis, OR: Oregon State University Press.
McCurdy T; Zelenka MP; Lawrence PM; Houston RM; Burton R. (1999). Acid aerosols in the Pittsburgh Metropolitan area. Atmos Environ, 33,
5133-5145.
McDonald DG. (1983). The effects of H+upon the gills of freshwater fish. Canadian Journal of Zoology, 61,691-703.
McDonald DG; Milligan CL. (1988). Sodium transport in the brook trout, Salvelinus fontinalis: effects of prolonged lowpH exposure in the
presence and absence of aluminum. Can J Fish Aquat Sci, 45,1606-1613.
McDowell WH; Magill AH; Aitkenhead-Peterson JA; Aber JD; Merriam JL; Kaushal SS. (2004). Effects of chronic nitrogen amendment on
dissolved organic matter and inorganic nitrogen in soil solution. For Ecol Manage, 196,29-41.
McFiggans G; Plane JMC; Allan B; Carpenter LJ; Coe H; O'Dowd C. (2000). A modeling study of iodine chemistry in the marine boundary
layer. J Geophys Res, 105,14371-14385.
McFiggans G; Coe H; Burgess R; Allan J; Cubison M; Alfarra MR; Saunders R; Saiz-Lopez A; Plane JMC; Wevill D; Carpenter LJ; Rickard AR;
Monks PS. (2004). Direct evidence for coastal iodine particles from Laminaria macroalgae: Linkage to emissions of molecular iodine.
Atmos Chem Phys, 4,701-713.
McFiggans G. (2005). Marine aerosols and iodine emissions. (Arising from O'Dowd et al. (2002) Nature 417,632-636). Nature, 433, e13, doi:
10.1038/nature03372.
McFrederick QS; Kathilankal JC; Fuentes JD. (2008). Air pollution modifies floral scent trails. Atmos Environ, 42,2336-2348.
McGurk MD; Landry F; Tang A; Hanks CC. (2006). Acute and chronic toxicity of nitrate to early life stages of lake trout (Salvelinus
namaycush).and lake whitefish (Coregonus clupeaformis). Environ Toxicol Chem, 25,2187-2196.
McHale MR; Cirmo CP; Mitchell MJ; McDonnell JJ. (2004). Wetland nitrogen dynamics in an Adirondack forested watershed. Hydrological
Processes, 18,1853-1870.
McKane RB; Johnson LC; Shaver GR; Nadelhoffer KJ; Rastetter EB; Fry B; Giblin AE; Kielland K; Kwiatkowski BL; Laundre JA; Murray G.
(2002). Resource-based niches provide a basis for plant species diversity and dominance in arctic tundra. Nature, 415,68-72.
McKeen SA; Liu SC. (1993). Hydrocarbon ratios and photochemical history of air masses. Geophys Res Lett, 20, 2363-2366.
McKeen SA; Liu SC; Hsie E-Y; Lin X; Bradshaw JD; Smyth S; Gregory GL; Blake DR. (1996). Hydrocarbon ratios during PEM-WEST A: a
model perspective. J Geophys Res, 101,2087-2109.
McKenna DS; Konopka P; GrooR, J-U; Gunther G; Miiller R; Spang R; Offermann D; Orsolini Y. (2002). A new chemical Lagrangian model of
the stratosphere (CLaMS). 1. Formulation of advection and mixing. J Geophys Res, 107, doi: 10.1029/2000JD000114.
McKnight DM; Smith RL; Bradbury JP; Baron JS; Spaulding S. (1990). Phytoplankton dynamics in 3 Rocky Mountain Lakes, Colorado USA.
Arct Antarct Alp Res, 22,264-274.
McLain JET; Martens DA. (2006). Moisture controls on trace gas fluxes in semiarid riparian soils. Soil Sci Soc Am J, 70,367-377.
McLaughlin SB; Tjoelker MJ. (1992). Growth and physiological changes in red spruce saplings associated with acidic deposition at high
elevations in the southern Appalachians, USA. For Ecol Manage, 51,43-51.
McLaughlin S; Percy K. (1999). Forest health in North America: some perspectives on actual and potential roles of climate and air pollution.
Water Air Soil Pollut, 116,151-197.
McLaughlin SB; Wimmer R. (1999). Tansley review no.104: Calcium physiology and terrestrial ecosystem processes. New Phytol, 142,373-
417.
McNeil BE; Read JM; Driscoll CT. (2007). Foliar nitrogen responses to elevated atmospheric nitrogen deposition in nine temperate forest
canopy species. Environ Sci Technol, 41, 5191-5197.
McNeil BE; Read JM; Sullivan TJ; McDonnell TC; Fernandez IJ; Driscoll CT. (2008). The spatial pattern of nitrogen cycling in the Adirondack
Park, New York. Ecol Appl, 18,438-452.
McNicol DK; Bendell BE; McAuley D. (1987). Avain Trophic Relationships and Wetland Acidity. Transcr N Amer Wildlife Natural Resources
Conference, 52,619-627.
McNicol DK; Mallory ML; WedelesCHR. (1995). Assessing biological recovery of acid-sensitive lakes in Ontario, Canada. Water Air Soil Pollut,
85,457-462.
McNicol DK. (2002). Relation of lake acidification and recovery to fish, common loon and common merganser occurrence in Algoma Lakes.
Water Air Soil Pollut: Focus, 2,151-168.
McNulty SG; Aber JD. (1993). Effects of chronic nitrogen additions on nitrogen cycling in a high-elevation spruce-fir stand. Can J For Res, 23,
1252-1263.
McNulty SG; Aber JD; Newman SD. (1996). Nitrogen saturation in a high elevation New England spruce-fir stand. For Ecol Manage, 84,109-
121.
McNulty SG; Boggs J; Aber JD; Rustad L; Magill A. (2005). Red spruce ecosystem level changes following 14 years of chronic N fertilization.
For Ecol Manage, 219, 279-291.
McNulty SG; Cohen EC; Myers JAM; Sullivan TJ; Li H. (2007). Estimates of critical acid loads and exceedances for forest soils across the
conterminous United States. Environ Pollut, 149,281-292.
McPhaden MJ; Zebiak SE; Glantz MH. (2006). ENSO as an integrating concept in earth science. Science, 314,1740-1745.
R-59
-------
McRoberts RE; McWilliams WH; Reams GA; Schmidt TL; Jenkins JC; O'Neill KP; Miles PD; Brand GJ. (2004). Assessing sustainability using
data from the Forest Inventory and Analysis program of the United States Forest Service. J Sustain For, 18,23-46.
McWilliams PG; Potts WTW. (1978). The effects of pH and calcium concentrations on gill potentials in the brown trout, Salmotrutta. Journal of
Comparative Physiology, 126,277-286.
Meagher JF; Cowling EB; Fehsenfeld FC; ParkhurstWJ. (1998). Ozone formation and transport in southeastern United States: overview of the
SOS Nashville/Middle Tennessee Ozone Study. J Geophys Res, 103,22213-22223.
Mebust MR; Eder BK; Binkowski FS; Roselle SJ. (2003). Models-3 community multiscale air quality (CMAQ).model aerosol component. 2.
Model evaluation. J Geophys Res, 108, doi: 10.1029/2001 JD001410.
Meehl GA; Washington WM; Collins WD; Arblaster JM; Hu A; Buja LE; Strand WG; Teng H. (2005). How much more global warming and sea
level rise? Science, 307,1769-1772.
Meili M. (1991). Fluxes, pools and turnover of mercury in Swedish forest lakes Water Air Soil Pollut, 56,719-727.
Meiwes KJ; Khanna PK. (1981). Distribution and cycling of sulphur in the vegetation of two forest ecosystems in an acid rain environment.
Plant Soil, 60, 369-375.
Meixner T; Fenn M. (2004). Biogeochemical budgets in a Mediterranean catchment with high rates of atmospheric N deposition - importance of
scale and temporal asynchrony. Biogeochemistry, 70,331-356.
Meixner T; Fenn ME; Wohlgemuth P; Oxford M; Riggan P. (2006). N Saturate symptoms in chaparral catchments are not reversed by
prescribed fire. Environ Sci Technol, 40, 2887-2894.
Melack JM; Stoddard JL; Ochs CA. (1985). Major ion chemistry and sensitivity to acid precipitation of Sierra Nevada lakes. Water Resour Res,
21,27-32.
Melack JM; Sickman JO; Leydecker A. (1988). Comparative analyses of high-altitude lakes and catchments in the Sierra Nevada: susceptibility
to acidification. (Final report). Sacramento, CA: California Air Resources Board.
Melack JM; Cooper SC; Jenkins TM; Barmuta L Jr; Hamilton S; Kratz K; Sickman J; Soiseth C. (1989). Chemical and biological characteristics
of Emerald Lake and the streams in its watershed, and the response of the lake and streams to acidic deposition. (Final report).
Sacramento, CA: California Air Resources Board.
Melack JM; Stoddard JL. (1991). Sierra Nevada: Unacidified, very dilute waters and mildly acidic atmospheric deposition. In: Charles DF,
Christie S (Eds.), Acidic deposition and aquatic ecosystems: regional case studies (pp. 503-530). New York: Springer-Verlag.
Melillo JM; Aber JD; Muratore JM. (1982). Nitrogen and lignin control of hardwood leaf litter decomposition dynamics. Ecology, 63,621-662.
Melillo JM; Steudler PA; Aber JD; Newkirk K; Lux H; Bowles FP; Catricala C; Magill A; Ahrens T; Morrisseau S. (2002). Soil Warming and
Carbon-Cycle Feedbacks to the Climate System (Vol. 298, pp. 2173-2176).
Mendelssohn IA. (1979). Nitrogen Metabolism in the Height Forms of Spartina Alterniflora in North Carolina. Ecology, 60,574-584.
Mendoza-Dominguez A; Russell AG. (2000). Iterative inverse modeling and direct sensitivity analysis of a photochemical air quality model.
Environ Sci Technol, 3,4974-4981.
Mendoza-Dominguez A; Russell AG. (2001a). Emission strength validation using four-dimensional data assimilation: application to primary
aerosol and precursors to ozone and secondary aerosol. J Air Waste Manage Assoc, 51,1538-1550.
Mendoza-Dominguez A; Russell AG. (2001b). Estimation of emission adjustments from the application of four-dimensional data assimilation to
photochemical air quality modeling. Atmos Environ, 35,2879-2894.
Meng F-R; Cox RM; Arp PA. (1994). Fumigating mature spruce branches with SO2 - Effects on net photosynthesis and stomatal conductance.
Can J For Res, 24,1464-1471.
Meng Z; Seinfeld JH. (1996). Time scales to achieve atmospheric gas-aerosol equilibrium for volatile species. Atmos Environ, 30,2889-2900.
Mennen MG; Van Elzakker BG; Van Putten EM; Uiterwijk JW; Regts TA; Van Hellemond J; Wyers GP; Otjes RP; Verhage AJL; Wouters LW.
(1996). Evaluation of automatic ammonia monitors for application in an air quality monitoring network. Atmos Environ, 30,3239-3256.
Meyer MW; Evers DC; Daulton T; Braselton WE. (1995). Common loons (Gaviaimmei) nesting on lowpH lakes in nothern Wisconsin have
elevated blood mercury content. Water Air Soil Pollut, 80,871-880.
Meyer JL; Swank WT. (1996). Ecosystem management challenges ecologists. Ecol Appl, 6,738-740.
Meyers TP; Sickles JE; Dennis R; Russell K; Galloway J; Church T. (2001). Atmospheric nitrogen deposition to coastal estuaries and their
watersheds. In: Valigura RM, Alexander RB, Castro MS, Greening H, Meyers T, Paerl H, Turner RE (Eds.), Nitrogen loading in coastal
water bodies: An atmospheric perspective. (Series: Coastal and Estuarine Studies, Volume 57; pp 53-76). Washington, DC: American
Geophysical Union.
Michalski G; Meixner T; Fenn M; Hernandez L; Sirulnik A; Allen E; Thiemens M. (2004). Tracing Atmospheric Nitrate Deposition in a Complex
Semiarid Ecosystem Using Δ17O. Environ Sci Technol, 38, 2175-2181.
Michel TJ; Saros JE; Interlandi SJ; Wolfe AP. (2006). Resource requirements of four freshwater diatom taxa determined by in situ growth
bioassays using natural populations from alpine lakes. Hydrobiologia, 568,235-243.
Mierle G. (1990). Aqueous inputs of mercury to Precambrian Shield Lakes in Ontario. Environ Toxicol Chem, 9,843-851.
Mierle G; Ingram R. (1991). The role of humic substances in the mobilization of mercury from watersheds. Water Air Soil Pollut, 56,349-357.
Mikan CJ; Zak DR; Kubiske ME; Pregitzer KS. (2000). Combined effects of atmospheric CO2 and N availability on the belowground carbon and
nitrogen dynamics of aspen mesocosms. Oecologia, 124,432-445.
Milberg P; Lamont BB; Perez-Fernandez MA. (1999). Survival and growth of native and exotic composites in response to a nutrient gradient.
Plant Ecology, 145,125-132.
Milchunas DG; Lauenroth WK. (1995). Inertia in plant community structure: State changes after cessation of nutrient-enrichment stress. Ecol
Appl, 5, 452-458.
Milford JB; Gao D; Sillman S. (1994). Total reactive nitrogen (NOjsub y}).as an indicator of the sensitivity of ozone to reductions in hydrocarbon
and NOjsub x} emissions. J Geophys Res; VOL 99 ; ISSUE: D2 ; PBD: 20 Feb 1994, pp. 3533-3542 ; PL:.
R-60
-------
Milindalekha J; Vashkin VN; Towpprayoon S. (2001). Calculation and mapping of sulfur critical loads for terrestrial ecosystems of Thailand.
Water Air Soil Pollut, 130,1265-1270.
Millennium Ecosystem Assessment. (2003). Ecosystems and Human Well-being: A Framework for Assessment. Washington, DC: Island Press.
Miller AE; Bowman WD. (2002). Variation in nitrogen-15 natural abundance and nitrogen uptake traits among co-occuring alpine species: Do
species partition by nitrogen form? Oecologia, 130, 609-616.
Miller EK; Blum JD; Friedland AJ. (1993). Determination of soil exchangeable-cation loss and weatheirng rates using Sr isotopes. Nature, 362,
438-441.
Miller EK; McDonald K. (2006). Assessment of forest sensitivity to nitrogen and sulfur deposition in Maine. (Report to the Maine Department of
Environmental Protection). http://www.maine.aov/dep/air/acidrain/ME-Forest-Mappina-Report-2007-01-26.pdf. Augusta, ME;
Conference of New England Governors and Eastern Canadian Premiers, Forest Mapping Group ; Bureau of Air Quality; Maine
Department of Environmental Protection.
Miller PR; Longbotham GJ; Longbotham CR. (1983). Sensitivity of selected western conifers to ozone. Plant Dis, 67,1113-1115.
Miller JD; Lindsay BE. (1993). Willingness to pay for a state gypsy moth control program in New Hampshire: A contingent valuation case study.
J Econ Entomol, 86, 828-837.
Miller PR; McBride JR., (Eds.), (1999). Oxidant air pollution impacts in the Montane forests of southern California: A case study of the San
Bernardino Mountains. (Series: Ecological Studies: Volume 134) New York, NY: Springer-Verlag.
Miller WW, Johnson DW; Loupe TM; Sedinger JS; Carroll EM; Murphy JD; Walker RF; Glass D. (2006). Nutrients Flow from Runoff at Burned
Forest Site in Lake Tahoe Basin. California Agriculture, 60,65-71.
Mills KH; Chalanchuk SM; Mohr LC; Davies IJ. (1987). Responses offish populations in Lake 223 to 8 years of experimental acidification. Can
J Fish Aquat Sci, 44,114-125.
Minnich RA; Barbour MG; Burk JH; Fernau RF. (1995). Sixty Years of Change in Californian Conifer Forests of the San Bernardino Mountains.
Conserv Biol, 9,902-914.
Minnich RA; Dezzani RJ. (1998). Historical decline of coastal sage scrub in the Riverside-Perris Plain, California. Western Birds, 29,366-391.
Minocha R; Shortle WC; Lawrence GB; David MB; Minocha SC. (1997). Relationships among foliar chemistry, foliar polyamines, and soil
chemistry in red spruce trees growing across the northeastern United States. Plant Soil, 191,109-122.
Mitchell MJ; Harrison RB; Fitzgerald JW; Johnson DW; Lindberg SE; Zhang Y; Autry A. (1992a). Sulfur distribution and cycling in forest
ecosystems. In: Johnson DW, Lindberg SE (Eds.), Atmospheric deposition and forest nutrient cycling: A synthesis of the integrated
forest study. (Series: Billings WD; Golley F; Lange OL; Olson JS; Remmert H (Eds.) Ecological studies: Analysis and synthesis,
Volume 91; pp. 90-129). New York, NY: Springer-Verlag.
Mitchell, M. J; Lindberg, S. E; Johnson, D. W; Lindberg, S. E. (1992b). Sulfur, chemistry, deposition, and cycling in forests. In: Johnson DW,
Lindberg SE (Eds.) Atmospheric deposition and forest nutrient cycling: A synthesis of the integrated forest study. (Series: Billings WD;
Golley F; Lange OL; Olson JS; Remmert H (Eds.) Ecological studies: Analysis and synthesis, Volume 91; pp. 72-149). New York, NY:
Springer-Verlag.
Mitchell MJ; Driscoll CT; Kahl JS; Likens GE; Murdoch PS; Pardo LH. (1996). Climatic control of nitrate loss from forested watersheds in the
northeast United States. Environ Sci Technol, 30,2609-2612.
Mitchell MJ; Piatek KB; Christopher S; Mayer B; Kendall C; McHale PJ. (2006). Solute sources in stream water during consecutive fall storms
in a northern hardwood forest watershed: a combined hydrological, chemical and isotopic approach. Biogeochemistry, 78,217-246.
Mitchell PJ; Branfireun BA; Kolka, RK. (2008). Assessing sulfate and carbon controls on net methylmercury production in peatlands: An in situ
mesocosm approach. Appl Geochem, 23,503-518.
Mitchell RJ; Sutton MA; Truscott A-M; Leith ID; Cape JN; Pitcairn CER; Van Dijk N. (2004). Growth and tissue nitrogen of epiphytic Atlantic
bryophytes: effects of increased and decreased atmospheric N deposition. Functional Ecology, 18,322-329.
Mitchell RJ; Truscot AM; Leith ID; Cape JN; Dijk NV; Tang YS; Fowler D; Sutton MA. (2005). A study of the epiphytic communities of Atlantic
oak woods along an atmospheric nitrogen deposition gradient. J Ecol, 93,482-492.
Mitsch W; Gosselink J. (1986). Wetlands, (1st Edition). New York, NY: Van Nostrand Reinhold International.
Mitsch W; Gosselink J. (2000). Wetlands, (3rd Edition). New York, NY: John Wiley & Sons.
Mitsch W; Gosselink J. (2007). Wetlands, (4th Edition). New York, NY: John Wiley & Sons.
Moayeri M; Meng FR; Arp PA; Foster NW. (2001). Evaluating critical soil acidification loads and exceedances for a deciduous forest at the
Turkey Lakes Watershed. Ecosystems, 4, 555-567.
Mocker D; Jung K; Forstel H; Schuurmann G. (1998). Isotopic and enzymatic investigations into the assimilation and effect of NO2 on C3 and
C4 plants. Journal of Appl Bot, 72,186-190.
Mohn J; Schurmann A; Hagedorn F; Schleppi P; Bachofen R. (2000). Increased rates of denitrification in nitrogen-treated forest soils. For Ecol
Manage, 137,113-119.
Moldan F; Wright RF; Ferrier RC; Andersson Bl; Hultberg H. (1998). Simulating the Gardsjon covered catchment experiment with the MAGIC
model. In: Hultberg H, Skeffington RA (Eds.), Experimental Reversal of Acid Rain Effects. The Gardsjon Roof Project, 466 pp. (pp.
p.351-362). Chichester, UK: Wiley and Sons.
Molot LA; Dillon PJ; LaZerte BD. (1989). Changes in ionic composition of streamwater during snowmelt in central Ontario. Can J Fish Aquat
Sci, 46,1658-1666.
Momen B; Lawrence GB; Nierzwicki-Bauer SA; Sutherland JW; Eichler L; Harrison JP; Boylen CW. (2006). Trends in summer chemistry linked
to productivity in lakes recovering from acid deposition in the Adirondack Region of New York. Ecosystems, 9,1306-1317.
Moncoulon D; Probst A; Party J. (2004). Weathering, atmospheric deposition and vegetation uptake: role for ecosystem sensitivity to acid
deposition and critical load. Comptes Rendus Geoscience, 336,1417-1426.
R-61
-------
Monteith DT; Stoddard JL; Evans CD; de Wit HA; Forsius M; Hogasen T; Wilander A; Skjelkvale BL; Jeffries DS; Vuorenmaa J; Keller B;
Kopacek J; Vesely J. (2007). Dissolved organic carbon trends resulting from changes in atmospheric deposition chemistry. Nature,
450, 537-540.
Montgomery DR; Buffington JM. (1998). Channel processes, classification, and response. In: Naiman, R. J; Bilby, R., eds. River ecology and
management: Lessons from the Pacific coastal ecoregion. (pp. 13-42). New York: Springer-Verlag.
Moore DRJ; Keddy PA; GaudetCL; Wisheu IC. (1989). Conservation of wetlands: Do infertile wetlands deserve a higher priority? Biological
Conservation, 47,203-217.
Moore JC; McCann K; Setala H; De Ruiter PC. (2003). Top-down is bottom-up: does predation in the rhizosphere regulate aboveground
dynamics? Ecology, 84,846-857.
Moore JD; Ouimet R. (2006). Ten-year effect of dolomitic lime on the nutrition, crown vigor, and growth of sugar maple. Can J For Res, 36,
1834-1841.
Moore MT; Lizotte JRE; Cooper CM; Smith JS; Knight SS. (2004a). Survival and growth of Hyalella azteca exposed to three Mississippi Oxbow
Lake sediments. Bull Environ Contam Toxicol, 72,777-783.
Moore TR; Trofymow JA; Taylor B; Prescott C; Camire C; Duschene L; Fyles J; Kozak L; Kranabetter M; Morrison I; Siltanen M; Smith S; Titus
B; Visser S; Wein R; Zoltai S. (1999). Litter decomposition rates in Canadian forests. Glob Change Biol, 5,75-82.
Moore T; Blodau C; Turunen J; Roulet N; Richard PJH. (2004b). Patterns of nitrogen and sulfur accumulation and retention in ombrotrophic
bogs, eastern Canada. Glob Change Biol, 11,356-367
Moorthi S; Suarez MJ. (1992). Relaxed Arakawa-Schubert: a parameterization of moist convection for general circulation models. Mon Weather
Rev, 120,978-1002.
Morel FMM; Hering JG. (1993). Principles and Applications of Aquatic Chemistry. New York, NY: Wiley-lnterscience.
Morey ER; Shaw WD. (1990). An economic model to assess the impact of acid rain: A characteristics approach to estimating the demand for
and benefits from recreational fishing. In: Link AN; Smith VK (Eds.), Advances in Applied Micro-Economics, Volume 5; Recent
developments in the modeling of technical change, and modeling demand for and valuation of recreations resources, (pp 195-216).
Greenwich, CT: JAI Press, Inc.
Morgan C; Owens N. (2001). Benefits of water quality policies: the Chesapeake Bay. Ecol Econ, 39,271-284.
Morley SA; Karr JR. (2002). Assessing and restoring the health of urban streams in the Puget Sound Basin. Conserv Biol, 16,1498-1509.
Morris DP; Lewis WM Jr. (1988). Phytoplankton nutrient limitation in Colorado mountain lakes. Freshw Biol, 20,315-327.
Morris JT. (1991). Effects of nitrogen loading on wetland ecosystems with particular reference to atmospheric deposition. Annual Review of
Ecological Systems, 22,257-279.
Morth CM; Torssander P; Kjonaas OJ; Stuanes AO; Moldan F; Giesler R. (2005). Mineralization of organic sulfur delays recovery from
anthropogenic acidification. Environ Sci Technol, 39, 5234-5240.
Mosier A; Schimel D; Valentine D; Branson K; Parton W. 1991. Methane and nitrous-oxide fluxes in native, fertilized and cultivated grasslands.
Nature, 350, 330-332.
Moss MR. (1978). Sources of sulfur in the environment; the global sulfur cycle. In Nriagu JO (Ed.), Sulfur in the environment, Part 1: The
atmospheric cycle (pp. 23-50). New York: Wiley.
Mount DR; Hockett JR; Gern WA. (1988). Effect of long-term exposure to acid, aluminum, and low calcium on adult brook trout(Salvelinus
fontinalis). 2. Vitellogenesis and osmoregulation. Can J Fish Aquat Sci, 45,1633-1642.
Mudd JB; Kozlowski TT., (Eds.) (1975). Responses of plants to air pollution. (Physiological ecology, A series of monographs, texts, and
treatises). New York, NY: Academic Press.
Mueller DK; Ruddy BC; Battaglin WA. (1997). Logistic model of nitrate in streams of the upper-midwestern United States. J Environ Qual, 26,
1223-1230.
Mueller PK; Collins JF. (1980). Development of a particulate sulfate analyzer. (Report no. P-1382F, NTIS no EPRI-EA-1492). Palo Alto, CA:
Electric Power Research Institute.
Muir PS; McCune B. (1988). Lichens, free growth, and foliar symptoms of air pollution: are the stories consistent? J Environ Qual, 17,361-370.
Mulder J; Stein A. (1994). The solubility of aluminum in acidic forest soils: Long-term changes due to acid deposition. Geochim Cosmochim
Acta, 58, 85-94.
Mulder J; van Breemen N; Eijck HC. (1989). Depletion of soil aluminium by acid deposition and implications for acid neutralization. Nature, 337,
247-249.
Mulholland PJ; Valett HM; Webster JR; Thomas SA; Cooper LW; Hamilton SK; Peterson B. (2004). Stream denitrification and total nitrate
uptake rates measured using a field 15N tracer addition approach. Limnol Oceanogr, 49,809-820.
Mulholland PJ; Helton AM; Poole GC; Hall RO Jr; Hamilton SK; Peterson BJ; Tank JL; Ashkenas LR; Cooper LW; Dahm CN; Dodds WK;
Findlay SEG; Gregory SV; Grimm NB; Johnson SL; McDowell WH; Meyer JL; Valett HM; Webster JR; Arango CP; BeaulieuJJ; Bernot
MJ; Burgin AJ; CrenshawCL; Johnson LT; Niederlehner BR; O'Brien JM; Potter JD; Sheibley RW; Sobota DJ; Thomas SM. (2008).
Stream denitrification across biomes and its response to anthropogenic nitrate loading. Nature, 452, 202-206.
Mullen JK; Menz FC. (1985). The effect of acidification damages on the economic value of the Adirondack fishery to New York anglers. Am J
Agric Econ, 67,112-119.
Muller BT, B;Rennenberg, H. (1996). Interaction between atmospheric and pedospheric nitrogen nutrition in spruce (Picea abies L. Karst)
seedlings. Plant, Cell, and Environment, 19,345-355.
Munger JW; Wofsy SC; Bakwin PS; Fan S-M; Goulden ML; Daube BC; Goldstein AH. (1996). Atmospheric deposition of reactive nitrogen
oxides and ozone in a temperate deciduous forest and a subarctic woodland. 1. Measurements and mechanisms. J Geophys Res,
101,12639-12657.
R-62
-------
Munger JW; Fan S-M; Bakwin PS; Goulden ML; Goldstein AH; Colman AS; Wofsy SC. (1998). Regional budgets of nitrogen oxides from
continental sources: Variations of rates for oxidation and deposition with season and distance from source regions. J Geophys Res,
103, 8355-8368.
Muniz IP. (1991). Freshwater acidification: its effects on species and communities of freshwater microbes, plants and animals. 97B, 227-254.
Munoz MSS; Rossi MJ. (2002). Heterogeneous reactions of HNO3 with flame soot generated under different combustion conditions. Reaction
mechanism and kinetics. Phys Chem Chem Phys, 4, 5110-5118.
Munoz-Hincapie MJ; Morell JM; Corredor JE. (2002). Increase of nitrous oxide flux to the atmosphere upon nitrogen addition to red mangroves
sediments. Mar Pollut Bull, 44, 992-996.
Munro DC; Gupta UC. (1969). Influence of sulfur, molybdenum and phosphorus on chemical composition and yields of Brussels sprouts and of
molybdenum on sulfur contents of several plant species grown in the greenhouse. Soil Sci, 107,114-118.
Munson RK; Gherini SA. (1991). Processes influencing the acid-base chemistry of surface waters. In: Charles DF; Christie S (Eds.), Acidic
deposition and aquatic ecosystems: Regional case studies, (pp 9-34). New York: Springer-Verlag.
Munson RK; Gherini SA. (1993a). Analysis of the mineral acid-base components of acid-neutralizing capacity in Adirondack Lakes. Water
Resour Res, 29,881-890.
Munthe J; Bodaly RAD; Branfireun BA; Driscoll CT; Gilmour CC; Harris R; Horvat M; Lucotte M; Malm 0.(2007). Recovery of Mercury-
Contaminated Fisheries. AMBIO: A Journal of the Human Environment, 36,33-44.
Murdoch PS; Burns DA; Lawrence GB. (1998). Relation of climate change to the acidification of surface waters by nitrogen deposition. Environ
Sci Technol, 32,1642-1647.
Murdoch PS; Stoddard JL. (1992). The role of nitrate in the acidification of streams in the Catskill Mountains of New York. Water Resour Res,
28, 2707-2720.
Murdoch PS; Stoddard JL. (1993). Chemical characteristics and temporal trends in eight streams of the Catskill Mountains. Water Air Soil
Pollut, 67, 367-395.
Murdoch PS; Shanley JB. (2006a). Detection of water quality trends at high, median, and low flow in a Catskill Mountain stream, New York,
through a new statistical method. Water Resour Res, 42, doi: 10.1029/2004WR003892.
Murdoch PS; Shanley JB. (2006b). Flow-specific trends in river-water quality resulting from the effects of the Clean Air Act in three mesoscale,
forested river basins in the northeastern United States through 2002. Environ Monit Assess, 120,1-25.
Murphy JD, Johnson DW; Miller WW; Walker RF; Blank RR. (2006). Prescribed fire effects on forest floor and soil nutrients in a Sierra Nevada
ecosystem. Soil Sci, 171,181-199.
Murphy JD; Johnson DW; Miller WW; Walker RF; Carroll EF; Blank RR. (2006). Wildfire Effects on Soil Nutrients and Leaching in a Tahoe
Basin Watershed. J Environ Qual, 35,479-489.
Murphy JJ; Delucchi MA; McCubbin DR; Kim HJ. (1999). The cost of crop damage caused by ozone air pollution from motor vehicles. J Environ
Manage, 55,273-289.
Musselman RC; Hudnell L; Williams MW; Sommerfeld RA. (1996). Water chemistry of Rocky Mountain Front Range aquatic ecosystems (No.
Research Paper RM-RP-325). Ft. Collins, CO: Rocky Mountain Forest and Range Experiment Station; U.S. Forest Service;
U.S. Department of Agriculture (USDA).
Myhre G; Berglen TF; Myhre CE; Isaksen ISA. (2004). The radiative effect of the anthropogenic influence on the stratospheric sulfate aerosol
layer. Tellus Ser B, 56,294-299.
Mylonda S. (1996). Sulphur dioxide emissions in Europe 1880—1991 and their effect on sulphur concentrations and depositions. Tellus, 48,
662-689.
Myneni RB; Keeling CD; Tucker CJ; Asrar G; Nemani RR. (1997). Increased plant growth in the northern high latitudes from 1981 to 1991.
Nature, 386, 698-702.
Nadelhoffer K; Downs M; Fry B; Magill A; Aber J. (1999a). Controls on N retention and exports in a forested watershed. Environ Monit Assess,
55(1), 187-210.
Nadelhoffer KJ; Emmett BA; Gundersen P; Kjonaas OJ; Koopmans CJ; Schleppi P; Tietema A; Wright RF. (1999b). Nitrogen deposition makes
a minor contribution to carbon sequestration in temperate forests. Nature, 398,145-148.
Nadelhoffer KJ. (2000). The potential effects of nitrogen deposition on fine-root production in forest ecosystems. New Phytol, 147,131-139.
Nadelhoffer KJ; Colman BP; Currie WS; Magill AH; Aber JD. (2004). Decadal-scale fates of 15N tracers to oak and pine stands under ambient
and elevated N inputs at the Harvard Forest (USA). For Ecol Manage, 196,89-107.
NADP/NTN (2004). http://nadp.sws.uiuc.edu. Annual 2004 isopleth maps. http://nadp.sws.uiuc.edu/isopleths/maps2004/2004Maps.pdf.
http://nadp.sws.uiuc.edu/isopleths/annualmaps.asp. Champaign, IL; Program Office, National Atmospheric Deposition Program
(NADP) / National Trends Network (NTN).
Nagao I; Matsumoto K; Tanaka H. (1999). Sunrise ozone destruction found in the sub-tropical marine boundary layer. Geophys Res Lett, 26,
3377-3380.
Nagase H; Ose Y; Sato T; Ishikawa T. (1984). Mercury methylation by compounds in humic material. Sci Total Environ, 32,147-156.
Nagy KA; Henen BT; Vyas DB. (1998). Nutritional quality of native and introduced food plants of wild desert tortoises. J Herpetol, 32,260-267.
Najjar RG; Walker HA; Anderson PJ; Barron EJ; Bord RJ; Gibson JR; Kennedy VS; Knight CG; Megonigal JP; O'Connor RE; Polsky CD; Psuty
NP; Richards BA; Sorenson LG; Steele EM; Swanson RS. (2000). The potential impacts of climate change on the mid-Atlantic coastal
region. Clim Res, 14, 219-233.
Nakamura K; Kondo Y; Chen G; Crawford JH; Takegawa N; Koike M; Kita K; Miyazaki Y; Shetter RE; Lefer BL; Avery M; Matsumoto J. (2003).
Measurement of NO2 by the photolysis conversion technique during the Transport and Chemical Evolution Over the Pacific (TRACE-
P) campaign. J Geophys Res, 108D24.
NAPAP. (1990). 1990 integrated assessment report. Washington, DC: National Acid Precipitation Assessment Program (NAPAP).
R-63
-------
NAPAP. (1991). Experience and legacy of NAPAP. Report of the Oversight Review Board of the National Acid Precipitation Assessment
Program (No. PB-92-128230/XAB). Washington, DC: National Acid Precipitation Assessment Program (NAPAP); Council on
Environmental Quality.
NAPAP. (1998). NAPAP biennial report to Congress: An integrated assessment. Silver Spring, MD: National Acid Precipitation Assessment
Program (NAPAP); Committee on Environment and Natural Resources (CENR) of the National Science and Technology Council
(NSTC).
NAPAP. (2005). National acid precipitation assessment program report to Congress: An integrated assessment.
http://www.esrl.noaa.aov/csd/aars/reports/napapreport05.pdf. Silver Spring, MD: National Acid Precipitation Assessment Program
(NAPAP); Committee on Environment and Natural Resources (CENR) of the National Science and Technology Council (NSTC).
NARSTO. (2005). Improving emission inventories for effective air quality management across North America. A NARSTO assessment. (Report
no. NARSTO-05-001). http://www.narsto.org/Section.src?SID=8. Pasco, WA: The NARSTO Emission Inventory Assessment Team
Nash TH; Sigal LL. (1999). Epiphytic lichens in the San Bernardino mountains in relation to oxidant gradients. In: Miller PR, McBride JR (Eds.),
Oxidant air pollution impacts on the montane forests of southern California: A case study of the San Bernardino mountains. Ecological
Studies, 134, (pp. 223-234). New York, NY: Springer-Verlag.
Nasholm T; Ekblad A; Nordin A; Giesler R; Hogberg M; Hogberg P. (1998). Boreal forest plants take up organic nitrogen. Nature, 392:914-916.
Naumova YY; Offenberg JH; Eisenreich SJ; Meng QY; Polidori A; Turpin BJ; Weisel CP; Morandi MT; Colome SD; Stock TH; Winer AM;
Alimokhtari S; Kwon J; Maberti S; Shendell D; Jones J; Farrar C. (2003). Gas/particle distribution of polycyclic aromatic hydrocarbons
in coupled outdoor/indoor atmospheres. Atmos Environ, 37,703-719.
Neary AJ; Mistry E; Vanderstar L. (1987). Sulphate relationships in some central Ontario forest soils. Can J Soil Sci, 67, 341-352.
Neary BP; Dillon PJ. (1988). Effects of sulphur deposition on lake-water chemistry in Ontario, Canada. Nature, 333,340-343.
Neary DG; Klopatek CC; DeBano LF; Ffolliot PF. (1999). Fire effects on belowground sustainability: A review and synthesis. For Ecol Manage,
122,51-71.
Needelman M; Kealy MJ. (1995). Recreational swimming benefits of New Hampshire lake water quality policies: An application of a repeated
discrete choice model. Agricultural and Resource Economics Review, 24,78-87.
Neff JS; Bowman WD; Holland EA; Fisk MC; Schmidt SK. (1994). Fluxes of nitrous oxide and methane from nitrogen-amended soils in a
Colorado alpine ecosystem. Biogeochemistry, 27,23-33.
Neff JC; Townsend AR; Gleixner G; Lehman SJ; Turnbull J; Bowman WD. (2002). Variable effects of nitrogen additions on the stability and
turnover of soil carbon. Nature, 419,915-917.
NEG/ECP Forest Mapping Group. (2001). Protocol for assessment and mapping of forest sensitivity to atmospheric S and N deposition. Acid
rain action plan. (Action item 4: forest mapping research project). http://www.ecosvstems-research.com/fmi/Protocol.pdf. Paper
presented at the The Conference of the New England Governors (NEG) and Eastern Canadian Premiers (ECP). Boston, MA: New
England Governors/Eastern Canadian Premiers.
Nellemann C; Thomsen MG. (2001). Long-term changes in forest growth: potential effects of nitrogen depsoition and acidification. Water Air
Soil Pollut, 128,197-205.
Nelson PO. (1991). Cascade Mountains: Lake chemistry and sensitivity of acid deposition. In: Charles DF, Christie S (Eds.), Acidic deposition
and aquatic ecosystems: regional case studies (pp. 531-563). New York: Springer-Verlag.
Neuman JA; Huey LG; Dissly RW; Fehsenfeld FC; Flocke F; Holecek JC; Holloway JS; Hubler G; Jakoubek R; Nicks DK Jr; Parrish DD;
Ryerson TB; Sueper DT; Weinheimer AJ. (2002). Fast-response airborne in-situ measurements of HNO3 during the Texas 2000 Air
Quality Study. J Geophys Res, 107,10.1029/2001 JD001437.
Newman DH. (1987). An econometric analysis of the southern softwood stumpage market: 1950-1980. For Sci,33,932-945.
Newton I; Hass MB. (1988). Pollutants in Merlin eggs and their effects on breeding. Br Birds, 81,258-269.
Nicholls KH; Hopkins GJ; Standke SJ; Nakamoto L. (2001). Trends in total phosphorus in Canadian nearshore waters of Laurentian Great
Lakes: 1976-1999. Great Lakes Res, 27,402-422.
Nicks DK; Benner RL. (2001). Subminute measurements of SO2 and low parts per trillion by volume mixing ratios in the atmosphere. J
Geophys Res, 106, 27769-27776.
Nieboer E; Richardson DHS; Puckett KJ; Tomassini FD. (1976). The phytotoxicity of sulphur dioxide in relation to measurable responses in
lichens. In: Mansfield T (Ed.), Effects of Air Pollutions on Plants (pp 61). Cambridge, U.K: Cambridge University Press.
Niering WA. (1998). Forces that shaped the forests of the northeastern United States. Northeast Nat, 5,99-110.
Nikitas C; Clemitshaw KC; Oram DE; Penkett SA. (1997). Measurements of PAN in the polluted boundary layer and free troposphere using a
luminol-N02 detector combined with a thermal converter. J Atmos Chem, 28,339-359.
Nilles MA; Conley BE. (2001). Changes in the chemistry of precipitation in the United States, 1981-1998. Water Air Soil Pollut, 130,409-414.
Nilsson J. (Ed.) (1986). Critical loads of nitrogen and sulphur. Environmental report 1986:11). Copenhagen, Denmark: Council of Ministers.
Nilsson J; Grennfelt P. (1988). Critical loads for sulphur and nitrogen. In: Nilsson J; Grennfelt P (Eds.), Report from UNECE/Nordic Council
workshop held in Skokloster, Sweden; March 19-24,1988. (Miljorapport: 1988:15; pp 225-268). Copenhagen, Denmark: Published by
Nordic Council of Ministers.
Nilsson LO; Wiklund K. (1995). Indirect effects of N and S deposition on a Norway spruce ecosystem. An update of findings within the Skogaby
project. Water Air Soil Pollut, 85,1613-1622.
Nilsson L; Giesler R; Baath E; Wallander H. (2004). Growth and biomass of mycorrhizal mycelia in coniferous forests along short natural
nutrient gradients. NewPhytol, 165, 613-622.
Nilsson SI; Miller JG; Miller JD. (1982). Forest growth as a possible cause of soil and water acidification: an examination of the concepts.
Oikos, 39,40-49.
R-64
-------
Nilsson SI. (1993). Acidification of Swedish ologotrophic lakes - interactions between deposition, forest growth and effects on lake-water
quality. Ambio, 22, 272-276.
Nixon S. W. (1995). Coastal marine eutrophication: a definition, social causes, and future concerns. Ophelia, 41,199-219.
Nodvin SC; Miegroet H; Lindberg SE; Nicholas NS; Johnson DW. (1995). Acidic deposition, ecosystem processes, and nitrogen saturation in a
high elevation Southern Appalachian watershed. Water Air Soil Pollut, 85,1647-1652.
Nolin MP; Stewart HT; Currier PN; Estabrook RH. (2004). Total Maximum Daily Load (TMDL) for 65 Acid Impaired New Hampshire Ponds
(Final). New Hampshire Department of Environmental Services Document Number: R-WD-0417.
Norby RJ; Weerasuriya Y; Hanson PJ. (1989). Induction of nitrate reductase activity in red spruce needles by NO2 and HNO3 vapor. Can J For
Res, 19,889-896.
Norby RJ. (1998). Nitrogen deposition: a component of global change analyses. New Phytol, 139,189-200.
Norby RJ; Cotrufo MF. (1998). Global change - a question of litter quality. Nature, 396,17-18.
Norby RJ; Cotrufo MF; Ineson P; O'Neill EG; Canadell JG. (2001). Elevated CO2, litter chemistry, and decomposition: a synthesis. Oecologia,
127,153-165.
Nordin A; Nasholm T; Ericson L. (1998). Effects of simulated N deposition on understorey vegetation of a boreal coniferous forest. Functional
Ecology, 12, 691-699.
Nordin A; Strengbom J; Witzell J; Nasholm T; Ericson L. (2005). Nitrogen depositon and the biodiversity of boreal forests: implications for the
nitrogen critical load. Ambio, 34,20-24.
Norton SA; Akielaszek JJ; Haines TA; Stromborg KL; Longcore JR. (1982). Bedrock geologic control of sensitivity of aquatic ecosystems in the
United States to acidic deposition, Report no NP-4901887; pp 1-13). Fort Collins, CO; National Atmospheric Deposition Program
(NADP).
Norton SA; Wright RF; Kahl JS; Scofield JP. (1992). The MAGIC simulation of surface water acidification at, and first year results from, the
Bear Brook Watershed Manipulation, Maine, USA. Environ. Pollut, 77,279-286.
Norton SA; Kahl JS; Fernandez I J; Rustad LE; Scofield JP; Haines TA. (1994). Response of the West Bear Brook watershed, Maine, USA, to
the addition of (NH4)2S04:3 year results. For Ecol Manage, 68,61 -73.
Norton S; Kahl J; Fernandez I. (1999a). Altered soil-soil water interactions inferred from stream water chemistry at an artificially acidified
watershed at Bear Brook Watershed, Maine USA. Environ Monit Assess, 55,97-111.
Norton S; Kahl J; Fernandez I; Haines T; Rustad L; Nodvin S; Schofield J; Strickland T; Erickson J; Wignington P Jr; Lee J. (1999b). The Bear
Brook Watershed, Maine (BBWM), USA. Environ Monitor Assess, 55,7-51.
Nosko P; Brassard P; Kramer JR; Kershaw KA. (1988). The effect of aluminum on seed germination and early seedling establishment, growth,
and respiration of white spruce (P/'cea glauca). Can J Bot, 66,2305-2310.
Notholt J; Hjorth J; Raes F. (1992a). Formation of HNO2 on aerosol surfaces during foggy periods in the presence of NO and NO2. Atmos
Environ, 26,211-217.
Notholt J; Hjorth J; Raes F; Schrems O. (1992b). Simultaneous long path field measurements of HNO2, CH20 and aerosol. Ber Bunsen Ges
Phys Chem, 96,290-293.
Nouchi I; Mayumi H; Yamazoe F. (1984). Foliar injury response of petunia and kidney bean to simultaneous and alternate exposures to ozone
and PAN. Atmos Environ, 18,453-460.
Novak M; Bottrell SH; Prechova E. (2001). Sulfur isotope inventories of atmospheric deposition, spruce forest floor and living Sphagnum along
a NW-SE transect across Europe. Biogeochemistry, 53,23-50.
Novak M; Kirchner JW; Fottova D; Prechova E; Jackova I; Kram P; Hruska J. (2005). Isotopic evidence for processes of sulfur retention/release
in 13 forested catchments spanning a strong pollution gradient (Czech Republic, central Europe). Glob Biogeochem Cycles, 19, doi:
10.1029/2004GB002396.
Novak M; Mitchell MJ; Jackoval; Buzek F; Schweigstillova J; Erbanova L; Prikryl R; Fottova D. (2007). Processes Affecting Oxygen Isotope
Ratios of Atmospheric and Ecosystem Sulfate in Two Contrasting Forest Catchments in Central Europe. Environ Sci Technol, 41,703-
709.
Nowak RS; Ellsworth DS; Smith SD. (2004). Functional responses of plants to elevated atmospheric C02-do photosynthetic and productivity
data from FACE experiments support early predictions? New Phytol, 162,253-280.
NPS. (2000). Federal land manager's air quality related values workgroup (FLAG) phase I report. Lakewood, CO; Air Resources Division;
National Park Service (NPS), U.S. Department of the Interior.
NPS. (2004). Federal land manager's critical loads for sulfur and nitrogen workshop. Denver, CO; Air Quality Branch; U.S. Fish and Wildlife
Service; U.S. Forest Service, U.S. Department of Agriculture (USDA). Lakewood, CO; Air Resource Management Program. Air
Resources Division; National Park Service (NPS); U.S. Department of the Interior.
NRC. (1992). Rethinking the ozone problem in urban and regional air pollution. Washington, DC: National Research Council (NRC); The
National Academies Press.
NRC. (1998). Research priorities for airborne particulate matter. I. Immediate priorities and a long-range research portfolio. Washington, DC:
National Research Council (NRC); The National Academies Press.
NRC. (2000). Clean coastal waters: Understanding and reducing the effects of nutrient pollution. Washington, DC: National Research Council
(NRC); The National Academies Press.
NRC. (2002). Estimating the public health benefits of proposed air pollution regulations. Washington, DC: National Research Council (NRC);
The National Academies Press.
NRC. (2004). Air quality management in the United States. Washington, DC: National Research Council (NRC); The National Academies
Press.
R-65
-------
Nriagu JO. (1978). Production and uses of sulfur. In: Nriagu JO (Ed.), Sulfur in the environment, Part 1: The atmospheric cycle (pp. 1-21). New
York: Wiley.
Nriagu JO; Holdway DA; Corker RD. (1987). Biogenic sulfur and the acidity of rainfall in remote areas of Canada. Science, 237,1189-1192.
NSTC. (2000). Integrated assessment of hypoxia in the northern Gulf of Mexico. Washington, DC; Committee on Environment and Natural
Resources; U.S. National Science and Technology Council (NSTC).
Nunnermacker LJ; Imre D; Daum PH; Kleinman L; Lee Y-N; Lee JH; Springston SR; Newman L; Weinstein-Lloyd J; Luke WT; Banta R; Alvarez
R; Senff C; Sillman S; Holdren M; Keigley GW; Zhou X. (1998). Characterization of the Nashville urban plume on July 3 and July 18,
1995. J Geophys Res, 103, 28129-28148.
Nussbaum S; Vonballmoos P; Gfeller H; Schluneggerz UP; Fuhrer J; Rhodes D; Brunold C. (1993). Incorporation of atmospheric (N02)-N-15-
nitrogen into free amino-acids by Norway spruce picea-abies (I) karst. Oecologia, 94,408-414.
Nydick KR; Lafrancois BM; Baron JS; Johnson BM. (2003). Lake-specific responses to elevated atmospheric nitrogen deposition in the
Colorado Rocky Mountains, USA. Hydrobiologia, 510,103-114.
Nydick KR; Lafrancois BM; Baron JS; Johnson BM. (2004a). Nitrogen regulation of algal biomass, productivity, and composition in shallow
mountain lakes, Snowy Range, Wyoming, USA. Can J Fish Aquat Sci, 61,1256-1268.
Nydick KR; Lafrancois BM; Baron JS. (2004b). NO3 uptake in shallow, oligotrophic, mountain lakes: the influence of elevated NO3
concentrations. J North Am Benthol Soc, 23,397-415.
Nykanen H; Vasander H; Huttunen JT; Martikainen PJ. (2002). Effect of experimental nitrogen load on methane and nitrous oxide fluxes on
ombrotrophic boreal peatland. Plant Soil, 242,147-155.
Odman MT; Ingram CI (1996). Multiscale air quality simulation platform (MAQSIP): Source code documentation and validation. (MCNC
Technical report ENV-96TR002-v1.0). Research Triangle Park, NC; MCNC Environmental Modeling Center @ the North Carolina
Supercomputing Center.
Odman MT; Russell AG. (1999). Mass conservative coupling of non-hydrostatic meteorological models with air quality models. In: Grying S-E;
Batchvarova E (Eds.), Air pollution modeling and its application XIII. New York, NY: Plenum Press.
Odman MT; Boylan JW; Wilkinson JG; Russell AG; Mueller SF; Imhoff RE; Doty KG; Norris WB; McNider RT. (2002). SAMI air quality
modeling. (Final report). Asheville, NC; Southern Appalachian Mountains Initiative (SAMI).
O'Dowd C; McFiggans G; Creasey DJ; Pirjola L; Hoell C; Smith MH; Allan BJ; Plane JMC; Heard DE; Lee JD; Pilling MJ; Kulmala M. (1999).
On the photochemical production of new particles in the coastal boundary layer. Geophys Res Lett, 26,1707-1710.
O'Dowd CD; Jimenez JL; Bahreini R; Flagan RC; Seinfeld JH; Hamerl K; Pirjola L; Kulmala M; Jennings SG; Hoffmann T. (2002). Marine
aerosol formation from biogenic iodine emissions. Nature, 417,632-636.
Odum EP. (1989). Ecology and our endangered life-support systems, (pp 58-66,108-118). Sunderland, MA: Sinauer Associates, Inc.
Oesch S; Faller M. (1997). Environmental effects on materials: The effect of the air pollutants SO2, NO2, NO and 0-3 on the corrosion of
copper, zinc and aluminium. A short literature survey and results of laboratory exposures. Corros Sci, 39,1505-1530.
Officer CB; Biggs RB; Taft JL; Cronin LE; Tyler MA; Boynton WR. (1984). Chesapeake Bay anoxa. Origin, development and significance.
Science, 223, 22-27.
Officer CB; Ryther JH. (1980). The possible importance of silicon in marine eutrophication. Marine Ecology Progress Series, 3,83-91.
OhnoT; Sucoff El; Erich MS; Bloom PR; Buschena CA; Dixon RK. (1988). Growth and nutrient content of red spruce seedlings in soil amended
with aluminum. J Environ Qual, 17, 666-672.
0kland J; 0kland KA. (1986). The effects of acid deposition on benthic animals in lakes and streams. Experientia, 42,471-486.
Oliver BG; Thurman EM; Malcolm RL. (1983). The contribution of humic substances to the acidity of colored natural waters. Geochim
Cosmochim Acta, 47, 2031-2035.
Ollinger SV; Lathrop RG; Ellis JM; Aber JD; Lovett GM; Millham SE. (1993). A spatial model of atmospheric deposition for the northeastern
U.S. Ecol Appl, 3,459-472.
Ollinger SV; Aber JD; Reich PB. (1997). Simulating ozone effects on forest productivity: Interactions among leaf-, canopy-, and stand-level
processes. Ecol Appl, 7,1237-1251.
Ollinger SV; Smith ML; Martin ME; Hallett RA; Goodale CL; Aber JD. (2002). Regional variation in foliar chemistry and N cycling among forests
of diverse history and composition. Ecology, 83,339-355.
Olsen Y; Agusti S; Andersen T; Duarte CM; Gasol JM; Gismervik I; Heiskanen AS; Hoell E; Kuuppo P; Lignell R; Reinertsen H; Sommer U;
Stibor H; Tamminen T; Vadstein O; Vaque D; Vidal M. (2006). A comparative study of responses in planktonic food web structure and
function in contrasting European coastal waters exposed to experimental nutrient addition. Limnol Oceanogr, 51,488-503.
Omernik JM; Powers CF. (1983). Map Supplement: Total Alkalinity of Surface Waters-A National Map. Annals of the Association of American
Geographers, 73,133-136.
Op Den Camp HJM; Kartal B; Guven D; van Niftrik LAMP; Haaijer SCM; van der Star WRL; Pas-Schoonen KTvd; Cabezas A; Ying Z; Schmid
MC; Kuypers MMM; van de Vossenberg J; Harhangi HR; Picioreanu C; van Loosdrecht MCM; Kuenen JG; Strous M; Jetten MS.
(2006). Global impact and application of the anaerobic ammonium-oxidizing (anammox) .bacteria. Biochemical Society Transactions,
34,174-178.
Ordonez C; Richter A; Steinbacher M; Zellweger C; Null H; Burrows JP; Prevot ASH. (2006). Comparison of 7 years of satellite-borne and
ground-based tropospheric NO2 measurements around Milan, Italy. J Geophys Res, 111, doi: 10.1029/2005JD006305.
Oren R; Ellsworth DS; Johnsen KH; Phillips N; Ewers BE; Maier C; Schafer KVR; McCarthy H; Hendrey G; McNulty SG; Katul GG. (2001). Soil
fertility limits carbon sequestration by forest ecosystems in a C02-enriched atmosphere. Nature, 411,469-472.
Oreskes N; Shrader-Frechette K; Belitz K. (1994). Verification, validation, and confirmation of numerical models in the earth sciences. Science,
263, 641-646.
Ormerod SJ; Tyler SJ; Lewis JMS. (1985). Is the breeding distribution of dippers influenced by stream acidity. Bird Study, 32,32-39.
R-66
-------
Ormerod SJ; Allinson N; Hudson D; Tyler SJ. (1986). The distribution of breeding dippers (Cindus cinclus) [L.]; Aves) in relation to stream
acidity in upland Wales. Freshw Biol, 16,501-507.
Ormerod SJ; Tyler SJ. (1987). Dippers (Cinclus cinclus) and grey wagtails (Motacilla cinerea) as indicators of stream acidity in upland Wales.
In: Diamond AW; Filion FL (Eds.), The value of birds. (ICBP technical publication no. 6 pp 191-208). Cambridge, UK: International
Council for Bird Preservation (ICBP).
Ormerod SJ; Boole P; McCahon CP; Weatherley NS; Pascoe D; Edwards RW. (1987). Short-term experimental acidification of a Welsh UK
stream comparing the biological effects of hydrogen ions and aluminum. Freshwater Biol, 17,341-356.
Ormerod SJ; Tyler SJ. (1991). Exploitation of prey by a river bird, the dipper Cinclus cinclus (L.), along acidic and circumneutral streams in
upland Wales. Freshwater Biol, 25,105-116.
Orth RJ; Moore KA. (1984). Distribution and abundance of submerged aquatic vegetation in Chesapeake Bay: an historical prespective.
Estuaries, 7,531-540.
Osthoff HD; Brown SS; Ryerson TB; Fortin TJ; Lerner BM; Williams EJ; Pettersson A; Baynard T; Dube WP; Ciciora SJ; Ravishankara AR.
(2006). Measurement of atmospheric NO2 by pulsed cavity ring-down spectroscopy. J Geophys Res, 111, doi:
10.1029/2005JD006942.
Ott LE; Pickering KE; Stenchikov GL; Huntrieser H; Schumann U. (2007). Effects of lightning NOx production during the July 21 European
Lightning Nitrogen Oxides Project storm studied with a three-dimensional cloud-scale chemical transport model. J Geophys Res, 112,
doi: 10.1029/2006JD007365.
Ouimet R; Arp PA; Watmough SA; Aherne J; Demerchant I. (2006). Determination and mapping critical loads of acidity and exceedances for
upland forest soils in Eastern Canada. Water Air Soil Pollut, 172,57-66.
Padgett P; Allen EB. (1999). Differential responses to nitrogen fertilization in native shrubs and exotic annuals common to Mediterranean
coastal sage scrub of California. Plant Ecology, 144,93-101.
Padgett PE; Allen EB; Bytnerowicz A; Minnich RA. (1999). Changes in soil inorganic nitrogen as related to atmospheric nitrogenous pollutants
in southern California. Atmos Environ, 33,769-781.
Padro J. (1996). Summary of ozone dry deposition velocity measurements and model estimates over vineyard, cotton, grass and deciduous
forest in summer. Atmos Environ, 30,2363-2369.
Paerl HW. (1985). Enhancement of marine primary production by nitrogen-enriched acid rain. Nature, 315,747-749.
Paerl H. (1995). Coastal eutrophication in relation to atmospheric nitrogen deposition: current perspectives. Ophelia, 41,237-259.
Paerl H. (1997). Coastal eutrophication and harmful algal blooms: importance of atmospheric deposition and groundwater as "new" nitrogen
and other nutrient sources. Limnol Oceanogr, 42,1154-1162.
Paerl H; Pinckney J; Fear J; Peierls B. (1998). Ecosystem responses to internal and watershed organic matter loading: Consequences for
hypoxia in the eutrophying Neusse River Estuary, NC, USA. Mar Ecol Prog Ser, 166,17-25.
Paerl HW; Pinckney JL; Steppe TF. (2000). Cyanobacterial-bacterial mat consortia: examining the functional unit of microbial survival and
growth in extreme environments. Environmental Microbiology, 2,11-26.
Paerl HW; Boynton WR; Dennis RL; Driscoll CT; Greening HS; Kremer JN; Rabalais NN; Seitzinger SP. (2001a). Atmospheric deposition of
nitrogen in coastal waters: Biogeochemical and ecological implications. In: Valigura RW; Alexander RB; Castro MS; Meyers TP; Paerl
HW; Stacey PE; Turner RE (Eds.), Nitrogen loading in coastal water bodies: an atmospheric perspective. (Coastal and estuarine
series; Volume 57, pp 11-52). Washington, DC: American Geophysical Union.
Paerl HW; Bales JD; Ausley LW; Buzzelli CP; Crowder LB; Eby LA; Go M; Peierls BL; Richardson TL; Ramus JS. (2001b). Ecosystem impacts
of three sequential hurricanes (Dennis, Floyd, and Irene) on the United States' largest lagoonal estuary, Pamlico Sound, NC. Proc Natl
Acad Sci USA, 98, 5655-5611.
Paerl HW. (2002). Connecting atmospheric deposition to coastal eutrophication. Environ Sci Technol, 36,323A-326A.
Paerl HW; Dennis RL; Whitall DR. (2002). Atmospheric deposition of nitrogen: Implications for nutrient over-enrichment of coastal waters.
Estuaries, 25,677-693
Paerl HW; Valdes LM; Pinckney JL; Piehler MF; Dyble J; Moisander PH. (2003). Phytoplankton photopigments as indicators of estuarine and
coastal eutrophication. Bioscience, 53,953-964.
Paerl HW; Valdes LM; Peierls BL; Adolf JE; Harding LW. (2006). Anthropogenic and climatic influences on the eutrophication of large estuarine
ecosystems. Limnol Oceanogr, 51,448-462.
Palmer PI; Abbot DS; Fu T-M; Jacob DJ; Chance K; Kurosu TP; Guenther A; Wiedinmyer C; Stanton JC; Pilling MJ; Pressley SN; Lamb B;
Sumner AL. (2006). Quantifying the seasonal and interannual variability of North American isoprene emissions using satellite
observations of the formaldehyde column. J Geophys Res, 111,10.1029/2005JD006689.
Palmer RE; Klauda RJ; Lewis TE. (1988). Comparative sensitivities of bluegill, channel catfish and fathead minnow to pH and aluminum.
Environ Toxicol Chem, 7,505-516.
Palmer SM; Driscoll CT; Johnson CE. (2004). Long-term trends in soil solution and stream water chemistry at the Hubbard Brook Experimental
Forest: relationship with landscape position. Biogeochemistry, 68, 51-70.
Palmqvist K. (2000). Carbon economy in lichens. New Phytol, 148,11-36.
Palmqvist K; Campbell D; Ekblad A; Johansson H. (1998). Photosynthetic capacity in relation to nitrogen content and its partitioning in lichens
with different photobionts. Plant, Cell and Environment, 21,361-372.
Palmqvist K; Dahlman L. (2006). Responses of the green algal foliose lichen Platismatia glauca to increased nitrogen supply. New Phytol, 171,
343-356.
Pandis SN. (2004). Atmospheric aerosol processes. In: McMurry PH; Shepard M; Vickery JS (Ed.), Particulate Matter Science for Policy
Makers: A NARSTO Assessment (Chapter 3). Cambridge, UK: Cambridge University Press.
R-67
-------
Papen H; Daum M; Steinkamp R; Butterbach-Bahl K. (2001). N2O- and Cm-fluxes from soils of a N-limited and N-fertilized spruce forest
ecosystem of the temperate zone. J Appl Bot, 75,159-163.
Paquet J; Belanger L. (1997). Public acceptability thresholds of clearcutting to maintain visual quality of boreal balsam fir landscapes. For Sci,
43,46-55.
Pardo LH; Driscoll CT. (1993). A critical review of mass balance methods for calculating critical loads of nitrogen for forest ecosystems.
Environmental Review, 1,145-156.
Pardo LH; Driscoll CT; Likens GE. (1995). Patterns of nitrate loss from a chronosequence of clear-cut watersheds. Water Air Soil Pollut, 85,
1659-1664.
Pardo LH; Driscoll CT. (1996). Critical loads for nitrogen deposition: case studies at two northern hardwood forests. Water Air Soil Pollut, 89,
105-128.
Pardo LH; Kendall C; Pett-Ridge JP; Change CCY. (2004). Evaluating the source of streamwater nitrate using 615N and 6180 in nitrate in two
watersheds in New Hampshire, USA. Hydrological Processes, 18,2699-2712.
Pardo LH; Duarte N. (2007). Assessment of effects of acidic deposition on forested ecosystems in Great Smoky Mountains National Park using
critical loads for sulfur and nitrogen. htto://www2.nature.nps.aov/air/Pubs/pdf/GSMN CL Report 080830.pdf. Lakewood, CO; Air
Resources Division; National Park Service; U.S. Department of the Interior.
Park J; Mitchell MJ; McHale PJ; Christopher SF; Myers TP. (2003). Interactive effects of changing climate and atmospheric deposition on N
and S biogeochemistry in forested watershed of the Adirondack Mountains, New York State. Glob Chang Biol, 9,1602-1619.
Park RJ; Stenchikov GL; Pickering; Dickerson RR; Allen DJ; Kondragunta S. (2001). Regional air pollution and its radiative forcing: studies with
a single column chemical and radiation transport model. J Geophys Res, 106,28751-28770.
Park TA; Bowker JM; Leeworthy VR. (2002). Valuing Snorkeling Visits to the Florida Keys With Stated and Revealed Preference Models. J
Environ Manage, 65,301-312.
Parker DR; Zelazny LW; Kinraide TB. (1989). Chemical speciation and plant toxicity of aqueous aluminum. In: Lewis TE (Ed.), Environmental
chemistry and toxicology of aluminum (pp. 117-145). American Chemical Society.
Parker JL; Fernandez IJ; Rustad LE; Norton SA. (2001). Effects of nitrogen enrichment, wildfire, and harvesting on forest-soil carbon and
nitrogen. Soil Sci Soc Am J, 65,1248-1255.
Parker KE. (1988). Common loon reproduction and chick feeding on acidified lakes in the Adirondack Park, New York. Canadian Journal of
Zoology, 66, 804-810.
Parkhurst DL; Appelo CAJ. (1999). User's Guide to PHREEQC (Version 2): A computer program for speciation, batch-reaction, one-
dimensional transport, and inverse geochemical calculations. (Water-Resources Investigations Report). Denver, CO: U.S. Geological
Survey.
Parrish DD; Trainer M; Buhr MP; Watkins BA; Fehsenfeld FC. (1991). Carbon monoxide concentrations and their relation to concentrations of
total reactive oxidized nitrogen at two rural U.S. sites. J Geophys Res, 96,9309-9320.
Parrish DD; Trainer M; Young V; Goldan PD; Kuster WC; Jobson BT; Fehsenfeld FC; Lonneman WA; Zika RD; Farmer CT; Riemer DD;
Rodgers MO. (1998). Internal consistency tests for evaluation of measurements of anthropogenic hydrocarbons in the troposphere. J
Geophys Res, 103, 22339-22359.
Parrish DD; Fehsenfeld FC. (2000). Methods for gas-phase measurements of ozone, ozone precursors and aerosol precursors. Atmos Environ,
34,1921-1957.
Parrish DD; Trainer M; Hereid D; Williams EJ; Olszyna KJ; Harley RA; Meagher JF; Fehsenfeld FC. (2002). Decadal change in carbon
monoxide to nitrogen oxide ratio in U.S. vehicular emissions. J Geophys Res, 107, doi: 10.1029/2001 JD000720.
Parton WJ; Schimel DS; Ojima DS; Cole CV. (1994). A general model for soil organic matter dynamics: sensitivity to litter chemistry, texture
and management. In: Bryant RB; Arnold RW (Eds.), Quantitative Modeling of Soil Forming Processes (pp. 147-167). Madison, Wl: Soil
Sci Society of America.
Parton WJ; Hartman M; Ojima D; Schimel D. (1998). DAYCENT and its land surface submodel: Description and testing. Global Planet Change,
19,35-48.
Parton W; Silver WL; Burke IC; Grassens L; Harmon ME; Currie WS; King JY; Adair EC; Brandt LA; Hart SC; Fasth B. (2007). Global-scale
similarities in nitrogen release patterns during long-term decomposition. Science, 315,361-364.
Paschke MW; McLendon T; Redente EF. (2000). Nitrogen availability and old-field succession in a shortgrass steppe. Ecosystems, 3,144-158.
Paul AJ; Leavitt PR; Schindler DW; Hardie AK. (1995). Direct and indirect effects of predation by a calanoid copepod (subgenus:
Hesperodiaptomus) and of nutrients in a fishless alpine lake. Can J Fish Aquat Sci, 52,2628-2638.
Pauli D; Peintinger M; Schmid B. (2002). Nutrient enrichment in calcareous fens: effects on plant species and community structure. Basic and
Applied Ecology, 3,255-266.
Paulsen SG; Larsen DP; Kaufmann PR; Whittier TR; Baker JR; Peck D; McGue J; Hughes RM; McMullen D; Stevens D; Stoddard JL;
Lazorchak J; Kinney W; Selle AR; Hjort R. (1991). EMAP: Surface waters monitoring and research strategy, fiscal year. Corvallis, OR;
Environmental Research Laboratory; Office of Research and Development; U.S. Environmental Protection Agency.
Paz-Gonzalez A; Taboada MT. (2000). Nutrient variability from point sampling on 2 meter grid in cultivated and adjacent forest land. Comm.
Soil Sci Plant Anal, 31,2135-2146.
Pearcy RW; Bjorkman O; Caldwell MM; Keeley JE; Monson RK; Strain BR. (1987). Carbon gain by plants in natural environments. Bioscience
37:21-29.
Pearson J; Stewart RG. (1993). Atmospheric ammonia deposition and its effects on plants. Tansley Review No 56, New Phytol, 283-305.
Peierls BL; Caraco NF; Pace ML; Cole JJ. (1991). Human influence on river nitrogen. Nature 350,386-387.
Peierls BL; Paerl HW. (1997). The bioavailability of atmospheric organic nitrogen deposition to coastal phytoplankton. Limnol Oceanogr, 42,
1819-1823.
R-68
-------
Pembrook H. (2004). 2004 update: Calculating critical loads of acidity and exceedances for acid-impaired lakes in Vermont using the steady-
state water chemistry (SSWC).model: Water Quality Division of the DEC; Vermont Agency of Natural Resources.
Perdue EM; Reuter JH; Parrish RS. (1984). A statistical model of proton binding by humus. Geochim Cosmochim Acta, 48,1257-1263.
Perez I; Wooldridge PJ; Cohen RC. (2007). Laboratory evaluation of a novel thermal dissociation chemiluminescence method for in situ
detection of nitrous acid. Atmos Environ, 41,3993-4001.
Perrini G; Tomasello M; Librando V; Minniti Z. (2005). Nitrated polycydic aromatic hydrocarbons in the environment: formation, occurrences
and analysis. Annates de Chimie: Science des Materiaux, 95,567-577.
Peterjohn WT; Adams MB; Gilliam FS. (1996). Symptoms of nitrogen saturation in two central Appalachian hardwood forest ecosystems.
Biogeochemistry, 35, 507-522.
Peterjohn WT; Correll DL. (1984). Nutrient dynamics in an agricultural watershed: Observation on the role of a riparian forest. Ecology, 65,
1466-1475.
Peters K; Eiden R. (1992). Modelling the dry deposition velocity of aerosol particles to a spruce forest. Atmospheric Environment 26A: 2555-
2564.
Peterson BJ; Wollheim WM; Mulholland PJ; Webster JW; Meyer JL; Tank JL; Marti E; Bowden WB; Valett HM; Hershey AE; McDowell WH;
Dodds WK; Hamilton SK; Gregory SV; Morrall DD. (2001). Control of nitrogen export from watersheds by headwater streams. Science,
292, 86-89.
Peterson DG; Rowe RD; Schulze WD; Russell GW; Boyce RR; Elliott SR; Hurd B. (1987a). Valuation of Visual Forest Damages from Ozone.
In: Improving Accuracy and Reducing Costs of Environmental Benefit Assessments. (Energy & Resource Consultants report to U.S.
Environmental Protection Agency). Boulder, CO; Center for Economic Analysis, University of Colorado.
Peterson DL; Arbaugh MJ; Wakefield VA; Milter PR. (1987b). Evidence for growth reduction in ozone injured Jeffrey pine (Pinus jeffreyi Grev.
and Balf.). in Sequoia and Kings Canyon National Parks. J Air Pollut Control Assoc, 37,908-912.
Peterson DL; Arbaugh MJ; Robinson LJ. (1991). Regional growth changes in ozone-stressed ponderosa pine (Pinus ponderosa).in the Sierra
Nevada, California, USA. Holocene, 1,50-61.
Peterson DL; Sullivan TJ; Eilers JM; Brace S; Horner D; Savig K; Morse D. (1998a). Assessment of air quality and air pollutant impacts in
national parks of the Rocky Mountains and northern Great Plains. (Report NPS/CCSOUW/NRTR-98/19).
http://www2.nature.nps.gov/air/Pubs/reqionPark.cfm. Denver, CO; Air Resources Division; National Park Service, U.S. Department of
the Interior.
Peterson EB; Neitlich P. (2001). Impacts of two coal-fired power plants on lichen communities of northwestern Colorado. Washington, DC:
Forest Health Monitoring Program; U.S. Forest Service; U.S. Department of Agriculture (USDA).
Peterson G; Allen CR; Holling CS. (1998b). Ecological resilience, biodiversity, and scale. Ecosystems,1,6-18.
Peterson RH; Daye PG; Metcalfe JL. (1980). Inhibition of Atlantic salmon (Salmo salar) hatching at low pH. Can J Fish Aquat Sci, 37,770-774.
Peterson RH; Martin-Robichaud DJ. (1986). Growth and major inorganic cation budgets of Atlantic salmon alevins at three ambient acidities.
Trans Am Fish Soc, 115, 220-226.
Peterson RH; Van Eechhaute L. (1992). Distributions of Ephemeroptera, Plecoptera, and Trichoptera of three maritime catchments differing in
pH. Freshwater Biol, 27,65-78.
Peterson SA; Van Sickle J; Herlihy AT; Hughes GH. (2007). Mercury concentration in fish from streams and rivers throughout the western
United States. Environ Sci Technol, 41,58-65.
Petriconi GL; Papee HM. (1972). On the photolytic separation of halogens from sea water concentrates. Water Air Soil Pollut, 1,117-131.
Pfister GG; Emmons LK; Hess PG; Honrath R; Lamarque J-F; Val Martin M; Owen RC; Avery MA; Browell EV; Holloway JS; Nedelec P; Purvis
R; Ryerson TB; Sachse GW; Schlager H. (2006). Ozone production from the 2004 North American boreal fires. J Geophys Res, 111,
doi: 10.1029/2006JD007695.
Phillips K. (1990). Where have all the frogs and toads gone? Bioscience, 40,422-424.
Phillips RL; Whalen SC; Schlesinger WH. (2001). Influence of atmospheric CO2 enrichment on nitrous oxide flux in a temperate forest
ecosystem. Glob BiogeochemCydes, 15, 741-752.
Phillips RL; Zak DR; Holmes WE; White DC. (2002). Microbial community composition and function beneath temperate trees exposed to
elevated atmospheric carbon dioxide and ozone. Oecologia, 131, 236-244.
Phillips RP; Yanai RD. (2004). The effects of AICI3 additions on rhizosphere soil and fine root chemistry of sugar maple (Acer saccharum).
Water Air Soil Pollut, 159, 339-356.
Piao S; Ciais P; Friedlingstein P; Peylin P; Reichstein M; Luyssaert S; Margolis H; Fang J; Barr A; Chen,A; Grelle A; Hollinger DY; Laurila T;
Lindroth A; Richardson AD; Vesala T. (2008). Net carbon dioxide losses of northern ecosystems in response to autumn warming.
Nature, 451,49-52.
Pickering KE; Thompson AM; Dickerson RR; Luke WT; McNamara DP; Greenberg JP; Zimmerman PR. (1990). Model calculations of
tropospheric ozone production potential following observed convective events. J Geophys Res, 95,14049-14062.
Pickering KE; Thompson AM; Scala JR; Tao W-K; Simpson J; Garstang M. (1991). Photochemical ozone production in tropical squall line
convection during NASA Global Tropospheric Experiment/Amazon Boundary Layer Experiment 2A. J Geophys Res, 96,3099-3114.
Pickering KE; Thompson AM; Scala JR; Tao W-K; Dickerson RR; Simpson J. (1992a). Free tropospheric ozone production following
entrainment of urban plumes into deep convection. J Geophys Res, 97,17985-18000.
Pickering KE; Thompson AM; Scala JR; Tao W-K; Simpson J. (1992b). Ozone production potential following convective redistribution of
biomass burning emissions. J Atmos Chem , 14,297-313.
Pickering KE; Thompson AM; Tao W-K; Kucsera TL. (1993). Upper tropospheric ozone production following mesoscale convection during
STEP/EMEX. J Geophys Res, 98, 8737-8749.
R-69
-------
Pickering KE; Thompson AM; Tao W-K; Rood RB; McNamara DP; Molod AM. (1995). Vertical transport by convective clouds: comparisons of
three modeling approaches. Geophys Res Lett, 22,1089-1092.
Pickering KE; Thompson AM; Wang Y; Tao W-K; McNamara DP; Kirchhoff WJH; Heikes BG; Sachse GW; Bradshaw JD; Gregory GL; Blake
DR. (1996). Convective transport of biomass burning emissions over Brazil during TRACE A. J Geophys Res, 101,23993-24012.
Pickering KE; Wang Y; Tao W-K; Price C; Miiller J-F. (1998). Vertical distributions of lightning NOx for use in regional and global chemical
transport models. J Geophys Res, 103, 31203-31216.
Pickering KE; Thompson AM; Kim H; DeCaria AJ; Pfister L; Kucsera TL; Witte JC; Avery MA; Blake DR; Crawford JH; Heikers BG; Sachse
GW; Sandholm ST; Talbot RW. (2001). Trace gas transport and scavenging in PEM-Tropics B South Pacific Convergence Zone
convection. J Geophys Res, 106,32591-32602.
Pike LH. (1978). The importance of epiphytic lichens in mineral cycling. Bryologist, 81,247-257.
Pikelnaya, O; Hurlock, S. C; Trick, S; Stutz, J. (2005). Measurements of reactive iodine species on the Isles of Shoals, Gulf of Maine. EOS
Trans. Am. Geophys. Union 86, A14A-06. http://www.aau.ora/meetinas/fm05/fm05-sessions/fm05 A14A.html.
Pikelnaya O; Hurlock SC; Trick S; Stutz J. (2007). Intercomparison of multiaxis and long-path differential optical absorption spectroscopy
measurements in the marine boundary layer. J Geophys Res, 112, doi: 10.1029/2006JD007727.
Pilcher H. (2005). Pipe Dreams. Nature, 437,1227-1228.
Pilegaard K; Skiba U; Ambus P; Beier C; Bruggemann N; Butterbach-Bahl K; Dick J; Dorsey J; Duyzer J; Gallagher M; Gasche R; Horvath L;
Kitzler B; Leip A; Pihlatie MK; Rosenkranz P; Seufert G; Vesala T; Westrate H; Zechmeister-Boltenstern S. (2006). Factors controlling
regional differences in forest soil emission of nitrogen oxides (NO and N2O). Biogeosciences, 3,651-661.
Pimentel AS; Arbilla G. (1997). Simulation of the urban atmosphere chemistry polluted by alcohol fueled automobiles. Quimica Nova, 20,252-
260.
Pimentel D; Wilson C; McCullum C; Huang R; Dwen P; Flack J; Tran Q; Saltman T; Cliff B. (1997). Economic and environmental benefits of
biodiversity. Bioscience, 47,747-757.
Pimm SL. (1984). The complexity and stability of ecosystems. Nature, 307,321-326
Pinay G; Black VJ; Planty-Tabacchi AM; Gumiero B; Decamps H. (2000). Geomorphic control of denitrification in large river floodplain soils.
Biogeochemistry, 50,163-182.
Pinay G; Roques L; Fabre A. (1993). Spatial and temporal patterns of denitrification in a riparian forest. J Appl Ecol, 30,581-591.
Pinckney JL; Richardson TL; Millie DF; Paerl HW. (2001). Application of photopigment biomarkers for quantifying microalgal community
composition and in situ growth rates. Organic Geochemistry, 32, 585-595.
Pinder RW; Strader R; Davidson CI; Adams PJ. (2004). A temporally and spatially resolved ammonia emission inventory for dairy cows in the
United States. Atmos Environ, 38,3747-3756.
Pinder RW; Adams PJ; Pandis SN; Gilliland AB. (2006). Temporally resolved ammonia emission inventories: current estimates, evaluation
tools, and measurement needs. J Geophys Res, 111, D16310, doi:10.1029/2005JD006603.
Pinder RW; Adams PJ; Pandis SN. (2007). Ammonia emission controls as a cost-effective strategy for reducing atmospheric particulate matter
in the Eastern United States. Environ Sci Technol, 41,380-386.
Pinto JP; Turco RP; Toon OB. (1989). Self-limiting physical and chemical effects in volcanic eruption clouds. J Geophys Res, 94,11165-11174.
Pinto JP; Bruhl C; Thompson AM. (1993). The current and future envirionmental role of atmospheric methane: model studies and uncertainties,
pp. 514-551. In: Khalil MAK (Ed.), Atmospheric methane sources, sinks, and role in global change (Vol. 13, NATO ASI Series): NATO.
Pippin M; Bertman S; Thornberry T; Town M; Carroll MA; Sillman S. (2001). Seasonal variations of PAN, PPN, and O3 at the upper Midwest
PROPHET site [03]. J Geophys Res, 106,24457-24463.
Pitcairn CER; Fowler D; Leith ID; Sheppard LJ; Sutton MA; Kennedy V; Okello E. (2003). Bioindicators of enhanced nitrogen deposition.
Environ Pollut, 126,353-361.
Pitchford M. (2008). Carbon analysis workshop summary. Presented at the 2008 IMPROVE Steering Committee Meeting, held in Okefenokke
NWR, GA; October 28 - 29,2008
Pitts JN Jr; Biermann HW; Atkinson R; Winer AM. (1984). Atmospheric implications of simultaneous nighttime measurements of N03 radicals
and HONO. Geophys Res Lett, 11,557-560.
Pitts JN Jr. (1987). Nitration of gaseous polycyclic aromatic hydrocarbons in simulated and ambient urban atmospheres: a source of mutagenic
nitroarenes. Atmos Environ, 21,2531-2547.
Plafkin JL; Barbour MT; Porter KD; Gross SK; Hughes RM. (1989). Rapid bioassessment protocols for use in streams and rivers: benthic
macroinvertebrates and fish. (Report no. EPA 440-89-001). Washington, DC; Office of Water Regulations and Standards; U.S.
Environmental Protection Agency.
Poe GL; Boyle KJ; Bergstrom JC. (2000). A meta-analysis of contingent values for groundwater quality in the United States. Paper presented at
the American Agricultural Economics Association Conference held in Tampa, FL; July 30-August2,2000. (No 21871).
http://aaeconsearch.umn.edU/bitstream/21871/1/spQ0po03.pdf
Poff NL; Allan JD; Bain MB; Karr JR; Prestegaard KL; Richter BD; Sparks RE; Stromberg JC. (1997). The natural flow regime: a paradigm for
river conservation and restoration. Bioscience, 47,769-784.
Pokharel SS; Bishop GA; Stedman DH. (2002). An on-road motor vehicle emissions inventory for Denver: An efficient alternative to modeling.
Atmos Environ, 36,5177-5184.
Polasky S; Costello C; Solow A. (2005). Economics of Biodiversity. In: Maler K-G, Vincent JR (Eds.), Handbook of Environmental Economics
(Vol. 3, pp. 1517-1560). Amsterdam, The Netherlands: Elsvier Publishing.
Pollman CD; Sweets PR. (1990). Hindcasting of pre-cultural ANC and pH of seepage lakes in the Adirondacks, Upper Midwest, and Florida.
(Report no. EPA 89007B1/RES-1). Corvallis, OR; U.S. Environmental Protection Agency.
R-70
-------
Pollman CD; Canfield DE. (1991). Florida, hydrologic and biogeochemical controls on seepage lake chemistry. In: Charles DF; Christie S
(Eds.), Acidic deposition and aquatic ecosystems: Regional case studies, (pp 367-416). New York: Springer-Verlag,.
Ponnamperuma FN. (1972). The chemistry of submerged soils. Adv Agron, 24, 29-96.
Poor JP; Pessagno KL; Paul RW. (2006). Exploring the hedonic value of ambient water quality: A local watershed-based study. Ecol Econ, 60,
797-806.
Poppe D; Wallasch M; Zimmermann J. (1993). The dependence of the concentration of OH on its precursors under moderately polluted
conditions: a model study. J Atmos Chem, 16,61-78.
Porter E; BlettT; Potter DU; Huber C. (2003). Protecting resources on federal lands: Implications of critical loads for atmospheric deposition of
nitrogen and sulfur. Bioscience, 55,603-612.
Posch M; Reinds GJ; de Vries W. (1993). SMART -A simulation model for acidification's regional trends: Model description and user manual.
(Report no 477). Helsinki, Finland: Mimeograph Series of the National Board of Waters and the Environment.
Posch M; De Smet PAM; Hettelingh J-P; Downing RJ (Eds.) (1995a). Calculation and mapping of critical thresholds in Europe. (Status report
1995; Coordination Center for Effects; RIVM report no. 259101004). Bilthoven, The Netherlands: National Institute of Public Health
and the Environment (RIVM). http://www.mnp.nl/bibliotheek/rapporten/259101004.pdf.
Posch M; de Vries W; Hettelingh JP. (1995b). Critical loads of sulfur and nitrogen. Calculation and mapping of critical loads in Europe, Status
Report, 31-41.
Posch M; Kamari J; Forsius M; Henriksen A; Wilander A. (1997). Environmental auditing: Exceedance of critical loads for lakes in Finland,
Norway and Sweden: reduction requirements for acidifying nitrogen and sulfur deposition. Environ Manage, 21,291-304.
Posch M; DeSmet PAM; Hettelingh JP; Downing RJ. (2001). Calculation and mapping of critical thresholds in Europe. (Status report 2001;
RIVM report no. 259101010). http://www.mnp.nl/cce/lmaaes/SR01 tcm42-16441 .pdf. Bilthoven, The Netherlands: Coordination
Center for Effects, National Institute of Public Health and the Environment (RIVM).
Posch M; Reinds GJ. (2003). VSD - User Manual of the Very Simple Dynamic soil acidification model. Bilthoven, The Netherlands:
Coordination Center for Effects, RIVM.
Posch M; Hettelingh J-P; Slootweg J (Eds.). (2003). Manual for dynamic modelling of soil response to atmospheric deposition. (Working Group
on Effects of the Convention on Long-range Transboundary Air Pollution; report 259101012).
http://www.mnp.nl/cce/lmaqes/ModMan1 tcm42-16497.pdf. Bilthoven, The Netherlands: National Institute for Public Health and the
Environment (RIVM).
Possanzini M; De Santis F; Di Palo V. (1999). Measurements of nitric acid and ammonium salts in lower Bavaria. Atmos Environ, 33, 3597-
3602.
Poth MA; Fenn ME. (1998). Mature ponderosa pine nutrient use and allocation responses to air pollution. Riverside, CA; U.S. Forest Service;
U.S. Department of Agriculture (USDA).
Powell JFF; McKeown BA. (1986). The effects of acid exposure on the ion regulation and seawater adaptation of Coho salmon (Oncorhynchus
kisutch) parrs and smolts. Comp Biochem Physiol C: Comp Pharmacol, 83,45-52.
Power SA; Green ER; Barker CG; Bell JNB; Ashmore MR. (2006). Ecosystem recovery: heathland response to a reduction in nitrogen
deposition. Glob Chang Biol, 12,1241-1252.
Prather MJ; Jacob DJ. (1997). A persistent imbalance in HOx and NOx photochemistry in the upper troposphere driven by deep tropical
convection. Geophys Res Lett, 24,3189-3192.
Pregitzer KS; Burton AJ; Mroz GD; Liechty HO; Macdonald NW. (1992). Foliar sulfur and nitrogen along an 800-km pollution gradient. Can J
For Res, 22,1761-1769.
Pregitzer KS; Zak DR; Maziasz J; DeForest J; Curtis PS; Lussenhop J. (2000). Interactive effects of atmospheric CO2 and soil-N availability on
fine roots of Populus tremuloides. Ecol Appl, 10,18-33.
Pregitzer KS; Zak DR; Burton AJ; Ashby JA; MacDonald NW. (2004). Chronic nitrate additions dramatically increase the export of carbon and
nitrogen from northern hardwood ecosystems. Biogeochemistry, 68,179-197.
Pregitzer KS; Burton AJ; Zak DR; Talhelm AF. (2008). Simulated chronic nitrogen deposition increases carbon storage in Northern Temperate
forests. Glob Chang Biol, 14,142-153.
Preston SD; Brakebill JW. (1999). Application of spatially reference regression modeling for the evaluation of total nitrogen loading in the
Chesapeake Bay Watershed: U.S. Geological Survey.
Prevot ASH; Staehelin J; Kok GL; Schillawski RD; Neininger B; Staffelbach T; Neftel A; Wernli H; Dommen J. (1997). The Milan photooxidant
plume. J Geophys Res, 102:23375-23388.
Price C; Penner J; and Prather M. (1997). NOx from lightning. 1. Global distribution based on lightning physics, J Geophys Res, 102,5929—
5941.
Priscu JC; Axler RP; Goldman CR. (1985). Nitrogen metabolism of the shallow and deep water phytoplankton in asubalpine lake. Oikos, 45,
137-147.
Probert ME; Samosir SSR. (1983). Sulfur in non-flooded tropical soils. In: Blair GJ, Till AR (Eds.), Sulfur in south-east Asian and South Pacific
agriculture (pp. 15-27). Armidal, N.S.W. Australia: The University of New England; The Australian Development Association Bureau
and the Sulphur Institute.
Providoli I; Bugmann H; Siegwolf R; Buchmann N; Schleppi P. (2005). Flow of deposited inorganic N in two Gleysol-dominated mountain
catchments traced with 15N03~ and 15NH4+. Biogeochemistry, 76,453-475.
Pryor SC; Sorensen LL. (2000). Nitric Acid-Sea Salt Reactions: Implications for Nitrogen Deposition to Water Surfaces. Journal of Applied
Meteorology, 39,725-731.
Pryor SC; Barthelmie RJ; Sorensen LL; Jensen B. (2001). Ammonia concentrations and fluxes over a forest in the midwestern USA. Atmos
Environ, 35, 5645-5656.
R-71
-------
Pszenny MP; Keene WC; Jacob DJ; Fan S; Maben JR; Zetwo MP; Springer-Young M; Galloway JN. (1993). Evidence of inorganic chlorine
gases other than hydrogen chloride in marine surface air. Geophys Res Lett, 20,699-702.
Pszenny MP; Moldanova J; Keene WC; Sander R; Maben JR; Martinez M; Crutzen PJ; Perner D; Prinn RG. (2004). Halogen cycling and
aerosol pH in the Hawaiian marine boundary layer. Atmos Chem Phys, 4,147-168.
Pszenny MP; Fischer EV; Russo RS; Sive BC; Varner RK. (2007). Estimates of CI atom concentrations and hydrocarbon kinetic reactivity in
surface air at Appledore Island, Maine (USA), during International Consortium for Atmospheric Research on Transport and
Transformation/Chemistry of Halogens at the Isles of Shoals. J Geophys Res, 112, D10S13, doi:10.1029/2006JD007725.
Puhe J. (2003). Growth and development of the root system of Norway spruce (P/'cea abies) in forest stands—A review. For Ecol Manage, 175,
253-273.
Pun BK; Wu SY; Seigneur C; Seinfeld JH; Griffin RJ; Pandis SN. (2003). Uncertainties in Modeling Secondary Organic Aerosols: Three-
Dimensional Modeling Studies in Nashville/Western Tennessee. Environ Sci Technol, 37,3647-3661.
Pyzik L; Caddick J; Marx P. (2004). Chesapeake Bay: Introduction to an ecosystem. (Report no EPA-903-R-04-003 // CBP/TRS 232/00).
Annapolis, MD: Chesapeake Bay Program (CBP). & Philadelphia, PA; Region III; Regional Center for Environmental Information;
U.S. Environmental Protection Agency (EPA).
QiaoZ, Murray, F. (1998). Improvement of the $15 N Dilution Method for Estimation of Absorption of NOxby Plants Supplied with $15 N-
Labelled Fertilizer. NewPhytol, 138,13-18.
Quist ME; Nasholm T; Lindeberg J; Johannisson C; Hogbom L; Hogberg P. (1999). Responses of a nitrogen-saturated forest to a sharp
decrease in nitrogen input. J Environ Qual, 28,1970-1977.
Rabalais NN; Turner RE; Justic D; Dortch Q; Wiseman WJ; Gupta BKS. (1996). Nutrient changes in the Mississippi River and system
responses on the adjacent continental shelf. Estuaries, 19,386-407.
Rabalais NN. (1998). Oxygen depletion in coastal waters. (NOM state of the coast report).
http://oceanservice.noaa.gov/websites/retiredsites/sotc pdf/HYP.PDF. Silver Springs, MD: National Oceanic and Atmospheric
Administration (NOM).
Rabalais NN. (2002). Nitrogen in aquatic ecosystems. Ambio, 31,102-112.
Raddum GG; Fiellheim A. (1984). Acidification and early warning organisms in freshwater in western Norway. Verh IntVer Theor Angew
Limnol, 22,1973-1980.
Raddum GG; Brettum P; MatzowD; Nilssen JP; Skov A; Svealy T; Wright RF. (1986). Liming the acid lake Hovvatn, Norway: a whole lake
study. Water Air Soil Pollut, 31,721-763.
Raddum GG; Skjelvale BL. (2001). Critical limit of acidifying compounds to invertebrates in different regions of Europe. Water Air Soil Pollut,
130,825-830.
Radke LF; Hegg DA; Hobbs PV; Nance JD; Lyons JH; Laursen KK; Weiss RE; Riggan PJ; Ward DE. (1991). Particulate and trace gas
emissions from large biomass fires in North America. In: Levine JS (Ed.) Global biomass burning: atmospheric, climatic, and
biospheric implications, (pp 209-224.). Cambridge, MA: MIT Press.
Rago PJ; Wiener JG. (1986). Does pH affect fish species richness when lake area is considered? Trans Am Fish Soc, 115,438-447.
Rainey SM; Nadelhoffer KJ; Silver WL; Downs MR. (1999). Effects of chronic nitrogen additions on understory species in a red pine planation.
Ecol Appl, 9, 949-957.
Raison RJ; McGarity JW. (1980). Some effects of plant ash on the chemical properties of soils and aqueous suspensions. Plant Soil, 55, 339-
352.
Raivonen M; Keronen P; Vesala T; Kulmala M; Hari P. (2003). Measuring shoot-level NOx flux in field conditions: The role of blank chambers.
Boreal Environment Research, 8,445-455.
Ramazan KA; Syomin D; Finlayson-Pitts BJ. (2004). The photochemical production of HONO during the heterogeneous hydrolysis of NO2.
Phys Chem Chem Phys, 6,3836-3843.
Ramdahl T; Zielinska B; Arey J; Atkinson R; Winer AM; Pitts JN Jr. (1986). Ubiquitous occurrence of 2-nitrofluoranthene and 2-nitropyrene in
air. Nature, 321,425-427.
Randall A; Kriesel W. (1990). Evaluating national policy proposals by contingent valuation. In Johnson RL; Johnson GV (Eds.), Economic
valuation of natural resources: Issues, theory, and applications, (pp 153-178). Boulder, CO: Westview Press.
Rapport DJ; Whitford WG. (1999). How Ecosystems Respond to Stress. Bioscience, 49,193-203.
Rashid GH. (1987). Effects of fire on soil carbon and nitrogen in a Mediterranean oak forest of Algeria. Plant Soil, 103,89-93.
Rattner BA; Haramis GM; Chu DS; Bunck CM; Scanes CG. (1987). Growth and physiological condition of black ducks reared on acidified
wetlands. Can J Zool, 65,2953-2958.
Rattray G; Sievering H. (2001). Dry deposition of ammonia, nitric acid, ammonium, and nitrate to alpine tundra at Niwot Ridge, Colorado.
Atmos Environ, 35,1085-1109.
Ravishankara AR. (1997). Heterogeneous and multiphase chemistry in the troposphere. Science, 276,1058-1065.
Reader JP; Dalziel TRK; Morris R. (1988). Growth, mineral uptake and skeletal calcium deposition in brown trout, Salmo trutta L., yolk-sac fry
exposed to aluminum and manganese in soft acid water. Journal of Fish Biology, 32,607-624.
Reavie ED; Smol JP. (2001). Diatom-environmental relationships in 64 alkaline southwestern Ontario (Canada).lakes: a diatom-based model
for water quality reconstructions. J Paleolimnol, 25,27-44.
Reckhow KH; Black RW; Stockton TB Jr; Vogt JD; Wood JG. (1987). Empirical models offish response to lake acidification. Can J Fish Aquat
Sci, 44,1432-1442.
Redbo-Torstensson P. (1994). The demographic consequences of nitrogen fertilization of a population of sundew, Drosera rotundifolia. Acta
Botanica Neerlandica, 43,175-188.
Redfield AC. (1958). The biological control of chemical factors in the environment. Am Sci, 46, 205-221.
R-72
-------
Regina K; Nykanen H; Maljanen M; Silvola J; Martikainen PJ. (1998). Emissions of N2O and NO and net nitrogen mineralization in a boreal
forested peatland treated with different nitrogen compounds. Can J For Res, 28,132-140.
Rehfuess K-E; Agren Gl; Andersson F; Cannell MGR; Friend A; Hunter I; Kahle H-P; Prietzel J; Spiecker H. (1999). Relationships between
recent changes of growth and nutrition of Norway spruce, Scots pine, and European beech forests in Europe. (EFI working paper 19).
Joensuu, Finland: European Forest Institute.
Reich PB. (1987). Quantifying plant response to ozone: a unifying theory. Tree Physiol, 3,63-91.
Reich PB; Grigal DF; Aber JD; Gower ST. (1997a). Nitrogen mineralization and productivity in 50 hardwood and conifer stands on diverse soils.
Ecology, 78, 335-347.
Reich PB; Walters MB; Ellsworth DS. (1997b). From tropics to tundra: global convergence in plant functioning. Proc Natl Acad Sci USA, 94,
13730-13734.
Reiners WA. (1986). Complementary models in ecology. The American Naturalist, 27,59-73.
Reisen F; Arey J. (2005). Atmospheric reactions influence seasonal PAH and nitro-PAH concentrations in the Los Angeles basin. Environ Sci
Technol, 39,64-73.
Reithmeier C; Sausen R. (2002). ATTILA: atmospheric tracer transport in a Lagrangian model. Tellus, 54B, 278-299.
Renberg I, Korsman T; Anderson NJ. (1993). A temporal perspective of lake acidification in Sweden. Ambio, 22, 264-271.
Rennenberg H. (1984). The fate of excess sulfur in higher plants. Annual Review of Plant Physiology, 35,121-153.
Resh VH; Norris, RH; Barbour MT. (1995). Design and implementation of rapid assessment approaches for water resource monitoring using
benthic macroinvertebrates. Aust J Ecol, 20,108-121.
RestrepoC; Zimmerman R; Thurston G; Clemente J; Gorczynski J; Zhong M; Blaustein M; Chen LC. (2004). A comparison of ground-level air
quality data with New York State Department of Environmental Conservation monitoring stations data in South Bronx, New York.
Atmos Environ, 38,5295-5304.
Reuss JO. (1980). Simulation of soil nutrient losses due to rainfall acidity. Ecol Modell, 11,15-38.
Reuss JO. (1983). Implications of the calcium-aluminum exchange system for the effect of acid precipitation on soils. J Environ Qual, 12,591 -
595.
Reuss JO; Johnson DW. (1985). Effect of soil processes on the acidification of water by acid deposition. J Environ Qual, 14,26-31.
Reuss JO; Johnson DW. (1986). Acid deposition and the acidification of soils and waters. (Series: Billings WD; Golley F; Lange OL; Olson JS;
Remmert H (Eds.), Ecological studies: Analysis and synthesis; Volume 59). New York, NY: Springer-Verlag.
Reuss JO; Vertucci FA; Musselman RC; Sommerfeld RA. (1995). Chemical fluxes and sensitivity to acidification of two high-elevation
catchments in southern Wyoming. J Hydrol, 173,165-189.
Reuter JE; Loeb SL; Axler RP; Carlton RG; Goldman CR. (1985). Transformation of nitrogen following an epilimnetic nitrogen fertilization in
Castle Lake, CA. 1. epilithic periphyton responses. Arch Hydrobiol, 102,425-433.
Reuter JE; Loeb SL; Goldman CR. (1986). Inorganic nitrogen uptake by epilithic periphyton in a N-deficient lake. Limnol Oceanogr, 31,149-
160.
Reuter JE; Axler RP. (1992). Physiological characterisisics of inorganic nitrogen uptake by spatially separate algal communities in a nitrogen-
deficient lake. Freshwater Biol, 27, 227-236.
Richardson DHS; Cameron RP. (2004). Cyanolichens: their response to pollution and possible management strategies for their conservation in
notheastern North America. Northeastern Naturalist, 11,1-22.
Richardson TL; Pinckney JL; Paerl HW. (2001). Responses of estuarine phytoplankton communities to nitrogen form and mixing using
microcosm bioassays. Estuaries, 24,828-839.
Richter A; Burrows JP; Nuss H; Granier C; Niemeier U. (2005). Increase in tropospheric nitrogen dioxide over China observed from space.
Nature, 437,129-132.
Richter DD; Markewitz D; Wells CG; Allen HL; April R; Heine PR; Urrego B. (1994). Soil Chemical Change during Three Decades in an Old-
Field Loblolly Pine (Pinus taeda L.) Ecosystem. Ecology, 75,1463-1473.
Richter DD; Markewitz D; Heine PR; Jin V; Raikes J; Tian K; Wells CG. (2000). Legacies of agriculture and forest regrowth in the nitrogen of
old-field soils. For Ecol Manage, 138,233-248.
Richter DD Jr; Markewitz D. (2001). Understanding soil change: Soil sustainability over millennia, centuries, and decades. Cambridge, UK:
Cambridge University Press.
Riddell J; Nash lii TH; Padgett P. (2008). The effect of HNO3 gas on the lichen Ramalina menziesii. Flora - Morphology, Distribution, Functional
Ecology of Plants, 203,47-54.
Ridley BA; Dye JE; Walega JG; Zheng J; Grahek FE; Rison W. (1996). On the production of active nitrogen by thunderstorms over New
Mexico. J Geophys Res, 101,20985-21005.
Riegman R. (1992). Phaeocystis blooms and eutrophication of the continental coastal zones of the North Sea. Marine Biology, 112,479-484.
Riggan PJ; Lockwood RN; Lopez EN. (1985). Deposition and processing of airborne nitrogen pollutants in Mediterranean-type ecosystems of
southern California. Environ Sci Technol, 19,781-789.
Riggan PJ; Lockwood RN; Jacks PM; Colver CG; Weirich F; DeBano LF; Brass JA. (1994). Effects of fire severity on nitrate mobilization in
watersheds subject to chronic atmospheric deposition. Environ Sci Technol, 28, 369-375.
Rillig MC; Allen MF. (1998). Arbuscular mycorrhizae of Gutierrezia sarothrae and elevated carbon dioxide: evidence for shifts in C allocation to
and within the mycobiont. Soil Biol. Biochem, 30, 2001-2008.
Rimmer CC; McFarland KP. (2005). Part I. Demographic monitoring of montane forest birds on Mt. Mansfield. In: Faccio SD; Rimmer CC
(Eds.), 2004-2005 report to the Vermont monitoring cooperative. (VINS Technical Report 05-02). Woodstock, VT; Conservation
Biology Department; Vermont Institute of Natural Science (VINS).
R-73
-------
Rimmer CC; McFarland KP; Evers DC; Miller EK; Aubry Y; Busby D; Taylor RJ. (2005). Mercury concentrations in Bicknell's thrush and other
insectivorous passerines in montane forests of northeastern North America. Ecotoxicology, 14, 223-240.
RMCC. (1990).The 1990 Canadian long-range transport of air pollutants and acid deposition assessment report, Part 4: Aquatic effects.
Ottawa, Ontario, Canada: Federal/Provincial Research and Monitoring Coordinating Committee (RMCC).
Roberts BJ; Howarth RW. (2006). Nutrient and light availability regulate the relative contribution of autotrophs and heterotrophs to respiration in
freshwater pelagic ecosystems. Limnol Oceanogr, 51,288-298.
Roberts JM; Williams J; Baumann K; Buhr MP; Goldan PD; Holloway J; Hubler G; Kuster WC; McKeen SA; Ryerson TB; Trainer M; Williams
EJ; Fehsenfeld FC; Bertman SB; Nouaime G; Seaver C; Grodzinsky G; Rodgers M; Young VL. (1998). Measurements of PAN, PPN,
and MPAN made during the 1994 and 1995 Nashville Intensives of the Southern Oxidant Study: Implications for regional ozone
production from biogenic hydrocarbons. J Geophys Res, 103,22473-22490.
Roberts JM; Flocke F; Stroud CA; Hereid D; Williams E; Fehsenfeld F; Brune W; Martinez M; Harder H. (2002). Ground-based measurements
of peroxycarboxylic nitric anhydrides (PANs) during the 1999 Southern Oxidants Study Nashville Intensive. J Geophys Res, 107.
Roberts JM; Jobson BT; Kuster W; Goldan P; Murphy P; Williams E; Frost G; Riemer D; Apel E; Stroud C. (2003). An examination of the
chemistry of peroxycarboxylic nitric anhydrides and related volatile organic compounds during Texas Air Quality Study 2000 using
ground-based measurements. J Geophys Res, 108,4495.
Roberts JM; Flocke F; Chen G; Gouw J; Holloway JS; Hubler G; Neuman JA; Nicks DK; Nowak JB; Parrish DD. (2004). Measurement of
peroxycarboxylic nitric anhydrides (PANs).during the ITCT 2K2 aircraft intensive experiment. J Geophys Res, 109.
Robertson GP; Crum JR; Ellis BG. (1993). The spatial variability of soil resources following long-term disturbance. Oecologia, 96,451-456.
Rochelle BP; Church MR. (1987). Regional patterns of sulphur retention in watersheds of the eastern U.S. Water Air Soil Pollut, 36, 61-73.
Rochelle BP; Church MR; David MB. (1987). Sulfur retention at intensively studied sites in the U.S. and Canada. Water Air Soil Pollut, 33, 73-
83.
Rodgers MO; Davis DD. (1989). A UV-photofragmentation/laser-induced fluorescence sensor for the atmospheric detection of HONO. Environ
Sci Technol, 23,1106-1112.
Rodriguez L; Macias F. (2006). Eutrophication trends in forest soils in Galicia (NW Spain).caused by the atmospheric deposition of nitrogen
compounds. Chemosphere, 63,1598-1609.
Roelofs JGM. (1986). The effect of airborne sulphur and nitrogen deposition on aquatic and terrestrial heathland vegetation. Experientia, 42,
372-377.
Roelofs JGM. (1991). Inlet of alkaline river water into peaty lowlands: effects on water quality and Stratiotes abides L. stands. Aquatic Botany,
39, 267-293.
Roem WJ; Berendse F. (2000). Soil acidity and nutrient supply ratio as possible factors determining changes in plant species diversity in
grassland and heathland communities. Biological Conservation, 92,151 -161.
Romansic JM; Diez KA; Higashi EM; Blaustein AR. (2006). Effects of nitrate and the pathogenic water mold Saprolegnia on survival of
amphibian larvae. Dis Aquat.Org, 68, 235-243.
Rose C; Axler RP. (1998). Uses of alkaline phosphatase activity in evaluating phytoplankton community phosphorous deficiency.
Hydrobiologia, 361,145-156.
Rosemond AD; Reice SR; Elwood JW; Mulholland PJ. (1992). The effects of stream acidity on benthic invertebrate communities in the south-
eastern United States. Freshwater Biol, 27,193-209.
Rosenberger RS; Loomis JB. (2001). Benefit transfer of outdoor recreation use values: A technical document supporting the Forest Service
Strategic Plan (2000 revision). (General technical report RMRS-GTR-72). Fort Collins, CO; Rocky Mountain Research Station; U.S.
Forest Service; U.S. Department of Agriculture (USDA).
Rosenqvist IT. (1978). Acid precipitation and other possible sources for acidification of rivers and lakes. Sci Total Environ, 10,271-272.
Ross DS; Lawrence GB; Fredriksen G. (2004). Mineralization and nitrification patterns at eight northeastern USA forested research sites. For
Ecol Manage, 188,317-335.
Rosseland BO; Skogheim OK. (1984). A comparative study on salmonid fish species in acid aluminum-rich water. II. Physiological stress and
mortality of one- and two-year-old fish. (Report no. 61; pp. 186-194). Drottningholm, Sweden: Institute of Freshwater Research.
Rosseland BO. (1986). Ecological effects of acidification on tertiary consumers: fish population response. Water Air Soil Pollut, 30,451-460.
Rosseland BO; EldhusetTD; Staurnes M. (1990). Environmental effects of aluminum. Environ Geochem Health, 12,17-27.
Rosseland BO; Staurnes M. (1994). Physiological Mechanisms for Toxic Effects and Resistance to Acidic Water: An Ecophysiological and
Ecotoxicological Approach. Acidification of Freshwater Ecosystem: Implications for the Future, 227-245.
Roth PM; Blumenthal DL; Roberts PT; Watson JG; Yocke M; Souten D; Ireson RG; Chinkin LR; Whitten GZ; Daly C; Smith TB. (1988). A
proposed concept and scope for the San Joaquin Valley Air Quality Study. (Report no. STI96040-710FR1). Santa Rosa, CA: Sonoma
Technology, Inc.
Roulet N; Moore TR. (2006). Environmental chemistry - Browning the waters. Nature, 444,283-284.
Rouse JD., Bishop CA., Struger J. (1999). Nitrogen pollution: an assessment of its threat to amphibian survival. Environ Health Perspect, 107,
799-803.
Rouss E; Ahti T. (1989). Systematics of some reindeer lichens (Cladonia subg. Cladina) in the Southern Hemisphere. - Lichenologist 21: 29-44.
Rowe EC; Moldan F; Emmett BA; Evans CD; Hellsten S. (2005). Model chains for assessing impacts of nitrogen on soils, waters and
biodiversity: A review. (Report no C02887). Brighton, UK: Centre for Ecology and Hydrology, Natural Environment Research Council.
RTI International. (2001). Evaluation of Waquoit Bay models to estimate the water quality and ecological improvements resulting from
atmospheric nitrogen reduction regulatory scenarios. Research Triangle Park, NC: U.S. Environmental Protection Agency, Office of Air
Quality Planning and Standards, contract no. 68-D-99-024.
R-74
-------
Ruckert GV; Giani A. (2004). Effect of nitrate and ammonium on the growth and protein concentration of Microcystis viridis Lemmermann
(Cyanobacteria). Revista Brasileira de Botanica, 27,325-331.
Ruddell EJ; Gramann JH. (1989). The psychological utility of visual penetration in near-view forest scenic-beauty models. Environment and
Behavior, 21,393-412.
Rudek J; Paerl HW; Mallin MA; Bates PW. (1991). Seasonal and hydrological control of phytoplankton nutrient limitation in the lower Neuse
River Estuary, North Carolina. Mar Ecol Prog Ser, 75,133-142.
Rueth HM; Baron JS. (2001). Englemann spruce nitrogen dynamics across a nitrogen deposition gradient in Colorado, USA. Ekologia, 20,43-
49.
Rueth HM; Baron JS. (2002). Differences in Englemann Spruce Forest Biogeochemistry East and West of the Continental Divide in Colorado,
USA. Ecosystems, 5,45-57.
Rueth HM; Baron JS; Allstott EJ. (2003). Responses of Engelmann spruce forests to nitrogen fertilization in the Colorado Rocky Mountains.
EcolAppI, 13, 664-673.
Rundle SD; Hildrew AG. (1990). The distribution of micro-anthropods in some southern English streams: The influence of physicochemistry.
Freshwater Biol, 23,411-431.
Runn P; Johansson N; Milbrink G. (1977). Some effects of low pH on the hatchability of eggs of perch, Perca fluviatilis L. Zoon, 5,115-125.
Russell A; Dennis R. (2000). NARSTO critical review of photochemical models and modeling. Atmos Environ, 34,2283-2324.
Russell CA; Kosola KR; Paul EA; Robertson GP. (2004). Nitrogen cycling in poplar stands defoliated by insects. Biogeochem, 68, 365-381.
Russell KM; Keene WC; Maben JR; Galloway JN; Moody JL. (2003). Phase-partitioning and dry deposition of atmospheric nitrogen at the mid-
Atlantic U.S. coast. J Geophys Res, 108, doi: 10.1029/2003JD003736.
Russell LM; Mensah AA; Fischer EV; Sive BC; Varner RK; Keene WC; Stutz J; Pszenny AAP. (2006). Nanopartide growth following
photochemical a- and (3-pinene oxidation at Appledore Island during International Consortium for Research on Transport and
Transformation/Chemistry of Halogens at the Isles of Shoals 2004. J Geophys Res, 112, doi: 10.1029/2006JD007736.
Russow R; Spott O; Stange CF. (2008). Evaluation of nitrate and ammonium as sources of NO and N2O emissions from black earth soils
(Haplic CherNOzem) based on N-15 field experiments. Soil Biol Biochem, 40,380-391.
Rustad LE; Kahl JS; Norton SA; Fernandez IJ. (1994). Underestimation of dry deposition by throughfall in mixed northern hardwood forests. J
Hydrol, 162,319-336.
Ryerson TB; Buhr MP; Frost GJ; Goldan PD; Holloway JS; Hubler G; Jobson BT; Kuster WC; McKeen SA; Parrish DD; Roberts JM; Sueper
DT; Trainer M; Williams J; Fehsenfeld FC. (1998). Emissions lifetimes and ozone formation in power plant plumes. J Geophys Res,
103, 22569-22583.
Ryerson TB; Trainer M; Holloway JS; Parrish DD; Huey LG; Sueper DT; Frost GJ; Donnelly SG; Schauffler S; Atlas EL; Kuster WC; Goldan
PD; Hubler G; Meagher JF; Fehsenfeld FC. (2001). Observations of ozone formation in power plant plumes and implications for ozone
control strategies. Science, 292,719-723.
Ryerson TB; Williams EJ; Fehsenfeld FC. (2000). An efficient photolysis system for fast-response NO2 measurements. J Geophys Res, 105,
26447-26461.
Rysgaard S; Glud RN; Risgaard-Petersen N; Dalsgaard T. (2004). Denitrification and anammox activity in Arctic marine sediments. Limnol
Oceanogr, 49,1493-1502.
Rysgaard S; Risgaard-Petersen N; Sloth NP; Jensen K; Nielsen LP. (1994). Oxygen regulation of nitrification and denitrification in sediments.
Limnol Oceanogr, 39,1643-1652.
Saarnio S; Silvola J. (1999). Effects of increased CO2 and N on CH4 efflux from a boreal mire: a growth chamber experiment. Oecologia, 119,
349-356.
Saarnio S; Saarinen T; Vasander H; Silvola J. (2000). A moderate increase in the annual CH4 efflux by raised CO2 or NH4NO3 supply in a
boreal oligotrophic mire. Glob Chang Biol, 6,137-144.
Saarnio S; Jarvio S; Saarinen T; Vasander H; Silvola J. (2003). Minor changes in vegetation and carbon gas balance in a boreal mire under a
raised CO2 or NH4NO3 supply. Ecosystems, 6,46-60.
Saathoff H; Naumann KH; Riemer N; Kamm S; Mohler O; Schurath U; Vogel H; Vogel B. (2001). The loss of NO2, HNO3, N03/N205, and
HO2/HOONO2 on soot aerosol: A chamber and modeling study. Geophys Res Lett, 28,1957-1960.
Sabaratnam S; Gupta G; Mulchi C. (1988). Nitrogen dioxide effects on photosynthesis in soybean. J Environ Qual, 17,143-146.
Sabbioni C; Bonazza A; Zappia G. (2002). Damage on hydraulic mortars: The Venice Arsenal. Journal of Cultural Heritage, 3, 83-88.
Sabbioni C; Zappia G; Riontinoa C; Blanco-Varela MT; Aguilerac J; Puertasc F; Van Balen K; Toumbakari EE. (2001). Atmospheric
deterioration of ancient and modern hydraulic mortars. Atmos Environ, 35,539-548.
Sadler K; Lynam S. (1988). The influence of calcium on aluminium-induced changes in the growth rate and mortality of brown trout. Salmo
trutta, 171-179.
Saiz-Lopez A; Plane JMC. (2004). Novel iodine chemistry in the marine boundary layer. Geophys Res Lett, 31, doi: 10.1029/2003GL019215.
Saiz-Lopez A; Plane JMC; Shillito JA. (2004). Bromine oxide in the mid-latitude marine boundary layer. Geophys Res Lett, 31, doi:
10.1029/2003GL018956.
Sakamaki F; Hatakeyama S; Akimoto H. (1983). Formation of nitrous acid and nitric oxide in the heterogeneous dark reaction of nitrogen
dioxide and water vapor in a smog chamber. IntJChem Kinetics, 15,1013-1029.
Sakugawa H; Kaplan IR. (1989). H2O2 and O3 in the atmosphere of Los Angeles and its vicinity: factors controlling their formation and their role
as oxidants of S02. J Geophys Res, 94,12957-12973.
Salmeen I; Durisin AM; Prater TJ; Riley T; Schuetzle D. (1982). Contribution of 1 -nitropyrene to direct-acting Ames assay mutagenicities of
diesel particulate extracts. Mutation Research / DNA Repair, 104,17-23.
R-75
-------
SAMAB. (1996). The Southern Appalachian Assessment. Aquatics technical report. KNOxville, TN; Southern Appalachian Man and the
Biosphere (SAMAB) Cooperative, http://samab.ora/saa/reports/aauatic/aauatic.html.
Samet JM; Bodurow CC (Eds.). (2008). Improving the presumptive disability decision-making process for veterans. Washington, DC: Institute
of Medicine of the National Academies, National Academics Press.
Samie F; Tidblad J; Kucera V; Leygraf C. (2007). Atmospheric corrosion effects of HNO3—Comparison of laboratory-exposed copper, zinc and
carbon steel. Atmos Environ, 41,4888-4896.
Samson JC; Shenker J. (2000). The teratogenic effects of methylmercury on early development of the zebrafish, Danio Rerio. Aquat Toxicol,
48, 343-354.
Sander R; Rudich Y; Von Glasow R; Crutzen PJ. (1999). The role of BrN03 in marine tropospheric chemistry: A model study. Geophys Res
Lett, 26, 2858-2860.
Sander R; Keene WC; Pszenny A. AP; Arimoto R. Ayers GP; Chainey JM; Crutzen PJ; Duce RA; Huebert BJ; MaenhautW; Turekian VC; Van
Dingenan, R. (2003). Inorganic bromine in the marine boundary layer: a critical review. Atmos Chem Phys, 3,1301-1336.
Sanz MJ; GriesC; Nash TH III. (1992). Dose-response relationships for SO2 fumigations in the lichens Ebernia prunastri (l).Ach, Ramalina
fraxinea (L.). Ach New Phytol, 122, 313-319.
Sarmiento JL; Gruber N. (2006). Ocean Biogeochemical Dynamics. Princeton, NJ: Princeton University Press.
Saros JE; Interlandi SJ; Wolfe AP; Engstrom DR. (2003). Recent changes in the diatom community structure of lakes in the Beartooth
Mountain Range, USA. Arct Antarct Alp Res, 35,18-23.
Saros JE; Michel TJ; Interlandi SJ; Wolfe AP. (2005). Resource requirements of Asterionella formosa and Fragilaria crotonensis in oligotrophy
alpine lakes: implications for recent phytoplankton community reorganizations. Can J Fish Aquat Sci, 62,1681-1689.
Sasaki JC; Arey J; Eastmond DA; Parks KK; Grosovsky AJ. (1997). Genotoxicity induced in human lymphoblasts by atmospheric reaction
products of naphthalene and phenanthrene. Mutation Research / DNA Repair, 393, 23-35.
Sawyer RF; Harley RA; Cadle SH; Norbeck JM; Slott R; Bravo HA. (2000). Mobile sources critical review: 1998 NARSTO assessment. Atmos
Environ, 34,2161-2181.
Saxe H. (1986). Effects of NO, NO2, andC02 on net photosynthesis, dark respiration, and transpiration of pot plants. New Phytol, 103,185-
197.
SaxeH. (1986). Stomatal-dependent and stomatal-independent uptake ofNOx. New Phytol, 103:199-205.
Scala JM; Garstang M; Tao W-K; Pickering KE; Thompson AM; Simpson J; Kirchhoff VWJH; Browell EV; Sachse GW; L. TA; Gregory GL;
Rasmussen RW; Khalil MAK. (1990). Cloud draft structure and trace gas transport. J Geophys Res, 95,17015-17030.
Schaberg PG; DeHayes DH; Hawley GJ; Strimbeck GR; Cumming JR; Murakami PF; Borer CH. (2000). Acid mist, soil Ca and Al alter the
mineral nutrition and physiology of red spruce. Tree Physiol, 20,73-85.
Schaberg PG; DeHayes DH; Hawley GJ; Murakami PF; Strimbeck GR; McNulty SG. (2002). Effects of chronic N fertilization on foliar
membranes, cold tolerance, and carbon storage in montane red spruce. Can J Forest Res, 32,1351-1359.
Schaeffer SM; Billings SA; Evans RD. (2003). Responses of soil nitrogen dynamics in a Mojave Desert ecosystem to manipulations in soil
carbon and nitrogen availability. Oecologia, 134,547-553.
Schauer C; Niessner R; Poschl U. (2004). Analysis of nitrated polycyclic aromatic hydrocarbons by liquid chromatography with fluorescence
and mass spectrometry detection: air particulate matter, soot, and reaction product studies. Anal Bioanal Chem, 378,725-736.
Schelske CL. (1991). Historical nutrient enrichment of Lake Ontario: paleolimnological evidence. Can J Fish Aquat Sci, 48,1529-1538.
Scheuhammer AM. (1988). Chronic dietary toxicity of methylmercury in the zebra finch, Poephila guttata. Bull Environ Contam Toxicol, 40,123-
130.
Scheuhammer AM; Blancher PJ. (1994). Potential risk to common loons (Gavia /mmerj.from methylmercury exposure in acidified lakes.
Hydrobiologia, 279/280,445-455.
Scheuhammer AM; Meyer MW; Sandheinrich MB; Murray MW. (2007). Effects of environmental methylmercury on the health of wild birds,
mammals, and fish. Ambio, 36,12-18.
Schichtel BA; Gebhart KA; Barna MG; Malm WC; Green MC. (2004). Big Bend regional aerosol and visibility observational (BRAVO).study
results: Air quality data and source attribution analyses results from the National Park Service. Ft Collins, CO; Cooperative Institute for
Research in the Atmosphere (CIRA); Colorado State University.
Schichtel BA; Gebhart KA; Malm WC; Barna MG; Pitchford ML; Knipping EM; Tombach IH. (2005). Reconciliation and interpretation of Big
Bend National Park particulate sulfur source apportionment: Results from the Big Bend regional aerosol and visibility observational
study-part I. J Air Waste Manage Assoc: 55,1709-1725.
Schier GA. (1985). Response of red spruce and balsam fir seedlings to aluminum toxicity in nutrient solutions. Can J For Res, 15,29-33.
Schimel DS; House Jl; Hibbard KA; Bousquet P; Ciais P; Peylin P; Braswell BH; Apps MJ; Baker D; Bondeau A; Canadell J; Churkina G;
Cramer W; Denning AS; Field CB; Friedlingstein P; Goodale C; Heimann M; Houghton RA; Melillo JM; Moore Bill; Murdiyarso D;
Noble I; Pacala SW; Prentice IC; Raupach MR; Rayner PJ; Scholes RJ; Steffen WL; Wirth C. (2001). Recent patterns and
mechanisms of carbon exchange by terrestrial ecosystems. Nature, 414,169-172.
Schimel JP; Bennett J. (2004). Nitrogen mineralization: challenges of a changing paradigm. Ecology, 85,597-602.
Schindler DW; Armstrong FAJ; Brunskill GJ; Holmgren SK. (1971). Eutrophication of Lake 227, Experimental Lakes Area, northwestern
Ontario, by addition of phosphate and nitrate. J Fish Res Board Can, 28,1763-1782.
Schindler DW. (1974). Eutrophication and recovery in experimental lakes: implications for lake management. Science, 184,897-899.
Schindler DW. (1980). The effect of fertilization with phosphorous and nitrogen versus phosphorus alone on eutrophication of experimental
lakes. Limnol Oceanogr, 25,1149-1152.
Schindler DW; Mills KH; Malley DF; Findlay MS; Schearer JA; Davies I J; Turner MA; Lindsey GA; Cruikshank DR. (1985). Long-term
ecosystem stress: Effects of years of experimental acidification. Science, 228,1395-1401.
R-76
-------
Schindler DW. (1986). The significance of in-lake production of alkalinity. Water Air Soil Pollut, 30,931-944.
Schindler DW. (1988). Effects of acid rain on freshwater ecosystems. Science, 239,149-157.
Schindler DW. (1990). Experimental perturbations of whole lakes as tests of hypotheses concerning ecosystem structure and function. Oikos,
57,25-41.
Schindler DW; Bayley SE. (1993). The biosphere as in increasing sink for atmospheric carbon: Estimates from increased nitrogen deposition.
Glob Biogeochem Cycles, 7,717-733.
Schlegel H; Amundson RG; Huttermann A. (1992). Element distribution in red spruce (Picea rubens).fine roots; evidence for aluminum toxicity
atWhiteface Mountain. Can J For Res, 22,1132-1138.
Schlesinger WH; Andrews JA. (2000). Soil respiration and the global carbon cycle. Biogeochemistry, 48,7-20.
Schlesinger WH; Ward TJ; Anderson J. (2000). Nutrient losses in runoff from grassland and shrubland habitats in southern New Mexico: II.
Field plots. Biogeochem, 49,69-86.
Schlesinger WH; Lichter J. (2001). Limited carbon storage in soil and litter of experimental forest plots under increased atmospheric CO2.
Nature, 411,466-469.
Schlesinger WL; Hartley AE. (1992). A global budget for atmospheric NH3. Biogeochemistry, 15,191-211.
Schnell S; King GM. (1994). Mechanistic analysis of ammonium inhibition of atmospheric methane consumption in forest soils. Appl Environ
Microbiol, 60, 3514-3521.
Schnitzer M; Skinner SIM. (1963). Organo-metallic interactions in soils: 2. Reactions between different forms of iron and aluminum and the
organic matter of a podzol Bh horizon. Soil Sci, 96,181-186.
Schnoor JL; Nikolaidis NP; Glass GE. (1986). Lake resources at risk to acidic deposition in the Upper Midwest. J WATER POLLUT CONTROL
FED, 58,139-148.
Schnug E. (1997). Significance of sulphur for the quality of domesticated plants. Sulphur Metabolism in Higher Plants, 3,109-130.
Schofield CL; Trojnar JR. (1980). Aluminum toxicity to brook trout (Salvelinus fontinalis) in acidified waters. In: Toribara TY; Miller MW; Morrow
PE (Eds.), Polluted rain: Proceedings of the 12th Rochester international conference on environmental toxicity held in Rochester, NY;
May 1979. (Series: Hollaender A; Probstein RF; Welch BL (Eds.), Environmental science research; Volume 17, pp 341-66). New York,
NY: Plenum Press.
Schofield CL; Driscoll CT. (1987). Fish species distribution in relation to water quality gradients in the North Branch of the Moose River Basin.
Biogeochemistry, 3,63-85.
Schofield CL. (1993). Habitat suitability for Brook Trout (Salvelinus fontinalis) reproduction in Adirondack lakes. Water Resour Res, 29,875-
879.
Schofield CL; Keleher C. (1996). Comparison of brook trout reproductive success and recruitment in an acidic Adirondack lake following whole
lake liming and watershed liming. Biogeochem, 32, 323-337.
Schofield KA; Pringle CM; Meyer JL; Sutherland AB. (2001). The importance of crayfish in the breakdown of rhododendron leaf litter. Freshw
Biol, 46,1-14.
Scholes RJ; Noble IR. (2001). Climate change - Storing carbon on land. Science, 294,1012-1013.
Schopp W; Posch M; Mylonda S; Johansson M. (2003). Long-term development of acid deposition (1880—2030) in sensitive freshwater
regions in Europe. Hydrol Earth Syst Sci, 7,436-446.
Schreck CB. (1981). Stress and rearing of salmonids. Aquaculture, 28,241-249.
Schreck CB. (1982). Stress and compensation in teleostean fishes: response to social and physical factors. In: Pickering AD (Ed.), Stress and
fish (pp. 295-321). London: Academic Press.
SchrimpfW; Lienaerts K; Mtiller KP; Rudolph J; Neubert R; SchuRler W; Levin I. (1996). Dry deposition of peroxyacetyl nitrate (PAN):
Determination of its deposition velocity at night from measurements of the atmospheric PAN and 222 Radon concentration gradient.
Geophys Res Lett, 23,3599-3602.
Schroder WH; Bauch J; Endeward R. (1988). Microbeam analysis of Ca exchange and uptake in the fine roots of spruce: influence of pH and
aluminum. Trees, 2,96-103.
Schubert SD; Rood RB; Pfaendtner J. (1993). An assimilated dataset for earth science applications. Bulletin of the American Meteorological
Society, 74, 2331-2342.
Schuetzle D. (1983). Sampling of vehicle emissions for chemical analysis and biological testing. Environ Health Perspect, 47,65-80.
Schultz MG; Jacob DJ; Bradshaw JD; Sandholm ST; Dibb JE; Talbot RW; Singh HB. (2000). Chemical NOx budget in the upper troposphere
over the tropical South Pacific. J Geophys Res, 105,6669-6679.
Schulze ED; Hogbberg L; van Oene H. (2000). Interactions between the carbon and nitrogen cycle and the role of biodiversity: A synopsis of a
study along a north-south transect through Europe. In: Schultze ED (Ed.), Carbon and Nitrogen Cycling in European Forest
Ecosystems (Vol. 142, pp. 468-492). Berlin: Springer.
Schwab JJ; Li Y; Bae MS; Demerjian KL; Hou J; Zhou X; Jensen B; Pryor SC. (2007). A laboratory intercomparison of real-time gaseous
ammonia measurement methods. Environ Sci Technol, 41,8412-8419.
Schweizer B; Arndt U. (1990). CO2/H2O gas exchange parameters of one- and two-year-old needles of spruce and fir. Environ Pollut, 68,275-
292.
Schwinning S; Starr Bl; Wojcik NJ; Miller ME; Ehleringer JE; Sanford RL Jr. (2005). Effects of nitrogen deposition on an arid grassland in the
Colorado Plateau cold desert. Journal of Rangeland Ecology and Management, 58,565-574.
Scott G; Crunkilton RL. (2000). Acute and chronic toxicity of nitrate to fathead minnows (Pimephales promelas), Ceriodaphnia dubia, and
Daphnia magna. Environ Toxicol Chem, 19,2918-2922.
Scott MG; Hutchinson TC; Feth MJ. (1989a). A comparison of the effects on Canadian boreal forest lichens of nitric and sulphuric acids as
sources of rain acidity. NewPhytol, 111,663-671.
R-77
-------
Scott MG; Hutchinson TC; Feth MJ. (1989b). Contrasting responses of lichens and Vaccinium angustifblium to long-term acidification of a
boreal forest ecosystem. Can J Bot, 67, 579-588.
Scott MC; Helfman GS. (2001). Native invasions, homogenization, and the mismeasure of integrity offish assemblages. Fisheries, 26,6-15.
Scudlark JR; Church TM. (1993). Atmospheric input of inorganic nitrogen to Delaware Bay. Estuaries, 16,747-759.
Seaman NL. (2000). Meteorological modeling for air quality assessments. Atmos Environ, 34,2231-2259.
Seastedt TR; Vaccaro L. (2001). Plant species richness, productivity, and nitrogen and phosphorus limitations across a snowpack gradient in
alpine tundra, Colorado, USA. Arct Antarct Alp Res, 33,100-106.
Seastedt TR; Bowman WD; Caine TN; McKnight D; Townsend A; Williams MW. (2004). The landscape continuum: a model for high elevation
ecosystems. Bioscience, 54,111-121.
Seely B; Lajtha K. (1997). Application of a 15N tracer to simulate and track the fate of atmospherically deposited N in the coastal forests of the
Waquoit Bay watershed, Cape Cod, Massachusetts. Oecologia, 112, 393-402.
Segschneider HJ; Hutzen H; Forstel H; Rockel P; Wildt J. (1993). Uptake of (N02)-N-15 and metabolic transfer of the (N02)-N-15 nitrogen to
various nitrogen fractions of sunflowers. Isotopenpraxis Environ Health Stud, 29,51-57.
Segschneider H-J; Wildt J; Forstel H. (1995). Uptake of 15N02 by sunflower (Helianthus-annuus).during exposures in light and darkness:
quantities, relationship to stomatal aperture and incorporation into different nitrogen pools within the plant. NewPhytol, 131,109-119.
Seinfeld JH; Pandis SN. (1998). Atmos Chem Phys: From Air Pollution to Climate Change. New York, NY: John Wiley-lnterscience Publishers.
Seip HM. (1980). Acidification of freshwater - sources and mechanisms. Paper presented at the Ecological impact of acid precipitation.
Proceedings of an International Conference, SNSF Project, Sandetjord, Norway, Oslo, Norway.
Seitzinger SP. (1988). Denitrification in freshwater and coastal marine ecosystems: ecological and geochemical significance. Limnol Oceanogr,
33, 702-724.
Seitzinger SP; Nielsen LP; Caffrey J; Christensen PB. (1993). Denitrification measurements in aquatic sediments: a comparison of three
methods. Biogeochemistry, 23,147-167.
Seitzinger SP; Styles RV; Boyer EW; Alexander R; Billen G; Howarth RW; Mayer B; Van Breemen N. (2002). Nitrogen retention in rivers: model
development and application to watersheds in the northeastern USA. Biogeochemistry, 57/58,199-237.
Seitzinger S; Harrison JA; Bohlke JK; Bouwman AF; Lowrance R; Peterson B; Tobias C; Van Drecht G. (2006). Denitrification across
landscapes and waterscapes: A synthesis. Ecol Appl, 16,2064-2090.
Shan JP; Morris LA; Hendrick RL. 2001. The effects of management on soil and plant carbon sequestration in slash pine plantations. J Appl
Ecol, 38, 932-941.
Shanley JB; Chalmers A. (1999). The effect of frozen soil on snowmelt runoff at Sleepers River, Vermont. Hydrological Processes, 13,1843-
1857.
Shanley J; Mayer B; Mitchell M; Michel R; Bailey S; Kendall C. (2005). Tracing Sources of Streamwater Sulfate During Snowmelt Using S and
O Isotope Ratios of Sulfate and 35S Activity. Biogeochemistry, 76,161-185.
Sharpe WE; Kimmel WG; Young ES Jr; DeWalle DR. (1983). In-situ bioassays offish mortality in two Pennsylvania streams acidified by
atmospheric deposition. Northeast Environ Sci, 2,171-178.
Shashoua Y. (2006). Inhibiting the inevitable; current approaches to slowing the deterioration of plastics. Macromol Symp, 238, 67-77.
Shaver GR; Billings WD. (1975). Root production and root turnover in a wet tundra ecosystem, Barrow, Alaska. Ecology, 56,401-409.
Shaver GR; Chapin FS. (1980). Response to fertilization by various plant growth forms in an Alaskan tundra: nutrient accumulation and growth.
Ecology, 61, 662-675.
Shaver GR; Melillo JM. (1984). Nutrient budgets of marsh plants: Efficiency concepts and relation to availability. Ecology, 65,1491-1510
Shaver GR; Johnson LC; Cades DH; Murray G; Laundre JA; Rastetter EB; Nadelhoffer KJ; Giblin AE. (1998). Biomass and CO2 flux in wet
sedge tundras: Responses to nutrients, temperature, and light. Ecol Monogr, 68, 75-97.
Shaver GR; Bret-Harte SM; Jones MH; Johnstone J; Gough L; Laundre J; Chapin FS. (2001). Species composition interacts with fertilizer to
control long-term change in tundra productivity. Ecology, 82,3163-3181.
Shaw MR; Zavaleta ES; Chiariello NR; Cleland EE; Mooney HA; Field CB. (2002). Grassland responses to global environmental changes
suppressed by elevated CO2. Science, 298,1987-1990.
Shaw WD. (1989). Valuing the effect of acidification damages on the Adirondack fishery: Comment. Am J Agric Econ, 71,217-221.
Shepard JP; Mitchell MJ; Scott, TJ; Zhang YM; Raynal DJ. (1989). Measurements of wet and dry deposition in a northern hardwood forest.
Water Air Soil Pollut, 48, 225.238.
Shepson PB; Bottenheim JW; Hastie DR; Venkatram A. (1992). Determination of the relative ozone and PAN deposition velocities at night.
GeophysRes Lett, 19,1121-1124.
Shepson PB; Mackay E; Muthuramu K. (1996). Henry's Law constants and removal processes for several atmospheric beta-hydroxy alkyl
nitrates. Environ Sci Technol, 30, 3618-3623.
Sherrod SK; Seastedt TR. (2001). Effects of the northern pocket gopher (Thomomys talpoides).on alpine soil characteristics, Niwot Ridge, CO.
Biogeochemistry, 55,195-218.
Shindell DT; Faluvegi G; Stevenson DS; Krol MC; Emmons LK; Lamarque J-F; Petron G; Dentener FJ; Ellingsen K; Schultz MG; Wild O;
Amann M; Atherton CS; Bergmann DJ; Bey I; Butler T; Cofala J; Collins WJ; Derwent RG; Doherty RM; Drevet J; Eskes HJ; Fiore AM;
Gauss M; Hauglustaine DA; Horowitz LW; Isaksen ISA; Lawrence MG; Montanaro V; Miiller J-F; Pitari G; Prather MJ; Pyle JA; Rast S;
Rodriguez JM; Sanderson MG; Savage NH; Strahan SE; Sudo K; Szopa S; Unger N; Van Noije TPC; Zeng G. (2006). Multimodel
simulations of carbon monoxide: Comparison with observations and projected near-future changes. J Geophys Res, 111, doi:
10.1029/2006JD007100.
Shinn JH. (1978). A critical survey of measurements of foliar deposition of airborne sulfates and nitrates. Paper presented at the 71st annual
meeting of the Air Pollution Control Association; June; Houston, TX, Pittsburgh, PA.
R-78
-------
Shogren JF; Shin SY; Hayes DJ; Kliebenstein JB. (1994). Resolving differences in willingness to pay and willingness to accept. Am Econ Rev,
84, 255-270.
Shorter JH; Herndon S; Zahniser MS; Nelson DD; Wormhoudt J; Demerjian KL; Kolb CE. (2005). Real-time measurements of nitrogen oxide
emissions from in-use New York City transit buses using a chase vehicle. Environ Sci Technol, 39 (20), 7991-8000.
Shortle WC; Smith KT. (1988). Aluminum-induced calcium deficiency syndrome in declining red spruce. Science, 240,1017-1018.
Shortle WC; Smith KT; Minocha R; Lawrence GB; David MB. (1997). Acidic deposition, cation mobilization, and biochemical indicators of stress
in healthy red spruce. J Environ Qual, 26,871-876.
Shriner DS; Henderson GS. (1978). Sulfur distribution and cycling in a deciduous forest watershed. J Environ Qual, 7,392-397.
Shugart HH; McLaughlin SB. (1985). Modeling SO2 effects on forest growth and community dynamics. In: Winner WE; Mooney HA; Goldstein
RA (Ed.), Sulfur dioxide and vegetation: Physiology, ecology, and policy issues, (pp 478-491). Stanford, CA: Stanford University
Press.
Shy CM; Love GJ. (1980) Recent evidence on the human health effects of nitrogen dioxide. In: Lee SD (Ed.), Nitrogen oxides and their effects
on health: [papers from the symposium "Biological studies of environmental pollutants, part II. Health effects of nitrogen oxides"]; April
1979; Honolulu, HI. Ann Arbor, Ml: Ann Arbor Science Publishers, Inc; 291-305.
Sickles JE; Hodson LL; McClenny WA; Paur RJ; Ellestad TG; Mulik JD; Anlauf KG; Wiebe HA; Mackay Gl; Schiff HI; Bubacz DK. (1990). Field
comparison of methods for the measurement of gaseous and particulate contributors to acidic dry deposition, Atmos Environ 24,155—
165.
Siciliano SD; O'Driscoll NJ; Tordon R; Hill J; Beauchamp S; Lean DRS. (2005). Abiotic production of methylmercury by solar radiation. Environ
Sci Technol, 39,1071-1077.
Sickman JO; Melack JM; Stoddard JL. (2002). Regional analysis of inorganic nitrogen yield and retention in high-elevation ecosystems of the
Sierra Nevada and Rocky Mountains. Biogeochem, 57/58,341-374.
Sickman JO; Melack JM; Clow DW. (2003a). Evidence for nutrient enrichment of high-elevation lakes in the Sierra Nevada, California. Limnol
Oceanogr, 48,1885-1892.
Sickman JO; Leydecker A; Chang CCY; Kendall C; Melack JM; Lucero DM; Schimel J. (2003b). Mechanisms underlying export of N from high-
elevation catchments during seasonal transitions. Biogeochem, 64,1-24.
Siegfried CA; Bloomfield JA; Sutherland JW. (1989). Planktonic rotifer community structure in Adirondack lakes: Effects of acidity, trophic
status, and related water quality characteristics. In: Sutherland JW (Ed.), Field studies of the biota and selected water chemistry
parameters in 50 Adirondack mountain lakes. Albany, NY: New York State Department of Environmental Conservtion.
Siegwolf RTW; Matyssek R; Sauer M; Maurer S; Gunthardt-Goerg MS; Schmutz P; Bucher JB. (2001). Stable isotope analysis reveals
differential effects of soil nitrogen and nitrogen dioxide on the water use efficiency in hybrid poplar leaves. New Phytol, 149,233-246.
Sievering H. (1999). Nitrogen deposition and carbon sequestration. Nature, 400,629-630.
Sievering H; Fernandez I; Lee J; Horn J; Rustad L. (2000). Forest canopy uptake of atmospheric nitrogen deposition at eastern U.S. conifer
sites: carbon storage implications? Glob Biogeochem Cycles, 14,1153-1159.
Sievering H; Rusch D; Marquez L. (1996). Nitric acid, particulate nitrate and ammonium in the continental free troposphere: Nitrogen deposition
to an alpine tundra ecosystem. Atmos Environ, 30, 2527-2537.
Sigal LL; Nash TH III. (1983). Lichen Communities on Conifers in Southern California Mountains: An Ecological Survey Relative to Oxidant Air
Pollution. Ecology, 64,1343-1354.
Sigiienza C. (2000). Nitrogen deposition and soil microorganisms of Artemisia californica and exotic grasses in southern California,
(dissertation). Riverside, CA: University of California.
Sigiienza C; Corkidi L; Allen EB (2006). Feedbacks of soil inoculum of mycorrhizal fungi altered by N deposition on the growth of a native shrub
and an invasive annual grass. Plant Soil, 286,153-165.
Sikiotis D; Kirkitsos P. (1995). The adverse effects of nitrates on stone monuments. Sci Total Environ, 171,173-182.
Sillman S; Logan J; Wofsy S. (1990). The sensitivity of ozone to nitrogen oxides and hydrocarbons in regional ozone episodes. J Geophys Res,
95,1851.
Sillman S. (1995). The use of NOy, H2O2 and HNO3 as indicators for ozone-NOx-hydrocarbon sensitivity in urban locations. J Geophys Res,
100,14175-14188.
Sillman S; Al-Wali Kl; Marsik FJ; Nowacki P; Samson PJ; Rodgers MO; Garland LJ; Martinez JE; Stoneking C; Imhoff R; Lee J-H; Newman L;
Weinstein-Lloyd J; Aneja VP. (1995). Photochemistry of ozone formation in Atlanta, GA: Models and measurements. Atmos Environ,
29, 3055-3066.
Sillman S; He D; Pippin MR; Daum PH; Imre DG; Kleinman LI; Lee JH; Weinstein-Lloyd J. (1998). Model correlations for ozone, reactive
nitrogen, and peroxides for Nashville in comparison with measurements: Implications for 03-NOx-hydrocarbon chemistry. J Geophys
Res, 103, 22629-22644.
Sillman S. (2000). Ozone production efficiency and loss of NOx in power plant plumes: Photochemical model and interpretation of
measurements in Tennessee. J Geophys Res, 105,9189-9202.
Sillman S; He D-Y. (2002). Some theoretical results concerning 03-NOx-VOC chemistry and NOx-VOC indicators. J Geophys Res, 107.
Silvola J; Saarnio S; Foot J; Sundh I; Greenup A; Heijmans M; Ekberg A; Mitchell E; van Breeman N. (2003). Effects of elevated CO2 and N
deposition on CH4 emissions from European mires. Glob Biogeochem Cycles, 17,1068-1079.
Siminon HA; Kretser WA; Bath DW; Olson M; Gallagher J. (1993). In situ bioassays of brook trout (Saiveiinus fontinaiis) andblacknose dace
(,Rhinichthys atratuius) in Adirondack streams affected by episodic acidification. Can J For Res, 50, 902-912.
Simonsson M. (1998). Aluminium solubility related to secondary solid phases in upper B horizons with spodic characteristics. The European
Journal of Soil Sci, 49,317-326.
R-79
-------
Simpson AJ; Lam B; Diamond ML; Donaldson DJ; Lefebvre BA; Moser AQ; Williams AJ; Larin N; Kvasha MP. (2006). Assessing the organic
composition of urban surface films using nuclear magnetic resonance spectroscopy. Chemosphere, 63,142-152.
Simpson KW; Bode, RW; Colquhoun JR. (1985). The macroinvertebrate fauna of an acid-stressed headwater stream system in the Adirondack
Mountains, New York. Freshwater Biol, 15,671-681.
Singh HB; Herlth D; Kolyer R; Salas L; Bradshaw JD; Sandholm ST; Davis DD; Crawford J; Kondo Y; Koike M; Talbot R; Gregory GL; Sachse
GW; Browell E; Blake DR; Rowland FS; Newell R; Merrill J; Heikes B; Liu SC; Crutzen PJ; Kanakidou M. (1996). Reactive nitrogen
and ozone over the western Pacific: distribution, partitioning, and sources. J Geophys Res, 101,1793-1808.
Singh HB; Kasting JF. (1988). Chlorine-hydrocarbon photochemistry in the marine troposphere and lower stratosphere. J Atmos Chem , 7,261 -
285.
Singh HB; Salas L; Herlth D; Kolyer R; Czech E; Avery M; Crawford JH; Pierce RB; Sachse GW; Blake DR; Cohen RC; Bertram TH; Perring A;
Wooldridge PJ; Dibb J; Huey G; Hudman RC; Turquety S; Emmons LK; Flocke F; Tang Y; Carmichael GR; Horowitz LW. (2007).
Reactive nitrogen distribution and partitioning in the North American troposphere and lowermost stratosphere. J Geophys Res, 112,
doi: 10.1029/2006JD007664.
Sinsabaugh RL; Zak DR; Gallo M; Lauber C; Amonette R. (2004). Nitrogen deposition and dissolved organic carbon production in northern
temperate forests. Soil Biol Biochem, 36,1509-1515.
Sitaula BK; Bakken LR; Abrahamsen G. (1995). Nutrient balance in Scots pine (Pinus sylvestris L) forest. 3: Fluxes of N2O from lysimeter as
influenced by nitrogen input. Water Air Soil Pollut, 85,1155-1159.
Skeffington RA; Wilson EJ. (1988). Excess nitrogen deposition: Issues for consideration. Environ Pollut, 54,159-184.
Skiba U; Fowler D; Smith KA. (1997). Nitric oxide emissions from agricultural soils in temperate and tropical climates: sources, controls and
mitigation options. Nutr Cycling Agroecosyst, 48,139-153.
Skiba U; Sheppard LJ; Pitcairn CER; Van Dijk S; Rossall MJ. (1999). The effect of N deposition on nitrous oxide and nitric oxide emissions
from temperate forest soils. Water Air Soil Pollut, 116,89-98.
Skinner MW; Pavlik B. M. (Eds.) (1994). California Native Plant Society's inventory of rare and endangered vascular plants of California.
Sacramento, CA: California Native Plant Society.
Skjelkvale BL; Evans CD; Larssen T; Hindar A; Raddum GG. (2003). Recovery from acidification in European surface waters: A view to the
future. Ambio, 30,170-175.
Skjelkvale BL; Stoddard JL; Andersen T. (2001). Trends in Surface Water Acidification in Europe and North America (1989-1998). Water Air
Soil Pollut, 130,787-792.
Skjelkvale BL; Wright RF; Henriksen A. (1998). Norwegian lakes show widespread recovery from acidification; results from national surveys of
lake-water chemistry. Hydrol and Earth Sci, 2,555-562.
Skjelkvale BL; Andersen T; Halvorsen GA; Raddum GG; Heegaard E; Stoddard J; Wright RF. (2000). The 12-year report: acidification of
surface water in Europe and North America; trends, biological recovery and heavy metals.
http://www.niva.no/svmfoni/RappArkiv5.nsf/URL/C125730900463888C1256FB80053D51D/SFILE/4208 200dpi.pdf. (ICP Waters
report 52/2000). Oslo, Norway: Norwegian Institute for Water Research.
Skogheim OK; Rosseland BO. (1986). Mortality of smolt of Atlantic salmon Salmo-salar L., at low levels of aluminum in acidic softwater. Bull
Environ Contam Toxicol, 37, 258-265.
Skyllberg U. (1999). pH and solubility of aluminium in acidic forest soils: a consequence of reactions between organic acidity and aluminium
alkalinity. European Journal of Soil Science, 50,95-106.
Smidt S. (1994). Gefahrdung von Waldbaumen durch organische Luftschadstoffe [Endangering of forest trees by organic air pollutants], Z
Pflanzenkranks Pflanzenschutz, 101,423-445.
Smith AM; Keene WC; Maben JR; Pszenny AAP; Fischer E; Stohl A. (2007). Ammonia sources, transport, transformation, and deposition in
coastal New England during summer. J Geophys Res, 112,10.1029/2006JD007574.
Smith GD; Clark RJH. (2002). The role of H2S in pigment blackening. J Cult Heritage, 3,101-105.
Smith GF; Nicholas NS. (1999). Post disturbance spruce-fir forest stand dynamics at seven disjunct sites. Castanea, 64,175-186.
Smith LK; Voytek MA; Bohlke JK; Harvey JW. (2006a). Denitrification in nitrate-rich streams: Application of N2A and 15N-tracer methods in
intact cores. Ecol Appl, 16,2191-2207.
Smith LV; deVries KL. (1993). Mechanical properties of polymeric fibres exposed to stress in a NOx environment. Polymer, 34,546-550.
Smith MD; Crowder LB. (2005). Valuing Ecosystem Services with Fishery Rents: A Lumped-Parameter Approach to Hypoxia in the Neuse
River Estuary. (Fondazione Eni Enrico Mattei (FEEM).Nota di Lavoro (Natural Resources Management (NRM).Working Papers -
FEEM no 2005.115). http://ideas.repec.ora/p/fem/femwpa/2005.115.html. Milan, Italy; Fondazione Eni Enrico Mattei (FEEM).
Smith M-L; Ollinger SV; Martin ME; Aber JD; Hallett RA;Goodale CL. (2002). Direct estimation of aboveground forest productivity through
hyperspectral remote sensing of canopy nitrogen. Ecol Appl, 12,1286-1302.
Smith RA; Schwarz GE; Alexander RB. (1997). Regional interpretation of water-quality monitoring data. Water Resour Res, 33,2781-2798.
Smith SM; Lee KD. (2006). Responses ofperiphyton to artificial nutrient enrichment in freshwater kettle ponds of Cape Cod National Seashore.
Hydrobiologia, 571,201-211.
Smith VH; Tilman GD; Nekola JC. (1999). Eutrophication: impacts of excess nutrient inputs on freshwater, marine, and terrestrial ecosystems.
Environ Pollut, 100,179-196.
Smith VH. (2006). Responses of estuarine and coastal marine phytoplankton to nitrogen and phosphorus enrichment. Limnol Oceanogr, 51,
377-384.
Smith VK; Desvousges WH. (1986). Measuring Water Quality Benefits. Boston, MA: Kluwer-Nijhoff.
Smith VK; Palmquist RB; Jakus P. (1991). Combining Farrell frontier and hedonic travel cost models for valuing estuarine quality. Rev Econ
Stat, 73, 694-699.
R-80
-------
Smith VK; Huang J.-C. (1995). Can markets value air quality? A meta-analysis of hedonic property value models. J Polit Econ, 103, 209-27.
Smith VK; Pattanayak SK. (2002). Is meta-analysis a Noah's Ark for non-market valuation? Environ Resour Econ, 22,271-296.
Smith VK; Pattanayak S; Houtven GV. (2006b). Structural benefits transfer: An example using VSL estimates. Ecol Econ, 60,361-371.
Smith VK. (2007). Reflections on the Literature. Rev Environ Econ Policy, 1,152-165.
Smith WH. (1981). Air pollution and forests: Interactions between air contaminants and forest ecosystems: New York, NY: Springer-Verlag.
Smith WH. (1990). Forests as sinks for air contaminants: soil compartment. In: Smith WH (Ed.), Air pollution and forests: Interactions between
air contaminants and forest ecosystems. 2nd edition. (Springer series on Environmental management; pp 113-146). New York, NY:
Springer-Verlag.
Smith WH; Staskawicz BJ. (1977). Removal of atmospheric particles by leaves and twigs of urban trees: some preliminary observations and
assessment of research needs. Environmental Management 1:317-330.
Smock LA; Gazzera SB. (1996). Effects of experimental episodic acidification on a southeastern USA blackwater stream. Journal of Freshwater
Ecology, 11,81-90.
Smolders AJP; Roelofs JGM. (1996). The roles of internal iron hydroxide precipitation, sulphide toxicity and oxidizing ability in the survival of
Stratiotes aloides roots at different iron concentrations in sediment pore water. New Phytol, 133,253-260.
Smolders AJP; Lamers LPM; Moonen M; Zwaga K; Roelofs JGM. (2001). Controlling phosphate release from phosphate-enriched sediments
by adding various iron compounds. Biogeochemistry, 54,219-228.
Smolders AJP; Lamers LPM; den Hartog C; Roelofs JGM. (2003). Mechanisms involved in the decline of Stratiotes aloides L. The Netherlands:
sulphate as a key variable Hydrobiologia, 506,603-610.
Smolders AJP; Lamers LPM; Lucassen E; Van DerVelde G; Roelofs JGM. (2006). Internal eutrophication: How it works and what to do about
it: A review. Chemistry and Ecology, 22,93-111.
Sobczak WV; Findlay S; Dye S. (2003). Relationship between DOC variability and nitrate removal in and upland stream: An experimental
approach. Biogeochemistry, 62,309-327.
Soderstrom H; Hajslova J; Kocourek V; Siegmund B; Kocan A; Obiedzinski MW; Tysklind M; Bergqvist PA. (2005). PAHs and nitrated PAHs in
air of five European countries determined using SPMDs as passive samplers. Atmos Environ, 39,1627-1640.
Sollins P; Grier CC; McCorison FM; Cromack K Jr; Fogel R; Fredriksen RL. (1980). The internal element cycles of an old-growth Douglas-fir
ecosystem in western Oregon. Ecol Monogr, 50,261-285.
Solomon PA; Larson SM; Fall T; Cass G. (1988). Basinwide nitric acid and related species concentrations observed during the Claremont
Nitrogen Species Comparison Study. Atmos Environ, 22,1587-1594.
Solomon PA; Salmon LG; Fall T; Cass GR. (1992). Spatial and temporal distribution of atmospheric nitric acid and particulate nitrate
concentrations in the Los Angeles area. Environ Sci Technol, 26,1596-1601.
Solomon P; Cowling E; Hidy G; Furiness C. (2000). Comparison of scientific findings from major ozone field studies in North America and
Europe. Atmos Environ, 34,1885-1920.
Solomon PA; Hopke PK. (2008). A Special Issue of JA&WMA Supporting Key Scientific and Policy-and Health-Relevant Findings from EPA's
Particulate Matter Supersites Program and Related Studies: An Integration and Synthesis of Results. J Air Waste Manage Assoc, 58,
137.
Solomon S. (1999). Stratospheric ozone depletion: A review of concepts and history. Rev Geophys, 37, 275-316.
Solomon S. (2007). Climate Change 2007: The Physical Science Basis: Contribution of Working Group I to the Fourth Assessment Report of
the Intergovernmental Panel on Climate Change: Cambridge University Press.
Soltic P; Weilenmann M. (2003). NO2/NO emissions of gasoline passenger cars and light-duty trucks with Euro-2 emission standard. Atmos
Environ, 37, 5207-5216.
Soulsby C; Turnbull D; Langan SJ; Owen R; Hirst D. (1995). Long-term trends in stream chemistry and biology in northeast Scotland: evidence
for recovery. Water Air Soil Pollut, 85,689-694.
Soussana JF; Allard V; Pilegaard K; Ambus P; Amman C; Campbell C; Ceschia E; Clifton-Brown J; Czobel S; Domingues R; Flechard C;
Fuhrer J; Hensen A; Horvath L; Jones M; Kasper G; Martin C; Nagy Z; Neftel A; Raschi A; Baronti S; Rees RM; Skiba U; Stefani P;
Manca G; Sutton M; Tubaf Z; Valentini R. (2007). Full accounting of the greenhouse gas (CO2, N2O, CH4) budget of nine European
grassland sites. Agric Ecosyst Environ, 121,121-134.
Spalding MG; Forrester DJ. (1991). Effects of parasitism and disease on the nesting success of colonial wading birds (Ciconiiformes) in
southern Florida. (Project NG88-008). Tallahassee, FL: Prepared by Department of Infectious Diseases; College of Veterinary
Medicine; University of Florida: for Nongame Wildlife Program ; Florida Game and Fresh Water Fish Commission.
Sparks JP; Roberts JM; Monson RK. (2003). The uptake of gaseous organic nitrogen by leaves: a significant global nitrogen transfer process.
Geophys Res Lett, 30, doi: 10.1029/2003GL018578.
Spash CL. (1997). Assessing the economic benefits to agriculture from air pollution control. J Econ Surv, 11,47-70.
Spiecker H; Meielikainen K; Kohl M; Skovsgaard J. (1996). Growth trends in European forests: Studies from 12 countries. Report 5. Springer-
Verlag, Berlin and New York: European Forest Research Institute.
Spoelstra J; Schiff SL; Elgood RJ; Semkin RG; Jeffries DS. (2001). Tracing the sources of exported nitrate in the Turkey Lakes watershed
using 15N/14N and 180/160 isotopic ratios. Ecosystems, 4,536-544.
Spurny KR. (1989). On the release of asbestos fibers from weathered and corroded asbestos cement products. Environ Res, 48,100-116.
Spurny KR. (1999). Trends in the chemical analysis of aerosols. In: Spurny KR (Ed.), Analytical chemistry of aerosols. Boca Raton, FL: Lewis
Publishers.
St. Clair SB; Carlson JE; Lynch JP. (2005). Evidence for oxidative stress in sugar maple stands growing on acidic, nutrient imbalanced forest
soils. Oecologia, 145,258-269.
R-81
-------
St. Louis VL; Breebaart L; Barlow JC. (1990). Foraging behaviour of tree swallows over acidified and nonacidic lakes. Can J Zool, 68,2385-
2392.
St. Louis VL; Rudd JWM; Kelly CA; Beaty KG; Bloom NS; Flett RJ. (1994). Importance of wetlands as sources of methyl mercury to boreal
forest ecosystems. Can J Fish Aquat Sci, 51,1065-1076.
St. Louis VL; Rudd JWM; Kelly CA; Beaty KG; Flett RJ; Roulet NT. (1996). Production and loss of methylmercury and loss of total mercury from
boreal forest catchments containing different types of wetlands. Environ Sci Technol, 30,2719 -2729.
Stacey PR; Wright MD. (2001). A proficiency testing scheme for aromatic hydrocarbons in air by the manual thermal desorption-GC method: A
comparison of laboratory performance with the uncertainty requirements of the European Union Ambient Air Directive. Journal of
Environmental Monitoring, 3,425-431.
Stacey PE; Greening JS; Kremer JN; Peterson D; Tomasko DA. (2001). Contribution of atmospheric nitrogen deposition to U.S. estuaries:
Summary and conclusions. In: Valigura RW; Alexander RB; Castro MS; Meyers TP; Paerl HW; Stacey PE; Turner RE., (Eds), Nitrogen
loading in coastal water bodies: An atmospheric perspective. Washington, DC: American Geophysical Union; 187-226.
Staffelbach T; Neftel A; Blatter A; Gut A; Fahrni M; Stahelin J; Prevot A; Hering A; Lehning M; Neininger B; Baumle M; Kok GL; Dommen J;
Hutterli M; Anklin M. (1997). Photochemical oxidant formation over southern Switzerland 1. results from summer 1994. J Geophys
Res, 102, 23345-23362.
Stanford G; Dzienia S; Vander Pol RA. (1975). Effect of temperature on denitrification rate in soils. Soil Science Society Amer Proc, 39,867-
870.
Stange F; Butterbach-Bahl K; Papen H; Zechmeister-Boltenstern S; Li C; Aber J. (2000). Composition and Chemistry-A process-oriented model
of N2O and NO emissions from forest soils, 2, Sensitivity analysis and validation (Paper 1999JD900948). J Geophys Res-Part D-
Atmospheres, 105,4385-4398.
Stanley EH; Short RA; Harrison JW; Hall Rl; Wiedenfeld RC. (1990). Variation in nutrient limitation of lotic and lentic algal communities in a
Texas (USA) river. Hydrobiologia, 206,61-71.
Staudt AC; Jacob DJ; Ravetta F; Logan JA; Bachiochi D; Krishnamurti TN; Sandholm ST; Ridley BA; Singh HB; Talbot B. (2003). Sources and
chemistry of nitrogen oxides over the tropical Pacific. J Geophys Res, 1088239.
Stauffer RE. (1990). Granite weathering and the sensitivity of alpine lakes to acid deposition. Limnol Oceanogr, 35(5), 1112-1134.
Stauffer RE; Wittchen BD. (1991). Effects of silicate weathering on water chemistry in forested, upland, felsic terrain of the USA. Geochim
Cosmochim Acta, 55, 3253-3271.
Stedman DH; Bishop G; Peterson JE; Guenther PL. (1991). On-road CO remote sensing in the Los Angeles Basin. (Final report). Sacramento,
CA; California Air Resources Board (ARB).
Stehr JW; Dickerson RR; Hallock-Waters KA; Doddridge BG; Kirk D. (2000). Observations of NOy , CO, and S02 and the origin of reactive
nitrogen in the eastern United States. J Geophys Res, 105,3553-3563.
Stein AF; Lamb D. (2003). Empirical evidence for the low- and high-NOx photochemical regimes of sulfate and nitrate formation. Atmos
Environ, 37, 3615-3625.
Stein AF; Lamb D; Draxler RR. (2000). Incorporation of detailed chemistry into a three-dimensional Lagrangian-Eulerian hybrid model:
application to regional fropospheric ozone. Atmos Environ, 34,4361-4372.
Steinbacher M; Zellweger C; Schwarzenbach B; Bugmann S; Buchmann B; Ordonez C; Prevot ASH; Hueglin C. (2007). Nitrogen oxide
measurements at rural sites in Switzerland: Bias of conventional measurement techniques. J Geophys Res, 112,1-13.
Steltzer H; Bowman WD. (1998). Differential influence of plant species on soil N transformations within the alpine tundra. Ecosystems, 1,464-
474.
Stemmler K; Ammann M; Donders C; Kleffmann J; George C. (2006). Photosensitized reduction of nitrogen dioxide on humic acid as a source
of nitrous acid. Nature, 440,195-198.
Stenchikov G; Dickerson R; Pickering K; Ellis WJ; Doddridge B; Kondragunta S; Poulida O; Scala J; Tao W-K. (1996). Stratosphere-
troposphere exchange in a midlatitude mesoscale convective complex. 2. Numerical simulation. J Geophys Res, 101, 6837-6851.
Sterner RW; Elser JJ. (2002). Ecological stoichiometry: the biology of elements from molecules to the biosphere. Princeton, NJ: Princeton
University Press.
Steudler PA; Bowden RD; Mellilo JM; Aber JD. (1989). Influence of nitrogen fertilization on methane uptake in temperate forest soil. Nature,
341,314-316.
Stevens CJ; Dise NB; Mountford OJ; Gowing DJ. (2004). Impact of nitrogen deposition on the species richness of grasslands. Science, 303,
1876-1878.
Stevens CJ; Laughlin RJ; Malone JP. (1997). Measuring the contributions of nitrification and denitrification to the flux of nitrous oxide from soil.
Soil Biol Biogeochem, 29,139-151.
Stevens RK; Mulik JD; O'Keefe AE; Krost KJ. (1971). Gas chromatography of reactive sulfur gases in air at the parts-per-billion level. Anal
Chem, 43, 827-831.
Stevens RK; O'Keeffe AE; Ortman GC. (1969). Absolute calibration of a flame photometric detector to volatile sulfur compounds at sub-part-
per-million levels. Environ Sci Technol, 3,652-655.
Stevenson DS; Dentener FJ; Schultz MG; Ellingsen K; van Noije TPC; Wild O; Zeng G; Amann M; Atherton CS; Bell N; Bergmann DJ; Bey I;
Butler T; Cofala J; Collins WJ; Derwent RG; Doherty RM; Drevet J; Eskes HJ; Fiore AM; Gauss M; Hauglustaine DA; Horowitz LW;
Isaksen ISA; Krol MC; Lamarque JF; Lawrence MG; Montanaro V; Mueller JF; Pitari G; Prather MJ; Pyle JA; Rast S; Rodriguez JM;
Savage NH; Shindell DT; Strahan SE; Sudo K; Szopa S. (2006). Multimodel ensemble simulations of present-day and near-future
tropospheric ozone. J Geophys Res, 111, doi: 10.1029/2005JD006338.
Stevenson FJ (Ed.). (1982). Origin and distribution of nitrogen in soil. Madison, Wl: American Society of Agronomy.
R-82
-------
Steyn DJ; Bottenheim JW; Thomson RB. (1997). Overview of tropospheric ozone in the Lower Fraser Valley, and the Pacific '93 field study.
Atmos Environ, 31,2025-2036.
Stienen H; Bauch J. (1988). Element content in tissues of spruce seedlings from hydroponic cultures simulating acidification and
deacidification. Plant Soil, 106,231-238.
Stith J; Dye J; Ridley B; Laroche P; Defer E; Baumann K; Hubler G; Zerr R; Venticinque M. (1999). NO signatures from lightning flashes. J
Geophys Res, 101,16081-16089.
Stitt M; Krapp A. (1999). The interaction between elevated carbon dioxide and nitrogen nutrition: the physiological and molecular background.
Plant Cell Environ, 22, 583-621.
Stockwell W; Middleton P; Chang J; Tang X. (1990). The second generation Regional Acid Deposition Model chemical mechanism for regional
air quality modeling. J Geophys Res, 95,16343-16367.
Stoddard JL. (1990). Plan for converting the NAPAP aquatic effects long-term monitoring (LTM) project to the temporally integrated monitoring
of ecosystems (TIME) project. (International Report). Corvallis, OR; U.S. Environmental Protection Agency.
Stoddard JL. (1991). Trends in Catskill stream water quality: evidence from historical data. Water Resour Res, 27,2855-2864.
Stoddard JL; Murdoch PS. (1991). Catskill Mountains: An overview of the impact of acidifying pollutants on aquatic resources. In: Charles DF
(Ed.), Acidic deposition and aquatic ecosystems: Regional case studies, (pp 237-271). New York: Springer-Verlag, Inc.
Stoddard JL; Kellogg JH. (1993). Trends and patterns in lake acidification in the state of Vermont: Evidence from the Long-Term Monitoring
project. Water Air Soil Pollut, 67,301-317.
Stoddard JL. (1994). Long-term changes in watershed retention of nitrogen: its causes and aquatic consequences. In Baker LA (Ed.),
Environmental chemistry of lakes and reservoirs, (pp. 223-284). Washington, D.C.: American Chemical Society.
Stoddard JL. (1995). Episodic acidification during snowmelt of high elevation lakes in the Sierra Nevada mountains of California. Water Air Soil
Pollut, 85, 353-358.
Stoddard JL; Urquhart NS; Newell AD; Kugler D. (1996). The Temporally Interated Monitoring of Ecosystems (TIME) project design 2.
Detection of regional acidification trends. Water Resour Res, 32,2529-2538.
Stoddard JL; Driscoll CT; Kahl S; Kellogg J. (1998a). A regional analysis of lake acidification trends in the Northeastern U.S., 1982-1994.
Environ Monit Assess, 51,399-413.
Stoddard JL; Driscoll CT; Kahl S; Kellogg J. (1998b). Can site-specific trends be extrapolated to the regional level? A lake acidification example
for the northeastern U.S. Ecol Appl, 8,288-299.
Stoddard JL; Jeffries DS; Lukewille A; Clair TA; Dillon PJ; Driscoll CT; Forsius M; Johannessen M; Kahl JS; Kellogg JH; Kemp A; Mannio J;
Monteith DT; Murdoch PS; Patrick S; Rebsdorf A; Skjelkvale BL; Stainton MP; Traaen T; Van Dam H; Webster KE; Wieting J;
Wilander A. (1999). Regional trends in aquatic recovery from acidification in North American and Europe. Nature, 401,575-578.
Stoddard J; Kahl JS; Deviney FA; DeWalle DR; Driscoll CT; Herlihy AT; Kellogg JH; Murdoch PS; Webb JR; Webster KE. (2003). Response of
surface water chemistry to the Clean Air Act Amendments of 1990 (No. EPA620/R-03/001). Research Triangle Park, NC; National
Health and Environmental Effects Research Laboratory; Office of Research and Development; U.S. Environmental Protection Agency.
Stoermer EF; Smol JP (Eds.). (1999). The diatoms: Applications for the environmental and earth sciences. New York, NY: Cambridge
University Press.
Stohlgren TJ; Binkley D; Chong GW; Kalkhan MA; Schell LD; Bull KA; Otsuki Y; Newman G; Bashkin M; Son Y. (1999). Exotic plant species
invade hot spots of native plant diversity. Ecol Monogr, 69, 25-46.
Stoklasa J. (1923). Die Beschadigungen der Vegetation durch Rauchgase und Fabriksexhalation [Damage to vegetation from flue-gas and
factory emissions], (pp 101-131). Berlin, Germany: Urban and Schwarzenberg.
Stolte W; McCollin T; Noordeloos A; Riegman R. (1994). Effects of nitrogen source on the size distribution within marine phytoplankton
populations. J Exp Mar Bio Ecol, 184,83-97.
Stolzenburg MR; Hering SV. (2000). Method for the automated measurement of fine particle nitrate in the atmosphere. Environ Sci Technol, 34,
907-914.
Stottlemyer R; Troendle CA. (1992). Nutrient concentration pattern in streams draining alpine and subalpine catchments, Fraser Experimental
Forest, Colorado. J Hydrol, 140,179-208.
Stottlemyer R; Troendle CA; Markowitz D. (1997). Changes in snowpack, soil water, and stream water chemistry with elevation during 1990,
Fraser Experimental Forest, Colorado. J Hydrol, 195,114-136.
Strain BR; Bazzaz FA. (1983). Terrestrial plant communities. In: Lemon ER (Ed.), CO2 and plants, (pp 177-222). Boulder, CO: Westview.
Strandberg H. (1998). Reactions of copper patina compounds—I. Influence of some air pollutants. Atmos Environ, 32,3511-3520.
Streets DG; Bond TC; Carmichael GR; Fernandes SD; Fu Q; He D; Klimont Z; Nelson SM; Tsai NY; Wang MQ; Woo JH; Yarber KF. (2003). An
inventory of gaseous and primary aerosol emissions in Asia in the year 2000. J Geophys Res-Atmos, 108.
Strengbom J; Nordin A; Nasholm T; Ericson L. (2001). Slow recovery of boreal forest ecosystem following decreased nitrogen input. Functional
Ecology, 15, 451-457.
Strengbom J; Walheim M; Nasholm T; Ericson L. (2003). Regional differences in the occurrence of understorey species reflect nitrogen
deposition in Swedish forests. Ambio, 32,91-97.
Stribley DP; Read DJ. (1980). The biology of mycorrhiza in the Ericacae. VII. The relationship between mycorrhizal infection and the capacity to
utilize simple and complex organic nitrogen sources. New Phytol, 86,365-371.
Strickland TC; Holdren GRJ; Ringold PL; Bernard D; Smythe K; Fallon W. (1993). A national critical loads framework for atmospheric
deposition effects assessment: I. method summary. Environ Manage, 17,329-334.
Stoklasa J. (1923). Die Beschadigungen der Vegetation durch Rauchgase und Fabriksexhalation [Damage to vegetation from flue-gas and
factory emissions], (pp 101-131). Berlin, Germany: Urban and Schwarzenberg.
R-83
-------
Stursova M; Crenshaw CL; Sinsabaugh RL. (2006). Microbial Responses to Long-Term N Deposition in a Semiarid Grassland. Microbial
Ecology, 51,90-98.
Stutz J; Hebestreit K; Alicke B; Piatt U. (1999). Chemistry of halogen ozides in the troposphere: Comparison of model calculations with recent
field data. J Atmos Chem , 34,65-85.
Stutz J; Kim ES; Piatt U; Bruno P; Perrino C; Febo A. (2000). UV-visible absorption cross sections of nitrous acid. J Geophys Res, 105,14585-
14592.
Stutz J; Ackermann R; Fast JD; Barrie L. (2002). Atmospheric reactive chlorine and bromine at the Great Salt Lake, Utah. Geophys Res Lett,
29, doi: 10.1029/2002GL014812.
Stutz J; Alicke B; Ackermann R; Geyer A; Wang S; White AB; Williams EJ; Spicer CW; Fast JD. (2004a). Relative humidity dependence of
HONO chemistry in urban areas. J Geophys Res, 109, doi: 10.1029/2003JD004135.
Stutz, J; Alicke, B; Ackermann, R; Geyer, A; White, A; Williams, E. (2004).Vertical profiles of NO3, N2O5,03, and NOx in the nocturnal boundary
layer: 1. Observations during the Texas Air Quality Study 2000. J Geophys Res, 109, doi: 10.1029/2003JD004209.
Sucoff El; Thornton FC; Joslin JD. (1990). Sensitivity of tree seedlings to aluminum: I. Honeylocust. J Environ Qual, 19,163-171.
Suding KN; Collins SL; Gough L; Clark C; Cleland EE; Gross KL; Milchunas DG; Pennings S. (2005). Functional- and abundance-based
mechanisms explain diversity loss due to N fertlization. Proc Natl Acad Sci USA, 102,4387-4392.
Suding KN; Miller AE; Bechtold H; Bowman WD. (2006). The consequence of species loss on ecosystem nitrogen cycling depends on
community compensation. Ocecologia, 149,141-149.
Suh HH; Spengler JD; Koutrakis P. (1992). Personal exposures to acid aerosols and ammonia. Environ Sci Technol, 26,2507-2517.
Suh HH; Koutrakis P; Spengler JD. (1994). The relationship between airborne acidity and ammonia in indoor environments. Journal of
Exposure Science and Environmental Epidemiology, 4,1-23.
Sullivan TJ; Seip HM; Muniz IP. (1986). A comparison of frequently used methods for the determination of aqueous aluminum. Intern J Environ
Anal Chem, 26,61-75.
Sullivan TJ; Eilers JM; Church MR; Blick DJ; Eshleman KN; Landers DH; DeHaan MS. (1988). Atmospheric wet sulphate deposition and
lakewater chemistry. Nature, 331,607-609.
Sullivan TJ; Driscoll CT; Gherini SA; Munson RK; Cook RB; Charles DF; Yatsko CP. (1989). Influence of aqueous aluminium and organic acids
on measurement of acid neutralizing capacity in surface waters. Nature, 338,408-410.
Sullivan TJ. (1990). Historical changes in surface water acid-base chemistry in response to acidic deposition. (State of the Science Report no
SOS/T 11). Washington, DC; National Acid Precipitation Assessment Program (NAPAP).
Sullivan TJ; Charles DF; Smol JP; Cumming BF; Selle AR; Thomas DR; Bernert JA; Dixit SS. (1990a). Quantification of changes in lakewater
chemistry in response to acidic deposition. Nature, 345,54-58.
Sullivan TJ; Kugler DL; Small MJ; Johnson CB; Landers DH; Rosenbaum BJ; Overton WS; Kretser WA; Gallagher J. (1990b). Variation in
Adirondack, New York, lakewater chemistry as a function of surface area. Water Resour Bull, 26,167-176.
Sullivan TJ; Turner RS; Charles DF; Cumming BF; Smol JP; Schofield CL; Driscoll CT; Cosby BJ; Birks HJ; Uutala AJ; Kingston JC; Dixit SS;
Bernert JA; Ryan PF; MarmorekDR. (1992). Use of historical assessment for evaluation of process-based model projections of future
environmental change: Lake acidification in the Adirondack mountains, New York, USA. Environ Pollut, 77,253-262.
Sullivan TJ. (1993). Whole ecosystem nitrogen effects research in Europe. Environ Sci Technol, 27,1482-1486.
Sullivan TJ; Charles DF. (1994). The feasibility and utility of a paleolimnology/paleoclimate data cooperative for North America. J Paleolimnol,
10,265-273.
Sullivan TJ; Eilers JM. (1994). Assessment of deposition levels of sulfur and nitrogen required to protect aquatic resources in selected sensitive
regions of North America (Final Report no PB-97-109441/XAB prepared by E&S Environmental Chemistry, Inc; prepared for Technical
Resources, Inc. and U.S. Environmental Protection Agency). Corvallis, OR; Environmental Research Laboratory; U.S. Environmental
Protection Agency.
Sullivan TJ; Cosby BJ; Driscoll CT; Charles DF; Hemond HF. (1996a). Influence of organic acids on model projections of lake acidification.
Water Air Soil Pollut, 91, 271-282.
Sullivan TJ; McMartin B; Charles DF. (1996b). Re-examination of the role of landscape change in the acidification of lakes in the Adirondack
Mountains, New York. Sci. Total Environ, 183, 231-248.
Sullivan TJ; Eilers JM; Cosby BJ; Vache KB. (1997). Increasing role of nitrogen in the acidification of surface waters in the Adirondack
Mountains, New York. Water Air Soil Pollut, 95,313-336.
Sullivan TJ; Cosby BJ. (1998). Modeling the concentration of aluminium in surface waters. Water Air Soil Pollut, 105,643-659.
Sullivan TJ; Cosby BJ; Bernert JA; Eilers JM. (1998). Model evaluation of dose/response relationships and critical loads for nitrogen and sulfur
deposition to the watersheds of Lower Saddlebag and White Dome Lakes. Report No 97-10-01 to USDA Forest Service E&S
Environmental Chemistry Inc, Corvallis, OR.
Sullivan TJ. (2000a). Aquatic effects of acidic deposition. Boca Raton, FL: Lewis Publishers.
Sullivan TJ. (2000b). Chemical dose-response relationships and critical loads. In: Sullivan, TJ, Aquatic Effects of Acidic Deposition. (Chapter
5). Boca Raton, FL: CRC Press.
Sullivan TJ; Peterson DL; BlanchardCL; Tannenbaum SJ. (2001). Assessment of air quality and air pollutant impacts in class I national parks
of California. Report no NPS/CCSOUW/NRTR-00/xx). Denver, CO; Air Resources Division; National Park Service; U.S. Department of
the Interior.
Sullivan TJ; Cosby BJ; Webb JR; Snyder KU; Herlihy AT; Bulger AJ; Gilbert EH; Moore D. (2002a). Assessment of the effects of acidic
deposition on aquatic resources in the Southern Appalachian Mountains. (Report Prepared by E&S Environmental Chemistry, Inc.
Corvallis, Oregon), http://www.esenvironmental.com/sami download.htm. Asheville, NC; Southern Appalachian Mountains Initiative
(SAM I).
R-84
-------
Sullivan TJ; Johnson DW; Munson RK; Joslin JD. (2002b). Assessment of the effects of acidic deposition on forest resources in the Southern
Appalachian Mountains. Asheville, NC; Southern Appalachian Mountains Initiative (SAMI).
Sullivan TJ; Cosby BJ; Laurence JA; Dennis RL; Savig K; Webb JR; Bulger AJ; Scruggs M; Gordon C; Ray J; Lee H; HogsettWE; Wayne H;
Miller D; Kern JS. (2003).Assessment of air quality and related values in Shenandoah National Park. (Technical report no
NPS/NERCHAL/NRTR-03/090). Philadelphia, PA; Northeast Region; National Park Service; U.S. Department of the Interior.
Sullivan TJ; Cosby BJ; Herlihy AT; Webb JR; Bulger AJ; Snyder KU; Brewer P; Gilbert EH; Moore DL. (2004). Regional model projections of
future effects of sulfur and nitrogen deposition on streams in the southern Appalachian Mountains. Water Resour Res, 40, W02101
doi:02110.01029/02003WR001998.
Sullivan TJ; Cosby BJ. (2005). Testing, improvement, and confirmation of a watershed model of acid-base chemistry. Water Air Soil Pollut, 85,
2607-2612.
Sullivan TJ; Cosby BJ; Tonnesen KA; ClowDW. (2005). Surface water acidification responses and critical loads of sulfur and nitrogen
deposition in Loch Vale Watershed, Colorado. Water Resour Res, 41, doi: 10.1029/2004WR003414.
Sullivan TJ; Fernandez IJ; Herlihy AT; Driscoll CT; McDonnell TC; Nowicki NA; Snyder KU; Sutherland JW. (2006a). Acid-base characteristics
of soils in the Adirondack Mountains, New York. Soil Sci Soc Am J, 70,141-152.
Sullivan TJ; Driscoll CT; Cosby BJ; Fernandez IJ; Herlihy AT; Zhai J; Stemberger R; Snyder KU; Sutherland JW; Nierzwicki-Bauer SA; Boylen
CW; McDonnell TC; Nowicki NA. (2006b). Assessment of the extent to which intensively studied lakes are representative of the
Adirondack Mountain region. (Final Report no 06-17).Corvallis, OR; prepared by Environmental Chemistry, Inc. for: Albany, NY;
Environmental Monitoring Evaluation and Protection Program of the New York State Energy Research and Development Authority
(NYSERDA).
Sullivan TJ; Cosby BJ; Herlihy AT; Driscoll CT; Fernandez I J; McDonnell TC; Boylen CW; Nierzwicki-Bauer SA; Snyder KU. (2007a).
Assessment of the extent to which intensively-studied lakes are representative of the Adirondack region and response to future
changes in acidic deposition. Water Air Soil Pollut, 185,279-291.
Sullivan TJ; Webb JR; Snyder KU; Herlihy AT; Cosby BJ. (2007b). Spatial distribution of acid-sensitive and acid-impacted streams in relation to
watershed features in the southern Appalachian mountains. Water Air Soil Pollut, 182,57-71.
Sullivan TJ; Cosby BJ; Webb JR; Dennis RL; Bulger AJ; Deviney JR., FA. (2007c). Streamwater acid-base chemistry and critical loads of
atmospheric sulfur deposition in Shenandoah National Park, Virginia. Environ Monit Assess, doi: 10.1007/s10661-007-9731-1.
Sullivan TJ; Cosby BJ; Webb JR; Dennis RL; Bulger AJ; Deviney FA Jr. (2008). Streamwater acid-base chemistry and critical loads of
atmospheric sulfur deposition in Shenandoah National Park, Virginia. Environ Monit Assess, 137,85-99.
Sundberg B; Nasholm T; Palmqvist K. (2001). The effect of nitrogen on growth and key thallus components in the two tripartite lichens,
Nephroma arcticum and Peltigera aphthosa. Plant, Cell and Environment, 24, 517-527.
Suns K; Hitchin G. (1990). Interrelationships between mercury levels in yearling yellow perch, fish condition and water quality. Water Air Soil
Pollut, 50, 255-265.
Sutton MA; Erisman JW; Dentener F; Moller D. (2008a). Ammonia in the environment: From ancient times to the present. Environ Pollut.
(Corrected proof; doi: 10:1016/j.envpol2008.03.013).
Sutton MA; Simpson D; Levy PE; Smith Rl; Reis S; van Oijen M; de Vries W. (2008b). Uncertainties in the relationship between atmospheric
nitrogen deposition and forest carbon sequestration. Glob Change Biol, 14,2057-2063.
Sutton MA; Dragosits U; Tang YS; Fowler D. (2000). Ammonia emissions from non-agricultural sources in the UK. Atmos Environ, 34,855-869.
Sutton RT; Hodson DLR. (2005). Atlantic Ocean forcing of North American and European summer climate. Science, 309,115-118.
Svensson J-E; Johansson L-G. (1993a). A laboratory study of the effect of ozone, nitrogen dioxide, and sulfur dioxide on the atmospheric
corrosion of zinc. J Electrochem Soc, 140,2210-2216.
Svensson J-E; Johansson L-G. (1993b). A laboratory study of the initial stages of the atmospheric corrosion of zinc in the presence of NaCI;
influence of SO2 and NO2. Corros Sci, 34,721-740.
Svensson JE; Johansson LG. (1996). The temperature-dependence of the S02-induced atmospheric corrosion of zinc; a laboratory study.
Corros Sci, 38, 2225-2233.
Svensson R; Ljungstroem E; Lindqvist O. (1987). Kinetics of the reaction between nitrogen dioxide and water vapour. Atmos Environ, 21,1529-
1539.
Sverdrup H; Warfvinge P. (1988). Weathering of primary silicate minerals in the natural soil environment in relation to a chemical weathering
model. Water Air Soil Pollut, 38,387-408.
Sverdrup H; de Vries W; Henriksen A. (1990). Mapping critical loads: a guidance to the criteria, calculations, data collection and mapping of
critical loads. (Environmental report 1990:14; NORD 1990:98). Copenhagen, Denmark: Nordic Council of Ministers.
Sverdrup H; Warfvinge P; Rabenhorst M; Janicki A; Morgan R; Bowman M. (1992). Critical loads and steady-state chemistry for streams in the
state of Maryland. Environ Pollut, 77,195-203.
Sverdrup H; Warfvinge P. (1993). The effect of soil acidification on the growth of trees, grass and herbs as expressed by the (Ca+ Mg+ K)/AI
ratio. Rep in Ecol & Eng, 2,1993.
Sverdrup H; de Vries W. (1994). Calculating critical loads for acidity with the simple mass balance method. Water Air Soil Pollut, 72,143-162.
Swackhamer DL; Paerl HW; Eisenreich SJ; Hurley J; Hornbuckle KC; McLachlan M; Mount D; Muir D; Schindler D. (2004). Impacts of
atmospheric pollutants on aquatic ecosystems. Issues in Ecology, 12,1-24.
Swain RE. (1949). Smoke fume investigations: A historical review. Ind Eng Chem, 41,2384-2388.
Swank WT. (1988). Stream chemistry responses to disturbance. In: Swank WT; Crossley DS (Eds.), Forest hydrology and ecology at Coweeta.
(pp 339-358). New York, NY: Springer-Verlag, Inc.
Swank WT; Vose JM. (1997). Long-term nitrogen dynamics of Coweeta forested watersheds in the southeastern United States of America.
Glob Biogeochem Cycles, 11,657-671.
R-85
-------
Sweets PR. (1992). Diatom paleolimnological evidence for lake acidification in the Trail Ridge region of Florida. Water Air Soil Pollut, 65,43-57.
Sweets PR; Bienert RW; Crisman TL; Binford MW. (1990). Paleoecological investigations of recent lake acidification in northern Florida. J
Paleolimnol, 4,103-137.
Swenson WA; McCormick JH; Simonson TD; Jensen KM; Eaton JG. (1989). Experimental acidification of Little Rock Lake (Wisconsin): Fish
research approach and early responses. Arch. Environ Contam Toxicol, 18,167-174.
Swietlik D; Faust M. (1984). Foliar nutrition of fruit crops. Horticulture Review 6:287-355.
Takemoto BK; Bytnerowicz A; Fenn ME. (2001). Current and future effects of ozone and atmospheric nitrogen deposition on California's mixed
conifer forests. For Ecol Manage, 144,159-173.
Talbot RW; Vijgen AS; Harriss RC. (1990). Measuring tropospheric HNO3: Problems and prospects for nylon filter and mist chamber
techniques. J Geophys Res, 95,7553-7561.
Tamm CO; Aronsson A; Popovic B. (1995). Nitrogen saturation in a long-term forest experiment with annual additions of nitrogen. In: Grennfelt
P; Rodhe H; Thornelof, E; Wisniewski J. (Eds.), Acid Reign '95? Proceedings from the 5th International Conference on Acidic
Deposition: Science & Policy, held in Goteborg, Sweden, June 26-30,1995. Water Air Soil Pollut, 85,1683-1688.
Tamm CO; Aronsson A; Popovic B; Flower-Ellis J. (1999). Optimum nutrition and nitrogen saturation in Scots pine stands. (Book Series: Studia
forestalia Suecica - Technical report no 206). Uppsala, Sweden; Department for Production Ecology; Faculty of Forest Services;
Swedish University of Agricultural Sciences.
Tanaka PL; Riemer DD; Chang S; Yarwood G; McDonald-Buller EC; Apel EC; Orlando JJ; Silva PJ; Jimenez JL; Canagaratna MR; Neece JD;
Mullins CB; Allen DT. (2003). Direct evidence for chlorine-enhanced urban ozone formation in Houston, Texas. Atmos Environ, 37,
1393-1400.
Tang IN; Munkelwitz HR. (1993). Composition and temperature dependence of the deliquescence properties of hygroscopic aerosols. ATMOS
ENVIRON PART A GER TOP, 27,467-473.
Tang IN; Tridico AC; Fung KH. (1997). Thermodynamic and optical properties of sea salt aerosols. J Geophys Res, 102,269-223.
Tank JL; Dodds WK. (2003). Nutrient limitation of epilithic and epixylic biofilms in ten North American streams. Freshw Biol, 48,1031-1049.
Tanner RL; D'Ottavio T; Garber R; Newman L. (1980). Determination of ambient aerosol sulfur using a continuous flame photometric detection
system. I. Sampling system for aerosol sulfate and sulfuric acid. Atmos Environ, 14,121-127.
Tanner RL; Kelly TJ; Dezaro DA; Forrest J. (1989). A comparison of filter, denuder, and real-time chemiluminescence techniques for nitric acid
determination in ambient air. Atmos Environ (1967), 23,2213-2222.
Tao W-K; Simpson J. (1993). The Goddard Cumulus Ensemble Model. Part I: model description. Terr Atmos Oceanic Sci (TAO), 4,35-71.
Tarnay L; Gertler AW; Blank RR; Taylor GE Jr. (2001). Preliminary measurements of summer nitric acid and ammonia concentrations in the
Lake Tahoe Basin air-shed: implications for dry deposition of atmospheric nitrogen. Environ Pollut, 113,145-153.
Taylor GE Jr; Hanson PJ; Baldocchi DD. (1988). Pollutant deposition to individual leaves and plant canopies: sites of regulation and
relationship to injury. In: Heck WW; Taylor OC; Tingey DT (Eds.), Assessment of crop loss from air pollutants, (pp 227-257). New
York, NY: Elsevier Applied Science.
Taylor GE Jr; Hanson PJ. (1992). Forest trees and tropospheric ozone: role of canopy deposition and leaf uptake in developing exposure-
response relationships. Agric Ecosyst Environ, 42,255-273.
Taylor GE Jr; Pitelka LF. (1992). Genetic diversity of plant populations and the role of air pollution. In: Barker JR; Tingey DT (Eds.), Air pollution
effects on biodiversity, (pp 111-130). New York, NY: Van Nostrand Reinhold.
Taylor OC; MacLean DC. (1970). Nitrogen oxides and the peroxyacyl nitrates. In: Jacobson JS; Hill AC (Eds.), Recognition of air pollution injury
to vegetation: A pictorial atlas, (pp E1-E14). Pittsburgh, PA: Air Pollution Control Association.
Teklemariam TA; Sparks JP. (2004). Gaseous fluxes of peroxyacetyl nitrate (PAN).into plant leaves. Plant Cell and Environment, 27,1149-
1158.
Teklemariam TA; Sparks JP. (2006). Leaf fluxes of NO and NO2 in four herbaceous plant species: the role of ascorbic acid. Atmos Environ, 40,
2235-2244.
Temple PJ; Riechers GH; Miller PR; LenNOx RW. (1992). Growth responses of ponderosa pine to long-term exposure to ozone, wet and dry
acidic deposition, and drought. Can J For Res, 23,59-66.
Temple PJ; Taylor OC. (1983). World-wide ambient measurements of peroxyacetyl nitrate (PAN).and implications for plant injury. Atmos
Environ, 17,1583-1587.
Temple PJ; Riechers GH; Miller PR. (1992). Foliar injury responses of ponderosa pine seedlings to ozone, wet and dry acidic deposition, and
drought. Environ Exp Bot, 32,101-113.
Tempter PH; Lovett GM; Weathers KC; Findlay SE; Dawson TE. (2005). Influence of tree species on forest nitrogen retention in the Catskill
Mountains, New York, USA. Ecosystems, 8,1-16.
ter Braak CJF; Wiertz J. (1994). On the statistical analysis of vegetation change: a wetland affected by water extraction and soil acidification.
Journal of Vegetation Science, 5, 361-372.
Tester PA; Varnam SM; Culver ME; Eslinger DL; Stumpf RP; Swift RN; Yungel JK; Black MD; Litaker RW. (2003). Airborne detection of
ecosystem responses to an extreme event: Phytoplankton displacement and abundance after hurricane induced flooding in the
Albemarle-Pamlico Sound System, North Carolina. Estuaries, 26,1353-1364.
Theodose TA; Bowman WD. (1997). Nutrient availability, plant abundance, and species diversity in two alpine tundra communities. Ecology,
78,1861-1872.
Thielman A; Prevot ASH; Griiebler FC; Staelhelin J. (2001). Empirical ozone isopleths as a tool to identify ozone production regimes. Geophys
Res Lett, 28, 2369-2372.
Thoene B; Schroder P; Papen H; Egger A; Rennenberg H. (1991). Absorption of Atmospheric NO2 by Spruce (Piceaabies L. Karst.) Trees: 1.
NO2 Influx and Its Correlation with Nitrate Reduction. New Phytol, 117, 575-585.
R-86
-------
Thomas MD; Hendricks HR; Hill GR. (1944). Some chemical reactions of sulfur dioxide after absorption by alfalfa and sugarbeets. Plant
Physiol, 19,212-226.
Thomas MD. (1951). Gas damage to plants. Annu Rev Plant Physiol, 2,293-322.
Thompson AM; Pickering KE; Dickerson RR; Ellis WG Jr; Jacob DJ; Scala JR; Tao W-K; McNamara DP; Simpson J. (1994). Convective-
transport over the central United States and its role in regional CO and O3 budgets. J Geophys Res, 99,18703-18711.
Thompson AM; Singh HB; Schlager H. (2000). Subsonic assessment ozone and nitrogen oxide experiment (SONEX) and pollution from aircraft
emissions in the North Atlantic Flight Corridor (POLINAT 2). J Geophys Res, 105,3595-3603.
Thompson CR; Hensel EG; Kats G; Taylor OC. (1970). Effects of continuous exposure of naval oranges to nitrogen dioxide. Atmos Environ, 4,
349-355.
Thorne RSJ; Williams WP; Gordon C. (2000). The macroinvertebrates of a polluted stream in Ghana. Journal of Freshwater Ecology, 15, 209-
217.
Thornton DC; Bandy AR. (1993). Sulfur dioxide and dimethyl sulfide in the central Pacific troposphere. J Atmos Chem ,17,1-13.
Thornton DC; Bandy AR; Blomquist BW; Anderson BE. (1996). Impact of anthropogenic and biogenic sources and sinks on carbonyl sulfide in
the North Pacific troposphere. J Geophys Res, 101,1873-1881.
Thornton DC; Brandy AR; Blomquist BW; Driedger AR; Wade TP. (1999). Sulfur dioxide distribution over the Pacific Ocean 1991-1996. J
Geophys Res, 104,5845-5854.
Thornton DC; Bandy AR; Tu FH; Blomquist BW; Mitchell GM; Nadler W; Lenschow DH. (2002). Fast airborne sulfur dioxide measurements by
atmospheric pressure ionization mass spectrometry (APIMS). J Geophys Res, 107, doi: 10.1029/2002JD02289.
Thornton FC; Schaedle M; Raynal DJ. (1986a). Effect of aluminum on the growth of sugar maple in solution culture. Can J For Res, 16,892-
896.
Thornton FC; Schaedle M; Raynal DJ. (1986b). Effects of aluminum on the growth, development, and nutrient composition of honeylocust
('Gleditia triacanthos L.) seedling. Tree Physiol, 2,307-316.
Thornton FC; Schaedle M; Raynal DJ; Zipperer C. (1986c). Effects of aluminum on honeylocust (Gleditsia triacanthos L.) seedlings in solution
culture. J Exp Bot, 37, 775-785.
Thornton FC; Schaedle M; Raynal DJ. (1987). Effects of aluminum on red spruce seedlings in solution culture. Environ Exp Bot, 27,489-498.
Thornton JA; Braban CF; Abbatt JPD. (2003). N2O5 hydrolysis on sub-micron organic aerosol: The effect of relative humidity, particle phase,
and particle size. Phys Chem Chem Phys, 5,4593-4603.
Throop HL. (2005). Nitrogen deposition and herbivory affect biomass production and allocation in an annual plant. Oikos, 111,91-100.
Throop HL; Lerdau MT. (2004). Effects of nitrogen deposition on insect herbivory: Implications for community and ecosystem processes.
Ecosystems, 7,109-133
Tie XX; Emmons L; Horowitz L; Brasseur G; Ridley, B; Atlas E; Stround C; Hess P; Klonecki A; Madronich S; Talbot R; Dibb J. (2003). Effect of
sulfate aerosol on tropospheric NOx and ozone budgets: Model simulations and TOPSE evidence. J Geophys Res, 108,
10.1029/2001 JD001508.
Tietema A; Beier C. (1995). A correlative evaluation of nitrogen cycling in the forest ecosystems of the EC projects NITREX and EXMAN. For
Ecol Manage, 71,143-151.
Tietema A; Emmett BA; Gundersen P; Kjonaas OJ; Koopmans CJ. (1998). The fate of 15N-labelled nitrogen deposition in coniferous forest
ecosystems. For Ecol Manage, 101,19-27.
Tietge JE; Johnson RD; Bergman HL. (1988). Morphometry changes in gill secondary lamellae of brook trout (Salvelinus fontinalis).after long-
term exposure to acid and aluminum. Can J Fish Aquat Sci, 45,1643-1648.
Tilman, D. (1981). Tests of resource competition theory using four species of Lake Michigan algae. Ecology, 62,802-815.
Tilman D. (1987). Secondary succession and the pattern of plant dominance along experimental nitrogen gradients. Ecol Monogr, 57,189-214.
Tilman D; Wedin D. (1991). Dynamics of nitrogen competition between successional grasses. Ecology, 72,1038-1049.
Tilman D; Downing JA. (1994). Biodiversity and stability in grasslands. Nature, 367,363-365.
Tilman D. (1996). Biodiversity: Population versus ecosystem stability. Ecology, 77,350-363.
Tilman D. (2000). Causes, consequences, and ethics of biodiversity. Nature, 405,208-211.
Tingey DT; HodsettWE; Henderson S. (1990). Definition of adverse effects for the purpose of establishing secondary national ambient air
quality standards. J Environ Qual, 19,635-639.
Tingey DT; Johnson MG; Phillips DL; Storm MJ. (1995). Effects of elevated CO2 and nitrogen on ponderosa pine fine roots and associated
fungal components. J Biogeogr, 22,281-287.
Tipping E; Berggren D; Mulder J; Woof C. (1995). Modelling the solid-solution distributions of protons, aluminium, base cations and humic
substances in acid soils. European Journal of Soil Science, 46,77-94.
Tipping E; Bass JAB; Hardie D; Haworth EY; Hurley MA; Wills G. (2002). Biological responses to the reversal of acidification in surface waters
of the English Lake District. Environ Pollut, 116,137-146.
Tokiwa H; Ohnishi Y. (1986). Mutagenicity and carcinogenicity of nitroarenes and their sources in the environment. Critical Reviews in
Toxicology, 17,23-60.
Tokiwa H; Nakanishi Y; Sera N; Hara N; Inuzuka S. (1998). Analysis of environmental carcinogens associated with the incidence of lung
cancer. Toxicology letters, 99,33-41.
Toman MA. (1998). Introduction to special Section: Forum on valuation of ecosystem services - Why not to calculate the value of the world's
ecosystem services and natural capital. Ecol Econ, 25,57-60.
Tomassen HBM; Smolders AJP; Lamers LPM; Roelofs JGM. (2003). Stimulated growth of Betula pubescens and Molinia caerulea on
ombrotrophic bogs: Role of high levels of atmospheric nitrogen deposition. J Ecol, 91,357-370.
R-87
-------
Tomassen HBM; Smolders AJP; Limpens J; Lamers LPM; Roelofs JGM. (2004). Expansion of invasive species on ombrotrophic bogs:
dessication or high N deposition? J Appl Ecol, 41,139-150.
Tonnesen GS; Dennis RL. (2000). Analysis of radical propagation efficiency to assess ozone sensitivity to hydrocarbons and NOx: 2. Long-
lived species as indicators of ozone concentration sensitivity. J Geophys Res, 105,9227-9241.
Tonnessen KA. (1991). The Emerald Lake Watershed Study: Introduction and site description. Water Resour Res, 27(7), 1537-1539.
Torfs K; Van Grieken R. (1996). Effect of stone thickness on run-off water composition and derived damage functions in ambient exposure
experiments. Atmos Environ, 30,1-8.
Toumi R. (1994). BrO as a sink for dimethylsulfide in the marine atmosphere. Geophys Res Lett, 21,117-120.
Townsend AR; Braswell BH; Holland EA; Penner JE. (1996). Spatial and temporal patterns in terrestrial carbon storage due to deposition of
fossil fuel nitrogen. Ecol Appl, 6,806-814.
Townsend AR; Howarth RW; Bazzaz FA; Booth MS; Cleveland CC; Collinge SK; Dobson AP; Epstein PR; Keeney DR; Mallin MA; Rogers CA;
Wayne P; Wolfe AH. (2003). Human health effects of a changing global nitrogen cycle. Front Ecol Environ, 1,240-246.
Townsend CR; HildrewAG; Francis J. (1983). Community structure in some southern English streams: the influence of physiochemical factors.
Freshw. Biol, 13,521-544.
Trabaud L. (1994). The effect of fire on nutrient losses and cycling in a Quercus coccifera garrigue (southern France). Oecologia, 99,379-386.
Trainer M; Parrish DD; Buhr MP; Norton RB; Fehsenfeld FC; Anlauf KG; Bottenheim JW; Tang YZ; Wiebe HA; Roberts JM; Tanner RL;
Newman L; BowerSOx VC; Meagher JF; Olszyna KJ; Rodgers MO; Wang T; Berresheim H; Demerjian KL; Roychowdhury UK. (1993).
Correlation of ozone with NOyin photochemically aged air. J Geophys Res, 98,2917-2925.
Trainer M; Parrish DD; Golday PD; Roberts J; Fehsenfeld FC. (2000). Review of observation-based analysis of the regional factors influencing
ozone concentrations. Atmos Environ, 34,2045-2061.
Treiman T. (2006). Are residents willing to pay for their community forests? Results of a contingent valuation survey in Missouri, USA. Urban
Studies, 43,1537-1548.
Treseder KK; Allen MF. (2000). Mycorrhizal fungi have a potential role in soil carbon storage under elevated CO2 and nitrogen deposition. New
Phytol, 147,189-200.
Treseder KK. (2004). The meta-analysis of mycorrhizal response to nitrogen, phosphorus, and atmospheric CO2 in field studies. New Phytol,
164,347-355.
Treshow M. (1980). Pollution effects on plant distribution. Environ Conserv, 7,279-286.
Trettin CC; Johnson DW; Todd DE Jr. (1999). Forest Nutrient and Carbon Pools at Walker Branch Watershed: Changes during a 21-Year
Period. Soil Sci Soc Am J, 63,1436-1448.
Treves K; Shragina L; Rudich Y. (2000). Henry's Law constants of some beta-, gamma-, and delta-hydroxy alkyl nitrates of atmospheric
interest. Environ Sci Technol, 34,1197-1203.
Tritton LM; Martin CW; Hornbeck JW; Pierce RS. (1987). Biomass and nutrient removals from commercial thinning and whole-tree clearcutting
of central hardwoods. Environ Manage, 11, 659-666.
Trojnar JR. (1977). Egg and larval survival of white suckers (Catostomus commersoni) at low pH. J Fish Res Board Can, 34,262-266.
Trombulak SC; Frissell CA. (2000). Review of ecological effects of roads on terrestrial and aquatic communities. Conserv Biol, 14,18-30.
Truman RA; Humphreys FR; Ryan PJ. (1986). Effect of varying solution ratios of Al toCa and Mg on the uptake of phosphorus by Pinus
radiata. Plant Soil, 96,109-123.
Tsai C-J; Huang H-Y. (1995). Atmospheric aerosol sampling by an annular denuder system and a high-volume PM10 sampler. Environ Int, 21,
283-291.
Turco RP; Toon OB; Whitten RC; Hamill P; Keesee RG. (1983). The 1980 eruptions of Mount St. Helens: Physical and chemical processes in
the stratospheric clouds. J Geophys Res, 88,5299-5319.
Turekian VC; Macko SA; Keene WC. (2001). Application of stable sulfur isotopes to differentiate sources of size-resolved particulate sulfate in
polluted marine air at Bermuda during spring. Geophys Res Lett, 28,1491-1494.
Turk JT; Spahr NE. (1991). Rocky Mountains. In :Charles DF; Christie S (Eds.), Acidic deposition and aquatic ecosystems: Regional case
studies (pp. 471-501). New York: Springer-Verlag, Inc.
Turk JT; Campbell DH; Spahr NE. (1992). Initial findings of synoptic snowpack sampling in Colorado Rocky Mountains. (Open file report 92-
645). Denver, CO; Geologic Survey;. U.S. Department of the Interior.
Turk JT; Campbell DH; Spahr NE. (1993). Use of chemistry and stable sulfur isotopes to determine sources of trends in sulfate of Colorado
lakes. Water Air Soil Pollut, 67,415-431.
Turner J; Lambert M J; Gessel SP. (1977). Use of foliage sulphate concentrations to predict response to urea application by Douglas-fir. Can J
For Res, 7,476-480.
Turner J; Lambert M J. (1980). Sulfur nutrition of forests. In: Shriner DS; Richmond CR; Lingberg SE., (Eds.), Atmospheric sulfur deposition:
Environmental impact and health effects. (Proceedings of the 2nd life sciences symposium held in Gatlinburg, TN; October 1979; pp
321-333). Ann Arbor, Ml: Ann Arbor Science Publishers, Inc.
Turner RE; Qureshi N; Rabalais NN; Dortch Q; Justic D; Shaw RF; Cope J. (1998). Fluctuating silicate: Nitrate ratios and coastal plankton food
webs. Proc Natl Acad Sci USA 95:13048-13051.
Turner RE; Stanley D; Brock D; Pennock J; Rabalais NN. (2001). A comparison of independent N-loading estimates for U.S. estuaries. In:
Valigura RA; Alexander RB; Castro MS; Meyers TP; Paerl HW; Stacey PE; Turner RE (Eds.), Nitrogen loading in coastal water bodies:
An atmospheric perspective. (AGU Coastal Estuaries Studies; Series Volume 57). Washington, DC: American Geophysical Union.
Turner RE; Stanley D; Brock D; Pennock J; Rabalais NN. (2001). A Comparison of Independent N-loading Estimates for U.S. Estuaries.
COASTAL AND ESTUARINE STUDIES, 107-118.
Turner RE. (2002). Element ratios and aquatic food webs. Estuaries, 25,694-703.
R-88
-------
Turner RS; Cook RB; Van Miegroet H; Johnson DW; Elwood JW; Bricker OP; Lindberg SE; Hornberger GM. (1991). Watershed and lake
processes affecting surface water acid-base chemistry. In: Irving PM (Ed.), Acidic deposition: state of science and technology report
no 10; volume II: Aquatic processes and effects. Washington, DC: The U.S. National Acid Precipitation Assessment Program
(NAPAP).
Turner RS; Ryan PF; Marmorek DR; Thornton KW; Sullivan TJ; Baker JP; Christensen SW; Sale MJ. (1992). Sensitivity to change for low-ANC
eastern U.S. lakes and streams and brook trout populations under alternative sulfate deposition scenarios. Environ Pollut, 77, 269-
277.
Turnipseed AA; Huey LG; Nemitz E; Stickel R; Higgs J; Tanner DJ; Slusher DL; Sparks JP; Flocke F; Guenther A. (2006). Eddy covariance
fluxes of peroxyacetyl nitrates (PANs) and NOy to a coniferous forest. J Geophys Res, 111,10.1029/2005JD006631.
Turnpenny AWH; Sadler K; Aston RJ; Milner AGP; Lynam S. (1987). The fish populations of some streams in Wales and northern England in
relation to acidity and associated factors. Journal of Fish Biology, 31,415-434.
Turunen J; Roulet NT; Moore TR; Richard PJH. (2004). Nitrogen deposition and increased carbon accumulation in ombrotrophic peatlands in
eastern Canada. Glob Biogeochem Cycles, 18, GB3002, doi:10.1029/2003GB002154.
Twilley RR; Kemp WM; StaverKW; Stevenson JC; Boynton WR. (1985). Nutrient enrichment of estuarine submersed vascular plant
communities. 1. Algal growth and effects on production of plants and associated communities. Marine Ecology Progress Series, 23,
179-191.
Twomey LJ; Piehler MF; Paerl HW. (2005). Phytoplankton uptake of ammonium, nitrate and urea in the Neuse River Estuarry, NC, USA.
Hydrobiologia, 533,123-134.
Tyler AC; Lambrinos JG; Grosholz ED. (2007). Nitrogen inputs promote the spread of an invasive marsh grass. Ecol Appl, 17,1886-1898.
Tyler SJ; Ormerod SJ. (1992). A Review of the Likely Causal Pathways Relating the Reduced Density of Breeding Dippers Cinclus cindus to
the Acidification of Upland Streams. Environ Pollut, 78,49-55.
U.K. AQEC. (2004). Nitrogen dioxide in the United Kingdom. London, UK; Department for Environment, Food and Rural Affairs. U.K. Air Quality
Expert Group (AQEC) http://www.defra.qov.uk/environment/airqualitv/publications/nitroqen-dioxide/index.htm.
Ulrich B; Mayer R; Khanna PK. (1980). Chemical changes due to acid precipitation in loess-derived soil in central Europe. Soil Sci, 130,193-
199.
Ulrich B. (1983). Soil acidity and its relations to acid deposition. In: Ulrich, B; Pankrath, J (Eds.), Effects of accumulation of air pollutants in
forest ecosystems. (Proceedings of a workshop held in Gottingen, Federal Republic of Germany; May 1982 pp. 127-146). Dordrecht,
The Netherlands: D. Reidel Publishing Company.
UNECE. (2004). Manual on methodologies and criteria for modeling and mapping critical loads and levels and air pollution effects, risks, and
trends, http://www.icpmappinq.org. Geneva, Switzerland; Convention on Long-Range Transboundary Air Pollution; United Nations
Economic Commission for Europe (UNECE)
UNEP (1999). Levine JS; Bobbe T; Ray N; Singh A; and Witt RG. Wildland Fires and the Environment: A Global Synthesis.
(UNEP/DEIAEW/TR.99-1). Nairobi, Kenya; Division of Environmental Information; Assessment and Early Warning (DEIA&EW); United
Nations Environment Programme (UNEP).
UNEP 2007
Unsworth MH; Wilshaw JC. (1989). Wet, occult and dry deposition of pollutants on forests. Agricultural and Forest Meteorology 47:221-238.
Urban NR; Bayley SE. (1986). The acid-base balance of peatlands: a short-term perspective. Water Air Soil Pollut, 30,791-800.
Urban NR; Eisenreich SJ. (1988). Nitrogen cycling in a forested Minnesota bog. Canadian Journal of Botany, 66,435-449.
Urquhart NS; Paulsen SG; Larsen DP. (1998). Monitoring for regional and policy-relevant trends over time. Ecol Appl, 8,246-257.
USDA. (1996). Southern Appalachian Man and the Biosphere (SAMAB), Southern Appalachian Assessment Aquatics Technical (Report 2 of
5). Atlanta, GA: Southern Region; U.S. Forest Service, U.S. Department of Agriculture (USDA).
U.S.CFR. (2006). Lower detectable limit—(1).technical definition. CFR40: §53.23c. Washington DC; U.S. Code of Federal Regulations.
U.S. Code. (1970a). Clean Air Act, §108, Air quality criteria and control techniques. Amended in 1977,1990. (Title 42§7408-108).
U.S. Code. (1970b). Clean Air Act, §109, National Ambient Air Quality Standards. Amended in 1977,1990. (Title 42§7408-109).
U.S. Congress (1990). Clean Air Act amendments of 1990 (PL 101-549). Washington DC.
U.S. Department of Health Education, and Welfare. (1970). Air quality criteria for sulfur oxides (No. AP-50A). Washington, DC; National Air
Pollution Control Administration; Public Health Service..
U.S. EPA. (1971). Air quality criteria for nitrogen oxides. (Report no. AP-84). Washington, DC; Air Pollution Control Office; U.S. Environmental
Protection Agency.
U.S. EPA. (1973). Effects of sulfur oxide in the atmosphere on vegetation: Revised chapter 5 of air quality criteria for sulfur oxides. (Report no.
EPA-R3-73-030). Research Triangle Park, NC; National Environmental Research Center, Office of Research and Development; U.S.
Environmental Protection Agency.
U.S. EPA. (1974). A preliminary report on asbestos in the Duluth, Minnesota area. Washington, DC; Office of Technical Analysis; U.S.
Environmental Protection Agency.
U.S. EPA. (1982a). Air quality criteria for oxides of nitrogen (No. EPA-600/8-82-026). Research Triangle Park, NC; Office of Health and
Environmental Assessment; Environmental Criteria and Assessment Office; Available from: NTIS, Springfield, VA; PB83-131011.
U.S. EPA. (1982b). Air quality criteria for particulate matter and sulfur oxides (Report no. EPA/600/8-82/029aF-cF; 3 Volumes). Research
Triangle Park, NC: Office of Health and Environmental Assessment, Environmental Criteria and Assessment Office; U.S.
Environmental Protection Agency.
U.S. EPA. (1982c). Review of the national ambient air quality standards for sulfur oxides: Assessment of scientific and technical information;
OAQPS staff paper. (Report no. EPA-450/5-82-007). Research Triangle Park, NC; Office of Air Quality Planning and Standards; U.S.
Environmental Protection Agency.
R-89
-------
U.S. EPA. (1984a). The acidic deposition phenomenon and its effects: Critical assessment review papers. Volume I: Atmospheric sciences
(No. EPA-600/8-83-016-AF). Washington, DC; Office of Acid Deposition; Environmental Monitoring, and Quality Assurance; National
Exposure Research Laboratory; Office of Research and Development; U.S. Environmental Protection Agency.
U.S. EPA. (1984b).The acidic deposition phenomenon and its effects: Critical assessment review papers. Volume II: Effects sciences. (Report
no. EPA-600/8-83-016-BF). Washington, DC; Office of Acid Deposition, Environmental Monitoring and Quality Assurance; National
Exposure Research Laboratory; Office of Research and Development; U.S. Environmental Protection Agency.
U.S. EPA. (1985). The acidic deposition phenomenon and its effects: Critical assessment review papers. Critical assessment document (No.
EPA/600/8-85/001). Washington, DC; Office of Acid Deposition; Environmental Monitoring, and Quality Assurance; National Exposure
Research Laboratory; Office of Research and Development; U.S. Environmental Protection Agency.
U.S. EPA. (1991). Guideline for regulatory application of the urban airshed model. (Report no. EPA-450/4-91-013). Research Triangle Park,
NC; Office of Air Quality Planning and Standards; U.S. Environmental Protection Agency.
U.S. EPA. (1993a). Air Quality Criteria for Oxides of Nitrogen (Report no. EPA/600/8-91/049aF-cF; 3 Volumes). Research Triangle Park, NC;
Environmental Criteria and Assessment Office; Office of Health and Environmental Assessment; U.S. Environmental Protection
Agency.
U.S. EPA. (1993b).Ambient air quality surveillance. (40CFR-Part 58: Federal Register 58FR8452 [pp 8452-8475]). Washington, DC; U.S.
Environmental Protection Agency.
U.S. EPA. (1995a). Acid deposition standard feasibility study. Report to Congress. (Report no EPA430-R-95-001a). Washington, DC; Office of
Air and Radiation; Environmental Protection Agency:.
U.S. EPA. (1995b). Review of the national ambient air quality standards for nitrogen dioxide: Assessment of scientific and technical information,
OAQPS staff paper.(Report no EPA-452/R-95-005). Research Triangle Park, NC; Office of Air Quality Planning and Standards; U.S.
Environmental Protection Agency.
U.S. EPA. (1996a). Air quality criteria for particulate matter (No. EPA/600/P-95/001aF-cF). Research Triangle Park, NC; National Center for
Environmental Assessment-RTP; Office of Research and Development; U.S. Environmental Protection Agency.
U.S. EPA. (1996b). Mercury study report to Congress. Volume. Ill: An assessment of exposure from anthropogenic mercury emissions in the
United States (Report no. EPA-452/R-96-001 c). Research Triangle Park, NC; Office of Air Quality Planning and Standards and Office
of Research and Development; U.S. Environmental Protection Agency.
U.S. EPA. (1996c). Air quality criteria for ozone and related photochemical oxidants. (Report no EPA/600/AP-93/004aF-cF; 3 Volumes).
Research Triangle Park, NC; National Center for Environmental Assessment-RTP; Office of Research and Development;
U.S. Environmental Protection Agency.
U.S. EPA. (1997a). Mercury study report to Congress. Volume VI: An ecological assessment for anthropogenic mercury emissions in the
United States. (Report no. EPA- 452/R-97-008). Research Triangle Park, NC; Office of Air Quality Planning and Standards and Office
of Research and Development; U.S. Environmental Protection Agency.
U.S. EPA. (1997b). National air pollutant emission trends 1990-1996. (Report no. EPA/454/R-97/011). Research Triangle Park, NC; Office of
Air Quality Planning and Standards; U.S. Environmental Protection Agency.
U.S. EPA. (1998). U.S. Environmental Protection Agency. (1998).SLAMS/NAMS/PAMS network review guidance. (Report no EPA-454/R-98-
003). Research Triangle Park, NC; Office of Air Quality Planning and Standards; U.S. Environmental Protection Agency.
U.S. EPA. (1999a). Getting started: Emissions inventory methods for PM-2.5. Technical report (Emission factor and inventory group; EIIP
document series; Volume IX). Research Triangle Park, NC; Office of Air Quality Planning and Standards; U.S. Environmental
Protection Agency.
U.S. EPA. (1999b). The benefits and costs of the Clean Air Act, 1990-2010. EPA report to Congress. (No. EPA-410-R-99-001). Washington,
DC; Office of Air and Radiation and Office of Policy; U.S. Environmental Protection Agency.
U.S. EPA. (2000a). Air quality criteria for carbon monoxide. (Report no. EPA/600/P-99/001F). Research Triangle Park, NC; National Center for
Environmental Assessment; Office of Research and Development; U.S. Environmental Protection Agency.
U.S. EPA. (2000b). Deposition of air pollutants to the great waters: Third report to Congress. (EPA-453/R-00-005)
http://www.epa.aov/air/oagps/ar8water/3rdrpt/index.html: Research Triangle Park, NC; Office of Air Quality Planning and Standards;
U.S. Environmental Protection Agency.
U.S. EPA. (2000c). Nutrient Criteria Technical Guidance Manual: Rivers and Streams. (Report no EPA-822-B00-022). Washington, DC; U.S.
Environmental Protection Agency.
U.S. EPA. (2000d). Nutrient Criteria Technical Guidance Manual: Lakes and Reservoirs. (EPA-822-B00-001). Washington, DC;
U.S. Environmental Protection Agency.
U.S. EPA. (2000e). National air pollutant emission trends 1990-1998. (Report no. EPA/454/R-O0/002). Research Triangle Park, NC; Office of
Air Quality Planning and Standards; U.S. Environmental Protection Agency.
U.S. EPA. (2001). National coastal condition report. (EPA-620/R-01/005). Washington, DC; Office of Water; Office of Research and
Development, U.S. Environmental Protection Agency.
U.S. EPA. (2002a). A framework for the economic assessment of ecological benefits. Washington, DC; Ecological Benefit Assessment
Workgroup; Social Sciences Discussion Group; Science Policy Council; U.S. Environmental Protection Agency.
U.S. EPA. (2002b). Inventory of U.S. greenhouse gas emissions and sinks: 1990-2000. (Report no EPA430-R-02-003). Washington, DC;
Office of Atmospheric Programs; U.S. Environmental Protection Agency.
U.S. EPA. (2003). Economic analyses of nutrient and sediment reduction actions to restore Chesapeake Bay water quality. Annapolis, MD;
Chesapeake Bay Program Office; Region III; U.S. Environmental Protection Agency.
R-90
-------
U.S. EPA. (2004). Air quality criteria for particulate matter. Volumes I and II (Report no. EPA/600/P-99/002aF; 2 Volumes). Research Triangle
Park, NC; National Center for Environmental Assessment-RTP; Office of Research and Development; U.S. Environmental Protection
Agency.
U.S. EPA. (2005a). Review of the national ambient air quality standards for particulate matter: policy assessment of scientific and technical
information. OAQPS staff paper. (Report no. EPA/452/R-05-005a). Research Triangle Park, NC; Office of Air Quality Planning and
Standards; U.S. Environmental Protection Agency.
U.S. EPA. (2005b). Technical assistance document (TAD) for precursor gas measurements in the NCore multi-pollutant monitoring network.
Version 4. (report no. EPA-454/R-05-003). Research Triangle Park, NC; Office of Air Quality Planning and Standards; U.S.
Environmental Protection Agency.
U.S. EPA. (2006a). 2002 National Emissions Inventory: Data and Documentation for the final 2002 point source national emissions inventory.
http://www.epa.aov/ttn/chief/net/2002inventorv.html. Research Triangle Park, NC; Emission Inventory and Analysis Group; Air Quality
and Analysis Division; U.S. Environmental Protection Agency.
U.S. EPA. (2006b). Air quality criteria for ozone and related photochemical oxidants (Report no. EPA/600/R-05/004aF - cF; 3 Volumes).
Research Triangle Park, NC; National Center for Environmental Assessment-RTP; Office of Research and Development; U.S.
Environmental Protection Agency.
U.S. EPA. (2006c). Multi-agency critical loads workshop: Sulfur and nitrogen deposition effects on freshwater and terrestrial ecosystems. Paper
presented at the Multi-agency critical loads workshop held at Charlottesville, VA; University of Virginia-Charlottesville. (Final Report no
EPA/68-W-03-02). Washington, DC; Clean Air Markets Division; U.S. Environmental Protection Agency.
U.S. EPA. (2006d). Wadeable streams assessment: A collaborative survey of the nation's streams. (Report no EPA-841-B-06-002).
Washington, DC; Office of Water; Office of Research and Development; U.S. Environmental Protection Agency.
U.S. EPA. (2006e).Acid rain program (ARP) 2005 progress report. Washington, DC; Office of Air and Radiation; Clean Air Markets Division;
EPA-430-R-06-015. Available: http://www.epa.gov/airmarkets/progress/docs/2005report.pdfI12 March, 2007],
U.S. EPA. (2007a). Integrated review plan for the secondary national ambient air quality standards for nitrogen dioxide and sulfur dioxide.
(Report no. EPA-452/R-08-006). Research Triangle Park, NC; U.S. Environmental Protection Agency.
U.S. EPA. (2007b). Inventory of U.S. greenhouse gas emissions and sinks: 1990-2005. (Report no. EPA 430-R-07-002). Washington, DC;
Office of Atmospheric Programs; U.S. Environmental Protection Agency.
U.S. EPA. (2008a). Integrated science assessment for oxides of nitrogen: Health criteria. (Report no. EPA/600/R-08/071). Research Triangle
Park, NC; National Center for Environmental Assessment; Office of Research and Development; U.S. Environmental Protection
Agency.
U.S. EPA. (2008b). Integrated science assessment for sulfur oxides: Health criteria. (Report no. EPA/600/R-08/047F). Research Triangle Park,
NC; National Center for Environmental Assessment; Office of Research and Development; U.S. Environmental Protection Agency.
U.S. EPA. (2008c). Risk and exposure assessment for the review of the secondary national ambient air quality standards foroxides of nitrogen
and oxides of sulfur (Draft). Research Triangle Park, NC; Office of Air Quality Planning and Standards; U.S. Environmental Protection
Agency.
U.S. FWS. (2005). Dahl TE (Ed.). Status and Trends of Wetlands in the Conterminous United States 1998 to 2004.
http://wetlandsfws.er.usas.gov/status trends/national reports/trends 2005 report.pdf. Washington, DC; Fisheries and Habitat
Conservation; U.S., Fish and Wildlife Service (FWS); Department of the Interior.
U.S. GAO. (2000). Acid Rain. Emissions trends and effects in the eastern United States (Report to Congressional Requesters: GAO/RCED-00-
47). Washington, DC; U.S. General Accounting Office (GAO).
Valiela I; Costa JE. (1988). Eutrophication of Buttermilk Bay, a Cape Cod coastal embayment: concentrations of nutrients and watershed
nutrient budgets. Environ Manage, 12,539-553.
Valiela I; Costa JE; Foreman K. (1990). Transport of groundwater-borne nutrients from watersheds and their effects on coastal waters.
Biogeochemistry, 10,177-197.
Valiela I; Foreman K; LaMontagne M; Hersh D; Costa J; Peckol P; DeMeo-Andreson B; D'Avanzo C; Babione M; Sham C; Brawley J; Lajtha K.
(1992). Couplings of watersheds and coastal waters: sources and consequences of nutrient enrichment in Waquoit Bay,
Massachusetts. Estuaries, 15,443-457.
Valigura RA; Alexander RB; Brock DA; Castro MS; Meyers TP; Paerl HW; Stacey PE; Stanley D. (2000). An assessment of nitrogen inputs to
coastal areas with an atmospheric perspective. In: Valigura RA; Alexander RB; Castro MS; Meyers TP; Paerl HW; Stacey PE; Turner
RE (Eds.), Nitrogen loading in coastal water bodies: An atmospheric perspective. (AGU Coastal Estuaries Studies; Series Volume 57).
Washington, DC: American Geophysical Union.
Vallano DM; Sparks JD. (2008). Quantifying foliar uptake of gaseous nitrogen dioxide using enriched foliar N values. New Phytol, 177,946-955.
Van Aalst RM. (1982). Dry deposition of NOx. In Schneider T, Grant L (Eds.), Air pollution by nitrogen oxides (pp. 263-270). Amsterdam, The
Netherlands: Elsevier Scientific Publishing Company.
Van Aardenne JA; Dentener FJ; Olivier JGJ; Klein Goldewijk CGM; Lelieveld J. (2001). A Idegree x Idegree resolution data set of historical
anthropogenic trace gas emissions for the period 1980-1990. Glob Biogeochem Cycles, 15,909-928.
van Breemen N; Boyer EW; Goodale CL; Jaworski NA; Paustian K; Seitzinger SP; Lajtha K; Mayer B; Van Dam D; Howarth RW; Nadelhoffer
KJ. (2002). Where did all the nitrogen go? Fate of nitrogen inputs to large watersheds in the northeastern USA Biogeochemistry,
57/58, 267-293.
van Breemen N; Finlay R; Lundstrom U; Jongmans AG; Giesler R; Olsson M. (2000). Mycorrhizal weathering: A true case of mineral plant
nutrition? Biogeochemistry, 49, 53-67.
van Breemen N; Mulder J; Driscoll C. (1983). Acidification and alkalinization of soils. Plant Soil, 75,283-308.
R-91
-------
van Breemen N; Van Dijk HFG. (1988). Ecosystem effects of atmospheric deposition of nitrogen in The Netherlands. Environ Pollut, 54,249-
274
van de Graaf A; Mulder A; de Brujin P; Jetten MSM; Robertson LA; Kuenen JG. (1995). Anaerobic oxidation of ammonium is a biologically
mediated process. Appl Environ Microbiol, 61,1246-1251.
van der Nat FWA; deBrouwer JFC; Middelburg JJ; Laanbroek HJ. (1997). Spatial distribution and inhibition by ammonium of methane oxidation
in intertidal freshwater marshes. Appl Environ Microbiol, 63,4734-4740.
Van Der Welle MEW; Smolders AJP; Op Den Camp HJM; Roelofs J; Lamers LPM. (2007). Biogeochemical interactions between iron and
sulphate in freshwater wetlands and their implications for interspecific competition between aquatic macrophytes. Freshw Biol, 52,
434-447.
Van der Werf GR; Randerson JT; Collate J; Giglio L. (2003). Carbon emissions from fires in tropical and subtropical ecosystems. Glob Change
Biol, 9, 547-562.
van Diepen L; Lilleskov E; Pregitzer K; Miller R. (2007). Decline of mycorrhizal fungi in northern hardwood forests exposed to chronic nitrogen
inputs. New Phytol, 176,175-183.
Van Dijk HFG; Roelofs JGM. (1988). Effects of excessive ammonium deposition on the nutritional status and condition of pine needles. Physiol
Plant, 73,494-501.
Van Dobben HF; de Bakker AJ. (1996). Re-mapping of lichen biodiversity in The Netherlands: Effects of decreasing SO2 and increasing NH3.
Acta Bot Neerl, 45,55-71.
van Dobben HF; Wolterbeek HT; Wamelink GW; Ter Braak CJ. (2001). Relationship between epiphytic lichens, trace elements and gaseous
atmospheric pollutants. Environ Pollut, 112,163-169.
Van Drecht G; Bouwman AF; Knoop JM; Beusen AHW; Meinardi CR. (2003). Global modeling of the fate of nitrogen from point and nonpoint
sources in soils, groundwater, and surface water. Glob Biogeochem Cycles, 17,1115.
Van Egmond K; BresserT; Bouwman L. (2002). The European nitrogen case. Ambio, 31,72-78.
van Griensven A; Bauwens W. (2001). Integral water quality modelling of catchments. Water Sci Technol, 43,321-328.
Van Groenigen K-J; Six J; Hungate BA; De Graaff M-A; Van Breemen N; Van Kessel C. (2006). Element interactions limit soil carbon storage.
Proc Natl Acad Sci USA, 103,6571-6574.
van Haluwyn C; van Herk CM. (2002). Bioindication: The community approach. In: Nimis PL; Scheidegger C; Wolseley PA (Eds.), Monitoring
with lichens-monitoring lichens (Vol. IV. Earth and environmental sciences, pp. 39-64). Dordrecht, The Netherlands: Kluwer Academic
Publishers.
van Herk CM. (1999). Mapping of ammonia pollution with epiphytic lichens in The Netherlands. Lichenologist, 31,9-20.
van Herk CM. (2001). Bark pH and susceptibility to toxic air pollutants as independent causes of changes in epiphytic lichen composition in
space and time. Lichenologist, 33,419-441.
Van Houtven G; Powers J; Pattanayak S. (2007). Valuing water quality improvements using meta-analysis: Is the glass half-full or half-empty
for national policy analysis? Resource and Energy Economics, 29,206-228.
Van Miegroet H; Cole DW; Foster NW. (1992a).Nitrogen distribution and cycling. In: Johnson DW; Lindberg SE (Eds.), Atmospheric deposition
and forest nutrient cycling: A synthesis of the integrated forest study. (Series: Billings WD; Golley F; Lange OL; Olson JS; Remmert H
(Eds.), Ecological studies: Analysis and synthesis; Volume 91, pp 178-196). New York, NY: Springer-Verlag.
Van Miegroet H; Cole DW; Foster NW. (1992b). Analysis of N cycles in polluted vs. unpolluted environment. In: Johnson DW; Lindberg SE
(Eds.), Atmospheric deposition and forest nutrient cycling: A synthesis of the integrated forest study. (Series: Billings WD; Golley F;
Lange OL; Olson JS; Remmert H (Eds.), Ecological studies: Analysis and synthesis; Volume 91, pp 199-202). New York, NY:
Springer-Verlag.
Van Miegroet H; Johnson DW. (1993). Nitrate dynamics in forest soils. In: Burt TP; Heathwaite AL; Trudgill ST (Eds.), Nitrate: Processes,
Patterns and Management, (pp 75-97). New York, NY: John Wiley & Sons Ltd.
Van Miegroet H; Moore PT; Tewksbury CE; Nicholas NS. (2007). Carbon sources and sinks in high-elevation spruce-fir forests of the
Southeastern U.S. For Ecol Manage, 238,249-260.
van Noije TPC; Eskes HJ; Dentener FJ; Stevenson DS; Ellingsen K; Schultz MG; Wild O; Amann M; Atherton CS; Bergmann DJ; Bey I;
Boersma KF; Butler T; Cofala J; Drevet J; Fiore AM; Gauss M; Hauglustaine DA; Horowitz LW; Isaksen ISA; Krol MC; Lamarque J-F;
Lawrence MG; Martin RV; Montanaro V; Miiller JF; Pitari G; Prather MJ; Pyle JA; Richter A; Rodriguez JM; Savage NH; Strahan SE;
Sudo K; Szopa S; van Roozendael M. (2006). Multi-model ensemble simulations of tropospheric NO2 compared with GOME retrievals
for the year 2000. Atmos Chem Phys, 6,2943-2979.
Van Sickle J; Baker JP; Simonin HA; Baldigo BP; Kretser WA; Sharpe WE. (1996). Episodic acidification of small streams in the northeastern
United States: Fish mortality in field bioassays. Ecol Appl, 6,408-421.
Van Sickle J; Church MR. (1995). Nitrogen Bounding Study Methods for Estimating the Relative Effects of Sulfur and Nitrogen Deposition on
Surface Water Chemistry. Corvallis, OR; Environmental Research Laboratory; National Health and Environmental Effects Research
Laboratory, Office of Research and Development; U.S. Environmental Protection Agency.
Van Tienhoven AM; Olbrick KA; Skoroszewski R; Taljaard J; Zunckel M. (1995). Application of the critical loads approach in South Africa.
Water Air Soil Pollut, 85,2577-2582.
Vautard R; Martin D; Beekman M; Drobinski P; Friedrich R; Jaubertie A; Kley D; Lattuati M; Moral P; Neininger B; Theloke J. (2002). Paris
emission inventory diagnostics from the ESQUIF airborne measurements and a chemistry transport model. J Geophys Res, 108, doi:
10.1029/2002JD002797.
Vaux HJ Jr; Gardner PD; Mills TJ. (1984). Methods for assessing the impact of fire on forest recreation. (No. General Technical Report PSW-
79). Berkeley, CA: Pacific Southwest Forest and Range Experiment Station; U.S. Forest Service; U.S. Department of Agriculture
(US DA).
R-92
-------
Venterea RT; Groffman PM; Verchot LV; Magill AH; Aber JD. (2004). Gross nitrogen process rates in temperate forest soils exhibiting
symptoms of nitrogen saturation. For Ecol Manage, 196.
Venterea RT; Groffman PM; Verchot LV; Magill AH; Aber JD; Steudler PA. (2003). Nitrogen oxide gas emissions from temperate forest soils
receiving long-term nitrogen inputs. Glob Chang Biol, 9,346-357.
Venterea RT; Groffman PM; Castro MS; Verchot LV; Fernandez IJ; Adams MB. (2004a). Soil emissions of nitric oxide in two forest watersheds
subjected to elevated N inputs. For Ecol Manage, 196,335-349.
Venterea RT; Groffman PM; Verchot LV; Magill AH; Aber JD. (2004b).Gross nitrogen process rates in temperate forest soils exhibiting
symptoms of nitrogen saturation. For Ecol Manage, 196,129-142.
Vertucci FA; Eilers JM. (1993). Issues in monitoring wilderness lake chemistry: a case study in the Sawtooth Mountains, Idaho. Environ Monit
Assess, 28, 277-294.
Vilche JR; Varela FE; Acuna G; Codaro E. N; Rosales B. M; Fernandez A; Moriena G. (1995). A survey of Argentinean atmospheric corrosion:
I. Aluminum and zinc samples. Corros Sci, 37,941-961.
Vitek L; Schwertner HA. (2007). The heme catabolic pathway and its protective effects on oxidative stress-mediated diseases. In: Advances in
Clinical Chemistry, Vol 43 (Vol. 43, pp. 1-57). San Diego: Elsevier Academic Press Inc.
Vitousek PM; Aber JD; Howarth RW; Likens GE; Matson PA; Schindler DW; Schlesinger WH; Tilman DG. (1997). Human alteration of the
global nitrogen cycle: sources and consequences. Ecol Appl, 7,737-750.
Vitousek PM; Howarth RW. (1991). Nitrogen limitation on land and in the sea: how can it occur? Biogeochemistry, 13,87-115.
Vitousek PM; Aber JD; Howarth RW; Likens GE; Matson PA; Schindler DW; Schlesinger WH; Tilman DG. (1997a). Human alteration of the
global nitrogen cycle: Sources and consequences. Ecol Appl, 7,737-750.
Vitousek PM; Mooney HA; Lubchenco J; Melillo JM. (1997b). Human domination of Earth's ecosystems. Science, 277,494-499.
Vitousek PM; Hedin, L; Matson, P; Fownes, J; Neff, J. C. (1998).Within-system element cycles, input-output budgets, and nutrient limitations.
In: Pace M; Groffman P (Eds.), Successes, limitations and frontiers in ecosystem science, (pp 432-451). New York, NY: Springer-
Verlag.
Vitt DH; Wieder K; Halsey LA; Turetsky M. (2003). Response of Sphagnum Aiscum to nitrogen deposition: A case study of ombrogenous
peatlands in Alberta, Canada. Bryologist, 106,235-245.
Vogt KA; Grier CC; Meier CE; Edmonds RL. (1982). Mycorrhizal role in net primary production and nutrient cycling in Abies amabilis
ecosystems in western Washington. Ecology, 63, 370-380.
Vogt KA; Dahlgren R; Ugolini F; Zabowski D; Moore EE; Zasoski R. (1987a). Aluminum, Fe, Ca, Mg, K, Mn, Cu, Zn and P in above- and
belowground biomass. I. Abies amabilis and Tsuga mertensiana. Biogeochem, 4,277-294.
Vogt KA; Dahlgren R; Ugolini F; Zabowski D; Moore EE; Zasoski R. (1987b). Aluminum, Fe, Ca, Mg, K, Mn, Cu, Zn and P in above- and
belowground biomass. II. Pools, and circulation in a subalpine Abies amabilis stand. Biogeochem, 4,295-311.
Vogt R; Crutzen PJ; Sander R. (1996). A mechanism for halogen release from sea-salt aerosol in the remote marine boundary layer. Nature,
383, 327-330.
Vogt R; Sander R; Von Glasow R; Crutzen PJ. (1999). Iodine chemistry and its role in halogen activation and ozone loss in the marine
boundary layer: A model study. J Atmos Chem , 32,375-395.
Vollenweider RA. (1968). Water management research: Scientific fundamentals of the eutrophication of lakes and flowing waters, with
particular reference to nitrogen and phosphorus as factors in eutrophication. (Report no DAS/CSI/68.27). Paris, France: Organisation
for Economic Co-operation and Development (OECD).
Von Ballmoos P; Nussbaum S; Brunold C. (1993). The Relationship of Nitrate Reductase Activity to Uptake and Assimilation of Atmospheric
15N02"Nitrogen in Needles of Norway Spruce (P/'cea abies [L.] Karst ). Isotopes in Environmental and Health Studies, 29, 59-70.
Von Glasow R; Sander R; Bott A; Crutzen PJ. (2002a). Modeling halogen chemistry in the marine boundary layer. 1. Cloud-free MBL. J
Geophys Res, 107, doi: 10.1029/2001 JD000942.
Von Glasow R; Sander R; Bott A; Crutzen PJ. (2002b). Modeling halogen chemistry in the marine boundary layer. 2. Interactions with sulfur
and cloud-covered MBL. J Geophys Res, 107, doi: 10.1029/2001 JD000943.
Von Glasow R; Von Kuhlmann R; Lawrence MG; Piatt U; Crutzen PJ. (2004). Impact of reactive bromine chemistry in the troposphere. Atmos
Chem Phys, 4,2481-2497.
Vose JM; Elliott KJ; Johnson DW; Walker RF; Johnson MG; Tingey DT. (1995). Effects of elevated CO2 and N fertilization on soil respiration
from Ponderosa pine (Pinus Ponderosa) in open-top chambers. Can J For Res, 25,1243-1251.
Vose JM; Swank WT. (1990). Preliminary estimates of foliar absorption of 15 N-labeled nitric acid vapor (HNO 3) by mature eastern white pine
(Pinus strobus). Can J For Res, 20,857-860.
Vutukuru S; Dabdub D. (2008). Modeling the effects of ship emissions on coastal air quality: A case study of Southern California. Atmos
Environ, 42, 3651-3764.
Waddell, E; Greenwood, R. (2006). Pacific Northwest nitrogen and sulfur deposition critical loads workshop; held at North Cascades
Environmental Learning Center in September; 2006. (Workshop summary report). Co-sponsored by: Northwest Clean Air Agency;
National Park Service; U.S. Forest Service; U.S. Geological Survey; U.S. Department of Agriculture (USDA).
http://nadp.sws.uiuc.edu/dad/PNWCL2006WS.pdf.
Wade AJ; Neal C; Whitehead PG; Flynn NJ. (2005). Modelling nitrogen fluxes from the land to the coastal zone in European systems: Q
perspective from the INCA project. J Hydrol, 304,413-429.
Wagner T; LeueC; Wenig M; Pfeilsticker K; Piatt U. (2001). Spatial and temporal distribution of enhanced boundary layer BrO concentrations
measured by the GOME instrument aboard ERS-2. J Geophys Res, 106,24225-24235.
Waiwood BA; Haya K. (1983). Levels of chorionase activity during embryonic development of Salmo salar under acid conditions. Bull Environ
Contam Toxicol, 30,511-515.
R-93
-------
Wake DB. (1991). Declining amphibian populations. Science, 253,860.
Walcek CJ; Stockwell WR; Chang JS. (1990). Theoretical estimates of the dynamic, radiative and chemical effects of clouds on tropospheric
trace gases. Atmospheric Research, 25,53-69.
Walcek CJ; Taylor GR. (1986). A theoretical method for computing vertical distributions of acidity and sulfate production within cumulus clouds.
Journal of the Atmospheric Sciences, 43,339-355.
Walega JG; Stedman DH; Shetter RE; Mackay Gl; Iguchi T; Schiff HI. (1984). Comparison of a chemiluminescent and a tunable diode laser
absorption technique for the measurement of nitrogen oxide, nitrogen dioxide, and nitric acid. Environ Sci Technol, 18,823-826.
Wales DL; Liimatainen VA. (1987). Preliminary assessment of the current impact and potential risk of acidic deposition on walleye populations
in Ontario. (Fisheries acidification series report no. 87-11). Toronto, ON, Canada; Ministry of Natural Resources.
Walker JT; Whitall DR; Robarge W; Paerl HW. (2004). Ambient ammonia and ammonium aerosol across a region of variable ammonia
emission density. Atmos Environ, 38,1235-1246.
Wall DH. (1999). Biodiversity and ecosystem functioning. Bioscience, 49,107-108.
Wall DH; Moore JC. (1999). Interaction underground: Soil biodiversity, mutualism, and ecosystem processes. Bioscience, 49,109-117.
Wallace JB; Webster JR; Lowe RL. (1992). High-gradient streams of the Appalachians. In: Hackney CT; Adams SM; Martin WH (Eds.),
Biodiversity of southern United States aquatic communities, (pp 133-191). New York, NY: John Wiley & Sons.
Walse C; Berg B; Sverdrup H. (1998). Review and synthesis on experimental data on organic matter decomposition with respect to the effect of
temperature, moisture and acidity. Environmental Review, 6,25-40.
Walsh RG; Bjonback RD; Aiken RA; Rosenthal DH. (1990). Estimating the public benefits of protecting forest quality. J Environ Manage, 30,
175-189.
Walsh RG; Ward FA; Olienyk JP. (1989). Recreational demand for trees in national forests. J Environ Manage, 28,255-268.
Wang C; Corbett JJ; Firestone J. (2007). Modeling Energy Use and Emissions from North American Shipping: Application of the Ship Traffic,
Energy, and Environment Model. Environ Sci Technol, 41,3226-3232.
Wang D; Lu WZ. (2006). Forecasting of ozone level in time series using MLP model with a novel hybrid training algorithm. Atmos Environ, 40,
913-924.
Wang LH; Milford JB; Carter WPL. (2000a). Reactivity estimates for aromatic compounds. Part 2. Uncertainty in incremental reactivities. Atmos
Environ, 4349-4360.
Wang LH; Milford JB; Carter WPL. (2000b). Reactivity estimates for aromatic compounds. Part 1: Uncertainty in chamber-derived parameters.
Atmos Environ, 34,4337-4348.
Wang X-K; Lu W-Z. (2006). Seasonal variation of air pollution index: Hong Kong case study. Chemosphere, 63,1261-1272.
Wang Y; Tao W-K; Pickering KE; Thompson AM; Kain JS; Adler RF; Simpson J; Keehn PR; Lai GS. (1996). Mesoscale model simulations of
TRACE A and preliminary regional experiment for storm-scale operational and research meteorology convective systems and
associated tracer transport. J Geophys Res, 101,24013-24027.
Wang Y; DeSilva AW; Goldenbaum GC; Dickerson RR. (1998). Nitric oxide production by simulated lightning: dependence on current, energy,
and pressure. J Geophys Res, 103,19149-19159.
Warby RAF; Johnson CE; Driscoll CT. (2005). Chemical recovery of surface waters across the northeastern U.S. from reduced inputs of acidic
deposition: 1984-2001. Environ Sci Technol, 39,6548-6554.
Warby RAF; Johnson CE; Driscoll CT. (2008). Continuing Acidification of Organic Soils across the Northeastern USA: 1984-2001. Soil Sci.
Soc. Am. J, 73, 274-284. doi:10.2136/sssaj2007.0016
Ward BB. (2003). Significance of anaerobic ammonium oxidation in the ocean. Trends in Microbiology, 11,408-410.
Warfvinge P; Falkengren-Greup U; Sverdrup H; Andersen B. (1993). Modelling long-term cation supply in acidified forest stands. Environ
Pollut, 80,209-221.
Warfvinge P; Sverdrup H. (1992). Calculating critical loads of acid deposition with PROFILE - a steady-state soil chemistry model. Water Air
Soil Pollut, 63,119-143.
Wargo PM. (1988). Amino nitrogen and phenolic constituents of bark of American beech, Fagus grandifolia, and infestation by beech scale,
Cryptococcus fagisuga. Eur J Forest Pathol, 18, 279-290.
Wargo PM; Tilley J; Lawrence G; David M; Vogt K; Vogt D; Holifield Q. (2003). Vitality and chemistry of roots of red spruce in forest floors of
stands with a gradient of soil Al/Ca ratios in the northeastern United States. Can J Forest Res, 33,635-652.
Warneck P. (1999a). Chemistry of the Natural Atmosphere (2nd edition). San Diego, CA: Academic Press.
Warneck P. (1999b). The relative importance of various pathways for the oxidation of sulfur dioxide and nitrogen dioxide in sunlit continental
fair weather clouds. Phys Chem Chem Phys, 1,5471-5483.
Watmough SA; Dillon PJ. (2003). Calcium losses from a forested catchment in south-central Ontario, Canada. Environ Sci Technol, 37,3085-
3089.
Watmough SA; Aherne J; Dillon PJ. (2005). Effect of declining lake base cation concentration on freshwater critical load calculations. Environ
Sci Technol, 39, 3255-3260.
Watmough SA; Dillon PJ. (2003). Do critical load models adequately protect forests? A case study in south-central Ontario. Can J For Res, 33,
1544-1556.
Watmough SA; Eimers MC; Aherne J; Dillon PJ. (2004). Climate effects on nitrate export from forested catchments in south-central Ontario.
Environ Sci Technol, 38, 2383-2388.
Watras CJ; Frost TM. (1989). Little Rock (Wisconsin): Perspecitives on an experimental ecosystem approach to seepage lake acidification.
Archives of environmental contamination and toxicology, 18,157-165.
Watras CJ; Morrison KA. (2008). The response of two remote, temperate lakes to changes in atmospheric mercury deposition, sulfate, and the
water cycle. Can J Fish Aquat Sci, 65.
R-94
-------
Watras CJ; Morrison KA; Regnell 0; Kratz TK. (2006). The methylmercury cycle in Little Rock Lake during experimental acidification and
recovery. Limnol Oceanogr, 51,257-270.
Watson JG; Chow JC; Lurmann FW; Musarra SP. (1994). Ammonium nitrate, nitric acid, and ammonia equilibrium in wintertime Phoenix,
Arizona. J Air Waste Manage Assoc, 44,405-412.
Watson JG; Fujita EM; Chow JC; Zielinska B; Richards LW; Neff W; Dietrich D. (1998). Northern front range air quality study. (Final report).
http://www.nfraqs.colostate.edu/index2. Fort Collins, CO; Cooperative Institute for Research in the Atmosphere; Colorado State
University.
Watts SH; Seitzinger SP. (2001). Denitrification rates in organic and mineral soils from riparian sites: a comparison of N2 flux and acetylene
inhibition methods. Soil Biol Biogeochem, 22,331-335.
Wayland M; McNicol DK. (1990). Status report on the effects of acid precipitation on common loon reproduction in Ontario: the Ontario Lakes
loon survey. Ottawa, ON, Canada: Canadian Wildlife Service.
Weathers KC; Lovett GM; Likens GE; Lathrop R. (2000). The Effect of Landscape Features on Deposition to Hunter Mountain, Catskill
Mountains, New York. Ecol Appl, 10,528-540.
Webb JR; Cosby BJ; Galloway JN; Hornberger GM. (1989). Acidification of native brook trout streams in Virginia. Water Resour Res, 25(6),
1367-1377.
Webb JR; Deviney FA; Galloway JN; Rinehart CA; Thompson PA; Wilson S. (1994). The acid-base status of native brook trout streams in the
mountains of Virginia. Charlottesville, VA; Department of Environmental Sciences; University of Virginia.
Webb JR; Cosby BJ; Deviney FA Jr; Eshleman KN; Galloway JN. (1995). Change in the acid-base status of an Appalachian Mountain
catchment following forest defoliation by the gypsy moth. Water Air Soil Pollut, 85,535-540.
Webb JR. (1999). Synoptic stream water chemistry. In: Bulger AJ; Cosby BJ; Dolloff CA; Eshleman KN; Webb JR; Galloway JN (Eds.),
Shenandoah National Park: Fish in sensitive habitats (SNP:FISH); final report vol. II, Ch. 3. Prepared for: National Park Service.
Charlottesville, VA; University of Virginia; Department of Environmental Sciences.
Webb JR; Cosby BJ; Deviney FA; Galloway JN; Maben SW; Bulger AJ. (2004). Are brook trout streams in western Virginia and Shenandoah
National Park recovering from acidification? Environ Sci Technol, 38,4091-4096.
Weber JH. (1993). Review of possible paths for abiotic methylation of mercury(ll) in the aquatic ecosystem. Chemosphere, 26,2063-2077.
Weber P; Nussbaum S; Fuhrer J; Gfeller H; Schlunegger UP; Brunold C; Rennenberg H. (1995). Uptake of atmospheric 15N02 and its
incorporation into free amino-acids in wheat (Triticum-aestivum). Physiologia Plantarum, 94,71-77.
Weber P; Thoene B; Rennenberg H. (1998). Absorption of atmospheric NO2 by spruce (Picea abies).trees. III. Interaction with nitrate reductase
activity in the needles and phloem transport. Plant Biology, 111,377-382.
Webster KE; Newell AD; Baker LA; Brezonik PL. (1990). Climatically induced rapid acidification of asoftwater seepage lake. Nature, 347, 374-
376.
Webster KE; Brezonik PL; Holdhusen BJ. (1993). Temporal trends in low alkalinity lakes of the Upper Midwest (1983-1989). Water Air Soil
Pollut, 67, 397-414.
Webster KL; Creed IF; Nicholas NS; Miegroet HV. (2004). Exploring interactions between pollutant emissions and climatic variability in growth
of red spruce in the Great Smoky Mountains National Park. Water Air Soil Pollut, 159, 225-248.
Webster PJ; Holland GJ; Curry JA; Chang H-R. (2005). Changes in tropical cyclone number, duration, and intensity in a warming environment.
Science, 309,1844-1846.
Wedemeyer GA; Barton BA; MeLeay DJ. (1990). Stress and acclimation. In: Schreck CB, Moyle PB (Eds.), Methods for fish biology (pp. 178-
196). Bethesda, MD: American Fisheries Society.
Wedin DA; Tilman D. (1996). Influence of Nitrogen Loading and Species Composition on the Carbon Balance of Grasslands. Science, 274,
1720-1723.
Weintraub MN; Schimel JP. (2005). Nitrogen cycling and the spread of shrubs control changes in the carbon balance of Arctic tundra
ecosystems. Bioscience, 55,408-415.
Weis W; Rotter V; Gottlein A. (2006).Water and element fluxes during the regeneration of Norway spruce with European beech: Effects of
shelterwood-cut and clear-cut. Water Air Soil Pollut, 224,304-317.
Weiss SB. (1999). Cars, cows, and checkerspot butterflies: Nitrogen deposition and management of nutrient-poor grasslands for a threatened
species. Conserv Biol, 13,1476-1486.
Weiss SB. (2006). Impacts of nitrogen deposition on California ecosystems and biodiversity (No. CEC-500-2005-165). Santa Barbara, CA:
Prepared for California Energy Commission Report by Bren School of Environmental Science, University of California.
Weissenrieder J; Kleber C; Schreiner M; Leygraf C. (2004). In Situ Studies of Sulfate Nest Formation on Iron. J Electrochem Soc, 151, B497.
Weitz AM; Keller M; Linder E; Crill PM. (1999). Spatial and temporal variability of nitrogen oxide and methane fluxes from a fertilized tree
plantation in Costa Rica. J Geophys Res-Atmos, 104, 30097-30107.
Wellburn AR. (1990). Why are atmospheric oxides of nitrogen usually phytotoxic and not alternative fertilizers? Tansley Review 24. New Phytol,
115, 395-429.
Wendel GJ; Stedman DH; Cantrell CA; Damrauer L. (1983). Luminol-based nitrogen dioxide detector. Anal Chem, 55,937-940.
Wesely ML; Hicks BB. (1977). Some factors that affect the deposition rates of sulfur dioxide and similar gases on vegetation. J Air Waste
Manage Assoc, 27,1110-1116.
Wesely ML. (1989). Parameterization of surface resistances to gaseous dry deposition in regional-scale numerical models. Atmos Environ, 23,
1293-1304.
Wesely ML; Lesht BM. (1989). Comparison of RADM dry deposition algorithms with a site-specific method for inferring dry deposition. Water
Air Soil Pollut, 44, 3-4.
Wesely ML; Hicks BB. (2000). A review of the current status of knowledge on dry deposition. Atmos Environ, 34,2261-2282.
R-95
-------
Westenbarger DV; Frisvold GB. (1995). Air pollution and farm-level crop yields: An empirical analysis of corn and soybeans. Agric Resour Econ
Rev, 24,156-165.
Westerling AL; Hildalgo HG; Cayan DR; Swetnam TW. (2006). Warming and earlier spring increase western U.S. forest wildfire activity.
Science, 313, 940-943.
Westman WE. (1977). How much are nature's services worth? Measuring the social benefits of ecosystem functioning is both controversial and
illuminating. Science, 197,960-964.
Westman WE; Preston KP; Weeks LB. (1985). SO2 effects on the growth of native plants. In: Winter WE; Mooney HA; Goldstein RA (Eds.),
Sulfur dioxide and vegetation: Physiology, ecology, and policy issues, (pp 264-279). Stanford, CA: Stanford University Press.
WetmoreCM. (1985). Lichens and air quality in Isle Royale National Park. Denver, CO; National Park Service ; U.S. Department of the Interior.
Wetzel RG. (2001). Limnology: lake and river ecosystems. San Diego, CA: Academic Press.
Wexler AS; Seinfeld JH. (1991). Second-generation inorganic aerosol model. Atmos Environ Part A General Topics, 25,2731-2748.
Whitall D; Bricker S. (2006). Assessment of eutrophication in estuaries: pressure-state-response and source apportionment. (USDA Forest
Service Proceedings RMRS-P-42CD) Silver Spring, MD: National Oceanic and Atmospheric Administration National Centers for
Coastal Ocean Science Center for Coastal Monitoring and Assessment. U.S. Forest Service; U.S. Department of Agriculture (USDA).
Whitall D; Castro M; Driscoll C. (2004). Evaluation of management strategies for reducing nitrogen loadings to four U.S. estuaries. Sci Total
Environ, 333,25-36.
Whitall DR; Paerl HW. (2001). Spatiotemporal variability of wet atmospheric nitrogen deposition to the Neuse River estuary, North Carolina. J
Environ Qual, 30,1508-1515.
White GN; Zelazny LW. (1986). Charge properties of soil colloids. In: Sparks DL (Ed.), Soil Physical Chemistry (pp. 39-81). Boca Raton, FL:
CRC Press.
White PS; Cogbill CV. (1992). Spruce-fir forests of eastern North America. In: Eagar C; Adams MB (Eds.) Ecology and decline of red spruce in
the eastern United States. (pp 3-39). New York, NY: Springer-Verlag.
Whitehead JC; Groothuis PA. (1992). Economic benefits of improved water quality: A case of North Carolina's Tar-Pamlico River. Rivers, 3,
170-178.
Whitehead JC; Haab TC; Huang J-C. (2000). Measuring recreation benefits of quality improvements with revealed and stated behavior data.
Resource and Energy Economics, 22,339-354.
Whitehead PG; Barlow J; Haworth EY; Adamson JK. (1997). Acidification in three Lake District tarns: Historical long term trends and modelled
future behaviour under changing sulphate and nitrate deposition. HydrolEarth System Sci, 1,197-204.
Whitehead PG; Reynolds B; Hornberger GM; Neal C; Cosby BJ; Paricos P. (1988). Modeling long term stream acidification trends in upland
Wales at Plynlimon. Hydrological Processes, 2,357-368.
Whiting MC; Whitehead DR; Holmes RW; Norton SA. (1989). Paleolimnological reconstruction of recent acidity changes in four Sierra Nevada
lakes. J Paleolimnol, 2,285-304.
Whitmore ME; Mansfield TA. (1983). Effects of long-term exposure to S02 and NO2 on Poa pratensis and other grasses. Environ Pollut (Series
A), 31, 217-235.
Whittier TR; Paulsen SB; Larsen DP; Peterson SA; Herlihy AT; Kaufmann PR. (2002). Indicators of ecological stress and their extent in the
population of northeastern lakes: a regional-scale assessment. Bioscience, 52,235-247.
Wiemeyer SN; Schmeling SK; Anderson A. (1987). Environmental pollutant and necropsy data for ospreys from the eastern United States,
1975-1982. J Wildl Dis, 23, 279-291.
Wiener JG; Krabbenhoft DP; Heinz GH; Scheuhammer AM. (2003). Ecotoxicology of mercury. In: Hoffman D J; Rattner BA; Burton GA; Cairns
J (Eds.), Handbook of Ecotoxicology. 2nd edition, (pp. 409-643). Boca Raton, FL: CRC Press.
Wiener JG; Knights BC; Sandheinrich MB; Jeremiason JD; Brigham ME; Engstrom DR; Woodruff LG; Cannon WF; Balogh SJ. (2006). Mercury
in Soils, Lakes, and Fish in Voyageurs National Park, Minnesota: Importance of Atmospheric Deposition and Ecosystem Factors.
Environ Sci Technol, 40, 6261.
Wigand C; McKinney RA; Charpentier MA; Chintala MM; Thursby GB. (2003). Relationships of nitrogen loadings, residential development, and
physical characteristics with plant structure in New England salt marshes. Estuaries and Coasts, 26,1494-1504.
Wigington PJ. (1990). Episodic Acidification of Surface Waters Due to Acidic Deposition. (NAPAP State of the Science (SOS).Technical Report
#12: Summary Compendium). Washington, DC; National Acid Precipitation Assessment Program (NAPAP).
Wigington PJ Jr; Davis TD; Tranter M; Eshleman KN. (1991). Episodic acidification of surface waters due to acidic deposition. In: Irving PM
(Ed.), Acidic deposition. (NAPAP State of science and technology report no 12, Volume II: Aquatic processes and effects).
Washington, DC; National Acid Precipitation Assessment Program (NAPAP).
Wigington PJ Jr; Baker JP; DeWalle DR; Kretser WA; Murdoch PS; Simonin HA; Van Sickle J; McDowell MK; Peck DV; Barchet WR. (1993).
Episodic acidification of streams in the northeastern United States: Chemical and biological results of the episodic response project
(Report no EPA/600/R-93/190). Corvallis, OR; Environmental Research Lab; National Health and Enviornmental Effects Research
Laboratory; Office of Research and Development; U.S. Environmental Protection Agency.
Wigington PJ; Baker JP; DeWalle DR; Kretser WA; Murdoch PS; Simonin HA; Van Sickle J; McDowell MK; Peck DV; Barchet WR. (1996a).
Episodic acidification of small streams in the northeastern United States: Episodic response project. Ecol Appl, 6, 374-388.
Wigington PJ Jr; DeWalle DR; Murdoch PS; Kretser WA; Simonin HA; Van Sickle J; Baker JP. (1996b). Episodic acidification of small streams
in the northeastern United States: Ionic controls of episodes. Ecol Appl, 6,389-407.
Wigington PJ. (1999). Episodic acidification: Causes, occurrence and significance to aquatic resources. In: Sharpe WE, Drohan JR (Eds.),
Proceedings of the 1998 Pennsylvania acidic deposition conference. Vol. 2: The effects of acidic deposition on aquatic ecosystems in
Pennsylvania. September 14-16; 1998. University Park, PA. (pp 1-5). University Park, PA; Environmental Resources Research
Institute; The Pennsylvania State University.
R-96
-------
Wilcox G; Decosta J. (1982). The effect of phosphorus and nitrogen addition on the algal biomass and species composition of an acidic lake.
Archiv fur Hydrobiologie, 94,393-424.
Williams EJ; Parrish DD; Fehsenfeld FC. (1987). Determination of nitrogen oxide emissions from soils: results from a grassland site in
Colorado, United States. JGR J Geophys Res Part D, Atmospheres, 92,2173-2179.
Williams EJ; Guenther A; Fehsenfeld FC. (1992). An inventory of nitric oxide emissions from soils in the United States. J Geophys Res, 97,
7511-7519.
Williams MR; Melack JM. (1997). Effects of prescribed burning and drought on the solute chemistry of mixed-conifer forest streams of the
Sierra Nevada, California. Biogeochem, 39,225-253.
Williams MW; Melack JM. (1991). Precipitation chemistry in and ionic loading to an alpine basin. Sierra Nevada Water Resour Res, 27,1563-
1574.
Williams MW; Bales RC; Brown AD; Melack JM. (1995). Fluxes and transformations of nitrogen in a high-elevation catchment, Sierra Nevada.
Biogeochemistry, 28,1-31.
Williams MW; Platts-Mills T; Caine N. (1996a). Landscape controls on surface water nitrate concentrations at catchment and regional scales in
the Colorado Rocky Mountains. Paper presented at the Chapman Conference: Nitrogen cycling in forested catchments, held in Sun
River, OR; September 16-20,1996.
Williams MW; Baron JS; Caine N; Sommerfeld R; Sanford JR. (1996b). Nitrogen saturation in the Rocky Mountains. Environ Sci Technol, 30,
640-646.
Williams MW; Bardsley T; Rikkers M. (1998). Overestimation of snow depth and inorganic nitrogen wetfall using NADP data, Niwot Ridge,
Colorado. Atmos Environ, 32,3827-33
Williams MW; Tonnessen KA. (2000). Critical loads for inorganic nitrogen deposition in the Colorado Front Range, USA. Ecol Appl, 10,1648-
1665.
Willig RD. (1976). Consumer's Surplus Without Apology. Am Econ Rev, 66,589-597.
Wilmot TR; Ellsworth DS; Tyree MT. (1995). Relationships among crown condition, growth, and stand nutrition in seven northern Vermont
sugarbushes. Can J For Res, 25,386-397.
Wilson MA; Carpenter SR. (1999). Economic valuation of freshwater ecosystem services in the United States: 1971-1997. Ecol Appl, 9,772-
783.
Wilson PR; Karl BJ; Toft RJ; Beggs JR; Taylor RH. (1998). The role of introduced predators and competitors in the decline of kaka (Nestor
meridionalis).populations in New Zealand. Biological Conservation, 83,175-185.
Wilson SD; Tillman D. (1991). Components of plant competition along an experimental gradient of nitrogen availability. Ecology, 72,1050-1065.
Winberry WT Jr; Ellestad T; Stevens B. (1999). Compendium method IO-4.2: determination of reactive acidic and basic gases and strong
acidity of atmospheric fine particles (< 2.5 (jm). In: Compendium of methods for the determination of inorganic compounds in ambient
air. (Report no. EPA/625/R-96/010a). Cincinnati, OH; Center for Environmental Research Information; U.S. Environmental Protection
Agency.
Winfrey MR; Rudd JWM. (1990). Environmental factors affecting the formation of methylmercury in lowpH lakes. Environ Toxicol Chem, 9,
853-869.
Winner WE; Atkinson CJ. (1986). Absorption of air pollution by plants, and consequences for growth. Trends Ecol Evol, 1,15-18.
Winterbourn MJ; Collier KJ. (1987). Distribution of benthic invertebrates in acid, brown water streams in the South Island of New Zealand.
Hydrobiologia, 153,255-286.
Wislicenus H. (1914). Sammlung von Abhandlungen uber Abgase und Rauschaden. Berlin, Germany; P Parey.
Wittig AE; Solomon PA. (2006). Preface: Special issue of Atmos Environ for particulate matter supersites program and related studies. Atmos
Environ, 40,179-181.
Witting PK; Zeng B; Wong M; McMahon AC; Rayner BS; Sapir AJ; Lowe HC; Freedman SB; Brieger DB. (In Press). Polymorphonuclear
leukocyte phagocytic function increases in plasminogen knockout mice. Thrombosis research, In Press, Corrected Proof.
Witz S; Eden RW; Wadley MW; Dunwoody C; Papa R; Torre KJ. (1990). Rapid loss of particulate nitrate, chloride and ammonium on quartz
fiber filters during storage. J Air Waste Manage Assoc, 40, 53-61.
Wodzinski RS; Labeda DP; Alexander M. (1977). Toxicity of SO2 and NOx: Selective inhibition of blue-green algae by bisulfite and nitrite. J Air
Pollut Control Assoc, 27,891-893.
Wold AP; Hershey AE. (1999). Spatial and temporal variability of nutrient limitation in 6 North Shore tributaries to Lake Superior. J North Am
Benthol Soc, 18,2-14.
Wolf I; Russow R. (2000). Different pathways of formation of N2O, N2 and NO in black earth soil. Soil Biol Biochem, 32,229-239.
Wolfe AP; Baron JS; CornettRJ. (2001). Anthropogenic nitrogen deposition induces rapid ecological changes in alpine lakes of the Colorado
Front Range (USA). J Paleolimnol, 25,1-7.
Wolfe AP; Van Gorpe AC; Baron JS. (2003). Recent ecological and biogeochemical changes in alpine lakes of Rocky Mountain National Park
(Colorado, USA): A response to anthropogenic nitrogen deposition. Geobiology, 1,153-168.
Wolfe AP; Cooke CA; Hobbs WO. (2006). Are current rates of atmospheric nitrogen deposition influencing lakes in the eastern Canadian
Arctic? Arct Antarct Alp Res, 38,465-476.
Wolfe MF; Schwarzbach S; Sulaiman RA. (1998). Effects of mercury on wildlife: A comprehensive review. Environ Toxicol Chem, 17,146-160.
Wolfe MH; Joslin JD. (1989). Honeylocust (Gleditsia triacanthos L.) root response to aluminum and calcium. Plant Soil, 119,181-185.
Wolff GT. (1993). On a NOx-focused control strategy to reduce O3. J Air Waste Manage Assoc, 43,1593-1596.
Wollmer H; Kottke I. (1990). Fine root studies in situ and in the laboratory. Environ Pollut, 68,383-407.
Wood CM; McDonald DG. (1982). Physiological mechanisms of acid toxicity to fish. In: Johnson RE (Ed.), Acid rain/fisheries (pp. 197-226).
Bethesda, MD: American Fisheries Society.
R-97
-------
Wood CM; McDonald DG. (1987). The physiology of acid/aluminum stress in trout. Ann Soc RZool Belg, 177, 399-410.
Wood CM. (1989). The physiological problems offish in acid waters. Acid Toxicity and Aquatic Animals, 125-152.
Wood CM; McDonald DG; Ingersoll CG; Mount DR; Johannsson OE; Landsberger S; Bergman HL. (1990). Effects of water acidity, calcium,
and aluminum on whole body ions of brook trout (Salvelinus fontinalis) continuously exposed from fertilization to swim-up: A study by
instrumental neutron activation analysis. Can J Fish Aquat Sci, 47,1593-1603.
Wood YA; Meixner T; Shouse, PJ; Allen, EB. (2006). Altered ecohydrologic response drives native Shrub loss under conditions of elevated
nitrogen deposition. J Environ Qual, 35,76-92.
Woodmansee RG. (1978). Additions and losses of nitrogen in grassland ecosystems. Bioscience, 28,448-453.
Woodward DF; Farag AM; Mueller ME; Little EE; Vertucci FA. (1989). Sensitivity of endemic Snake River cutthroat trout to acidity and elevated
aluminum. Trans Am Fish Soc, 118,630-641.
Woodward DF. (1991). Sensitivity of greenback cutthroat trout to acidic pH and elevated aluminum. Trans Am Fish Soc, 120,34-42.
Woodward RT; Wui Y-S. (2001). The economic value of wetland services: A meta-analysis. Ecol Econ 37, 257-270.
Woodwell GM. (1970). Effects of pollution on the structure and physiology of ecosystems: Changes in natural ecosystems caused by many
different types of disturbances are similar and predictable. Science, 168,429-433.
World Resources Institute (2000). World resources 2000-2001: People and ecosystems: The fraying web of life. Washington, DC; United
Nations; United Nations Environment Programme (UNEP); Development Programme (UNDP); World Bank; World Resources Institute
(WRI).
Wright RF; Snekvik E. (1978). Acid precipitation: chemistry and fish populations in 700 lakes in southernmost Norway. Verh Int Ver Theor
Angew Limnol, 20,765-775.
Wright RF; Lotse E; Semb A. (1988a). Reversibility of acidification shown by whole-catchment experiments. Nature, 334, 670-675.
Wright RF; Norton SA; Brakke DF; FrognerT. (1988b). Experimental verification of episodic acidification of freshwaters by sea salts. Nature,
334,422-424.
Wright RF; Cosby BJ; Flaten MB; Reuss JO. (1990). Evaluation of an acidification model with data from manipulated catchments in Norway.
Nature, 343,53-55.
Wright RF; Lotse E; Semb A. (1993). RAIN project: Results after 8 years of experimentally reduced acid deposition to a whole catchment. Can
J Fish Aquat Sci, 50, 258-268.
Wright RF; Cosby BJ; Ferrier RC; Jenkins A; Bulger AJ; Harriman R. (1994). Changes in the acidification of lochs in Galloway, southwestern
Scotland, 1979-1988: The MAGIC model used to evaluate the role of afforestation, calculate critical loads, and predict fish status. J
Hydrol, 161,257-285.
Wright RF; Emmett BA; Jenkins A. (1998). Acid deposition, land-use change and global change: MAGIC7 model applied to Risdalsheia,
Norway (RAIN and CLIMEX projects) and Aber, UK (NITREX project). Hydrol Earth System Sci, 2,385-397.
Wright RF; Larssen T; Camarero L; Cosby BJ; Ferrier RC; Helliwell R; Forsius M; Jenkins A; Kopacek J; MajerV; Moldan F; Posch M; Rogora
M; Schopp W. (2005). Recovery of acidified European surface waters. Environ Sci Technol, 39, 64A-72A.
Wulff A; Karenlampi L. (1996). Effects of long-term open-air exposure to fluoride, nitrogen compounds and S02 on visible symptoms, pollutant
accumulation and ultrastructure of Scots pine and Norway spruce seedlings. Trees-Struct Funct, 10,157-171.
Wyers GP; Otjes RP; Slanina J. (1993). A continuous-flow denuder for the measurement of ambient concentrations and surface-exchange
fluxes of ammonia. Atmos Environ, 27,2085-2090.
Xie SD; Hao JM; Zhou ZP; Qi L; Yin H. (1995). Assessment of critical loads in Liuzhou, China using static and dynamic models. Water Air Soil
Pollut, 85, 2401-2406.
Xu J; DuBois D; Pitchford M; Green M; Etyemezian V. (2006). Attribution of sulfate aerosols in Federal Class I areas of the western United
States based on trajectory regression analysis. Atmospheric Environment: 40,3433-3447.
Xun L; Campbell NER; Rudd JWM. (1987). Measurements of specific rates of net methyl mercury production in the water column and surface
sediments of acidified and circumneutral lakes. Can. J Fish Aquat Sci, 44,750-757.
Yan ND; Keller W; Somers KM; Pawson TW; Girard RE. (1996a). Recovery of crustacean zooplankton communities from acid and metal
contamination: comparing manipulated and reference lakes. Can J Fish Aquat Sci, 53,1301-1327.
Yan ND; Welsh PG; Lin H; Taylor DJ; Filion J-M. (1996b). Demographic and genetic evidence of the long-term recovery oWaphnia galeata
mendotae (Crustacea: Daphniidae) in Sudbury lakes following additions of base: the role of metal toxicity. Can J Fish Aquat Sci, 53,
1328-1344.
Yan XY; Akimoto H; Ohara T. (2003). Estimation of nitrous oxide, nitric oxide and ammonia emissions from croplands in East, Southeast and
South Asia. Glob Chang Biol, 9,1080-1096.
Yanai RD; Phillips RP; Arthur MA; Siccama TG; Hane EN. (2005). Spatial and Temporal Variation in Calcium and Aluminum in Northern
Hardwood Forest Floors. Water Air Soil Pollut, 160,109-118.
Yanai RD; Siccama TG; Arthur MA; Federer CA; Friedland AJ. (1999). Accumulation and depletion of base cations in forest floors in the
northeastern United States. Ecology, 80, 2774-2787.
Ye XM; Hao JM; Duan L; Zhou ZP. (2002). Acidification sensitivity and critical loads of acid deposition for surface waters in China. Sci Total
Environ, 289,189-203.
Yeakley JA; Coleman DC; Haines BL; Kloeppel BD; Meyer JL; Swank WT; Argo BW; Deal JM; Taylor SF. (2003) Hillslope nutrient dynamics
following upland riparian vegetation disturbance. Ecosystems, 6,154-167.
Yerrapragada SS; Chirra SR; Jaynes JH; Li S; Bandyopadhyay JK; Gauri KL. (1996). Weathering Rates of Marble in Laboratory and Outdoor
Conditions. Journal of Environmental Engineering, 122,856-863.
Yienger JJ; Levy H II. (1995). Empirical model of global soil-biogenic NOx emissions. J Geophys Res, 100,11447-11464.
R-98
-------
Ying Q; Kleeman MJ. (2006). Source contributions to the regional distribution of secondary particulate matter in California. Atmos Environ, 40,
736-752.
Ylimartimo A. (1991). Effects of foliar nitrogen, potassium and magnesium concentrations on the resistance of Scots pine seedlings, to
Scleroderris canker infection. Eur J Forest Pathol, 21,414-423.
Yoshida LC; Allen EB. (2001). Response to ammonium and nitrate by a mycorrhizal annual invasive grass and a native shrub in southern
California. American Journal of Botany, 88,1430-1436.
Young TF; Sanzone S. (Eds.) (2002). A framework for assessing and reporting on ecological condition: An SAB report. (Report no. EPA-SAB-
EPEC-02-009). http://www.epa.gov/sab/pdf/epec02009.pdf. Washington, DC; Science Advisory Board; U.S. Environmental Protection
Agency.
Young VL; Kieser BN; Chen SP; Niki H. (1997). Seasonal trends and local influences on nonmethane hydrocarbon concentrations in the
Canadian boreal forest. J Geophys Res, 102,5913-5918.
Zafiriou OC; True MB. (1979). Nitrate photolysis in seawater by sunlight. Mar Chem, 8,33-42.
Zak DR; Pregitzer, KS; Curtis PS; Vogel CS; Holmes WE; Lussenhop J. (2000a). Atmospheric CO2, soil-N availability, and allocation of
biomass and nitrogen by Populus tremuloides. Ecol Appl, 10,34-46.
Zak DR; Pregitzer KS; King JS; Holmes WE. (2000b). Elevated atmospheric CO2, fine roots and the response of soil microorganisms: A review
and hypothesis. New Phytol, 147,201-222.
Zak DR; Pregitzer KS; Curtis PS; Holmes WE. (2000c). Atmospheric CO2 and the composition and function of soil microbial communities. Ecol
Appl, 10,47-59.
Zak DR; Holmes WE; Finzi AC; Norby RJ; Schlesinger WH. (2003). Soil nitrogen cycling under elevated CO2: A synthesis of forest face
experiments. Ecol Appl, 13,1508-1514.
Zanis P; Trickl T; Stohl A; Wernli H; Cooper O; Zerefos C; Gaeggeler H; Schnabel C; Tobler L; Kubik PW; Priller A; Scheel HE; Kanter HJ;
Cristofanelli P; Forster C; James P; Gerasopoulos E; Deldoo A; Papayannis A; Claude H. (2003). Forecast, observation and modelling
of a deep stratospheric intrusion event over Europe. Atmos Chem Phys, 3,763-777.
Zappia G; Sabbioni C; Riontino C; Gobbi G; Favoni O. (1998). Exposure tests of building materials in urban atmosphere. Science of the Total
Environment, The, 224,235-244.
Zavaleta ES; Shaw MR; Chiariello NR; Thomas BD; Cleland EE; Field CB; Mooney HA. (2003). Grassland responses to three years of elevated
temperature, CO2, precipitation, and N deposition. Ecol Monogr, 73,585-604.
Zeigler I. (1975). The effects of SO2 pollution on plant metabolism. Residue Rev, 56,79.
Zhai J; Driscoll CT; Sullivan TJ; Cosby BJ. (2008). Regional application of the PnET-BGC model to assess historical acidification of Adirondack
Lakes. Water Resour Res, 44, doi: 10.1029/2006WR005532 .
Zhang GJ; McFarlane NA. (1995). Sensitivity of climate simulations to the parameterization of cumulus convection in the Canadian Climate
Centre general circulation model. Atmos Ocean, 33,407-446.
Zhang JZ; Millero FJ. (1991). The rate of sulfite oxidation in seawater. Geochim Cosmochim Acta, 55,677-685.
Zhang LH; Song CC; Zheng XH; Wang DX; Wang YY. (2007). Effects of nitrogen on the ecosystem respiration, CH4 and N2O emissions to the
atmosphere from the freshwater marshes in northeast China. Environmental Geology, 52,529-539.
Zhang XQ; McMurry PH. (1987). Theoretical analysis of evaporative losses from impactor and filter deposits. Atmos Environ, 21,1779-1789.
Zhang Y; Liu P; Pun B; Seigneur C. (2006). A comprehensive performance evaluation of MM5-CMAQ for the Summer 1999 Southern Oxidants
Study episode - Part I: Evaluation protocols, databases, and meteorological predictions. Atmos Environ, 40,4825-4838.
Zhao DW; Seip HM. (1991). Assessing effects of acid deposition in southwestern China using the MAGIC model. Water Air Soil Pollut, 60,83-
97.
Zhou X; Beine HJ; Honrath RE; Fuentes JD; Simpson W; Shepson PB; Bottenheim J. (2001). Snowpack photochemical production of HONO: A
major source of OH in the Arctic boundary layer in springtime. Geophys Res Lett, 28,4087-4090.
Zhou X; Civerolo K; Dai H. (2002a). Summertime nitrous acid chemistry in the atmospheric boundary layer at a rural site in New York state. J
Geophys Res, 107, doi: 10.1029/2001JD001539.
Zhou XL; He Y; Huang G; Thornberry TD; Carroll MA; Bertman SB. (2002b). Photochemical production of nitrous acid on glass sample
manifold surface. Geophys Res Lett, 29, doi 10.1029/2002GL015080.
Zhou X; Gao H; He Y; Huang G. (2003). Nitric acid photolysis on surfaces in low-NOx environments: significant atmospheric implications.
Geophys Res Lett, 30, doi:10.1029/2003GL018620.
Zielinska B; Arey J; Atkinson R; McElroy PA. (1989). Formation of methylnitronaphthalenes from the gas-phase reactions of 1- and 2-
methylnaphthalene with OH radicals and N2O5 and their occurrence in ambient air. Environ Sci Technol, 23,723-729.
Zillioux EJ; Porcella DB; Benoit JM. (1993). Mercury cycling and effects in freshwater wetland ecosystems. Environ Toxicol Chem, 12,2245-
2264.
Zimmermann J; Poppe D. (1993). Nonlinear chemical couplings in the tropospheric NOx-HOx gas phase chemistry. J Atmos Chem , 17,141 -
155.
Zingler J; Piatt U. (2005). Iodine oxide in the Dead Sea Valley: evidence for inorganic sources of boundary layer IO. J Geophys Res, doi:
10.1029/2004JD004993.
Zulla Y; Billett MF. (1994). Long-term changes in chemical weathering rates between 1949-50 and 1987 in forest soils from north-east
Scotland. Eur J Soil Sci, 45,327-335.
R-99
------- |