&EPA
United States
Environmental Protection
Agency
.'- J<
Response of
Surface Water
Chemistry to the
Clean Air Act
Amendments
of 1990
-------
EPA620/R-03/001
January 2003
Response of Surface Water Chemistry to
the Clean Air Act Amendments of 1990
U.S. Environmental Protection Agency
Office of Research and Development
National Health and Environmental Effects Research Laboratory
Research Triangle Park, NC 27711
-------
Notice
The research described in this report has been funded by the United States Environmental
Protection Agency. This document has been reviewed according to Agency policy and approved
for publication. Mention of trade names or commercial products does not constitute
endorsement or recommendation for use.
11
-------
Dear Reader:
EPA's Office of Research and Development and collaborators are issuing a report,
"Response of Surface Water Chemistry to the Clean Air Act Amendments of 1990,"
EPA/620/R-02/004. This report suggests that a market-based approach to pollution control
works. The positive results are a product of the 1990 Clean Air Act Amendments (CAAA),
which utilize the successful market-based cap and trade program on which President Bush's
Clear Skies Initiative is modeled.
The report concludes that the CAAA regulations have resulted in a large and widespread
decrease in the deposition of wet sulfur. The amount of wet sulfur deposited in lakes and
streams declined by approximately 40 percent in the 1990s for all of the regions. Regional
declines in surface water sulfate can be directly linked to declines in emissions and deposition of
sulfur that have occurred since the 1990 CAAA.
The 1990 CAAA required control measures for coal-fired power plants, industries, and
other sources in an effort to reduce sulfur emissions that contribute to the development of acid
rain. EPA and collaborators have conducted extensive monitoring and scientific assessment
since 1990 to determine whether control measures have reduced levels of acidity in lakes and
streams in five regions of the Northern and Eastern United States most affected by acid rain.
In three regions, one-quarter to one-third of lakes and streams previously affected by acid
rain are no longer acidic. That's good news for the environment and good news for the market-
based approach.
Sincerely
Paul Oilman, Ph.D.
Assistant Administrator
in
-------
Authors
John L. Stoddard1, Jeffrey S. Kahl2, Frank A. Deviney3, David R. DeWalle4, Charles T.
Driscoll, Alan T. Herlihy , James H. Kellogg , Peter S. Murdoch , James R. Webb
Katherine E. Webster10
and
1 Western Ecology Division
U.S. Environmental Protection Agency
200 SW 35th Street
Corvallis, OR 97333
stoddard j ohn@epa.gov
2 Senator George Mitchell Center
102 Norman Smith Hall
University of Maine
Orono, ME 04469
kahl@maine.edu
3 Department of Environmental Sciences
Clark Hall
University of Virginia
Charlottesville, VA 22903
4 Pennsylvania State University
106 Land and Water Research Bldg.
University Park, PA 16802
5 Dept. Civil & Environmental Engineering
Syracuse University
Syracuse, NY 13244-1190
6 Oregon State University
c/o U.S. EPA
200 SW 35th Street
Corvallis, OR 97333
7 Dept. Environmental Conservation
Water Quality Division
103 South Main Street
Waterbury,VT 05671-0408
8 U.S. Geological Survey
425 Jordan Road
Troy, NY 12180
9 Department of Environmental Sciences
Clark Hall
University of Virginia
Charlottesville, VA 22903
10 Department of Biological Sciences
Murray Hall
University of Maine
Orono, ME 04469
Reviewers
Richard Haeuber, U.S. EPA Office of Air and Radiation
Jacques Dupont, Quebec Ministry of the Environment
Dean Jeffries, Environment Canada
Kathy Tonnessen, National Park Service
Gary Lovett, Institute of Ecosystem Studies, Millbrook, NY
N. Scott Urquhart, Department of Statistics, Colorado State University
Acknowledgements
The quality of this report was improved significantly by the comments of the reviewers. The
authors greatly appreciate their efforts and constructive input. Supplemental funding from the
U.S. EPA through a USGS interagency agreement (DW14938374) made it possible for co-author
Kahl to participate in this effort.
IV
-------
Table of Contents
Executive Summary vii
Introduction 1
Objectives 3
Background 8
Data Used in this Report 12
Trends in Emissions and Deposition 20
Response of Surface Waters, 1990-2000 29
Conclusions 64
Terms and Acronyms 68
References 70
List of Figures
Figure A. Acid sensitive regions of the northern and eastern United States vii
Figure B. Summary of regional trends in surface water chemistry ix
Figure 1. Emission sources affected under the EPA Acid Rain Program 3
Figure 2. Acid sensitive regions of the northern and eastern United States 6
Figure 3. Acid sensitive regions of the eastern United States 7
Figure 4. Distribution offish species by lake pH 11
Figure 5. Trends in sulfur dioxide emissions following implementation of Phase 1 21
Figure 6. Trends in wet sulfate deposition in the eastern United States 21
Figure 7. Trends in wet deposition concentrations of sulfate during 1990-2000 23
Figure 8. Trends in nitrogen oxide emissions following implementation of Phase I 24
Figure 9. Trends in wetNCV deposition in the eastern United States 25
Figure 10. Cumulative frequency distribution of slopes for sulfate and nitrogen 26
Figure 11. Distribution of slopes for acidity and base cations trends 28
Figure 12. Annual precipitation in the study regions 30
Figure 13. The relationship between Gran ANC and [CA2+ + Mg2+] 31
Figure 14. Time series data 34
Figure 15. Time series data 35
Figure 16. Time series data 36
Figure 17. Time series data 37
Figure 18. Time series data 38
Figure 19. Regional sulfate trends in LTM network 40
Figure 20. Regional nitrate trends in LTM network 41
Figure 21. Regional ANC trends in LTM network 44
Figure 22. Regional hydrogen ion trends in LTM network 45
Figure 23. Regional [CA2++ Mg2+] trends in LTM network 47
Figure 24. Regional DOC trends in LTM network 48
Figure 25. Relationship between trends in DOC and trends in SO42" 49
Figure 26. Regional aluminum trends in LTM network 50
-------
Figure 27. Comparison of trends in SO42"in TIME probability sites 54
Figure 28. Comparison of trends in ANC in TIME probability sites 55
Figure 29. Comparison of trends in base cations in TIME probability sites 57
Figure 30. Comparison of % change in SO42" concentration 58
Figure 31. Relationship between summer and spring ANC values at LTM sites 61
Figure 32. Paleolimnological reconstruction of historical pH change 63
Figure 33. The interaction of SC>42", N(V and base cations in East Bear Brook 65
Figure 34. Regional trends, 1990-2000 66
List of Tables
Table 1. Sources of data and sample sizes for datasets 5
Table 2. Location of NADP/NTN wet deposition sites 14
Table 3. Typical analytical methods used for samples 19
Table 4. Regional trend results for atmospheric deposition 27
Table 5. Regional trend results for Long-Term Monitoring sites 39
Table 6. Slope of trends in Gran ANC 52
Table 7. Regional trend results for populations of sites in acid sensitive regions 53
Table 8. Estimates of change in number and proportion of acidic surface waters 62
VI
-------
EXECUTIVE SUMMARY
Response of Surface Water Chemistry to the Clean Air Act Amendments of 1990
Purpose of this report. Title IV of the 1990 Clean Air Act Amendments (CAAA) set target
reductions for sulfur and nitrogen emissions from industrial sources as a means of reducing the
acidity in deposition. One of the intended effects of the reductions was to decrease the acidity of
low alkalinity waters and thereby improve their biological condition. The purpose of this report
is to assess recent changes in surface water chemistry in the northern and eastern U.S., in
response to changes in deposition. The regions covered in this report are New England (sites in
Maine, New Hampshire, Vermont and Massachusetts), the Adirondack Mountains of New York,
the Northern Appalachian Plateau (New York, Pennsylvania and West Virginia), the Ridge and
Blue Ridge provinces of Virginia, and the Upper Midwest (Wisconsin and Michigan). The data
covered in this report are from 1990 through 2000, the period since the last major science review
by the National Acidic Precipitation Assessment Program (NAPAP).
Acid Sensitive Regions of the Northern and Eastern United States
i
Figure A. Acid sensitive regions of the northern and eastern United States; this report assesses
trends in surface waters in each of these regions.
vn
-------
Substantial reductions in emissions of sulfur have occurred in the past 30 years, with the rate of
decline accelerated by Phase I of the 1990 CAAA, implemented in 1995. Modest reductions in
nitrogen emissions have occurred since 1996. The key questions are (a) whether the declines in
emissions translate into reductions in acidic deposition; and (b) whether biologically relevant
water chemistry has improved in acid sensitive regions. The measures of expected recovery
include decreased acidity, sulfate, and toxic dissolved aluminum concentrations.
Anthropogenic acidity in atmospheric deposition. NOX and SOX from the combustion of fossil
fuels react with water in the atmosphere to produce acid rain, a dilute solution of nitric and
sulfuric acids. This acidity (and the acid anions sulfate and nitrate) may travel hundreds of miles
before being deposited on the landscape. The northern and eastern U.S. receives precipitation
with mean pH that ranges from 4.3 in Pennsylvania and New York, to 4.8 in Maine and the
Upper Midwest. The acidity (hydrogen ion concentration) in precipitation in the eastern U.S. is
at least twice as high as in pre-industrial times. Atmospheric deposition is one of the most
ubiquitous non-point sources of chemicals to ecosystems.
Acid-base status of surface waters. The 1984-86 EPA National Surface Water Survey (NSWS)
estimated the number of acidic waters at 4.2% of lakes and 2.7% of stream segments in acid-
sensitive regions of the North and East. Acidic waters are defined as having acid neutralizing
capacity (ANC) less than zero (i.e., no acid buffering capacity in the water), corresponding to a
pH of about 5.2.
This report addresses the recent chemical responses in the surface waters in five regions of the
North and East that are considered sensitive to acidic deposition. The data in this report are
largely from the EPA Long Term Monitoring (LTM) and the EPA Temporally Integrated
Monitoring of Ecosystems (TIME) projects, part of EMAP (Environmental Monitoring and
Assessment Program). The regions include lakes in the Adirondacks, central and northern New
England, and the upper Midwest. Sensitive regions with small streams are found in the mid-
Atlantic region, including the northern and central Appalachian Plateau and the Ridge and Blue
Ridge provinces. Surface waters in most other regions are not sensitive to the impacts of
acidification due to the nature of the local geology.
Recent changes in atmospheric deposition. We evaluated the changes in atmospheric
deposition from the five regions during 1990-2000, using National Atmospheric Deposition
Program (NADP) data. Sulfate declined significantly at a rate between -0.75 and -1.5
|ieq/L/year. There was a sharp drop in sulfate concentrations in 1995 and 1996, followed by a
modest increase in 1997-2000, in parallel with emissions. Nitrogen (nitrate + ammonium)
declined slightly in the Northeast, and increased slightly in the Upper Midwest; most of these
changes can be attributed to changes in nitrate deposition. Base cations in deposition, which are
important for the neutralization of acidity in precipitation and in watersheds, showed no
significant changes during the decade in the East, and increased slightly in the Upper Midwest.
These changes in deposition are a continuation of trends that pre-date the 1995 implementation
of Phase I of the CAAA, and are consistent with other recent published analyses of changes in
regional deposition patterns.
Recent changes in acid base status in surface waters All regions except the Ridge/Blue
Ridge province in the mid-Atlantic showed significant declines in sulfate concentrations in
Vlll
-------
surface waters, with rates ranging from -1.5 to -3 |ieq/L/year (Figure B). These declines were
consistent with the decline in sulfate in precipitation. Nitrate concentrations decreased in two
regions with the highest ambient nitrate concentration (Adirondacks, Northern Appalachian
Plateau) but were relatively unchanged in regions with low concentrations. Dissolved Organic
Carbon (DOC) increased in each region, potentially contributing natural organic acidity to offset
the recovery from decreased acidity and sulfate in deposition.
Acid neutralizing capacity is a key indicator of recovery, as it reflects the capacity of watersheds
to buffer inputs of acidity. We expect increasing values of either ANC or pH (or both) in
response to decreasing deposition of sulfur and nitrogen from the atmosphere. ANC increased in
three of the regions (Adirondacks, Northern Appalachian Plateau and Upper Midwest) at a rate
of+1 |ieq/L/year, despite a decline in base cations (calcium + magnesium) in each region (Figure
B). The decline in base cations offsets some of the decline in sulfate, and thus limits the increase
in ANC or pH. In the Adirondacks and Northern Appalachians, surface water ANC and pH both
increased significantly in the 1990s; toxic aluminum concentrations also declined slightly in the
Adirondacks. Regional surface water ANC did not change significantly in New England or in
the Ridge/Blue Ridge.
Regional Trends, 1990-2000
(in lakes and streams)
Sulfate (ueq/L/yr)
Nitrate (ueq/L/yr)
ANC (ueq/L/yr)
Hydrogen Ion (ueq/L/yr)
Base Cations (ueq/L/yr)
DOC (mg/L/yr)
Aluminum (ug/l_/yr)
-3-2-10 1 2
Slope of Trend
^^m New England Lakes
^^m Adirondack Lakes
i Northern Appalachian Streams
^^m Upper Midwest Lakes
i Ridge and Blue Ridge Streams
Figure B. Summary of regional trends in surface water chemistry in regions covered by this report.
IX
-------
Has the number of acidic waters changed? Modest increases in ANC have reduced the
number of acidic lakes and stream segments in some regions. We estimate that there are
currently 150 Adirondack lakes with ANC less than 0, or 8.1% of the population, compared to
13% (240 lakes) in the early 1990s. In the Upper Midwest, an estimated 80 of 250 lakes that
were acidic in mid-1980s are no longer acidic. TIME surveys of streams in the northern
Appalachian Plateau region estimated that 5,014 kilometers of streams (ca. 12%) were acidic in
1993-94. We estimate that 3,600 kilometers of streams, or 8.5%, remain acidic in this region at
the present time. In these three regions, approximately one-quarter to one-third of formerly
acidic surface waters are no longer acidic, although still with very low ANC. We find little
evidence of a regional change in the acidity status of New England or the Ridge/Blue Ridge
regions and infer that the numbers of acidic waters remain relatively unchanged. There is no
evidence that the number of acidic waters have increased in any region, despite a general decline
in base cations and a possible increase in natural organic acidity.
Do changes in deposition translate into changes in surface waters? A major goal of this
assessment is to evaluate the effectiveness of emission reductions in changing surface water
chemistry. We only make this assessment for sulfate because changes in the deposition of
nitrogen have been minor. In New England, the Adirondacks and the Northern Appalachians,
the percent declines in sulfate concentrations in precipitation were generally steeper than in
surface waters. This is largely as expected and suggests that for a majority of aquatic systems,
sulfate recovery exhibits a somewhat lagged response. However, the lakes and streams with the
steepest declines in sulfate had very similar rates to those in deposition, indicating that the most
responsive watersheds responded directly and rapidly to the sulfate decrease in deposition. As
expected, there was little correspondence between rates of sulfate decline in streams and
deposition in the Ridge and Blue Ridge provinces due to the adsorptive capacity of the soils in
the region. In the upper Midwest, the rate of decline in lakes was greater than the decline in
deposition, probably reflecting the residual effects of the drought of the late 1980s. Longer term,
we expect the chemistry of seepage lakes in the Upper Midwest to mirror the decline in
deposition, similar to the pattern seen in seepage lakes in New England that did not experience
the 1980s drought.
Complications for assessing recovery. Declines in atmospheric deposition of sulfate have led
to nearly universal declines in sulfate concentrations in surface waters. This response is one
simple measure of the intended recovery in surface waters and marks a success of the CAAA and
efforts by industry in reducing 862 emissions. However, the anticipated decrease in acidity
corresponding to the decline in sulfate has been modest.
It is important to recognize that recovery will not be a linear process. Moreover, the changes in
surface water chemistry reported here have occurred over very short periods relative to the
implementation of the CAAA emission reductions in 1995. The decline in sulfate is without
question due to the decline in emissions and deposition, but mechanisms producing other
changes are much less clear. Other responses in surface waters may be partially attributable to
factors other than atmospheric deposition, such as climate change and forest maturation. In
particular, some of the observed increase in ANC may result from decreases in nitrate
concentrations (e.g., in the Adirondacks and Northern Appalachian Plateau); changes in nitrate
are unrelated to changes in nitrogen deposition and are not expected to continue. If the trend
toward lower nitrate in surface water reverses, some of the gains in ANC may be lost.
-------
We can identify at least five factors that are important in determining the recovery, or lack of
recovery, in surface waters of the northern and eastern U.S. Continued long-term research and
monitoring will be necessary to understand the causes, effects, and trends in these processes.
1) Base cations. We report declining surface water concentrations of base cations (e.g.,
calcium, magnesium) in all of the glaciated regions in this report (the Ridge and Blue
Ridge region is the only non-glaciated region). At some individual sites, further
acidification has occurred because base cations are declining more steeply than sulfate.
While decreases in base cation loss from watersheds probably indicates slower rates of
soil acidification, they nonetheless limit the magnitude of surface water recovery.
Continued long-term research at acid-sensitive sites is needed to determine the cause and
effect of the relationship between base cations and sulfate and the effects of cation loss
on soil and surface water recovery.
2) Nitrogen. Continued atmospheric loading of nitrogen may be influencing the acid-base
status of watersheds in yet undetermined ways. Unlike sulfate, concentrations of
nitrogen in deposition have not changed substantially in 20 years. Also unlike sulfate,
most nitrogen deposited from the atmosphere is retained in watershed soils and
vegetation; nitrogen sequestration is not expected to continue adinfmitum (Stoddard
1994, Aber et al., in press). We report that surface water nitrate concentrations are
largely unchanged, except in two regions characterized by high nitrate concentrations a
decade ago (Adirondacks, Northern Appalachian Plateau). The mechanisms behind
these decreases in nitrate are not understood and could include climate change, forest
recovery from disturbance, and the effects of land-use history. Future increases in
nitrate concentrations in all regions are not improbable and would retard recovery if
other factors remain constant.
3) Natural organic acidity. Increases in dissolved organic carbon in acid-sensitive waters
may have contributed additional natural organic acidity to surface waters, complicating
our interpretation of the response in acidity. This factor is an important long-term
research question that is probably linked to complex issues including climate change and
forest maturation.
4) Climate. Climatic fluctuations induce variability in surface water chemistry and thus
obscure changes that we expect to result from declining acidic deposition. Climate or
climate-related processes may counteract recovery by producing declines in base cations
to offset a decline in sulfate or by inducing an increase in natural organic acidity. These
interactions of factors underscore the need to continue monitoring a subset of sensitive
systems so as to understand the full suite of drivers and responses in ecosystems.
5) Lag in response. Documentation of the response of watersheds to changes in atmospheric
deposition may take longer than the timeframe of available data. Recovery itself may
have an inherent lag time beyond the time scale of currently available monitoring data.
Moreover, the changes observed are not unidirectional. Uncertainty with respect to
timeframes can only be resolved with continued long-term data.
Indicators of recovery. A main goal of the Title IV of the CAAA is to decrease the acidity of
affected surface waters. Although decreases in acidity have occurred in several regions,
additional factors appear to point toward recovery, forecasting an improvement in biologically
XI
-------
relevant surface water chemistry. It is not yet clear if further reductions in emissions and
deposition will be necessary for widespread recovery to occur. These factors forecast the onset
of recovery:
a) Sulfate is an increasingly smaller percentage of total ion concentration in surface waters.
b) ANC has increased modestly in three of the five regions.
c) Dissolved Organic Carbon has increased regionally, perhaps toward a more natural pre-
industrial concentration as acidity decreases in surface waters.
d) Toxic aluminum concentrations appear to have decreased slightly in some sensitive systems.
Expectations for recovery. An important consideration for measuring the success of the CAAA
is to have appropriate expectations for the magnitude of potential recovery. Lakes inferred to
have been measurably acidified by atmospheric deposition were already marginally acidic,
typically with pH less than 6, before anthropogenic atmospheric pollution began more than 100
years ago. Therefore, full recovery of acidic lakes will not yield neutral pH. However, there is
evidence that DOC will increase during recovery, and both increasing DOC and increasing pH
values will lower the toxicity of aluminum. This change may allow recovery offish populations
to historical conditions even if pH remains low.
Recommendations. In the North and East, there is evidence of recovery from the effects of
acidic deposition. The complexities of ecosystem response - effects of forest health, soil status,
natural organic acidity, the relative importance of sulfur vs. nitrogen deposition, future
emission/deposition scenarios - make predictions of the magnitude and timing of further
recovery uncertain. The results of this trend analysis suggest two recommendations for
environmental monitoring:
1) Deposition monitoring: The analyses in this report depended heavily on the long-term
NADP/NTN program for monitoring the chemistry of precipitation. The future
assessment of deposition and aquatic trends will depend heavily on these data, and
therefore our recommendation is to maintain a national precipitation chemistry network.
2) Surface water monitoring: The effectiveness of current or future amendments to the
Clean Air Act can best be determined by monitoring the response of subpopulations of
sensitive surface waters through time. Long-term records provide the benchmark for
understanding trends in ecological responses. The reviewers of early drafts of this report
strongly urged the authors to recommend the continuation of the long-term research
programs upon which this report is based and the addition of biological monitoring to
begin documenting potential biotic recovery.
Future research. The data from these long-term sites will be invaluable for the evaluation of
the response of forested watersheds and surface waters to a host of research and regulatory issues
related to acidic deposition, including soil and surface water recovery, controls on nitrogen
retention, mechanisms of base cation depletion, forest health, sinks for sulfur in watersheds,
changes in DOC and speciation of aluminum, and various factors related to climate change. As
one reviewer of this report noted, " ...these sites have irreplaceable long-term data that should
constitute a 'research infrastructure' akin to an EPA laboratory. These sites will help address
many basic science issues in which EPA ORD has a continuing interest.... " Moreover, as
xn
-------
several of the reviewers observed, long-term data serve as the foundation for ecological research
and modeling. Without such data, our ability to ask the right questions is reduced, and our
ability to base the answers to these questions on actual data is likewise compromised.
Xlll
-------
INTRODUCTION
The existence of acid rain, more properly
called acidic deposition, has been known for
more than 100 years (Smith 1872).
Documentation of acidification of surface
waters began in Scandinavia (e.g., Oden
1968), although reports of acidic lakes date
back to the 1950s in North America
(Gorham, 1957). Recognition of the issue
became common in the U.S. in the early
1970s (Likens et al., 1972), with
identification of impacts on fish by the mid-
1970s (e.g., Schofield 1976).
Acidic deposition is a dilute solution of
sulfuric acid and nitric acids derived from
emissions of SC>2 and NOX, by-products of
the combustion of fossil fuels. The regions
most impacted by acidic deposition are
downwind of the documented sources of
industrial sulfur emissions in the
Midwestern states, and located in
geologically sensitive areas in the Northeast,
Mid-Atlantic and Upper Midwest regions.
Lake and stream chemical data from these
regions are the subject of this report.
The potential and documented effects of
acidic deposition include damage to building
materials, impacts to forests or crops,
leaching of soil nutrients, acidification of
surface waters, direct impacts on aquatic or
amphibious biota, and indirect impact on
other organisms in the same foodweb. The
effects of the precursor emissions that lead
to acidic deposition also include human
respiratory stress and decreased visibility,
both effects that would be reduced if sulfur
and nitrogen emissions were reduced.
This report assesses the responses of surface
waters in the five acid-sensitive regions of
the northern and eastern states: New
England (lakes); the Adirondack Mountains
(lakes); the northern Appalachian Plateau
(streams); the Ridge/Blue Ridge province
(streams); and the Upper Midwest (lakes).
With the exception of Florida, these regions
include the vast majority of acidic surface
waters in the U.S. Florida lakes are not
addressed in this report because the majority
of the acidity there is due to natural organic
acidity (Baker et al., 1990) and because data
have not been systematically collected there
for more than a decade.
There have been major changes in emissions
during the past century. Emissions of SC>2
are inferred to have increased from 9 million
metric tons in 1900 to a peak of 29 million
metric tons in 1973 (Husar et al., 1991),
declining to 18 million metric tons by 1998.
NOx emission are inferred to have increased
from 2.5 million metric tons in 1900 to 22
million metric tons around 1990, remaining
relatively constant for more than a decade.
As a result of the changes in emissions, the
ratio of S to N in deposition in the North and
East has changed from 2:1 to 1:1, with a
minor decrease in acidity. It is within the
context of these changes in deposition that
this report evaluates the responses in surface
waters since 1990.
Aquatic Effects Research in the U.S. A
consortium of federal agencies mobilized to
address the issues of acidic deposition
effects in 1980, creating the National
Atmospheric Precipitation Assessment
Program (NAPAP). The NAPAP effort
spent more than $500 million in the 1980s,
one of the major environmental research
efforts in U.S. history. The ecological
research under NAPAP greatly enhanced
our understanding of ecological and
hydrological processes relevant to
acidification, and set the stage for the
NAPAP assessment report, and the CAAA,
in 1990.
-------
This report makes no attempt to summarize
NAPAP research results from the 1980s
because many such efforts exist (e.g., the
State of Science and Technology series from
1990, including reports 9 through 15 on
surface water status, trends, and processes in
watersheds). Instead, this report presents
results as they relate to the patterns and
trends found in the data in the 1990s.
Title IV, Clean Air Act Amendments of
1990. Despite the complication of long-
range transport across political boundaries,
legislative actions in the United State and
Canada have resulted in reductions in acidic
emissions and deposition. The purpose of
Title IV (the Federal Acid Deposition
Control Program) of the Clean Air Act
Amendments of 1990 (Public Law 101-549,
amending the Clean Air Act of 1970) were
to reduce the adverse effects of acidic
deposition through reductions in emissions
of N and S acid precursors (National Acid
Precipitation Assessment Program 1998).
The 1990 CAAA enhanced the trend toward
decreased sulfur emissions and deposition
that had been underway since the early
1970s.
Title IV largely targets the emissions of
electric utilities that are estimated to account
for 70% of S emissions and 30% of N
emissions (Government Accounting Office
2000). These reduction targets were
implemented with Phase I in 1995, and
Phase II in 2000 (Figure 1). The goal is a 10
million ton reduction in SC>2, and 2 million
ton reduction in NOx, by 2010, compared to
the baseline year of 1980. Emissions of SC>2
from utilities were 10.6 million tons per year
in 2001, compared to the baseline years of
1980 (17.3 m tons) and 1990 (15.7 m tons).
The target decrease in S emissions
represents a 40% permanent cap compared
to 1980 emissions. The target decrease in N
is 10% compared to 1980.
Phase I affected 263 units at 110 mostly
coal-burning electric utility plants located in
21 eastern and mid-western states. An
additional 182 units joined Phase I of the
program as substitution or compensating
units, bringing the total of Phase I units to
445. Phase II tightened the annual
emissions limits imposed on these large,
higher-emitting plants and also set
restrictions on smaller, cleaner plants fired
by coal, oil, and gas, encompassing over
2,000 units in all. In 2001, program sources
emitted 10.6 million tons of SC>2 — nearly 7
million tons below 1980 levels, representing
a reduction of approximately 40%.
Title IV focuses on NOx-emitting coal-fired
electric utility boilers. As with the SC>2
emission reduction requirements, the NOx
program was implemented in two phases,
beginning in 1996 and 2000. The NOx
program embodies many of the same
principles of the SO2 trading program, with
flexibility in the method to achieve emission
reductions. However, it does not cap NOx
emissions as the SO2 program does, nor does
it utilize an allowance trading system. There
has been a reduction in NOx emissions from
Title IV affected sources from 5.5 million
tons in 1990 to 4.7 million tons in 2001.
Linkages to other environmental
stressors. Although the evaluation of the
effectiveness of the Clean Air Act
Amendments is important for understanding
the direct link to surface water chemistry
and fisheries, there are ancillary advantages
to controlling emissions and deposition. For
example, coastal eutrophication is a problem
of increasing importance for estuarine
ecosystems and marine productivity
(Jaworski et al., 1997, Paerl 2002, Nielsen
and Kahl, in press).
-------
Phase I Plants
(includes substitution/com pan sating units)
Phase II Plants
Figure 1. Emission sources affected under the EPA Acid Rain Program.
Similarly, the linkage to climate change is
impossible to ignore. Reductions in S and N
emissions may be coupled to reduced
emissions of greenhouse gases and mercury,
especially if the reductions are the result of
increased efficiency or conservation. In
addition, potential climate change may have
an impact on the response of ecosystems and
complicate interpretation of data collected
for purposes of evaluating the effectiveness
of the CAAA (Webster and Brezonik 1995,
Norton et al., in press).
OBJECTIVES
The overarching objective of this report is to
evaluate the chemical response of lakes and
streams in the northern and eastern U.S.
during the past decade. This response will
be evaluated relative to the well-documented
status and trends in precipitation chemistry.
For purposes of this report, we define the
northern and eastern U.S. as including all of
the acid sensitive regions north of the
Virginia/North Carolina border and east of
the Rocky Mountains. The time period of
our focus is 1990 to 2000.
-------
Our specific objectives are to:
1) Identify trends and regional patterns in
surface water chemistry using:
a. Statistical population assessments
from the U.S. EPA Environmental
Monitoring and Assessment
Program (EMAP), and the
Temporally Integrated Monitoring
of Ecosystems (TIME) project -
these data result from repeated
probability surveys of lakes in the
Adirondacks and New England and
of streams in the Mid-Atlantic and
can be expanded statistically to the
population(s) of surface waters in
each region;
b. Data from lakes and streams
identified to be the indicator
systems most responsive to changes
in deposition chemistry, including
the U.S. EPA Long-Term
Monitoring (LTM) project - these
data are from specially selected
sensitive surface waters, and
represent the behavior of the most
affected systems.
2) Evaluate linkages in changes in surface
waters, if any, to changes in deposition
that are related to regulatory goals.
These two objectives will assist in providing
the foundation for setting targets for future
emission reductions, based on goals and
expectations for future changes in aquatic
chemistry. The results of analyses reported
here will inform debates about re-
authorization of the current CAAA and
about any future emissions regulation.
Evaluating the response: a multi-aquatic
system approach. A key issue for the
evaluation of the CAAA is the relationship
between trends in deposition and trends in
surface water chemistry. Have regulatory
changes produced the intended results in
surface waters? The data may indicate a
direct relationship between variables, a
delayed response, or no apparent
relationship. The more direct the apparent
relationship and the more types of aquatic
systems in which the relationship occurs, the
more confidence we have that the trends in
deposition are the causal factor.
The data in this report represent many
locations and types of surface waters in the
North and East (Table 1, Figures 2, 3); we
group these sites into regions based on
similarities in their acid sensitivity and the
rates of acidic deposition to which they have
historically been exposed. Consistent
recovery across a regional population of
streams or lakes provides the strongest
evidence that trends are directly related to
acidic deposition controls. The regions
covered here contained 95% of the acidic
lakes, and 84% of the acidic streams, in the
1980s (Linthurst et al., 1986a, Landers et al.,
1988). We are therefore confident that we
have both represented the bulk of aquatic
ecosystems affected by acidic deposition
and that we report regionally significant
results.
Acidic or acidified? In an assessment of
recovery, it is important to distinguish
between acidic waters and acidified waters.
Acidic describes a condition that can be
defined and documented (i.e., Gran ANC <
0) and may be due either to the effects of
acidic deposition or to natural causes such as
organic acidity or the weathering of sulfur-
containing minerals in the watershed. In
contrast, acidified refers to the process of
acidification (an increase in acidity observed
through time) and does not require that the
water body be acidic. The process of
acidification has been observed in many
long-term monitoring records (e.g., Lake
Langtjern in Norway; Henriksen and Grande
2002), in short-term studies of episodic
acidification (e.g., Wigington et al., 1996),
and experimental manipulations of aquatic
ecosystems (Schindler 1988, Kahl et al.,
-------
1993a, Wright etal., 1993). Chronic
acidification can also be inferred from
reconstructions of pH based on the remains
of diatoms or chrysophytes collected in lake
sediment cores (Charles et al., 1990, Smol et
al., 1998). The term anthropogenically-
acidified implies that acidic deposition was
responsible for any increase in acidity; the
term acidified does not imply the cause.
Table 1. Sources of data and sample sizes for datasets used in this report, along with
estimates of the condition of surface waters in each region in the 1980s. 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.
Sources of Data
Statistical Surveys
New England Lakes4
Adirondack Lakes4
Appalachian Plateau Streams
Sensitive Surface Waters
New England Lakes
Adirondack Lakes
Northern Appalachian Streams
Upper Midwest Lakes
Ridge/Blue Ridge Streams
No. of sites1
30
43
31
24
48
9
38
69
Size of Population2
4,327 lakes
1,290 lakes
72,000 stream miles
N.A.
N.A.
N.A.
N.A.
N.A.
Percent acidic in
1980s3
5%
14%
6%
5%
14%
6%
3%
5%
1 Number of monitoring sites with data available for this report (1990-2000); locations of sites are
illustrated in Figures 2 and 3.
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 (Linthurst et al.,
1986b, 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.
What is recovery? Recovery from
anthropogenic acidification can be defined
in several ways. The most easily
documented is a biologically relevant
decrease in acidity, measured as an increase
in pH or Gran ANC, a measure of the
buffering capacity of the water. However,
an expectation of large increases in pH is
unrealistic based on historical information
for sensitive lakes. Today's acidic lakes were
marginally acidic in pre-industrial times
(typical pH less than 6, inferred ANC
probably less than 20 |ieq/L; Charles and
Norton 1986, Smol et al., 1998). Progress
toward recovery should be considered to be
progress toward pre-industrial chemical
conditions, rather than progress toward pH 7.
-------
Acid Sensitive Regions of the Northern and Eastern United States
Long-Term Monitoring (LTM) Sites
Adirondack* o w
Mountains f 8>0 "
Northern
f , ..Appalachian Plate
Figure 2. Location of acid-sensitive regions of the northern and eastern U.S. covered in this report and locations of individual LTM
sites used in trend analysis.
-------
Other indicators of recovery may be
forecasting recovery, although it is not clear
if further reductions in emissions and
deposition will be necessary for widespread
recovery to occur. These indicators of
recovery include decreased sulfate
concentrations which constitute an
increasingly smaller percentage of total ion
concentration in surface waters and
decreased aluminum concentrations. Higher
dissolved organic carbon (DOC)
concentrations have been reported
ubiquitously in surface waters of the
northern hemisphere in the past decade
(Evans and Monteith 2001, Skjelkvale et al.,
2001, Jeffries et al., 2002; this report).
Higher DOC may be an indicator of
recovery, due to a decline in the protonation
and precipitation of DOC from acidification
(Krug and Frink 1983, Davis et al., 1985), or
may be driven by climate change. These
factors will be discussed later in the context
of trends results for DOC and aluminum.
Acid Sensitive Regions
of the
Eastern United States
Temporally Integrated Monitoring
of Ecosystem (TIME) Sites
£ Appalachian Plateau
Figure 3. Location of acid-sensitive regions of the northern and eastern U.S. for which statistical survey
data are available in the 1990s, and locations of individual TIME sites used in trend analysis.
-------
BACKGROUND
Precipitation chemistry. Acid rain has
received much media coverage and research
attention since widespread recognition in the
late 1970s. Acidic deposition refers to
deposition of dilute acids from the
atmosphere to the landscape. The source of
the acids is the largely the combustion of
fossil fuels that produce waste by-products
including gases such as oxides of sulfur and
nitrogen. Ammonia (NH3) is a by-product
of some natural processes, as well as
agricultural sources (e.g., application of
nitrogen fertilizers; confined animal
feedlots). In its dissolved form (NH4+) it
contributes acidity to surface waters through
the process of nitrification (Stoddard 1994).
Oxidized sulfur and nitrogen gasses are acid
precursors in the atmosphere. For example,
SO2 reacts with water in the atmosphere to
yield sulfuric acid:
SO2 + H2O+
H2SO4
An analogous reaction of water with
nitrogen oxides, symbolized as NOX, yields
nitric acid (HNO3).
In addition to wet deposition (rain, snow,
and fog), acidic deposition includes the
deposition of dry, particulate, and gaseous
acid precursors that become acidic in contact
with moisture. This dry deposition is
difficult to quantify and expensive to
measure. Inferential methods indicate that
dry deposition represents 20% to 80% of the
total deposition of acids to the landscape,
depending on factors such as location,
season, and total rainfall (Sisterson et al.,
1990, Government Accounting Office
2000).
Natural sources can also contribute
additional acidity to precipitation (Likens et
al., 1983, Keene and Galloway 1984,
Lindberg et al., 1984). Natural emissions of
sulfuric, nitric, and hydrochloric acids occur
from wetlands and geologic sources. Major
natural sources of NOX include lightning and
soil microbes. Organic acidity may arise
from freshwater wetlands and coastal
marshes. It is these natural sources that lead
to the inference that pre-industrial
precipitation in forested regions had a pH
around 5.0 (Charlson and Rodhe 1982). If
true, then modern precipitation in the North
and East is two to three times more acidic
than pre-industrial.
The acidity of precipitation is still subject to
misunderstanding. Even in pristine
environments, precipitation pH is rarely
controlled by the CO2 reaction that has an
equilibrium pH of 5.6:
H2O + CO2 -» H2CO3
Because of the many sources of acidity in
precipitation, pH 5.6 is not the benchmark
normal pH against which the acidity of
modern precipitation should be compared.
Precipitation is a variable and complex
mixture of particulates and solutes derived
from local sources and long-range transport.
For example, in arid or partly forested
regions, dust from soil and bedrock typically
neutralizes both the natural and human
sources of acidity in precipitation, yielding a
solution that may be quite basic (pH greater
than 7). In the northeastern U.S. and eastern
Canada, annual precipitation pH ranges from
4.3 in Pennsylvania, New York, and Ohio,
to 4.8 in Maine and maritime Canada
(National Atmospheric Deposition
Program/National Trends Network 2002).
Effects of acidic deposition. Acidic
deposition and emissions of acid precursors
have been implicated as a risk to human
health, including lung and cardiovascular
system impacts (National Acid Precipitation
-------
Assessment Program 1998). These effects
are typically associated with air pollution
events and poor visibility caused by high
ozone or sulfate haze. Acidic deposition is
also known to cause degradation of building
materials and monuments, such as those
constructed from marble (National Acid
Precipitation Assessment Program 1998).
At the ecosystem-scale, impacts may occur
to the terrestrial portions of both forested
and agricultural landscapes. The most
severe and widespread ecosystem effects are
recognized to occur in surface waters, where
both long-term chronic acidification and
short-term temporary episodic acidifications
are of concern. The issue of surface water
acidification is the main subject of this
report.
Why do we care about lake and stream
chemistry? Surface water chemistry is a
direct indicator of the potential deleterious
effects of acidification on biotic integrity.
Because surface water chemistry integrates
the sum of processes upstream in a
watershed, it is also an indicator of the
indirect effects of watershed-scale impacts,
such as nitrogen saturation, forest decline, or
soil acidification. Admittedly, we do not
always know how to interpret the potentially
complex interactions of upstream processes
based solely on surface water chemistry, but
their dual interpretive nature makes them
invaluable in tracking surface water and
watershed recovery.
Chronic acidification. Surface waters
become acidic when the supply of acids
from atmospheric deposition and watershed
processes exceeds the capacity of watershed
soils and drainage waters to neutralize them
(Charles 1991). Surface waters are defined
as acidic if their acid neutralizing capacity
(ANC, analogous to alkalinity) is less than
0, corresponding to pH values less than
about 5.2. We are not addressing the
impacts of acid-mine drainage in this report,
and the sites used in this assessment exclude
those with significant sources of geologic
sulfur. The chemical conditions that define
acidity are that acid anion concentrations
(sulfate, nitrate, organic acids) are present in
excess of concentrations of base cations
(typically calcium or magnesium), the
products of mineral weathering reactions
that neutralize acidity in soil or rock.
The National Surface Water Survey
(NSWS) documented the status and extent
of chronic acidification during probability
surveys conducted from 1984 through 1988
in acid sensitive regions throughout the U.S.
(Linthurst et al., 1986b, Landers et al., 1987,
Kaufmann et al., 1988). By using statistical
techniques to select representative lakes and
streams in each region surveyed, the NSWS
estimated the chemical conditions of 28,300
lakes and 56,000 stream reaches in all of the
major acid-sensitive regions of the U.S.
(Baker et al., 1990). The data collected by
the NSWS have allowed us to focus our
monitoring efforts on those areas that
contain the vast majority of
anthropogenically acidified waters in the
U.S. and that are most likely to respond to
changes in deposition as a result of the
CAAA.
The NSWS concluded that 4.2% of lakes
larger than 4 hectares and 2.7% of stream
segments in the acid-sensitive regions were
acidic. The regions represented in this
report (Table 1) are estimated to contain
95% of the lakes and 84% of the streams
that have been anthropogenically acidified
in the U.S. The Adirondacks had the largest
proportion of acidic surface waters (14%) in
the NSWS; the Adirondack Lake Survey
Corporation sampled 1469 Adirondack lakes
greater than 0.5 ha in size in 1984-87, and
estimated that 26% were acidic (Driscoll et
al., 1991). The proportions of lakes
estimated by NSWS to be acidic were
smaller in New England and the Upper
Midwest (5% and 3%, respectively), but the
large numbers of lakes in these regions
-------
translate to several hundred acidic waters in
each region.
The Valley and Ridge province and
Northern Appalachian Plateau had 5% and
6% acidic sites, respectively. The only acid-
sensitive region not assessed in the current
report is Florida, where the high proportion
of naturally acidic lakes, and a lack of long-
term monitoring data, make assessment
problematic.
Episodic acidification. Short-term
decreases in pH and Gran ANC, lasting on
the order of hours to weeks, can occur
during high flow events in both lakes and
streams. Severe episodes can produce
conditions that are as deleterious to biota as
chronic acidification (Baker et al., 1996).
The mechanisms that produce acidic
episodes include dilution, and flushing of
N(V, SC>42" and/or organic acids from forest
soils (Kahl et al., 1992, Wigington et al.,
1996, Wigington 1999, Lawrence 2002).
Acidic deposition may contribute to episodic
acidification both by supplying nitrogen to
sensitive watersheds (producing pulses of
MV during high flow events), producing
hydrologically mobile stores of SC>42
through dry deposition and by lowering
baseline pH and ANC so that smaller
episodes are sufficient to produce acidic
conditions.
Biologically relevant Surface Water
Chemistry. The main cause for concern
over the effects of surface water
acidification in the U.S. and elsewhere is the
potential for detrimental biological affects
(Baker and Christensen 1991, National Acid
Precipitation Assessment Program 1998).
Typically, there is concern for biological
impact if the pH is less than 6. At low pH
values, aluminum may be present at
concentrations that are toxic to biota,
including sensitive life stages offish and
sensitive invertebrates. Aluminum is an
abundant and normally harmless component
of rocks and soils. However, it leaches from
silicate minerals when they come in contact
with low pH waters. While much of the
aluminum present in surface waters is
organically bound and relatively non-toxic,
certain inorganic species are highly toxic.
The best indicator of recovery in
biologically relevant chemistry would be a
decrease in concentrations of inorganic
monomeric aluminum, the most toxic form.
Decreases in total aluminum would also
suggest recovery, although the actual
magnitude of the improvement in chemical
conditions for biota would be unknown
because we would not know how much of
the decrease is due to inorganic vs. organic
forms of aluminum.
Biological effects. The biological effects of
acidification have been demonstrated in
laboratory and field bioassays (e.g., Baker et
al., 1996), with whole-ecosystem
acidification experiments (e.g., Schindler et
al., 1985), and through field surveys (e.g.,
Baker and Schofield 1982, Gallagher and
Baker 1990). A number of the species,
especially offish and macro-invertebrates,
that commonly occur in surface waters
susceptible to acidic deposition cannot
survive, reproduce or compete in acidic
waters. Sensitive species may be lost even
at moderate levels of acidity. For example,
some important zooplankton predators are
not found at pH levels below 5.6; sensitive
mayfly species (e.g., Baetis lapponicus) are
affected at pH levels near 6.0; and sensitive
fish species, such as the fathead minnow,
experience recruitment failure and extinction
at pH 5.6 to 5.9 (Baker and Christensen
1991).
Unfortunately, there are very few examples
of long-term monitoring data for biological
assemblages in acid-sensitive surface waters
(e.g., Henriksen and Grande 2002), and none
in the U.S. Therefore, many of our
conclusions about the effect of acidic
deposition on the distribution of sensitive
10
-------
species are based on relationships like the
one shown in Figure 4. This relationship
represents the sum total of the effects of
acidification, lake size, and other complex
interactions between the survival offish
species and the quality of their habitat.
Research in the Adirondacks has shown that
many lakes have always had marginal
spawning habitat for Eastern Brook Trout
(Schofield 1993), and at least some of the
currently fishless acidic lakes in the
Adirondacks may never have supported fish.
Given the multiple lines of evidence
available, it is very likely that acidic
conditions, both natural and anthropogenic,
limit the distribution offish, benthic
invertebrate, and zooplankton species, but a
lack of historical data make it difficult to
infer the magnitude of biological recovery
(if any) from acidic deposition in individual
lakes or streams.
Ecological response: is sulfate or nitrate
more deleterious? The NAPAP research
effort of the 1980s was focused largely on
sulfur as the problem behind the ecological
impacts of acidic deposition (Baker et al.,
1990, Church et al., 1990, Turner et al.,
1990). Conceptual models were developed
to understand regional responses (e.g.,
Galloway et al., 1983, Wright 1983).
Computer models of watershed
geochemistry enhanced our understanding of
sulfate in watersheds (Christophersen and
Wright 1981, Chen et al., 1984, Cosby et al.,
1985). As a result, sulfur biogeochemistry
16
14 -
W) HO J
CD 1^ "
'o
CD
»1(H
IT
CO
il 8
H—
O
CD 6
i
7 4
2
0
* High Elevation Lakes in Maine
o Maine Seepage Lakes
o
o
o o
O QO O O
• o o
o o o
O O CD •• CD • O«« O
COtDOO •OO •
O CO €D CB • OLE»«DO • O
4.0 4.5 5.0 5.5 6.0 6.5 7.0 7.5 8.0 8.5
PH
Figure 4. Distribution offish species by lake pH in high elevation and seepage lakes in Maine
(Kahl and Scott 1994). Y-axis represents the number offish species found in each lake.
11
-------
is relatively well understood qualitatively.
However, a decade after NAPAP, it is clear
that we cannot predict sulfate quantitatively
in most instances due to the complications
of other, less well understood, factors.
One of the key factors still the subject of
much research and scientific debate is the
role of nitrogen in watershed responses to
acidic deposition. The concept of nitrogen
saturation (Aber et al., 1989, Stoddard
1994) received increasing attention in the
1990s (Murdoch and Stoddard 1992,
Mitchell et al., 1996, Williams et al., 1996,
Aberetal., 1998). Nitrogen saturation is
defined as deposition of N to a watershed in
excess of the assimilative capacity of soils
and vegetation, resulting in the export of
NO3". Nitrate export can contribute to
acidification (especially episodic
acidification), mobilization of aluminum,
and leaching of cations from soils (Aber et
al., 1998, Fennetal., 1998).
In the U.S., nitrogen saturation has received
the most attention in the eastern states,
although it is not clear if chronic N-
saturation has occurred. Several forest
manipulation studies in New England show
increased mortality and declining growth of
softwoods in response to N additions,
leading to the suggestion that the widespread
spruce decline in the eastern U.S. may be
related to long-term N deposition (Aber et
al., 1995). Nitrate concentrations in streams
draining forests in the Northeast are higher
than in any other forested region in the
country and have increased by a factor of 3
to 4 since 1970 (Murdoch and Stoddard
1992, Aber et al., in press), although these
concentrations have begun to decrease based
on analyses in this report.
Our understanding of nitrogen cycling in
forested watersheds is complicated by
nitrogen's response to climatic variation,
land use history, forest disturbances (e.g.,
fire and insect defoliation) and drought
(Mitchell et al., 1996, Aber and Driscoll
1997, Goodale et al., 2000). Indeed, under
most conditions, nitrogen from deposition
functions as a fertilizer for N-limited forests.
Despite the complexities of the nitrogen
cycle in forested watersheds, the scientific
consensus seems to be that, in the absence of
substantial declines in nitrogen deposition,
future increases in surface water NOs
concentrations, particularly in the Northeast
and Mid-Atlantic U.S., are likely (Stoddard
1994, Aber et al., in press). The difficulties
inherent in controlling N emissions further
complicate this issue—emissions are from
both combustion sources (NOX) and
agricultural sources (NH3 and NOX), and
unlike S emissions, mobiles sources play a
much larger role than stationary sources. An
assessment of current status and trends, like
the one presented here, can serve as the
starting point for any future discussions of
future effects and regulations concerning
nitrogen.
DATA USED IN THIS REPORT
National Atmospheric Deposition
Program/National Trends Program
(NADP/NTN). This report relies heavily on
long-term records of wet deposition from the
NADP/NTN program
(http://nadp.sws.uiuc.edu). The
NADP/NTN network has collected data on
rates of atmospheric deposition (rainfall,
snowfall and major ion chemistry) since the
early 1980s. The network currently consists
of 245 sites and is an excellent example of
grass roots scientific cooperation creating an
enormously useful source of data.
In order to characterize deposition trends
with adequate confidence, we chose to use at
least nine NADP/NTN locations for each
region. None of the regional boundaries
shown in Figure 2 encloses an adequate
number of sites with long data records. We
therefore expanded the regional boundaries
12
-------
slightly to include deposition monitoring
stations within close enough proximity to
have a high probability of exhibiting trends
similar to those actually within the
boundaries. The list of NADP/NTN sites
used in this exercise are shown in Table 2.
We chose not to use dry deposition
estimates from programs such as the Clean
Air Status and Trends Network (CASTNet;
U.S. Environmental Protection Agency
2002) for several reasons. Perhaps the most
important of these is the need to compare
directly the concentrations of ions in surface
waters with the concentrations in deposition.
At present there is no straightforward
method to produce deposition concentrations
(e.g., SO42 in |ieq/L) from dry deposition
data. In addition, the limited spatial and
temporal coverage, and the high spatial
variability of CASTNet data, make them
difficult to use with any confidence in the
sort of regional analysis presented here.
There are, for example, no CASTNet sites
currently operating in the Adirondack
Mountains, and only two useable sites in the
Upper Midwest region, as we define it here.
13
-------
Table 2. Location of NADP/NTN wet deposition sites in each region.
Region
New England
Adirondacks
Northern Appalachians
Upper Midwest
NADP/NTN
Site Code
MA01
MA08
MA13
ME02
ME09
ME98
NH02
VT01
VT99
NY08
NY20
NY52
NY68
NY98
NY99
VT01
VT99
NY08
NY65
NY68
NY99
PA15
PA29
PA42
PA72
WV18
MI09
MI98
MI99
MN16
MN18
State
Massachusetts
Massachusetts
Massachusetts
Maine
Maine
Maine
New Hampshire
Vermont
Vermont
New York
New York
New York
New York
New York
New York
Vermont
Vermont
New York
New York
New York
New York
Pennsylvania
Pennsylvania
Pennsylvania
Pennsylvania
West Virginia
Michigan
Michigan
Michigan
Minnesota
Minnesota
NADP/NTN Site Name
North Atlantic Coastal Lab
Quabbin Reservoir
East
Bridgton
Greenville Station
Acadia National Park
Hubbard Brook
Bennington
Underbill
Aurora Research Farm
Huntington Wildlife Refuge
Bennett Bridge
Biscuit Brook
Whiteface Mountain
West Point
Bennington
Underbill
Auroroa Research Farm
Jasper
Biscuit Brook
West Point
Penn State
Kane Experimental Forest
Leading Ridge
Milford
Parsons
Douglas Lake
Raco
Chassell
Marcell Experimental Forest
Fernberg
14
-------
Region
Upper Midwest
Ridge and Blue Ridge
NADP/NTN
Site Code
MN23
MN27
WI09
WI25
WI28
WI36
WI37
WI98
WI99
KY22
MD03
NC45
PA15
PA29
PA42
PA72
TN11
VAOO
VA13
VA28
WV04
WV18
State
Minnesota
Minnesota
Wisconsin
Wisconsin
Wisconsin
Wisconsin
Wisconsin
Wisconsin
Wisconsin
Kentucky
Maryland
North Carolina
Pennsylvania
Pennsylvania
Pennsylvania
Pennsylvania
Tennessee
Virginia
Virginia
Virginia
West Virginia
West Virginia
NADP/NTN Site Name
Camp Ripley
Lamberton
Popple River
Suring
Lake Dubay
Trout Lake
Spooner
Wildcat Mountain
Lake Geneva
Lilley Cornett Woods
White Rock
Mt. Mitchell
Penn State
Kane Experimental Forest
Leading Ridge
Milford
Great Smoky Mountains National Park
Charlottesville
Horton's Station
Shenandoah National Park
Babcock State Park
Parsons
Lake and stream monitoring programs.
The core of the acid rain aquatic effects
monitoring effort in the eastern U.S. has
been two EPA programs: the Temporally
Integrated Monitoring of Ecosystems
(TIME) project (Stoddard 1990) and the
Long-Term Monitoring (LTM) project (Ford
etal., 1993, Stoddard et al., 1998b). Both
projects are operated cooperatively with
numerous collaborators in state agencies,
academic institutions and other federal
agencies. Augmenting these core programs
are ongoing surveys of lake sub-populations
especially sensitive to acidification or
recovery. These programs all have slightly
different objectives and structures, outlined
below.
TIME project: At the core of the TIME
project is the concept of a probability
sample where each sampling site is chosen
statistically to be representative of a target
population. In the Northeast (New England
and Adirondacks), this target population
consists of lakes likely to be responsive to
15
-------
changes in rates of acidic deposition (i.e.,
those with Gran ANC < 100 |ieq/L). In the
Mid-Atlantic, the target population is upland
streams with a high probability of
responding to changes in acidic deposition
(i.e., Northern Appalachian Plateau streams
with Gran ANC < 100 jieq/L).
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 (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 EMAP (Paulsen
et al., 1991, Herlihy et al., 2000). The
TIME project began sampling Northeast
lakes in 1991; for the purposes of this report,
we group these lakes into two acid-sensitive
regions, the Adirondacks and New England.
Data from 43 Adirondack lakes can be
extrapolated to the target population of low
ANC lakes in the region (there are ca. 1000
low ANC Adirondack lakes, out of a total
population of 1830 lakes with surface area >
1 ha). Data from 30 lakes (representing ca.
1500 low ANC lakes, out of a total
population of 6,800) form the basis for our
trend analysis in New England.
Probabilistic monitoring of Mid-Atlantic
streams began in 1993; this report utilizes
data from 30 low ANC streams in the
Northern Appalachian Plateau (representing
ca. 42,000 km of stream length, of which ca.
24,000 km are considered low ANC). In the
Ridge and Blue Ridge provinces, the sample
coverage from TIME was not sufficient to
conduct regional trend analysis; all of our
conclusions about this region (ca. 5,500 km
of low ANC streams, out of a total
population of 23,000 km) are therefore
based on the non-probabilistic Virginia
Trout Stream Survey described below.
There are no sample survey data available
from the Upper Midwest in the 1990-2000
time period. The ultimate goal of having a
probability program like TIME is to
determine not just how a representative
sample of lakes and/or streams is changing
through time, but to know whether the
proportion of the population that is acidic
has changed.
LTM project: As a complement to lake
and stream sampling in a statistical
population of lakes in TIME, the Long-Term
Monitoring project samples a subset of
sensitive lakes and streams with long-term
data, most dating back to the early 1980s.
These sites are sampled 3 to 15 times per
year. This information is used to
characterized how the most sensitive of
aquatic systems in each region are
responding to changing deposition, as well
as giving information on seasonal chemistry
and episodic acidification. In most regions,
a small number of higher ANC (e.g., Gran
ANC > 100 jieq/L) sites are also sampled
and help separate temporal changes due to
acidic deposition from those attributable to
other disturbances (e.g., climate change,
landuse change). 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. The details
of data from each region are given below.
New England lakes: The LTM project
collects quarterly data from lakes in Maine
(sampled by the University of Maine; Kahl
et al., 1991, Kahl et al., 1993b) and Vermont
(data collected by the Vermont Department
of Environmental Conservation; Stoddard
and Kellogg 1993, Stoddard et al., 1998a).
Data from 24 New England lakes are
available for the trend analysis reported here
(time period 1990-2000). In addition to
quarterly samples, a subset of these lakes
have outlet samples collected on a weekly
basis during the snowmelt season; these data
are used to characterize spring episodic
chemistry. The majority of New England
LTM lakes have mean Gran ANC values
16
-------
ranging from -20 to 100 |ieq/L; two higher
ANC lakes (100< Gran ANC < 200 jieq/L)
are also monitored.
Adirondack lakes: Our trend analysis
includes data from 48 Adirondack lakes,
sampled monthly by the Adirondack Lake
Survey Corporation (Driscoll and Van
Dreason 1993, Driscoll et al., 1995); a
subset of these lakes are sampled weekly
during spring snowmelt to help characterize
acidic episodes. The Adirondack LTM
dataset includes both seepage and drainage
lakes, most with Gran ANC values in the
range -50 to 100 |ieq/L; 3 lakes with Gran
ANC > 100 |ieq/L (but less than 200) are
also monitored.
Appalachian Plateau streams: Stream
sampling in the Northern Appalachian
Plateau is conducted roughly 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 from Pennsylvania (collected by
Pennsylvania State University; DeWalle and
Swistock 1994) are used in this report. All
of the Northern Appalachian LTM streams
have mean Gran ANC values in the range
-25 to 50 neq/L.
Upper Midwest lakes: Roughly 40 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 used in this report consist of
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 data
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 Survey (Webb et al., 1989,
Webb et al., 1994) 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 reported here. The data are
collected cooperatively with the University
of Virginia and the National Park Service.
Mean Gran ANC values for the Ridge/Blue
Ridge data range from -15 to 200 |ieq/L,
with 7 of the 69 sites exhibiting mean Gran
ANC > 100 |ieq/L.
Supplemental datasets: Lake surveys
conducted in the 1980s revealed that 70% of
the known acidic lakes in the state of Maine
were either in seepage lakes or high
elevation lakes (Kahl et al., 1991). We have
used data from these two sets of lakes, as
well as from the control stream at the Bear
Brook Watershed Manipulation Project
(Norton et al., 1994, Church 1999), in some
of the supplementary analyses presented in
this report.
Maine Seepage lakes: This dataset includes
data from 120 of the estimated 150 lakes in
Maine meeting the following criteria: 1)
location in a sand and gravel area mapped
by the U.S.G.S. or Maine Geological
Surveys; 2) depth at least 1 meter; and 3)
area at least 0.4 hectare (1 acre). Sampling
was conducted in 1986-87 and 1998-2000
and included at least one fall index sample
17
-------
for each lake. The population was reduced
to 87 lakes for this report by screening for
lakes with Gran ANC less than 100 |ieq/L.
Maine High Elevation lakes: Data from 90
high elevation lakes (above 600 meters
elevation), available annually for the periods
1986-88 and 1997-99, represent the vast
majority of lakes in Maine meeting this
criteria: 1) depth at least 1 meter; and 2) area
at least 0.4 hectare (1 acre). The population
was reduced to 64 lakes for this report by
screening for lakes with Gran ANC less than
100 |ieq/L.
Surface water classification. For purposes
of analysis and illustration, the surface water
data contained in this report were classified
according to average Gran ANC and DOC
values. For the period of record, acidic
waters were those with mean Gran ANC less
than or equal to zero. Low ANC waters were
those with Gran ANC greater than 0 but less
than or equal to 25 |ieq/L. Moderate ANC
waters were those with Gran ANC greater
than 25. The site selection process or the
definition of the statistical population
eliminated high ANC waters. The vast
majority of sites included here all have mean
Gran ANC less than 100 |ieq/L.
Sites were also classified by dissolved
organic carbon (DOC), with the intent to
evaluate the contributions of natural organic
acidity to the response of systems. Waters
with mean DOC less than or equal to 5 mg/L
were classified as low DOC. Sites with
DOC greater than 5 mg/L were high DOC.
The highest DOC lake (a seepage lake in
Wisconsin) had a mean DOC concentration
of 17 mg/L.
Additionally, seepage lakes were also
evaluated by silica concentration in order to
classify systems as perched or flow-through.
Seepage lakes with mean silica less than
Img/L were considered perched. The
significance of this classification is that
perched lakes are disconnected
hydrologically from groundwater (they
supply water to the local groundwater) and
have minimal watershed contributions to
their chemistry. Such lakes are expected to
be more responsive to changes in
atmospheric deposition. Seepage lake data
are available from Maine, the Adirondacks,
and Wisconsin.
Analytical methods and quality
assurance. The specific field protocols,
laboratory methods, and quality assurance
procedures are specific for each investigator.
This information is contained in the cited
publications of each research group and are
compiled in Newell et al., (1987). The
EMAP and TIME protocols and quality
assurance methods are generally consistent
with those of the LTM cooperators and are
detailed in Peck (1992) and in Table 3.
The data used in this report have been
screened for internal consistency among
variables including ion balance and
conductance balance. Samples with
unexplained variation in these variables
have been deleted. Sites with mean Gran
ANC greater than 200 jieq/L were deleted.
We excluded sites with chloride values that
were outliers in their region because high Cl"
is typically associated with human
development in the watershed. The Cl" and
associated Na+ would alter normal soil ion
exchange relationships, thus obscuring the
response to acidic deposition.
Sea-salt corrections. We did not
systematically correct surface water
chemistry for contributions from marine
aerosols in large part because the marine
influence was minor for most regions. The
exception was stream data from the Bear
Brook Watershed Manipulation which is of
close proximity to the Gulf of Maine and
where the detailed analyses of relationships
between SO42 and base cation behavior
required the greater precision provided by
sea salt correction. For these streams, and in
18
-------
a few other instances as noted in the text we
corrected surface water chemistry by
assuming all chloride was of marine origin
and correcting sulfate and base cations based
on the ratios found in seawater (Riley and
Chester 1971).
Table 3. Typical analytical methods used for samples reported here. Samples are filtered before
analysis, except in Adirondack LTM.
Variable
Technique
Protocol reference
pH, closed cell
pH, aerated
Specific conductance
True color
Gran ANC
Anions: Cl", NO3', SO42
Calcium
Magnesium
Sodium
Potassium
Aluminum, total
Dissolved Organic Carbon
Dissolved Inorganic Carbon
Ammonium
Silica
Electrode
Electrode
Wheatstone bridge
Spectrophotometer, 457.5 nm
Gran Titration
Ion chromatography
AAS with N20-acetylene flame
AAS with N20-acetylene flame
AAS with air-acetylene flame
AAS with air-acetylene flame
AAS with graphite furnace
IR C analyzer, persulfate oxidation
IR C analyzer, direct injection
Autoanalyzer
Autoanalyzer
Hillman, et al.6, EPA 19.05
Hillman, et al.6, EPA 5.05
EPA 120.12,EPA23.05
EPA 110.22
Hillman etal.6, EPA 5.05
EPA 300.01
EPA 215.12
EPA 242.12
EPA 258.12
EPA 273.12
EPA 200.94
EPA 415.12
EPA 13.05, OI, 1990s
EPA 9.05 and Bran & Luebbe 780-
86T7
EPA 22.05 and Bran & Luebbe 785-
86T7
AAS=atomic absorption spectrophotometry
IR=Infrared Spectrophotometry
Method references:
1 Methods for the Determination of Inorganic Substances in Environmental Samples, EPA 600/R-93/100,
1993.
2 Methods for Chemical Analysis of Water and Wastes, EPA 600/4-79/020, 1979, Revised 1983.
3 Standard Methods for Examination of Water and Wastewater, 18th ed. 1992.
4 Methods for the Determination of Metals in Environmental Samples, EPA 600/4-91/010, 1991,
Supplement 1, EPA 600/R-94/111, 1994.
5 Handbook of Methods for Acid Deposition Studies: Laboratory Analysis For Surface Water Chemistry,
EPA 600/4-87/026, 1987.
6 Hillman,D.C., J. Potter, and S. Simon, 1986. Analytical methods for the National Surface Water Survey,
Eastern Lake Survey. EPA/600/4-86/009, EPA Las Vegas.
7 Bran & Luebbe Manual
8 OI Analytical Manual, 1990
19
-------
We did not correct NADP/NTN data for
marine aerosols. Air masses move west to
east, meaning that marine contributions to
precipitation chemistry will be minor, with
the exception of seasonal effects in coastal
New England. Other sources are more
likely to contribute chloride than seawater,
and thus correction of sulfate and base
cations based on a seawater-source
assumption will likely lead to erroneous
corrections. In the instance of base cation
data that are important for watershed
function, there is no point in mathematically
removing cations presumed to arise from
marine aerosols because the source of the
cations is irrelevant to watersheds.
Statistical analyses. Numerous statistical
techniques are available to analyze trends in
time series like those presented here. For
trends at single monitoring sites, the
Seasonal Kendall Tau (SKT) test (Hirsch et
al., 1982, Hirsch and Slack 1984) is often
the method of choice because it deals well
with censored data and with data collected at
irregular intervals with marked seasonality
(Loftis and Taylor 1989). One weakness of
the SKT is that it does not compute a slope.
The regional analyses we present in this
report depend on the ability to calculate a
robust estimator of slope for each site.
Rather than utilizing a Sen estimator (Sen
1968), as is often done with the SKT, we
have chosen to use simple linear regression
(SLR) to calculate a trend slope for each
monitoring site. This has the added
advantage of allowing us to use identical
methods for detecting regional trends in both
the LTM and TIME data; the probability
data from TIME are not of sufficient sample
size (for individual sites) or sampling
frequency to make the choice of SKT
appropriate. While the significance of
individual tests conducted with SLR are
questionable (especially when repeated tests
are conducted within each region), the
slopes calculated for multiple sites within a
region represent a distribution of results
which can in turn be examined and analyzed
for patterns. Within each region, we test for
a significant regional trend by calculating
confidence limits about the median value in
the slope distribution (SAS Institute, Inc.,
1988, Altman et al. 2000) and testing
whether these confidence limits include
zero. For a distribution in which all of the
slopes are negative, for example, the median
value would be significantly less than zero,
indicating a significant regional downward
trend.
TRENDS IN EMISSIONS AND
DEPOSITION
Overview of Emissions and Deposition
Changes: Sulfate emissions declined
substantially in the 1990s (Figure 5); plants
regulated under Phase I of the CAAA
reduced emissions by 50%, while total
emissions nationwide were reduced by 30%
over 1980 levels.
20
-------
Total $02 Emissions: 1980
Total S02 Emissions: 1990
Total S02 Emissions: 1999
T»Ul UlHIty SOi E mlwtoits [thaw and lon«]
CHO-40D QUD1-UGH
I 1 401 800 • 1J01 - ZJOT
^]BD1 . Uflfl
Figure 5. Trends in sulfur dioxide emissions following implementation of Phase I of the Acid Rain
Program: Total state-level utility SO2 (1980, 1990, 1999).
Data from NADP/NTN indicate that sulfur
emissions declines have contributed to a
continuation of the long-term decline in
sulfur deposition (Figure 6) from the
previous decade (Holland et al., 1999,
Shannon 1999, Government Accounting
Office 2000, Lynch et al., 2000). Many
1989-1991
NADP/NTN sites exhibit declines in excess
of 45% since 1980. Compared to trends
from 1983-94, the decline in wet deposition
sulfate accelerated after 1995 at 74% of
NADP sites in the eastern U.S. presumably
as a result of Phase I of Title IV of the
CAAA (Lynch et al., 2000).
1995-1998
<5 8 11 14 17 20 23 26 29 32 >35 <5 8 11 11 17 20 23 26 29 32 >35
Figure 6. Trends in wet sulfate deposition in the eastern United States from NADP/NTN data (1989-
1991 vs. 1995-1998). Data are in kg/ha.
21
-------
Power plant SO2 emissions increased after
1996, perhaps because utilities overshot the
CAAA targets in 1995 (Government
Accounting Office 2000). This increase
translated into a measurable increase in
deposition in many NADP/NTN stations in
the regions of interest for surface water
response (Figure 7). The lowest SC>42"
concentration in precipitation occurred in
1995 or 1996, depending on the station
(Figure 7; Government Accounting Office
2000, National Atmospheric Deposition
Program/National Trends Network 2002).
Since 1980, total U.S. SO2 emissions have
decreased by about 5.5 million tons.
Decreases in NOX emissions were more
modest than those of sulfur with declines
noted in several states in the northeastern
corridor (Figure 6). Since 1990, total utility
NOX emissions (Phase I and II sources) were
reduced an average of 23% nationally
following implementation of Phase I of the
Acid Rain Program. However, electric
utilities contribute only about one-third of
total NOX emissions. Since total NOX
emissions from other sources have remained
relatively constant (motor vehicles and other
industrial sources also contribute
significantly), the reductions achieved under
the Acid Rain Program have not resulted in
a significant change in total NOX emissions.
22
-------
0)
•4-J
•2
Z3
CO
70
60 -
50 -
40 -
30 -
20 -
10
New England (n=9)
cr
(D
-------
1990
1999
Total Utility NOx Emissions (thousand tons)
~IO-JD I 1135-300
I 30 -75
I 75 -135
300 540
Figure 8. Trends in nitrogen oxide emissions following implementation of Phase I of the Acid Rain
Program: Total state-level utility NOX (1990 vs. 1999).
Relatively constant NOX emissions translate
to little change in wet NCV deposition in the
1990s (Figure 8). The relative lack of
change in NOs" deposition, relative to the
decline in SO42~, has changed the ratio of
S:N from 2:1 in 1980 to 1:1 in 2000 (Lynch
et al., 2000). This long-term flat trend in
nitrate deposition may have begun to shift
downward after the implementation of Phase
I for nitrate in 1996. Nevertheless, the
geographic pattern of nitrate deposition has
not changed substantially (Figure 9), in
contrast to the changing pattern for sulfate
(Figure 6).
Regional deposition trends for SC>42~ and
nitrogen: We evaluated the trends in wet
deposition in each region for which we
report trends in surface water response. Our
analysis uses precipitation concentrations as
reported by NADP/NTN (National
Atmospheric Deposition Program/National
Trends Network 2002).
Throughout this report, we illustrate the
range of trend behaviors in each region (for
both deposition and surface waters) through
the use of cumulative frequency
distributions (e.g., Figure 10). These plots
show the range of trend slopes in each
region with the minimum slope shown as the
value of the x-axis corresponding to the left
side of the regional curve and maximum
slope value equivalent to the x-value at the
right end of the curve. The median value for
each region which is the value generally
reported in the text and for which statistical
significance is calculated can be estimated
by finding the 50th percentile value for the
region (from the y-axis) and estimating the
corresponding value on the x-axis. For
example, the median change in SC>42
concentration in wet deposition in the
Northern Appalachian Plateau region
(Figure lOa) is ca. -1.6 |ieq/L/year.
24
-------
<6.0
10.8 12.1 H.O 15.6 17.2 18.8 20.4 >22.0 <6.0
9.2 10.8 12.4 H.O 15.6 17.2 18.8 20.4 >22.0
Figure 9. Trends in wet NO3" deposition in the eastern United States from NADP/NTN monitoring data
(1989-1991 vs. 1995-1998). Data are in kg/ha.
Sulfate concentrations declined substantially
in each of the regions at a median rate
between -0.8 and -1.5 |ieq/L/year for the
period 1990 to 2000 (Figure lOa). Nitrogen
(N(V plus NH4+) declined slightly in the
Northeast and Northern Appalachian Plateau
and was essentially unchanged in the Ridge
and Blue Ridge provinces. The median
change in wet deposition of nitrogen in the
Upper Midwest during the 1990s was
upward (Figure lOb); this was the only
region to exhibit an increase in
concentrations of either SC>42 or nitrogen
consistent with results reported by Lynch et
al., (2000).
An immediate goal of the Title IV of the
CAAA is to decrease atmospheric
deposition of SO42" by reducing sulfur
emissions. Wet deposition of SC>42" declined
significantly in every region assessed (Table
4), with median slopes of-0.8 and -1.5
|ieq/L/year for the period 1990-2000.
Nitrogen and N(V concentrations also
declined significantly in every eastern
region with the exception of the Ridge and
Blue Ridge provinces where declines were
not significant. These trends appear to be
driven by a decline in N(V in deposition
after 1995. Nitrogen and NOs" wet
deposition actually increased significantly in
the Upper Midwest.
Regional deposition trends for acidity and
base cations. It is clear from the emission
and deposition data that deposition closely
follows emissions for sulfur and perhaps for
nitrogen. This is a positive result for
evaluating the effectiveness of the CAAA.
The expectation is that these declines in acid
anions should result in a decline in acidity in
precipitation. However, most studies have
failed to find a commensurate decline in
acidity (hydrogen ion) corresponding to the
decline in SO42" deposition although the
declines in hydrogen ion are statistically
significant at many NADP stations (Lynch
et al., 2000). Some investigators have
attributed the lower than anticipated
response in hydrogen ion to decreased
deposition of base cations that can offset
declines in the wet deposition of sulfate
(Hedin et al., 1994, Lynch et al., 2000).
25
-------
Sulfate Concentration Trends in Wet Deposition
100
(a)
CD
Q
CD
CL
CD
"
E
3
O
New England
Adirondacks
No. Appalachians
Upper Midwest
Ridge/Blueridge
-2.5 -2.0 -1.5 -1.0 -0.5 0.0
Slope of Trend (peq/L/yr)
0.5 1.0
[NO" + NH4+] Concentration Trends in Wet Deposition
CD
Q
CD
o_
CD
•—
_CD
3
E
^
O
-1.0 -0.5 0.0 0.5
Slope of Trend (peq/L/yr)
1.0
Figure 10. Cumulative frequency distribution of slopes for SO42" (panel a, top) and nitrogen (panel b,
bottom) trends in wet deposition, from regional NADP/NTN data for the period 1990-2000. The
x-axis is the slope (rate of change in (ieq/L/yr). The y-axis is the cumulative percent of the
population; the median rate of change for each region corresponds to the value of the x-axis at the
50th percentile of the curve.
26
-------
Table 4. Regional trend results (1990-2000) for atmospheric deposition (wet-only annual concentration
data from NADP/NTN network) in acid sensitive regions. All units are (ieq/L/yr. Values are the
median slopes for each region with significance determined by calculating confidence intervals
around each regional median.
Region
New England
Adirondacks
Appalachian Plateau
Ridge/Blue Ridge Province
Upper Midwest
SO42"
-0.96**
_j 47**
-1.55**
-0.96**
-0.75**
Nitrogen
-0.26*
-0.37**
-0.41**
-0.16ns
+0.24**
NCV
-0.20**
-0.38**
-0.37**
-0.13ns
+0.21**
Base Cations
+0.02ns
+0.01 ns
+0.01 ns
+0.05 ns
+0.18**
Hydrogen
-0.81**
-1.48**
-1.54**
-0.94 ns
-0.57**
ns regional trend not significant (p > 0.05)
* p<0.05
**
p<0.01
We evaluated the deposition patterns for
base cations (Ca2+ plus Mg2+) and hydrogen
ion in each region. In watersheds, base
cations are weathering by-products of the
neutralization reactions that produce ANC.
It is not well understood if base cation
deposition from the atmosphere contributes
quantitatively important amounts of cations
to soils or to the neutralization capacity of
the watershed. A long-term decline in the
deposition of base cations (Hedin et al.,
1994) has been suggested as a possible
cause of declining base cation
concentrations in surface waters (Driscoll et
al., 1989). Widespread declines in surface
water base cations have offset decreases in
SC>42" concentrations, and minimized or
prevented recovery in ANC in many waters
(e.g., Stoddard et al., 1999). While median
slopes for base cation deposition trends were
positive in every eastern region (Figure
1 Ib), only in the Upper Midwest did wet
deposition concentrations of base cations
increase significantly.
The combination of significantly decreasing
SC>42" deposition (and to a lesser extent
declining nitrogen) and largely unchanged
base cations deposition, as reported in this
analysis, leads to an expectation of declining
acidity in wet deposition as well. In fact,
wet deposition of hydrogen ion has
decreased in every region (Figure 1 la) and
is significant in every region except the
Ridge and Blue Ridge provinces (Table 4).
This is contrary to previous analyses (e.g.,
Lynch et al., 2000) using data through 1997,
and can be attributed to the lack of trend in
base cation concentrations for the period
1990-2000.
27
-------
Hydrogen Ion Concentration Trends in Wet Deposition
100
(a)
0
E
0
Q_
0
E
^
O
New England
Adirondacks
No. Appalachians
Upper Midwest
Ridge/Blue Ridge
-2.5 -2.0 -1.5 -1.0 -0.5 0.0
Slope of Trend (|jeq/L/yr)
0.5 1.0
[Ca2+ + Mg2+] Concentration Trends in Wet Deposition
0
2
0
CL
0
E
^
O
-0.25 0.00 0.25 0.50
Slope of Trend (|jeq/L/yr)
0.75
Figure 11. Distribution of slopes for acidity (hydrogen ion; panel a, top) and base cations (Ca2+ +Mg2+)
(panel b, bottom) trends in regional NADP/NTN data, by region, for the period 1990-2000.
28
-------
What is background precipitation
chemistry? Miller (1999) analyzed the
relationship between NADP/NTN SO42" wet
deposition and sulfur emission inventories
for the period 1980-98, using data from New
England. We used his equations for the
relationship at inland NADP/NTN stations
to extrapolate SC>42" deposition to a
hypothetical condition of zero SC>2
emissions. The predicted range of
precipitation SO42" concentrations at zero
SC>2 emission was 1.9 to 4.4 kg/ha at the six
stations for which Miller developed
equations. The mean SC>42" deposition
intercept (i.e., at zero emissions) was 2.9
kg/ha. Assuming consistent climatic
conditions, this deposition would equate to 6
to 10 |ieq/L SC>42" in precipitation as
background based on current concentration-
deposition relationships.
A background concentration of 6 to 10
|ieq/L can be compared to the present range
of 20 to 60 |ieq/L of wet SO42" at
NADP/NTN sites in the eastern U.S., and 5
to 20 |ieq/L in the western U.S. The
precipitation pH represented by this
background SC>42" would be pH 5.0 to 5.2
not including acidity from nitrate or organic
acids. This analysis supports the inference
of Charlson and Rodhe (1982) that the pre-
industrial background precipitation pH in
the eastern U.S. was around 5.0 setting the
baseline against which to compare modern
chemistry. When other sources of acidity
are included, it is unlikely that regulatory
actions will be able to increase precipitation
pH in the eastern U.S. to mean values
greater than pH 5.0.
Regional trends in precipitation amount.
Climatic variation can be a major
complication for interpreting chemical
changes in surface waters. In particular,
more dilute concentrations of both acid
anions and base cations can be expected to
occur in wet years, and lower concentrations
may result during periods of drought, at least
in streams and drainage lakes. In
Midwestern seepage lakes, drought may
have the effect of isolating lakes from base-
rich groundwater, essentially creating
temporarily perched seepage lakes, and
short-term acidification (Webster et al.,
1990). During the period investigated in this
report, annual rainfall was relatively uniform
in all of the regions (Figure 12). None of
the trends in precipitation amount was
significant for the 1990-2000 time period,
and we have not considered trends in rainfall
as part of our analysis of surface water
trends. Seasonal wet and dry spells may
nonetheless impart extra noise to the data,
but we assume that the length of the data
record smoothes the noise relative to
seasonal variation. In short, we do not
consider that climate variation is a major
factor affecting the trends in the east and
northeast with one exception: recovery from
the drought of the 1980s in the Upper
Midwest (Webster and Brezonik 1995) is
believed to have influenced lake chemistry
in the early 1990s and is discussed later in
this document.
RESPONSE OF SURFACE WATERS,
1990-2000
Important response variables. Our
analysis of surface water response to
changing deposition focuses on the key
variables that play major roles in
acidification and recovery:
1) SO42" and NOs, the acid anions that,
when found in surface waters, are
often the indicators of acidic
deposition. Trends in the
concentrations of these anions are
inferred to reflect recent trends in
deposition (especially SO42) and in
ecosystem response to long-term
deposition (e.g., NOs ).
29
-------
ro
OJ
>s
E
o
o
"TO
o
0)
CL
"(6
13
C
160 -
140
120 -
100
80
60 -
160
140 -
120
100 -
80
60 -
160
140 -
120
100
80 -
60
160
140
120 -
100
80 -
60
160
140
120 -
100 -
80
60 -
New England (n=9)
O
0 ro
^ m ^ o t'
• o © o **
0 0
Adirondacks (n=9)
o
Jh,
• o
U u ^-J Q w
0 0
Appalachian Plateau (n=9) o
0 - ©
W "" r^ ^ n *^
o « ° °
Upper Midwest (n=14)
"— O f".> no o r>
u Q o w
O
o ^ ° ^
u u u o " © o
o
Ridge and Blue Ridge (n=13)
1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000
Year
Figure 12. Annual precipitation in the study regions (National Atmospheric Deposition
Program/National Trends Network 2002). Symbols represented mean values for sites in each region.
Locations of NADP/NTN sites in each region listed in Table 2.
30
-------
200
150
!£, 0
D)
200
150
100
50
-50 0 50 100 150 200
+
CN
CO
Q 0
^ 140
oT 120
5 100
£T 80
J* 60
+ 40
Aco 20
O 0
300
No. Appalachians
-40 -20 0 20 40
Upper Midwest
-40 -20 0 20 40 60 80 100
5
0 50 100 150 200
ANC (|Jeq/L)
200
150
100
50
0
New England
o
200
150
100
50
-50
0
50 100 150
Adirondacks
0
200
150
100
50
0
-50 0 50 100 150 200
No. Appalachians
-40 -20 0 20 40
CT
0)
*:
+
z
140
120
100
80
60
40
20 \
0
Upper Midwest
^> o
-40-20 0 20 40 60 80 100
300
0 50 100 150 200
ANC ((Jeq/L)
Figure 13. The relationships between Gran ANC and [Ca2+ + Mg2+], and between Gran ANC and [K+ +
Na+] in acid-sensitive regions.
31
-------
2) Base cations, the by-products of the
weathering reactions that neutralize
acids in watersheds. We have
chosen to evaluate the response of
the two divalent cations, calcium and
magnesium (henceforth [Ca2+ +
Mg2+]) because of the nature of their
behavior in soils; these two cations
are quantitatively the most important
for generation of Gran ANC (Figure
13). In most regions, the relationship
between [Ca2+ + Mg2+] and Gran
ANC is much stronger than between
potassium (K+) plus sodium (Na+).
The addition of [K+ + Na+] to the
cation-ANC relationship imparts
extra variability to the interpretation
of base cation response although the
patterns differ by region.
3) pH and Gran ANC, the measures of
acidity. These variables reflect the
outcome of interactions between
changing acid anions and trends in
base cations.
4) Aluminum, one of the most
important variables for biological
impacts due to its toxicity in the
ionic form. Changes in
concentrations and forms of
aluminum are largely driven by
changes in pH and DOC with lower
and less toxic aluminum
concentrations resulting from high
pH values and DOC concentrations.
5) Concentrations of Dissolved
Organic Carbon (DOC), as a
surrogate for organic acidity.
Organic acids are additional, largely
natural, sources of acidity in surface
waters.
Representative data from the regions.
The trends we report below are for the
period of focus for this report, 1990-2000.
Many of the monitoring sites for which we
report data have been operating since the
early 1980s, and examining the entire period
of record helps to put current trends into a
longer-term context. As detailed earlier,
SO42 deposition has been declining in the
northern and eastern U.S. since the 1970s,
and changes produced by the CAAA of
1990 overlay the changes already underway.
In order to illustrate this historical context,
we have chosen representative sites from
each region and present time series of the
important response variables outlined above.
Representative in this case refers to the
observed time trends; each site exhibits
trends in the 1990s that are typical of the
region.
Salmon Pond in eastern Maine (Figure 14) is
one of 24 drainage lakes sampled as part of
the LTM project in New England which
operates sites in Maine and Vermont.
Salmon Pond has been sampled since early
1982 pre-dating the beginning of the LTM
project by a year. It is sampled quarterly -
in the spring, summer, and fall.
Sampling at Darts Lake (Figure 15) in the
Adirondacks began with the RILWAS
project (Gherini et al., 1989) in 1982; it
became part of the LTM project in 1983 and
has been sampled monthly through the
1980s and 1990s.
The East Branch of the Neversink River
(Figure 16) is a famous trout fishing stream
in the Catskill Mountains of New York and
exhibits trends typical of the Northern
Appalachian Plateau in the 1990s. It has
been sampled as part of the LTM project
since 1984 with variable sampling
frequencies. The Neversink is sampled for
both baseflow and episodic chemistry.
Vandercook Lake in Wisconsin (Figure 17)
exhibits trends representative of the Upper
32
-------
Midwest. This seepage lake has been Shenandoah National Park. It became part
sampled quarterly in the spring, summer, of the LTM network in 1992 but has been
and fall since 1984. sampled as part of various programs funded
rr, c. . n- i • . i by the National Park Service since 1987. It
The Staunton River showing trends .J , , , ,,,..,
representative of the Ridge/Blue Ridge '? ™Ten^. s,a™Pled on a schedule similar to
region of Virginia (Figure 18) is located in the Neversink
33
-------
itf Cr
"5 9>
2 cr
.•H HI
E
3
'w
c
O>'
05 ;
60
55
50
45
40
15
10
5
+
E
3
^O
OS
O
0
110
100
90
80
70
70 H
50
40
30
6.5
U d
O ^3
Q E
6
4
2 -\
Salmon Pond (New England)
tH*
•• * % t.,
t
.>*„•%
•
-
*"
mm A A
'* • «• r*. .• 1*11J •
* • !•
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
Figure 14. Time series data for SO42 , NO3 , base cations [Ca2+ + Mg2+], Gran ANC, pH, and DOC in
Salmon Pond, Maine (New England region). Significant trends are indicated by trend lines.
Shaded box indicates time period of analyses reported here.
34
-------
Darts Lake (Adirondacks)
180
CO
0)
05 x
CD
E
3
|o
CO
O
I
Q.
O o>
Q E
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
Figure 15. Time series data for SO42 , NO3 , base cations [Ca2+ + Mg2+], Gran ANC, pH, and DOC in
Darts Lake, NY (Adirondack region). Significant trends are indicated by trend lines. Shaded
box indicates time period of analyses reported here.
35
-------
East Branch Neversink River (Appalachian Plateau)
3
CO
2 -
.-K 0)
E
3
'w
Q E
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
Figure 16. Time series data for SO42 , NO3 , base cations [Ca2+ + Mg2+], Gran ANC, pH, and DOC in the
E. Branch of the Neversink River, New York Catskills (Northern Appalachian Plateau region).
Significant trends are indicated by trend lines. Shaded box indicates time period of analyses.
36
-------
Vandercook Lake (Upper Midwest)
100
90 H
80
70
60
50
40
E "a-
.-e 03
E
3
'w
c
05
3
o
I
Q.
O D)
Q E
15
10
5
0
95
90
85 H
80
75
70
40 -
30
20
10
0
7.0
6.5
6.0
5.5
5.0
5
4
3
•• ••
•• - • — ••,
••
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
Figure 17. Time series data for SO42 , NO3 , base cations [Ca2+ + Mg2+], Gran ANC, pH, and
DOC in Vandercook Lake, Wisconsin (Upper Midwest region). Significant trends are indicated
by trend lines. Shaded box indicates time period of analyses reported here.
37
-------
Staunton River (Virginia)
o-
3
CO
£ -
.-H 0)
E
^3
-------
Trends in sulfate by region. SC>42
declined in surface waters in the glaciated
regions of the North and East by median
values between -2 and -4 jieq/L/year
(Figure 19) with the smallest changes in
New England and the largest declines in the
Upper Midwest. The only exception to the
pattern of declining SC>42 was in the
Ridge/Blue Ridge provinces which by virtue
of their soil characteristics were not
expected to show a rapid response to
changes in deposition (Cosby et al., 1986,
Church et al., 1989). The median SO42
increased slightly in Ridge and Blue Ridge
streams. All of the regional SC>42 trends are
highly significant (Table 5) and are
consistent with the trends reported
previously including other regions of North
America and Europe (Mattson et al., 1997,
Stoddard et al., 1999, Evans and Monteith
2001, Skjelkvale et al., 2001).
The marked declines in SC>42 concentrations
in the glaciated portions of the North and
East are almost certainly direct responses to
declining emissions and SC>42 deposition in
the 1990s and represent the most dramatic
effects of Title IV of the CAAA and
previous emissions regulations. These
changes in emissions and deposition
continue the trend in declining SC>42 that
has been occurring for almost three decades
(Stoddard et al., 1999).
The small increase in SC>42 concentrations
in the Ridge and Blue Ridge provinces
results from the SC>42 adsorption properties
typical of soils found throughout this
unglaciated region which have the effect of
both decoupling (in time) trends in
deposition and surface waters (Tables 4 and
5) and lowering the ambient concentrations
of SC>42 well below those found in
deposition (Figure 7). The future response
of SC>42 concentrations in this region will be
a critical element in governing any potential
recovery in this region.
Table 5. Regional trend results for Long-Term Monitoring sites for the period 1990 through
2000. Values are median slopes for set of sites in each region. Units for sulfate, nitrate, base
cations [Ca2+ + Mg2+], Gran ANC and hydrogen are (ieq/L/year. Units for DOC are mg/L/year.
Units for aluminum are (ig/L/year.
Region
New England Lakes
Adirondack Lakes
Appalachian Streams
Jpper Midwest Lakes
lidge/Blue Ridge Streams
SO42
-1.77**
-2.26**
-2.27*
-3.36**
+0.29**
NO3
+0.01ns
-0.47**
-1.37**
+0.02ns
-0.07**
Base
Cations
-1.48**
-2 29**
-3.40**
-1 42**
-0.01 ns
Gran
ANC
+0.1 lns
+1.03**
+0.79*
+1.07**
-0.07ns
Hydrogen
-0.01 ns
-0.19**
-0.08*
-0.01*
+0.01 ns
DOC
+0.03*
+0.06**
+0.03 ns
+0.06**
NA
Aluminum
+0.09ns
-1.12**
+0.56 ns
-0.06 ns
NA
ns regional trend not significant (p > 0.05)
* p<0.05
** p<0.01
NA insufficient data
39
-------
CD
O
CD
CL
CD
>
'-I—'
_co
^
E
^
O
100
Regional Sulfate Trends in LTM Network
New England Lakes
Adirondack Lakes
Appalachian Streams
Upper Midwest Lakes
Ridge and Blue Ridge Streams
-12 -10 -8 -6 -4 -2
Slope of Trend (|jeq/L/yr)
0
Figure 19. Cumulative frequency diagram (distribution) of slopes ((ieq/L/year) for SO42
concentrations in surface water monitoring sites, by region for the period 1990-2000.
40
-------
Regional Nitrate Trends in LTM Network
co
100
90
80
70 -I
CD
O
CD 60
CL
CD
50
I 4°
O 30
20 -
10 -
0
New England Lakes
Adirondacks Lakes
Appalachian Streams
Upper Midwest Lakes
Ridge/Blue Ridge Streams
-3
-2 -1 0
Slope of Trend (|jeq/L/yr)
Figure 20. Cumulative frequency diagram (distribution) of slopes ((ieq/L/year) forNO3
concentrations in surface water monitoring sites, by region for the period 1990-2000.
Trends in nitrate by region. In general,
trends in NOs were much smaller than
trends in SC>42 with the only ecologically
significant changes occurring in the two
regions with the highest ambient NOs
concentrations (Figure 20). Both lakes in
the Adirondacks and streams in the Northern
Appalachian Plateau exhibited small but
significant downward trends in NOs in the
1990s (Table 5). Both of these regions are
central to the debate over whether nitrogen
saturation is a legitimate threat to the health
of forests and surface waters (Stoddard
1994, Aberetal., 1998). While declining
NOs concentrations in these regions are a
positive development for these ecosystems,
we clearly do not know if these trends will
continue especially since they do not reflect
recent trends in emissions or deposition
(Table 4, Figures 7-9). The presence of
strong upward trends in NOs in these same
regions in the 1980s (Murdoch and Stoddard
1992, Stoddard 1994) suggests that trends
on the scale of a single decade probably
represent noise (variability) in a long-term
pattern of changing NOs leakage from
41
-------
forested watersheds controlled by factors
that will take many years to determine.
While great uncertainty exists and the time
scales of nitrogen saturation may be longer
than previously considered (e.g., centuries
rather than decades), the long-term retention
of nitrogen deposition in forested regions is
unlikely to continue indefinitely (Aber et al.,
in press).
In New England and the Upper Midwest,
where ambient NOs" concentrations are
much lower than in the Adirondacks and
Northern Appalachian Plateau (Figures 14-
17), NO3 concentrations in surface waters
were unchanged. The Ridge/Blue Ridge
province registered a small but significant
decrease in NOs during the 1990s, but
interpretation of trends for NO3 in this
region is complicated by an outbreak of
Gypsy Moths that also occurred during this
period. Forest defoliation by Gypsy Moths
was the cause of a pulse in NO3 export
from many streams in this region in the mid-
1990s (Eshleman et al., 1998).
Trends in Gran ANC and acidity, by
region. In a very real sense, increasing
trends in Gran ANC would be the strongest
signs of recovery from acidification in
response to the CAAA. There was a
tendency toward increasing Gran ANC
(Figure 21) in all of the glaciated regions of
the North and East (i.e., New England,
Adirondacks, Northern Appalachian Plateau
and Upper Midwest) and decreasing Gran
ANC in the Ridge/Blue Ridge province.
Only the regional increases in the
Adirondacks, Northern Appalachian Plateau,
and Upper Midwest were significant (Table
5). Median increases of ca. +1 |ieq/L/year
in the Northern Appalachian Plateau,
Adirondacks and Upper Midwest represent
significant ecological recovery from
acidification.
Previous regional analyses using data
through 1995 have highlighted the lack of
ANC recovery in the Adirondacks, Northern
Appalachian Plateau and Upper Midwest
(Stoddard et al., 1999). In the Adirondacks
and Northern Appalachians, many of the
sites exhibit the pattern shown for Darts
Lake in Figure 15 and the Neversink River
in Figure 16 with Gran ANC increasing after
an inflection point in the early- to mid-
1990s. The current Gran ANC trends (Table
5) in the Adirondacks and Northern
Appalachian Plateau appear to have reversed
the pattern of lack of recovery with further
acidification occurring in some cases that
occurred throughout the 1980s and early
1990s in this region.
In the Upper Midwest, the downward Gran
ANC trends reported by Stoddard et al.,
(1999) for the period 1990-95 appear to
have been a temporary departure from the
long-term recovery of ANC in this region.
The trends reported for the 1990s (Table 5)
are consistent with the long-term pattern of
increasing Gran ANC observed at individual
sites.
Stoddard et al., (1999) reported recovery in
Gran ANC in the New England region using
data through 1995. This trend is no longer
evident when all data from the 1990s are
examined (Table 5). The temporal pattern
for New England lakes appears to be the
reverse of that exhibited by lakes in the
Adirondacks, with long-term recovery in
Gran ANC having fallen off after a similar
inflection point in the early 1990s (Figure
14). The slight tendency for Gran ANC to
increase in this region is not significant, with
nearly as many lakes showing small
decreases (11) as show slight increases (13).
Gran ANC showed an insignificant
downward trend in the Ridge/Blue Ridge
province consistent with the small increases
in SC>42 observed. It is clear that while the
evidence for a widespread recovery in ANC
in the glaciated North and East is quite
42
-------
strong, no such evidence exists for the
unglaciated Ridge/Blue Ridge province.
Widespread changes in Gran ANC
translated into some small, but significant,
changes in acidity (hydrogen ion; Figure
22). The largest significant decreases in
hydrogen ion occurred in the Adirondacks
(Table 5), but on an equivalent basis they
were less than 20% of the changes observed
in Gran ANC in this region. The moderate
decrease observed in the Northern
Appalachians was likewise small relative to
the change in Gran ANC (< 10% of the
ANC change). All of the other regions
exhibited median slopes near zero, and only
the Upper Midwest trend was significant
(Table 5). To put these changes in hydrogen
ion in context, a decline in hydrogen
concentration of 1 |ieq/L/year in an acidic
lake with pH 4.7 would equate to a of pH
5.00 over 10 years. All of the changes
observed here are considerably smaller than
this.
Trends in base cations, by region. All of
the glaciated regions on the northern and
eastern U.S. exhibited declines in base
cation [Ca2+ + Mg2+] concentrations in the
range of-1.5 to -3.4 |ieq/L/year (Figure
23), and all of the regional trends were
highly significant (Table 5). One of the
most universal watershed responses to acidic
deposition is the mobilization of base
cations from soils (Galloway et al., 1983); as
rates of acidic deposition decline in the
North and East and the supply of acid anions
to watersheds soils decreases, the rates of
cation mobilization are also expected to
decrease. Lowered rates of cation
mobilization translate to downward trends in
surface water base cation concentrations, a
change widely observed in the northern
hemisphere for more than a decade. Keller
et al., (2001) suggested that a trend of
declining Ca2+ has been occurring for more
than a century in Ontario lakes.
As a broad generalization, SO42 has
decreased at a rate of approximately -2.5
|ieq/L/year (the mean of regional median
slopes), and NOs~at a rate of-0.5
|ieq/L/year, in surface waters of glaciated
terrain. These rates of change set an upper
limit to our expectation of ANC recovery of
+3 |ieq/L/yr (i.e., the sum of SO42 and NOs
trend magnitudes). The Gran ANC increase
is actually about one-third of this maximum,
+1 |ieq/L/year. The difference between the
observed Gran ANC trend and the maximum
trend estimated from rates of acid anion
change can be almost entirely explained by
regional declines in base cations; the
average regional median decline in [Ca2+ +
Mg2+] was about-2.0 |ieq/L/year.
43
-------
CD
O
Q_
(D
E
^
O
100
Regional ANC Trends in LTM Network
New England Lakes
Adirondacks Lakes
Appalachian Streams
Upper Midwest Lakes
Ridge and Blue Ridge Streams
-20246
Slope of Trend (|jeq/L/yr)
8
10
Figure 21. Cumulative frequency diagram (distribution) of slopes ((ieq/L/year) for Gran ANC
concentrations in surface water monitoring sites, by region for the period 1990-2000.
44
-------
CD
O
CD
CL
CD
>
'-i—i
_co
^
E
^
O
100
90
80
70 4
Regional Hydrogen Ion Trends in LTM Network
-1.5
New England Lakes
Adirondack Lakes
Appalachian Streams
Upper Midwest Lakes
Ridge and Blue Ridge Streams
-1.0 -0.5 0.0
Slope of Trend (|jeq/L/yr)
0.5
Figure 22. Cumulative frequency diagram (distribution) of slopes ((ieq/L/year) for hydrogen ion
concentrations in surface water monitoring sites, by region for the period 1990-2000.
45
-------
The decline in [Ca2+ + Mg2+] while not
unexpected limits the magnitude of ANC
recovery by offsetting some of the decline in
SO42 and NOs . Moreover, if the rate of
base cation decline is equal to the rates of
decline in acid anions, recovery will be
prevented. This is the pattern observed in
several regions of North America in
previous regional analyses (Stoddard et al.,
1998a, Stoddard et al., 1998b) using data
through the early 1990s. The observation
that acid anions, and particularly SO42 ,
concentrations are decreasing at a faster rate
than base cations in most regions (Table 5)
leads to an increase in Gran ANC, and is
consistent with the pattern that has been
observed in northern Europe since about
1990 (Stoddard etal., 1999).
Several researchers (e.g., Likens et al., 1996,
Lawrence et al., 1999, Driscoll et al., 2001)
have proposed that the recovery of surface
waters will be substantially delayed because
soil impoverishment induced by decades of
base cation leaching by the anions in acidic
deposition has reached critical levels. At
least one aspect of the current analysis
argues against this hypothesis. The regional
response of base cations to changing SO42
deposition is following expectations (e.g.,
[Ca2+ + Mg2+] is decreasing at lower rates
than SO42 in glaciated terrain and is
unchanged in unglaciated terrain) in every
region.
Trends in dissolved organic carbon
(DOC), by region. Another potential factor
influencing the responsive in ANC and
acidity is the general increase in DOC
(Figure 24). All regions with sufficient
DOC data exhibited increases in
concentrations, and all regional trends were
significant with the exception of the
Northern Appalachian Plateau (the region
with the lowest median DOC
concentrations). While the character of the
DOC found in these watersheds has not been
evaluated, some portion of the DOC is
organic acid. The median increase of 0.05
mg/L/year corresponds to an increase of
about 10% region-wide similar to trends
reported elsewhere in the Northern
Hemisphere (Evans and Monteith 2001,
Skjelkvale etal.,2001).
It is beyond the scope of an analysis of
monitoring data like the one presented here
to determine the mechanism for DOC
increase unequivocally. However, there are
several aspects of the DOC trends that
warrant further attention. First, increases in
DOC have now been reported from nearly
all regions of North America (this report)
and Europe (Skjelkvale et al., 2001). This
argues for a large-scale cause, and both
climate change (warmer temperatures are
expected to increase decomposition and
could potentially increase the export of DOC
from watersheds) and decreasing acidic
deposition are possible hypotheses. Several
researchers have proposed that at least part
of the increase in mineral acids (e.g., SO42
and NOs ) observed in acidifying
watersheds in the past (Rosenqvist 1978,
Krug and Frink 1983) resulted from the
replacement of natural organic acidity with
mineral acidity. They thus propose that the
process of acidification was accompanied by
a decrease in surface water DOC
concentrations. This hypothesis has
received some support from
paleolimnological studies where
reconstructions of historic DOC
concentrations show decreases during the
period of acidification in the 1900s (Davis et
al., 1985, Kingston and Birks 1990, Dixit et
al., 2001). David et al., (1999) were also
able to measure a decrease in organic
anions in response to the experimental
whole-water shed acidification experiment at
the Bear Brook Watershed in Maine (Norton
et al., 1999). In one of the few studies of
46
-------
DOC behavior during recovery from
acidification, Wright et al., (1993)
concluded that ANC increases in a small
watershed where rates of acidic deposition
were experimentally reduced were limited
by the increasing role of organic acids and
an overall increase in DOC.
Regional [Ca2+ + Mg2+] Trends in LTM Network
-------
-------
In a regional study of trends like the one
presented here one might predict that if this
hypothesis (the replacement of organic
acidity by mineral acidity during
acidification) is correct, then the current
rates of DOC increase should be correlated
to rates of SO42 decline. This is in fact
observed in the trend data for all lake
regions (Figure 25). The largest rates of
DOC increase are observed at the sites with
the strongest declines in surface water
SO42 . This relationship is strongest in high
DOC systems (i.e., those with DOC > 5
mg/L) but is also observed in low DOC
lakes.
CD
£
o
o
Q
o>
o>
c
TO
_C
O
New England, Adirondack and Upper Midwest Lakes
0.7
0.6 -
0.5 -
0.4
0.3 -
0.1 -
0.0 -
-0.1 -
-0.2 -
-0.3 -
-0.4
\
0
,0 0
• Low DOC Lakes
O High DOC Lakes
G
O
-12
-10
-8
-6
-4
-2
2"
Change in SO4" ((jeq/L/year)
Figure 25. Relationship between trends in DOC and trends in SO42~ concentrations for lakes. This graph
includes data from the Maine high elevation and seepage lakes in order to produce a relationship
with sufficient numbers of high DOC in small, responsive lakes.
49
-------
100
Regional Aluminum Trends in LTM Network
New England Lakes
Adirondack Lakes
Appalachian Streams
Upper Midwest Lakes
-20
-15 -10 -505
Slope of Trend (|jg/L/yr)
10
Figure 26. Cumulative frequency diagram (distribution) of slopes ((ig/L/year) for aluminum
concentrations in surface water monitoring sites, by region for the period 1990-2000. Data from the
Adirondacks are for inorganic monomeric aluminum; all others are total dissolved aluminum.
50
-------
Trends in aluminum by region.
Aluminum plays a major role in the biotic
impoverishment that often accompanies
surface water acidification (e.g., Baker and
Schofield 1982, Cleveland et al., 1986).
Decreasing trends in aluminum, particularly
in the most toxic form of aluminum
(inorganic monomeric aluminum), would
therefore be a biologically relevant indicator
of recovery. Only one of the regions
analyzed here exhibited regional decreases
in aluminum (Figure 26). In the
Adirondacks, the median decline in
inorganic monomeric aluminum was
approximately -1.0 |ig/L/year, a very small
number in terms of importance to biota.
However, this change is consistent with
regional increases in Gran ANC and
decreases in hydrogen ion. The Adirondack
trends are also consistent with reported
decreases in toxic aluminum in soil water
samples from the Hubbard Brook
Experimental Forest in New Hampshire
(Palmer and Driscoll 2002). No regional
trends other than the Adirondacks were
significant (Table 5).
Our analysis for aluminum is complicated
by the fact that the most consistently
collected data for aluminum are for total
dissolved aluminum which includes both
organic forms and polymeric inorganic
forms. Small changes in inorganic
monomeric aluminum which may make up
only a small portion of total dissolved
aluminum may be undetectable if the other
forms are relatively unchanged, or are
highly variable. We do not therefore
consider the general lack of trends in
regional aluminum concentrations
conclusive although they are generally
consistent with patterns of unchanged
hydrogen ion.
Relationship between ANC trends and
ANC class: It is reasonable to expect, based
on our conceptual understanding of the
process of acidification, that decreasing rates
of acidic deposition and decreasing surface
water concentrations of acid anions will
produce the greatest rates of recovery in the
sites that have undergone the most severe
acidification. One outcome of this
expectation would be larger upward trends
in acidic lakes and streams, small upward
trends in low ANC sites, and no change in
the relatively insensitive sites we classify as
moderate ANC. An analysis of Gran ANC
trends by ANC class designed to test this
expectation is shown in Table 6. Using data
from all of the sites in regions where
decreases in surface water SO42 and/or
NOs have occurred, we find that acidic
lakes and streams exhibit a highly
significant median increase in Gran ANC of
+1.3 |ieq/L/year. Low ANC sites show a
smaller, but nonetheless significant, median
ANC increase of+0.8 |ieq/L/year.
Moderate ANC sites, those with mean ANC
values greater than 25 |ieq/L, show no
significant change in Gran ANC.
51
-------
Table 6. Slopes of trends in Gran ANC in acidic, low ANC and moderate ANC sites for the period 1990-
2000. Analysis includes all sites in New England, Adirondacks, Appalachian Plateau and Upper
Midwest; Ridge and Blue Ridge sites excluded.
ANC Class
Acidic
(ANC < 0 |ieq/L)
Low ANC
(0 < ANC < 25 |ieq/L)
Moderate ANC
(25 < ANC < 200 |ieq/L)
Number of
Sites
26
51
43
Change in
Gran ANC (|ieq/L/yr)
+1.29**
+0.84**
+0 32 ns
ns trend not significant (p > 0.05)
* p<0.05
**
p<0.01
Regional trends in probability networks.
The data described thus far in this report
come from the LTM project and from other
monitoring networks with similar objectives.
The method of site selection utilized for
LTM varied among regions and was in part
driven by the existence of historical data
and/or ongoing data collection rather than
statistical frameworks to determine
population trends. Stoddard et al., (1998b)
placed LTM sites from the Northeast in the
lake classification system of Young and
Stoddard (1996) and found that they largely
covered the range in current status (e.g.,
current Gran ANC and base cation
concentrations) of the most sensitive
subpopulation(s) of lakes in the region. The
trend results, however, indicated two
different classes of response: lakes in New
England were increasing in ANC (based on
data for the period 1984-94), while those in
the Adirondacks were acidifying slightly,
even when the current status data indicated
they belonged in the same lake
subpopulation.
The data in this report offer the first
opportunity to extend the analysis of LTM
representativeness into the realm of trends,
as well as changing population status. By
comparing the magnitudes of trends
measured from repeated probability surveys
(i.e., from the TIME project) with those
observed in the same regions through LTM
sampling, we have an opportunity to verify
the validity of LTM trends.
In both the Adirondacks and New England,
population-wide results inferred from TIME
data show very similar trends in SC>42 to
those for LTM sites (Figure 27). Both
datasets suggest highly significant regional
decreases in SC>42 slightly larger than -2
|ieq/L/year in the Adirondacks and -1.8
|ieq/L/year in New England (Tables 5 and
7).
Both datasets suggest similar ranges of trend
slopes for Gran ANC (Figure 28). In the
Adirondacks, regional increases in Gran
ANC are significant (Tables 5 and 7) in both
the LTM sites and TIME population data.
In New England, only the population trend
in Gran ANC was significant (Table 7),
although both datasets suggest that New
England lakes are undergoing slight
52
-------
increases in Gran ANC. The greater
response in TIME lakes may be due to the
different length of record between the
networks. If increases in Gran ANC
accelerated in the late 1990s as they did in
the Adirondacks, then the TIME data may
well estimate a larger rate of change than
LTM data.
Table 7. Regional trend results for populations of sites in acid sensitive regions. Results from
TIME probability sites are extrapolated to regional target populations. Lake trends in the
Adirondacks and New England are for the period 1991-2001. Stream trends in the Northern
Appalachian Plateau are for the period 1993-2001. Values are weighted median slopes for all
lakes or streams in each region. Units for SO42 , NO3~, base cations [Ca2+ + Mg2+], Gran ANC
and hydrogen are (ieq/L/yr. Units for DOC are mg/L/yr. Units for aluminum (inorganic
monomeric) are (ig/L/yr.
Region
Adirondack Lakes
New England Lakes
Appalachian Streams
SO42
-2.10**
-1.88**
-0.64*
NO3
+0.01ns
+0.02ns
+0.04ns
Base
Cations
-1.22*
-1.57**
-0.32 ns
Gran
ANC
+0.56*
+0.40*
+0.34*
Hydrogen
-0.09 ns
+0.01ns
-0.01ns
DOC
+0.09*
+0.08*
+0.01ns
Aluminum
+0.66 ns
-1.94ns
+0.14ns
ns regional trend not significant (p > 0.05)
* p<0.05
**
p<0.01
53
-------
0>
a
0)
CL
0)
E
^
O
100
90
80
70
TIME Lakes
LTM Lakes
-6 -5 -4 -3 -2 -1 0
Slope of Trend (|jeq/L/yr)
Sulfate in Adirondack Lakes
TIME Lakes
LTM Lakes
-5 -4 -3 -2 -1 0
Slope of Trend (peq/L/yr)
Figure 27. Comparison of trends in SO42 in TIME probability sites (expressed as the
distribution of slopes in the target population) and LTM sites in New England and the
Adirondacks.
54
-------
0
E
0
CL
E
3
o
0)
E
0)
Q_
-------
TIME population estimates and LTM data
also lead to the same conclusions for
regional trends in [Ca2+ + Mg2+]
concentrations in the two regions (Figure
29). In New England, the distributions of
slopes are nearly identical with median
decreases of-1.6 |ieq/L/year in the TIME
population estimate and -1.5 |ieq/L/year the
LTM data; both regional trends are highly
significant (Tables 5 and 7). In the
Adirondacks both estimates are significant,
but the estimate based on TIME population
data (-1.2 |ieq/L/year) is only half that of the
LTM data (-2.3 |ieq/L/year). The reason for
this modest discrepancy again may be the
different time periods represented by the
data.
Overall, the conclusions drawn from either
TIME probability sampling and LTM
intensive sampling are the same. In New
England lakes, SC>42 is declining only
slightly more than base cations, and the
result is only a small increase in Gran ANC
(significant in the TIME data; insignificant
for LTM). In the Adirondacks, base cation
declines exceed SO42 declines by a large
amount, resulting in more substantial
recovery in Gran ANC.
Do trends in deposition translate into
trends in surface waters? A major goal of
this assessment report is to evaluate the
changes in surface water chemistry in
relation to CAAA reductions in deposition.
In particular, did the changes in deposition
result in commensurate changes in surface
water chemistry in the past decade? There
has been discussion in the scientific
literature about the recovery response for
SC>42~in surface waters: would the response
be rapid or delayed (Church et al., 1989,
Church 1999)?
It is not possible to compare directly
changes in SC>42 concentrations in surface
waters and deposition because of the poorly
quantified effects of dry deposition and
evapotranspiration. Both dry deposition and
the evaporative concentration of ions in
surface waters cause SC>42 concentrations in
lakes and streams to be typically 2-3 times
higher than in deposition (Sisterson et al.,
1990, Rustad et al., 1994). Thus, higher
concentrations in surface waters lead to the
likelihood of greater downward slopes for
SC>42 concentrations in surface waters than
in deposition. In contrast, the percent
change in SC>42 should be relatively
unaffected by ambient concentrations
assuming that dry deposition declines at the
same rate as wet deposition. In this section,
we present a comparison of the slopes of
change in both deposition and surface
waters, using the regional NADP/NTN and
LTM data already presented.
In New England, the Adirondacks and the
Northern Appalachian Plateau, percentage
decline in deposition of sulfate (expressed as
%SO42~) are generally steeper than in
surface waters, suggesting that most aquatic
systems are exhibiting a lagged response.
Apparently watershed soils retain enough
atmospheric SC>42 to reduce the rate
declines occurring in deposition (Figure 30).
Interestingly, the lakes and streams with the
steepest declines in %SC>42~ in these regions
have very similar rates to those in
deposition, suggesting that the most
responsive watersheds are acting essentially
as flow-through systems. As expected, there
is little correspondence between rates of
%SO42~ in streams and deposition in the
Ridge and Blue Ridge provinces (Figure
30), due to the adsorptive capacity of the
highly weathered soils in the region.
56
-------
100
TIME Lakes
LTM Lakes
-5 -4 -3 -2-10 1
Slope of Trend (peq/L/yr)
100 i
-5
[Ca2+ + Mg2+] in Adirondack Lakes
TIME Lakes
LTM Lakes
-4
-3
-2
-1
0
1
Slope of Trend (peq/L/yr)
Figure 29. Comparison of trends in base cations [Ca2+ + Mg2+] in TIME probability sites
(expressed as the distribution of slopes in the target population) and LTM sites in New England
and the Adirondacks.
57
-------
New England, 1990-2000
O
-6 -5 -4 -3 -2 -1 0
Percent change in SO42" per year
00
M—
O
0)
o
Q_
_
=3
E
^
O
100
80 -
Adirondacks, 1990-2000
-5 -4 -3 -2-10 1
Percent change in SO42" per year
to
&
00
O
0)
O
L_
0)
Q_
JS
=3
E
O
100
Appalachian Plateau, 1990-2000
-5 -4 -3 -2-10 1
Percent change in SO,2" per year
to
&
oo
M—
O
-I— '
c
0)
e
0)
Q_
=5
E
^
O
100
Upper Midwest, 1990-2000
Percent change in SO,2" per year
Ridge and Blue Ridge, 1990-2000
O
Percent change in SO4 per year
Figure 30. Comparison of % change in SO42 concentration in wet deposition and surface waters,
by region.
58
-------
In the upper Midwest, the rate of decline in
lakes is greater than the decline in
deposition, probably reflecting the residual
effects of the drought of the late 1980s. All
of the lakes with strong negative trends in
%SO42~ are seepage lakes with long water
residence times. These lakes had trends in
sulfate that reflected a decline from the
evaporative concentration that would have
occurred during the drought and the flushing
of sulfur likely stored in wetlands during the
drought (Dillon et al., 1997, Jeffries et al.,
2002). Some SO42~may also be stored in the
sediments of these lakes via sulfate
reduction (Baker et al., 1986), although the
rate of sulfate reduction is presumed to
decrease as the concentrations decrease.
We conclude that surface waters have
responded relatively rapidly to the decline in
sulfate deposition. There is reason to expect
that additional reductions in deposition
would result in additional declines in surface
water concentrations of sulfate.
Has there been a decline in the number of
acidic waters? Such probability samples
such as TIME lakes allow us to estimate the
number or proportion of surface waters in
each region that are acidic. Because we
have time series data in TIME, we have a
unique opportunity to assess the changes in
the number of acidic systems during the past
decade. We have combined analyses of the
TIME and LTM datasets to conduct this
analysis and have used data from the NSWS,
conducted in the mid 1980s, to estimate the
extent of acidification in the 1980s in
regions where TIME does not operate.
New England Lakes: During 1991 -94,
TIME population data indicated that there
were 386 lakes in New England with Gran
ANCXO |ieq/L, or 5.6% of the population.
Both TIME and LTM data from this region
suggest that only a small increase in Gran
ANC has occurred since that time period;
combined TIME/LTM results suggest a
change of ca. +0.3 jieq/L/year (Tables 5 and
7). If we extend this time trend for 10 years,
then sensitive lakes can be expected to have
increased in ANC by ca. 3 |ieq/L. In the
1991-94 TIME probability survey there
were 12 lakes with Gran ANC values
between 0 and -3 |ieq/L, and only these
lakes are estimated to have become non-
acidic by the present time (this estimate is
for the time period 2001-04, 10 years after
the original TIME estimate). We find little
evidence of a profound change in the acidity
status of this region with the proportion of
acidic lakes decreasing only from 5.6% to
5.5% over the previous 10 years (Table 8).
Adirondack Lakes: TIME population
estimates from 1991-94 indicate that 13.0%
of lakes, or 238 lakes, were acidic in the
early 1990s. If we apply an approximate
rate of change in Gran ANC of +0.8
|ieq/L/year to these estimates (based on
trend slopes for TIME and LTM data in this
region, Tables 5 and 7), then sensitive lakes
could be expected to have increased their
ANC values by 8 |ieq/L during the 1990s.
We estimate that at the present time there
are ca. 149 Adirondack lakes with Gran
ANC<0, or 8.1% of the population. This
represents a decrease of nearly 5%; roughly
38% of lakes that were acidic in the early
1990s are estimated to be non-acidic a
decade later.
Northern Appalachian Plateau streams:
Population estimates from TIME surveys of
streams in the Northern and Central
Appalachian Plateau ecoregions
(synonymous with the northern Appalachian
Plateau region covered by the LTM project)
indicate that 5,014 kilometers of streams
were acidic in 1993-94 (at the beginning of
TIME sampling); this represents 11.8% of
the stream length. The approximate rate of
change in Gran ANC in the region (Tables 5
and 7) is +0.7 jieq/L/year and extrapolated
59
-------
for a 10-year period amounts to an increase
in sensitive streams of+7 |ieq/L. We
estimate that roughly 3,600 kilometers of
stream (or 8.5%) remain acidic in this region
at the present time. This represents a ca.
28% decrease in acidic stream length over
the decade.
Upper Midwest Lakes: The TIME project
has not operated in the Upper Midwest, so
the best population estimates for this region
come from the 1984 Eastern Lake Survey
(Linthurst et al., 1986b). ELS sampling in
the Upper Midwest estimated that 251 lakes,
or 2.9% of the population, were acidic in
1984 (Baker et al., 1991). LTM data
suggest a rate of change of+1 |ieq/L/year in
this region (Table 5); extrapolated to the
present, this represents an increase of+18
|ieq/L of Gran ANC in sensitive lakes
between 1984 and 2002. In 1984, 80 lakes
had Gran ANC values less than -18 |ieq/L,
and all of these are now estimated to be non-
acidic. This represents a change from 2.9%
acidic in 1984, to 0.9% in 2002, an overall
reduction in the number of acidic lakes of
68%.
Ridge/Blue Ridge Province streams: While
we do not have adequate data from TIME to
calculate trends in this region, the data are
adequate to estimate the number of acidic
streams. TIME data indicate that fewer than
400 km of streams (< 1%) were acidic in the
Ridge and Blue Ridge region in 1993-94.
Because we have no evidence of a
significant change in Gran ANC in this
region (Table 5), we estimate that <1% of
Ridge/Blue Ridge streams remain acidic in
2002. There has probably been no change in
the number of acidic waters in this region in
the past decade.
Two important caveats in this analysis
should be noted. First, as shown in Table 6,
the rates of ANC increase (outside of the
Ridge and Blue Ridge provinces) are more
rapid in acidic lakes and streams than in low
ANC sites. Due to sample size constraints
(it is not possible to do the ANC class
analysis by region), the changes in Gran
ANC used in the analysis in Table 8 are
based on the median change in all sites in a
region. 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 here will be conservative.
It is also important to realize that the trends
we report here are for recovery from chronic
acidification. Our regression analysis
focuses on the central tendency of sites (e.g.,
their median or mean ANC over time) and
whether this central tendency is increasing
or decreasing. We know that most sites
exhibit seasonally lower ANC and pH
values than would be captured by the trend
analysis. In many cases, sites that are not
chronically acidic nonetheless undergo
short-term episodic acidification during
spring snowmelt or during intense rain
events. In order to estimate the relative
importance of episodic acidification in the
regions we cover in this report, we examined
the mean summer Gran ANC at each site
and compared it to the average minimum
ANC experienced each spring during the
period 1990-2000. The results of this
analysis are shown in Figure 31. On
average, spring ANC values in New
England, the Adirondacks and Northern
Appalachian Plateau are 30 jieq/L lower
than summer values. This implies that lakes
and streams in these regions would need to
recover to Gran ANC values above 30 jieq/L
before they could be expected not to
experience acidic episodes.
Estimating the magnitude of historical
lake acidification. It is widely established
that many present-day acidic lakes were at
least marginally acidic in pre-industrial
times (Davis and Anderson 1985, Sullivan
1990, Cumming et al., 1994, Smol et al.,
1998). The main method for determining
60
-------
pre-historical pH is using siliceous fossil
remains of diatom and chrysophyte algal
species in lake sediment cores. Cumming et
al., (1994) describes a classification system
for the low ANC lakes with four types of
response: 1) lakes that have not acidified; 2)
lakes that acidified in the mid-1900s; 3) low
pH lakes that acidified around 1900; 4)
naturally acidic lakes that only acidified
slightly. Two examples of these
acidification responses are shown (Figure
32) from New England. Little Long Pond,
with a present Gran ANC of 10 |ieq/L,
acidified by about 0.2 pH units after 1920.
Its neighbor Mud Pond (Gran ANC today of
-20 |ieq/L) acidified from about pH 5.4 to
4.9, beginning early in the century.
The estimates of the magnitude of lake
acidification range from fractions of a pH
unit to about 2.0 depending on the lake and
the region. Chrysophyte analyses tend to
infer a lower pH because they are most
abundant in the spring and therefore reflect
the lower pH spring condition and because
they may respond to episodic acidifications
(Cumming etal., 1994). The average
diatom-inferred pH decline in New England
lakes is about 0.3 pH units (Davis et al.,
1994). Sullivan et al., (1990) inferred
similar average changes for the
Adirondacks. Cumming et al., (1994)
examined 14 Adirondack lakes that had
acidified since 1900 (mean pre-industrial pH
= 5.6, range 4.9 to 6.2) and inferred a mean
chrysophyte-inferred pH decline of about
1.1 pH units. Kahl etal., (1989, 1991)
observed that no clearwater (low DOC)
lakes in Maine had pH less than 5 and
concluded that organic acidity was
necessary to acidify lakes to less than pH 5
at levels of acidic deposition in the 1980s.
200
g" 150
o
-z.
<
I
1
O)
Q.
CO
c
ro
0)
100
50
-50
• New England Lakes
O Adirondack Lakes
O Appalachian Streams
O
-50 0 50 100 150
Mean Summer ANC (|Jeq/L)
200
Figure 31. Relationship between summer and spring ANC values at LTM sites in New England, the
Adirondacks and Northern Appalachian Plateau. Values are mean summer values for each site during the
period 1990-2000 (horizontal axis) and mean spring minima for each site for the same time period. On
average, spring ANC values are 30 (ieq/L lower than summer values.
61
-------
Table 8. 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.
Region
New England
Adirondacks.
No. Appalachians
Ridge/Blue Ridge
Upper Midwest
Population
Size
6,834 lakes
1830 lakes
42,426 km
32,687 km
8,574 lakes
Number
Acidic1
386 lakes
238 lakes
5,014km
1,634km
251 lakes
%
Acidic2
5.6%
13.0%
11.8%
5.0%
2.9%
Time
Period of
Estimate
1991-94
1991-94
1993-94
1987
1984
Current
Rate of
ANC
change3
+0.3
+0.8
+0.7
-0.0
+1.0
Estimated
Number
Currently
Acidic
374 lakes
149 lakes
3,600 km
1,634 km
80 lakes
Current %
acidic
5.5%
8.1%
8.5%
5.0%
0.9%
% Change
in Number
of Acidic
Systems
-2%
-38%
-28%
0%
-68%
1 Number of lakes/streams with Gran ANC<0 in past probability survey (data collected at "Time Period of Estimate," in column 5)
""' Percent of population (from Column 2) with Gran ANC<0 in past probability survey (data collected at "Time Period of Estimate," in column 5)
Based on regional trends presented in this report, in ^eq/L/year
62
-------
Little Long Pond
Mud Pond
ro
0)
2000
1950
1900
1850
1800 -
1750
1700
1650
2000
1950
1900
1850
1800
1750
1700
1650
5.6 5.7 5.8 5.9 6.0 6.1 6.2
Diatom Inferred pH
4.84.95.05.1 5.25.35.45.5
Diatom Inferred pH
Figure 32. Paleolimnological reconstruction of historical pH change in two New England LTM
lakes, as inferred from diatom remains (from Davis et al., 1994).
With the exception of lakes in the heavily
impacted region around the Sudbury,
Ontario, studies of pre-industrial lake pH
agree that the pH of presently acidified lakes
was less than pH 6 with rare exceptions
before the onset of acidic deposition.
Therefore, we propose that pH 6 is an upper
limit for recovery of acidified lakes
equivalent to an Gran ANC of about 20-30
|ieq/L in all regions.
The interaction of S, N, and cations.
Interactions between trends in SO42 , NOs
and [Ca2+ + Mg2+] govern the magnitude of
trends in Gran ANC. At East Bear Brook in
Maine (Norton et al., 1999), both SO4
NOsliave declined significantly since
sampling began in 1998. If the main
2-
and
regional change in surface water chemistry
had been a decline in SC>4 without any other
changes in major ion chemistry or DOC,
then regional increases in ANC would have
followed. Instead of a simple ANC
response, there have been substantial
declines in base cations and modest
increases in DOC (Figure 33).
The relationship among these changes is
illustrated in the response of East Bear
Brook at BBWM (Figure 33).
Concentrations of SO42 and NOs have
declined in parallel which is presumably
coincidental (Figure 33a). The rate of
decline in base cations is faster than the
decline in SO42 (Figure 33b). The large
decline in base cations apparently results
63
-------
from a change in soil ion exchange
processes or from an increase in biomass
uptake because there is no corresponding
change in concentrations of silica, an
indicator of mineral weathering rates (Figure
33c). Silica concentrations in the LTM and
TIME networks were also statistically
unchanged during the 1990s.
The trend in ANC can also be estimated
from:
ANC = [Ca2+ + Mg2+ Na+ + K+] minus
unchanged in regions with low ambient
concentrations.
In this low DOC stream, the estimated and
measured Gran ANC agree closely
suggesting that acid anion inputs and
buffering by base cations are indeed the
likely controlling processes for ANC. It is
also possible to calculate the ANC change
that would have occurred if NOs had not
declined in 1990s (Figure 33d) by using the
equation above but holding NOs~ constant.
The resulting calculated ANC in 1998 would
be -20 jieq/L instead of zero. It is possible
that the observed regional decline in NO3
during the 1990s prevented additional
acidification in sensitive systems.
CONCLUSIONS
Trends in deposition and surface water
chemistry. The CAAA has been successful
in reducing the emissions of SC>2 and
deposition of SC>42 accelerating the trend in
place since about 1970. There was little
change in NOX emissions and deposition
until the late 1990s when a slight decrease
occurred due to Phase I controls on nitrogen
emissions beginning in 1996. In response,
SC>42 concentrations in surface waters
declined across all the glaciated regions of
the North and East (Figure 34). South of the
glaciated limit, SC>42 concentrations have
increased slightly due to the re-equilibration
of SC>42 retaining soils with changing
deposition. Nitrate concentrations exhibited
few significant trends: NCV decreased
slightly in regions with high NOs and was
Base cations, which are important for
neutralization of acidity in precipitation and
in watersheds, increased non-significantly in
atmospheric deposition during 1990-2000.
However, the 20-year decline in base cations
in surface waters continued, offsetting some
of the decline in surface SC>42 . As a result,
Gran ANC and pH have increased less than
might be expected from SC>42 trends alone.
Significant increases in Gran ANC were
recorded in the Adirondack Mountains, the
Northern Appalachian Plateau, and in the
Upper Midwest (Figure 34). Lakes in New
England show little sign of recovery in the
1990s, and streams in the Ridge/Blue Ridge
province may even be acidifying slightly.
Expectations for recovery. While the
widespread declines in SO42 reported here
for sensitive surface water regions
(exclusive of the Ridge/Blue Ridge region)
are important signs of the success of
emission control programs, the true test of
Title IV s effectiveness is a reduction in the
acidity, and the eventual biological
recovery, of lakes and streams in the U.S.
We report here that in some regions the
number of acidic lakes and streams has
declined significantly in the past 10 years
(Table 8). In particular, many lakes in the
Adirondacks and Upper Midwest, and
streams in the Northern Appalachian
Plateau, have achieved positive Gran ANC
values since implementation of the CAAA.
It is important to recognize, however, that
despite a roughly one-third reduction in the
number of acidic systems in these regions,
8% of Adirondack lakes and 8% of Northern
Appalachian streams remain acidic. It is
beyond the scope of this report to forecast
whether continued emissions reductions will
substantially reduce these numbers or
whether further additional regulations may
be necessary.
64
-------
c.
O
U— '
ro
"c
Q)
o
c
o
O
D)
Q
100
East Bear Brook, Maine
80 -
60 -
40
20 }
0
SO/* (ueq/L)
(a)
O O
N03~ (peq/L)
1988 1990 1992 1994 1996 1998 2000
120
100 -
80 -
60
(b)
SO,2'* (ueq/L)
1988 1990 1992 1994 1996 1998 2000
-2 100
80 -
60
40
20
(c)
A SBC*([JM)
A
Si (|JM)
1988 1990 1992 1994 1996 1998 2000
10
0
-10
-20
-30
measured Gran ANC (ueq/L)
(d)
calculated Gran ANC
with constant NO3~ (ueq/L)
1988 1990 1992
1994
Year
1996 1998 2000
Figure 33. The interaction of SO42 , NO3 and base cations in East Bear Brook, the reference
watershed at the Bear Brook Watershed in Maine.
65
-------
Uncertainties in the future behavior
and base cations make it very difficult to say
how long the current recovery trends will
continue.
It is also important to note that one region,
the Ridge/Blue Ridge province, has yet to
show any signs of recovery, either in SC>42
concentrations or in acidity. The proportion
of streams in this region that are acidic is
currently less than 1%, but this region has
the highest likelihood of undergoing further
acidification as the capacity of regional soils
to adsorb atmospheric SC>42 reaches
capacity. The small increase is surface
water SC>42 we report for this region should
be a red flag to regulators that this region
deserves further research and assessment.
Research needs. Scientists and regulators
have long noted the frustrations associated
with dealing with complex environmental
issues without long-term baseline data.
With the continuing evolution of the
chemistry of atmospheric deposition, data
such as are represented in the TIME and
LTM programs will be the only method of
assessing recovery of surface water
chemistry. LTM and TIME will also
provide baseline sites for future biological
assessments of recovery.
Regional Trends, 1990-2000
(in lakes and streams)
Sulfate (|jeq/L/yr)
Nitrate (|jeq/L/yr)
ANC (ueq/L/yr)
Hydrogen Ion (|jeq/L/yr)
Base Cations (|jeq/l_/yr)
DOC (mg/Uyr)
Aluminum (ug/L/yr)
-3 -2-10 1
Slope of Trend
••• New England Lakes
i i Adirondack Lakes
i i No. Appalachian Streams
••• Upper Midwest Lakes
••• Ridge/Blue Ridge Streams
Figure 34. Summary of surface water trends in SO42 , NO3, Gran ANC, base cations [Ca2+
Mg2+], DOC and aluminum, by region.
66
-------
Importantly, these data will also prove
useful for other relevant environmental
management issues such as base cation
depletion, nitrogen saturation, effects of
drought, forest health, and the myriad
potential ecosystem-level effects of climate
change. Increases in dissolved organic
carbon in most lake populations may have
contributed natural organic acidity to surface
waters, further complicating our
interpretation of the response to the CAAA.
This factor is an important long-term
research question that is not well
understood. These LTM and TIME sites are
core research locations that will provide
essential information for a number of
present and future research questions faced
by EPA. Their long-term data serve as the
foundation for other ecological research to
be conducted using the locations and their
data. Without such data, our ability to ask
the right questions is reduced, and our
ability to base the answers to these questions
on actual data is seriously compromised. It
is not possible to re-create or replace long-
term data once the data record is interrupted.
The more long-term data we accumulate, the
better we can differentiate competing
controls on surface water chemistry, and the
better we can identify slow and subtle
emerging trends, like the increased
contribution of DOC in surface waters.
67
-------
TERMS AND ACRONYMS USED IN THIS REPORT
acidic describes lake or stream water with Gran ANC less than zero, corresponding to
pH values less than ca. 5.5.
acidified describes lake or stream water with sufficient historical information, or
appropriate chemical relationships, to justify classification as a water that was
historically less acidic than shown by recent measurement. The term may also
apply to stream water which is temporarily acidified during a hydrologic episode.
ANC Acid Neutralizing Capacity - a theoretical estimate of the ability of water to
buffer acid (similar to buffering capacity or alkalinity). ANC may be either
calculated ('calculated ANC' = sum of cations - sum of anions) or measured
(Gran ANC) by titrating a sample of lake or stream water with a known
concentration of acid.
anion a negative charged ion, especially sulfate (SO4"2) and (nitrate NO3") of importance
to this report. Also see cation.
alkalinity See ANC
ALSC Adirondack Lake Survey Corporation
BBWM Bear Brook Watershed in Maine. The original EPA Watershed Manipulation
Project location for a paired-watershed experiment of enhanced S and N
deposition (in eastern Maine) begun in 1987, with the experimental treatment
continuing since 1989.
CaPMon The Canadian equivalent to the U.S. NADP.
cation A positively charged ion, especially hydrogen (H+), and the base cations calcium
(Ca+2), magnesium (Mg+2), potassium (K+) and sodium (Na+). See anion.
DOC Dissolved organic carbon
CAA Federal Clean Air Act of 1970.
CAAA Federal Clean Air Act Amendments of 1990.
CASTNet Clean Air Status and Trends Network, the U.S. EPA-operated dry deposition
network located primarily in the eastern U.S.
ELS Eastern Lake Survey. The 1984 EPA program which established the baseline
chemical condition of lakes in the eastern U.S. using a statistical survey of lake
subpopulations.
EMAP Environmental Monitoring and Assessment Program. The U.S. EPA monitoring
program begun in 1991, which includes TIME and LTM.
EMAP-SW EMAP-Surface Waters. The aquatic module within the larger environmental
monitoring mission of EMAP.
GAO General Accounting Office
68
-------
Gran ANC Gran Acid Neutralizing Capacity - a measured estimate of ANC, based on using
the Gran technique to find the inflection point in an acid-base titration of a water
sample (Gran 1952).
HELM High Elevation Lakes in Maine. The 90 lakes higher than 600 meters which, as a
population, are as acidified as lakes in the Adirondack NY region.
LTM The U.S. EPA Long-Term Monitoring project. This EPA monitoring program
began in 1983 and included low ANC lakes in Maine, New York (Adirondacks)
and Vermont, and low ANC streams in New York (Catskills), Pennsylvania and
Virginia. Before 1995 it also included lakes in Colorado and the Upper Midwest
(Wisconsin, Michigan and Minnesota).
NADP/NTN National Atmospheric Deposition Program/National Trends Network, the
U.S.G.S.-led entity that operates the 200+ site network of wet-only precipitation
collectors in the U.S.
NAPAP National Acidic Precipitation Assessment Program, the program under which
federally sponsored acid rain research was funded in the 1980s. NAPAP was
responsible for the 1990 assessment report and has continued a integrative role
into 2002.
NSS U.S. EPA National Stream Survey.
NSWS U.S. EPA National Surface Water Survey. Consisting of the Eastern Lake Survey
(see ELS), the Western Lake Survey (see WLS) and National Stream Survey (see
NSS).
PIRLA Paleolimnological Investigations of Recent Lake Acidification - a project
conducted in the 1980s using diatom remains in lake sediment cores to reconstruct
historic chemistry trends.
recovery A return to a more pre-industrial condition, in chemistry or biology. Chemically,
recovery would be considered an increase in alkalinity (ANC) and/or an increase
in pH, both equivalent to a decrease in acidity in surface water.
TIME The U.S. EPA Temporally Integrated Monitoring of Ecosystems project. TIME is
the component of EMAP targeted at monitoring subpopulations of lakes
(Northeast) and streams (Mid-Atlantic) that are most likely to respond to changes
in acidic deposition. TIME and EMAP both use probability techniques to select
sites so that the results can be statistically inferred to represent the population of
interest (e.g., acid sensitive lakes and/or streams in the targeted region).
USGS U.S. Geological Survey, Department of the Interior.
WLS Western Lake Survey. The 1986 EPA program which established the baseline
chemical condition of lakes in the western U.S. using a statistical survey of lake
subpopulations.
WMP U. S. EPA Watershed Manipulation Project. See BBWM.
69
-------
REFERENCES
Aber, I, W. McDowell, K. Nadelhoffer, A. Magill, G. Berntson, M. Kamakea, S. McNulty, W.
Currie, L. Rustad, and I. Fernandez. 1998. Nitrogen saturation in temperate forest
ecosystems. BioScience 48:921-934.
Aber, J. D., and C. T. Driscoll. 1997. Effects of land use, climate variation and N deposition on
N cycling and C storage in northern hardwood forests. Global Biogeochemical Cycles
11:639-648.
Aber, J. D., C. L. Goodale, S. V. Ollinger, M.-L. Smith, A. H. Magil, M. E. Martin, R. A.
Hallett, and J. L. Stoddard. In press. Is nitrogen deposition altering the nitrogen status of
Northeastern forests? BioScience.
Aber, J. D., A. Magill, S. G. McNulty, R. D. Boone, K. J. Nadelhoffer, M. Downs, and R.
Hallett. 1995. Forest biogeochemistry and primary production altered by nitrogen
saturation. Water Air and Soil Pollution 85:1665-1670.
Aber, J. D., K. J. Nadelhoffer, P. Steudler, and J. M. Melillo. 1989. Nitrogen saturation in
northern forest ecosystems. BioScience 39:378-386.
Altman, D., T. Bryant, M. Gardner, and D. Machin. 2000. Statistics with Confidence. BMJ
Books, London.
Baker, J. P., and S. W. Christensen. 1991. Effects of acidification on biological communities in
aquatic ecosystems. Pages 83-106 in D. F. Charles, editor. Acidic Deposition and
Aquatic Ecosystems. Regional Case Studies. Springer-Verlag, New York.
Baker, J. P., and C. L. Schofield. 1982. Aluminum toxicity to fish in acidic waters. Water Air
and Soil Pollution 18:289-309.
Baker, J. P., J. Van Sickle, C. J. Gagen, B. P. Baldigo, D. W. Bath, R. F. Carline, D. R. DeWalle,
W. A. Kretser, P. S. Murdoch, W. E. Sharpe, H. A. Simonin, and P. J. Wigington. 1996.
Episodic acidification of small streams in the northeastern United States: Effects on fish
populations. Ecological Applications 6:422-437.
Baker, L. A., P. L. Brezonik, and C. D. Pollman. 1986. Model of internal alkalinity generation:
Sulfate retention component. Water Air and Soil Pollution 31:89-94.
Baker, L. A., A. T. Herlihy, P. R. Kaufmann, and J. M. Eilers. 1991. Acidic lakes and streams
in the United States: The role of acidic deposition. Science:! 151-1154.
Baker, L. A., P. R. Kaufmann, A. T. Herlihy, and J. M. Eilers. 1990. Current Status of Surface
Water Acid-Base Chemistry. NAPAP State-of-Science/Technology Report No. 9,
National Acid Precipitation Assessment Program, Washington, DC.
Charles, D. F., M. W. Binford, E. J. Furlong, R. A. Kites, M. J. Mitchell, S. A. Norton, F.
Oldfield, M. J. Paterson, J. P. Smol, A. J. Uutala, J. R. White, D. R. Whitehead, and R. J.
Wise. 1990. Paleoecological investigation of recent lake acidification in the Adirondack
Mountains, New York. Journal of Paleolimnology 3:195-241.
Charles, D. F., and S. A. Norton. 1986. Paleolimnological evidence for trends in atmospheric
deposition of acids and metals. Pages 335-431 in Acid Deposition: Long-term Trends.
National Academy Press, Washington, D.C.
Charles, D. R., editor. 1991. Acidic Deposition and Aquatic Ecosystems. Regional Case
Studies. Springer-Verlag, New York.
Charlson, R. J., and H. Rodhe. 1982. Factors controlling the acidity of natural rainwater.
Nature 295:683-685.
70
-------
Chen, C. W., S. A. Gherini, R. J. M. Hudson, J. D. Dean, and R. A. Goldstein. 1984.
Development and calibration of the Integrated Lake-Watershed Acidification Study
model. Pages 175-203 in J. L. Schnoor, editor. Modeling of Total Acid Precipitation
Impacts. Butterworth Publishers, Stoneham, MA.
Christophersen, N., and R. F. Wright. 1981. Sulfate flux and a model for sulfate concentrations
in stream water at Birkenes, a small forested catchment in southernmost Norway. Water
Resources Research 17:377-389.
Church, M. R. 1999. The Bear Brook Watershed Manipulation Project: Watershed science in a
policy perspective, in S. A. Norton and I. J. Fernandez, editors. The Bear Brook
Watershed in Maine: A Paired Watershed Experiment — The First Decade (1987-1997).
Kluwer Academic Publishers.
Church, M. R., P. W. Shaffer, K. N. Eshleman, and B. P. Rochelle. 1990. Potential effects of
sulphur deposition on stream chemistry in the Southern Blue Ridge Province. Water Air
and Soil Pollution 50:39-48.
Church, M. R., K. W. Thornton, P. W. Shaffer, D. L. Stevens, B. P. Rochelle, G. R. Holdren, M.
G. Johnson, J. L. Lee, R. S. Turner, D. L. Cassell, D. A. Lammers, W. G. Campbell, C. I.
Liff, C. C. Brandt, L. H. Liegel, G. D. Bishop, D. C. Mortenson, S. M. Pierson, and D. D.
Schomoyer. 1989. Direct/Delayed Response Project: Future Effects of Long-Term
Sulfur Deposition on Surface Water Chemistry in the Northeast and Southern Blue Ridge
Province. Volume III. Level III Analyses and Summary of Results. U.S. Environmental
Protection Agency, Washington, D.C.
Cleveland, L., E. Little, S. Hamilton, D. Buckler, and J. Hunn. 1986. Interactive toxicity of
aluminum and acidity to early life stages of brook trout. Transactions of the American
Fisheries Society 115:610-620.
Cosby, B. J., G. M. Hornberger, R. F. Wright, and J. N. Galloway. 1986. Modelling the effects
of acid deposition: Control of long-term sulfate dynamics by soil sulfate adsorption.
Water Resources Research 22:1283-1291.
Cosby, B. J., R. F. Wright, G. M. Hornberger, and J. N. Galloway. 1985. Modeling the effects
of acid deposition: Assessment of a lumped parameter model of soil water and
streamwater chemistry. Water Resources Research 21:51 -63.
Gumming, B. F., K. A. Davey, J. P. Smol, and H. J. B. Birks. 1994. When did acid-sensitive
Adirondack lakes (New York, USA) begin to acidify and are they still acidifying?
Canadian Journal of Fisheries and Aquatic Science 51:1550-1568.
David, M., G. Vance, and J. Kahl. 1999. Chemistry of dissolved organic carbon at Bear Brook
Watershed, Maine: stream water response to (NH/t^SO/t. Environmental Monitoring and
Assessment 55:149-163.
Davis, R. B., and D. S. Anderson. 1985. Methods of pH calibration of sedimentary diatom
remains for reconstructing history of pH in lakes. Hydrobiologia 120:69-87.
Davis, R. B., D. S. Anderson, and F. Berge. 1985. Palaeolimnological evidence that lake
acidification is accompanied by loss of organic matter. Nature 316:436-438.
Davis, R. B., D. S. Anderson, S. A. Norton, P. R. Sweets, J. Ford, and J. S. Kahl. 1994.
Sedimented diatoms in northern New England lakes and their use as pH and alkalinity
indicators. Canadian Journal of Fisheries and Aquatic Science 51:1855-1876.
DeWalle, D. R., and B. R. Swistock. 1994. Causes of episodic acidification in five
Pennsylvania streams on the northern Appalachian Plateau. Water Resources Research
30:1955-1963.
71
-------
Dillon, P. J., L. A. Molot, and M. Putter. 1997. A note on the effects of El Nino-related drought
on the recovery of acidified lakes. International Journal of Environmental Monitoring
and Assessment 46:105-111.
Dixit, S. S., W. Keller, A. Dixit, and J. P. Smol. 2001. Diatom-inferred dissolved organic
carbon reconstructions provide assessments of pastUV-B penetration in Canadian Shield
lakes. Canadian Journal of Fisheries and Aquatic Sciences 58:543-550.
Driscoll, C. T., G. B. Lawrence, A. J. Bulger, T. J. Butler, C. S. Cronan, C. Eager, K. F. Lambert,
G. E. Likens, J. L. Stoddard, and K. C. Weathers. 2001. Acidic deposition in the
northeastern United States: Sources and inputs, ecosystem effects and management
strategies. BioScience 51:180-198.
Driscoll, C. T., G. E. Likens, L. O. Hedin, J. S. Eaton, and F. H. Bormann. 1989. Changes in the
chemistry of surface waters: 25-year results at the Hubbard Brook Experimental Forest.
Environmental Science and Technology 23:137-143.
Driscoll, C. T., R. M. Newton, C. P. Gubala, J. P. Baker, and S. W. Christensen. 1991.
Adirondack Mountains. Pages 133-202 in D. F. Charles, editor. Acidic Deposition and
Aquatic Ecosystems: Regional Case Studies. Springer-Verlag, New York, NY.
Driscoll, C. T., K. M. Postek, W. Kretser, and D. J. Raynal. 1995. Long-term trends in the
chemistry of precipitation and lake water in the Adirondack region of New York, USA.
Water Air and Soil Pollution 85:583-588.
Driscoll, C. T., and R. Van Dreason. 1993. Seasonal and long-term temporal patterns in the
chemistry of Adirondack lakes. Water Air and Soil Pollution 67:319-344.
Eshleman, K. N., R. P. Morgan II, J. R. Webb, F. A. Deviney, and J. N. Galloway. 1998.
Temporal patterns of nitrogen leakage from mid-Applachian forested watersheds: role of
insect defoliation. Water Resources Research 34:2005-2016.
Evans, C. D., and D. T. Monteith. 2001. Chemical trends at lakes and streams in the UK Acid
Waters Monitoring Network, 1988-2000: Evidence for recent recovery at a national scale.
Hydrology and Earth Systems Sciences 5:351-366.
Fenn, M. E., M. A. Poth, J. D. Aber, J. S. Baron, B. T. Bormann, D. W. Johnson, A. D. Lemly, S.
G. McNulty, D. F. Ryan, and R. Stottlemyer. 1998. Nitrogen excess in North American
ecosystems: predisposing factors, ecosystem responses, and management strategies.
Ecological Applications 8:706-733.
Ford, J., J. L. Stoddard, and C. F. Powers. 1993. Perspectives in environmental monitoring: an
introduction to the U.S. EPA Long-Term Monitoring (LTM) project. Water Air and Soil
Pollution 67:247-255.
Gallagher, J., and J. Baker. 1990. Current status of fish commuities in Adirondack lakes. Pages
3-11 to 13-48 in Adirondack Lakes Survey: An Interpretative Analysis of Fish
Communities and Water Chemistry, 1984-87. Adirondack Lake Survey Corporation,
Ray Brook, NY.
Galloway, J. N., S. A. Norton, and M. R. Church. 1983. Freshwater acidification from
atmospheric deposition of sulfuric acid: A conceptual model. Environmental Science
and Technology 17:541-545.
Gherini, S., R. Munson, E. Altwicker, R. April, C. Chen, N. Clesceri, C. Cronan, C. Driscoll, A.
Johannes, R. Newton, N. Peters, and C. Schofield. 1989. Regional Integrated Lake-
Watershed Acidification Study (RILWAS): Summary of Major Findings. Electric Power
Research Institute, Palo Alto, CA.
72
-------
Goodale, C. L., J. D. Aber, and W. H. McDowell. 2000. The long-term effects of disturbance
on organic and inorganic nitrogen export in the White Mountains, New Hampshire.
Ecosystems 3:433-450.
Government Accounting Office. 2000. Acid Rain: Emission Trends and Effects in the Eastern
United States. GAO/RCED-00-47, Goverment Accounting Office, Washington, DC.
Gran, G. 1952. Determination of the equivalence point in potentiometric titrations. ANALYST:
International Congress on Analytical Chemistry 77:661-671.
Hedin, L. O., L. Granat, G. E. Likens, T. A. Buishand, J. N. Galloway, T. J. Butler, and H.
Rodhe. 1994. Steep declines in atmospheric base cations in regions of Europe and North
America. Nature 367:351-354.
Henriksen, A., and M. Grande. 2002. Lake Langtjern — fish studies in the Langtjern area 1966-
2000. Report SNO 4537-2002, Norwegian Institute for Water Research, Oslo.
Herlihy, A. T., D. P. Larsen, S. G. Paulsen, N. S. Urquhart, and B. J. Rosenbaum. 2000.
Designing a spatially balanced, randomised site selection process for regional stream
surveys: The EMAP Mid-Atlantic Pilot Study. Environmental Monitoring and
Assessment 63:95-113.
Hirsch, R. M., and J. R. Slack. 1984. A nonparametric trend test for seasonal data with serial
dependence. Water Resources Research 20:727-732.
Hirsch, R. M., J. R. Slack, and R. A. Smith. 1982. Techniques of trend analysis for monthly
water quality analysis. Water Resources Research 18:107-121.
Holland, D. M., P. P. Principe, and J. E. Sickles II. 1999. Trends in atmospheric sulfur and
nitrogen species in the eastern United States for 1989-1995. Atmospheric Environment
33:37-49.
Husar, R. B., T. J. Sullivan, and D. F. Charles. 1991. Historical trends in atmospheric sulfur
deposition and methods for assessing long-term trends in surface water chemistry. Pages
65-82 in D. F. Charles, editor. Acidic Deposition and Aquatic Ecosystems: Regional
Case Studies. Springer-Verlag, New York.
Jaworski, N. A., R. W. Howarth, and L. I. Hetling. 1997. Atmospheric deposition of nitrogen
oxides onto the landscape contributes to coastal eutrophication in the northeast United
States. Environmental Science and Technology Scope 31:1995-2004.
Jeffries, D. S., R. G. Semkin, F. D. Beall, and J. Franklyn. 2002. Temporal trends in water
chemistry in the Turkey Lakes watershed, Ontario, Canada, 1982-1999. Water Air and
Soil Pollution Focus 2:5-22.
Kahl, J. S., S. A. Norton, C. A. Cronan, I. J. Fernandez, L. C. Bacon, and T. A. Haines. 1991.
Maine. Pages 203-235 in D. F. Charles, editor. Acidic Deposition and Aquatic
Ecosystems: Regional Case Studies. Springer-Verlag, New York, NY.
Kahl, J. S., S. A. Norton, I. J. Fernandez, K. J. Nadelhoffer, C. T. Driscoll, and J. D. Aber.
1993a. Experimental inducement of nitrogen saturation at the watershed scale.
Environmental Science and Technology 27:565-568.
Kahl, J. S., S. A. Norton, T. A. Haines, and R. B. Davis. 1993b. Recent trends in the acid-base
status of surface waters in Maine, U.S.A. Water Air and Soil Pollution 67:281-300.
Kahl, J. S., S. A. Norton, T. A. Haines, C. W. Fay, and R. B. Davis. 1989. The influence of
organic acidity on the chemistry of Maine surface waters. Water Air and Soil Pollution
46:221-233.
73
-------
Kahl, J. S., S. A. Norton, T. A. Haines, E. A. Rochette, R. H. Heath, and S. C. Nodvin. 1992.
Mechanisms of episodic acidification in low-order streams in Maine, USA.
Environmental Pollution 78:37-44.
Kahl, J. S., and M. Scott. 1994. High Elevation Lake Monitoring in Maine: 1986-89. Maine
Department of Environmental Protection, Augusta, Maine.
Kaufmann, P. R., A. T. Herlihy, J. W. Elwood, M. E. Mitch, W. S. Overton, M. J. Sale, J. J.
Messer, K. A. Cougan, D. V. Peck, K. H. Reckhow, A. J. Kinney, S. J. Christie, D. D.
Brown, C. A. Hagley, and H. I. Jager. 1988. Chemical Characteristics of Streams in the
Mid-Atlantic and Southeastern United States. Volume I: Population Descriptions and
Physico-Chemical Relationships. U.S. Environmental Protection Agency, Washington,
D.C.
Keene, W. C., and J. N. Galloway. 1984. Organic acidity in precipitation of North America.
Atmospheric Environment 18:2491-2498.
Keller, W., S. S. Dixit, and J. Heneberry. 2001. Calcium declines in northeastern Ontario lakes.
Canadian Journal of Fisheries and Aquatic Science 58:2011-2020.
Kingston, J. C., and H. J. B. Birks. 1990. Dissolved organic carbon reconstructions from diatom
assemblages in PIRLA project lakes, North America. Philosophical Transactions of the
Royal Society of London 327:279-288.
Krug, E. C., and C. R. Frink. 1983. Acid rain on acid soil: A new perspective. Science
221:520-525.
Landers, D. H., J. M. Eilers, D. F. Brakke, W. S. Overton, P. E. Kellar, W. E. Silverstein, R. D.
Schonbrod, R. E. Crowe, R. A. Linthurst, J. M. Omernik, S. A. league, and E. P. Meier.
1987. Characteristics of Lakes in the Western United States - Volume I: Population
Descriptions and Physico-Chemical Relationships. U.S. Environmental Protection
Agency, Washington, D.C.
Landers, D. H., W. S. Overton, R. A. Linthurst, and D. F. Brakke. 1988. Eastern Lake Survey:
Regional estimates of lake chemistry. Environmental Science and Technology 22:128-
135.
Larsen, D. P., K. W. Thornton, N. S. Urquhart, and S. G. Paulsen. 1994. The role of sample
surveys for monitoring the condition of the nation's lakes. Environmental Monitoring
and Assessment 32:101-134.
Larsen, D. P., and N. S. Urquhart. 1993. A framework for assessing the sensitivity of the EMAP
design. Pages 4.1-4.37 in D. P. Larsen and S. J. Christie, editors. EMAP-Surface Waters
1991 Pilot Report. U.S. Environmental Protection Agency, Corvallis, OR.
Lawrence, G. B. 2002. Persistent episodic acidification of streams linked to acid rain effects on
soil. Atmospheric Environment 36:1589-1598.
Lawrence, G. B., M. B. David, G. M. Lovett, P. S. Murdoch, D. A. Burns, J. L. Stoddard, B. P.
Baldigo, J. H. Porter, and A. W. Thompson. 1999. Soil calcium status and the response
of stream chemistry to changing acidic deposition rates. Ecological Applications 9:1059-
1072.
Likens, G. E., F. H. Bormann, and N. M. Johnson. 1972. Acid rain. Environment 14:33-40.
Likens, G. E., C. T. Driscoll, and D. C. Buso. 1996. Long-term effects of acid rain: Response
and recovery of a forest ecosystem. Science 272:244-246.
Likens, G. E., E. S. Edgerton, and J. N. Galloway. 1983. The composition and deposition of
organic carbon in precipitation. Tellus 35:16-24.
74
-------
Lindberg, S. E., J. M. Coe, and W. A. Hoffman. 1984. Dissociation of weak acids during Gran
plot free acidity titrations. Tellus 36:186-191.
Linthurst, R. A., D. H. Landers, J. M. Eilers, D. F. Brakke, W. S. Overton, E. P. Meier, and R. E.
Crowe. 1986a. Characteristics of lakes in the eastern United States - Volume I:
Population Descriptions, and Physico-Chemical Relationships. U.S. Environmental
Protection Agency, Washington, D.C.
Linthurst, R. A., D. H. Landers, J. M. Eilers, D. F. Brakke, W. S. Overton, E. P. Meier, and R. E.
Crowe. 1986b. Characteristics of Lakes in the Eastern United States. Volume I.
Population Descriptions and Physico-Chemical Relationships. U.S. Environmental
Protection Agency, Washington, D.C.
Loftis, J. C., and C. H. Taylor. 1989. Detecting acid precipitation impacts on lake water quality.
Environmental Management 13:529-539.
Lynch, J. A., V. Bowersox, and J. Grimm. 2000. Acid rain reduced in the eastern U.S.
Environmental Science and Technology 34:940-949.
Mattson, M. D., P. J. Godfrey, M. Walk, P. A. Kerr, and O. T. Zajicek. 1997. Evidence of
recovery from acidification in streams. Water Air and Soil Pollution 96:211-232.
Miller, P. 1999. Emissions-related Acidic Deposition Trends in Maine. Final report for the
Maine Ecological Assessment Project, EPA Office of Air and Radiation, Washington,
DC.
Mitchell, M. J., C. T. Driscoll, J. S. Kahl, G E. Likens, P. S. Murdoch, and L. H. Pardo. 1996.
Climatic control of nitrate loss from forested watersheds in the northeast United States.
Environmental Science and Technology 30:2609-2612.
Murdoch, P. S., and J. L. Stoddard. 1992. The role of nitrate in the acidification of streams in
the Catskill Mountain of New York. Water Resources Research 28:2707-2720.
Murdoch, P. S., and J. L. Stoddard. 1993. Chemical characteristics and temporal trends in eight
streams of the Catskill Mountains, New York. Water Air and Soil Pollution 67:367-395.
National Acid Precipitation Assessment Program. 1998. NAPAP Biennial Report to Congress:
An Integrated Assessment. National Acid Precipitation Assessment Program, Silver
Spring, Maryland.
National Atmospheric Deposition Program/National Trends Network. 2002. NADP/NTN
Annual Data Summary. Precipitation Chemistry in the United States. Illinois State
Water Survey, Champaign, Illinois.
Newell, A. D., C. F. Powers, and S. J. Christie. 1987. Analysis of Data from Long-term
Monitoring of Lakes. U.S. Environmental Protection Agency, Corvallis, OR.
Nielsen, M. G., and J. S. Kahl. In press. Nutrient export to estuaries as a function of land-use
and fire history. Journal of the American Water Resources Association.
Norton, S. A., B. J. Cosby, I. J. Fernandez, J. S. Kahl, and M. R. Church, in press. Inter-annual
and seasonal variations in soil-air CO2 cause variable watershed alkalinity generation.
Hydrologic and Earth Science Systems.
Norton, S. A., J. S. Kahl, I. J. Fernandez, L. E. Rustad, T. A. Haines, S. C. Nodvin, J. P. Scofield,
T. C. Strickland, P. J. Wigington, and J. Lee. 1999. The Bear Brook Watershed in Maine
(BBWM). Environmental Monitoring and Assessment 55:7-51.
Norton, S. A., J. S. Kahl, I. J. Fernandez, L. E. Rustad, J. P. Scofield, and T. A. Haines. 1994.
Response of the West Bear Brook watershed, Maine, USA, to the addition of (NH4 )2 SO4
: 3-year results. Forest Ecology and Management 68:61-73.
75
-------
Oden, S. 1968. The acidification of air and precipitation and its consequences in the natural
environment. Swedish National Research Council, Stockholm.
Paerl, H. W. 2002. Connecting atmospheric nitrogen deposition to coastal eutrophication.
Environmental Science and Technology August 1, 2002:232A-326A.
Palmer, S. M., and C. T. Driscoll. 2002. Decline in mobilization of toxic aluminium. Nature
417:242-243.
Paulsen, S. G., D. P. Larsen, P. R. Kaufmann, T. R. Whittier, J. R. Baker, D. V. Peck, J. McGue,
R. M. Hughes, D. McMullen, D. Stevens, J. L. Stoddard, J. Lazorchak, W. Kinney, A. R.
Selle, and R. Hjort. 1991. EMAP-Surf ace Waters Monitoring and Research Strategy.
Fiscal Year 1991. U.S. Environmental Protection Agency, Corvallis, OR.
Peck, D. V. 1992. Environmental Monitoring and Assessment Program: Integrated Quality
Assurance Project Plan for the Surface Waters Resource Group. EPA/600/X-91/080,
U.S. Environmental Protection Agency.
Riley, J. P., and R. Chester. 1971. Introduction to Marine Chemistry. Academic Press, New
York.
Rosenqvist, I. T. 1978. Acid precipitation and other possible sources for acidification of rivers
and lakes. Science of the Total Environment 10:271-272.
Rustad, L. E., J. S. Kahl, S. A. Norton, and I. J. Fernandez. 1994. Underestimation of dry
deposition by throughfall in mixed northern hardwood forests. Journal of Hydrology
162:319-336.
SAS Institute Inc. 1988. SAS/STAT User's Guide, Release 6.03 Edition. SAS Institute, Inc.,
Cary, NC.
Schindler, D. W. 1988. The effects of acid rain on freshwater ecosystems. Science 239:149-
157.
Schindler, D. W., K. H. Mills, D. F. Malley, D. L. Findlay, J. A. Shearer, I. J. Davies, M. A.
Turner, G. A. Linsey, and D. R. Cruikshank. 1985. Long-term ecosystem stress: The
effects of years of experimental acidification on a small lake. Science 228:1395-1401.
Schofield, C. L. 1976. Acid precipitation: Effects on fish. Ambio 5:228-230.
Schofield, C. L. 1993. Habitat suitability for Brook Trout (Salvelinus fontinalis) reproduction in
Adirondack lakes. Water Resources Research 29:875-879.
Sen, P. K. 1968. On a class of aligned rank order tests in two-way layouts. Annals of
Mathematics and Statistics 39:1115-1124.
Shannon, J. D. 1999. Regional trends in wet deposition of sulfate in the United States and SO2
emissions from 1980 through 1995. Atmospheric Environment 33:807-816.
Sisterson, D. L., V. C. Bowersox, A. R. Olsen, T. P. Meyers, and R. L. Vong. 1990. Deposition
monitoring: methods and results. State-of-Science/Technology Report No. 6, National
Acid Precipitation Assessment Program, Washington, DC.
Skjelkvale, B. L., J. L. Stoddard, and T. Andersen. 2001. Trends in surface water acidification
in Europe and North America (1989-1998). Water Air and Soil Pollution 130:787-792.
Smith, R. A. 1872. Air and Rain: The Beginnings of a Chemical Climatology. Longmans,
Green, London.
Smol, J. P., B. F. Cumming, A. S. Dixit, and S. S. Dixit. 1998. Tracking recovery pattens in
acidified lakes: a paleolimnological perspective. Restoration Ecology 6:318-326.
Stoddard, J. L. 1990. Plan for Converting the NAPAP Aquatic Effects Long-Term Monitoring
(LTM) Project to the Temporally Integrated Monitoring of Ecosystems (TIME) Project.
Internatl Report U.S. Environmental Protection Agency, Corvallis, OR.
76
-------
Stoddard, J. L. 1994. Long-term changes in watershed retention of nitrogen: its causes and
aquatic consequences. Pages 223-284 in L. A. Baker, editor. Environmental Chemistry
of Lakes and Reservoirs. American Chemical Society, Washington, DC.
Stoddard, J. L., C. T. Driscoll, J. S. Kahl, and J. Kellogg. 1998a. A regional analysis of lake
acidification trends for the northeastern U.S., 1982-1994. Environmental Monitoring and
Assessment 51:399-413.
Stoddard, J. L., C. T. Driscoll, S. Kahl, and J. Kellogg. 1998b. Can site-specific trends be
extrapolated to a region? An acidification example for the Northeast. Ecological
Applications 8:288-299.
Stoddard, J. L., D. S. Jeffries, A. Lukewille, T. A. Clair, P. J. Dillon, C. T. Driscoll, M. Forsius,
M. Johannessen, J. S. Kahl, J. H. Kellogg, A. Kemp, J. Mannio, D. Monteith, P. S.
Murdoch, S. Patrick, A. Rebsdorf, B. L. Skjelkvale, M. Stainton, T. Traaen, H. van Dam,
K. E. Webster, J. Wieting, and A. Wilander. 1999. Regional trends in aquatic recovery
from acidification in North America and Europe. Nature 401:575-578.
Stoddard, J. L., and J. H. Kellogg. 1993. Trends and patterns in lake acidification in the state of
Vermont: evidence from the Long-Term Monitoring project. Water Air and Soil
Pollution 67:301-317.
Stoddard, J. L., A. D. Newell, N. S. Urquhart, and D. Kugler. 1996. The TIME project design:
II. Detection of regional acidification trends. Water Resources Research 32:2529-2538.
Sullivan, T. J. 1990. Historical Changes in Surface Water Acid-Base Chemistry in Response to
Acidic Deposition. NAPAP State of Science and Technology Report 11, National Acid
Precipitation Assessment Program.
Turner, R. S., R. B. Cook, H. Van Miegroet, D. W. Johnson, J. W. Elwood, O. P. Bricker, S. E.
Lindberg, and G. M. Hornberger. 1990. Watershed and Lake Processes Affecting
Surface Water Acid-Base Chemistry. State-of-Science/Technology Report No. 10,
National Acid Precipitation Assessment Program, Washington, DC.
U.S. Environmental Protection Agency. 2002. Clean Air Status and Trends Network
(CASTNet): 2000 Annual Report. U.S. Environmental Protection Agency, Research
Triangle Park, NC.
Urquhart, N. S., S. G. Paulsen, and D. P. Larsen. 1998. Monitoring for regional and policy-
relevant trends over time. Ecological Applications 8:246-257.
Webb, J. R., B. J. Cosby, J. N. Galloway, and G. M. Hornberger. 1989. Acidification of native
brook trout streams in Virginia. Water Resources Research 25:1367-1377.
Webb, J. R., F. A. Deviney, J. N. Galloway, C. A. Rinehart, P. A. Thompson, and S. Wilson.
1994. The Acid-Base Status of Native Brook Trout Streams in the Mountains of
Virginia. Dept. Environmental Sciences, University of Virginia.
Webster, K. E., P. L. Brezonik, and B. J. Holdhusen. 1993. Temporal trends in low alkalinity
lakes of the Upper Midwest (1983-1989). Water Air and Soil Pollution 67:397-414.
Webster, K. E., and P. O. Brezonik. 1995. Climate confounds detection of chemical trends
related to acid deposition in Upper Midwest lakes in the USA. Water Air and Soil
Pollution 85:1575-1580.
Webster, K. E., A. D. Newell, L. A. Baker, and P. L. Brezonik. 1990. Climatically induced
rapid acidification of a softwater seepage lake. Nature 347:374-376.
Wigington Jr., P. J. 1999. Episodic acidification: causes, occurrence and significance to aquatic
resources. Chapter 1, pg. 1-5 in J. R. Drohan, editor. The Effects of Acidic Deposition
77
-------
on Aquatic Ecosystems in Pennsylvania. 1998 PA Acidic deposition Conference.
Environmental Resources Research Institute, University Park, PA.
Wigington, P. I, J. P. Baker, D. R. DeWalle, W. A. Kretser, P. S. Murdoch, H. A. Simonin, J.
Van Sickle, M. K. McDowell, D. V. Peck, and W. R. Barchet. 1996. Episodic
acidification of small streams in the northeastern United States: Episodic Response
Project. Ecological Applications 6:374-388.
Williams, M. W., J. S. Baron, N. Caine, R. Sommerfield, and R. Sanford. 1996. Nitrogen
saturation in the Rocky Mountains. Environmental Science and Technology 30:640-646.
Wright, R. F. 1983. Predicting Acidification of North American Lakes. Acid Rain Research
Report 4/1983.
Wright, R. F., E. Lotse, and A. Semb. 1993. RAIN Project: Results after 8 years of
experimentally reduced acid deposition to a whole catchment. Canadian Journal of
Fisheries and Aquatic Science 50:258-268.
Young, T. C., and J. L. Stoddard. 1996. The TIME project design: I. Classification of
Northeast lakes using a combination of geographic, hydrogeochemical, and multivariate
techniques. Water Resources Research 32:2517-2528.
78
------- |