EPA/600/R-20/278| September 2020| www.epa.gov/isa
United States
Environmental Protection
Agency
&EPA
Integrated Science Assessment
for Oxides of Nitrogen, Oxides
of Sulfur and Particulate Matter-
Ecological Criteria
Office of Research and Development
Center for Public Health & Environmental Assessment, Research Triangle Park, NC
-------
United States
Environmental Protection
#m Agency
EPA/600/R-20/278
September 2020
www.epa. gov/ncea/isa
Integrated Science Assessment
for Oxides of Nitrogen, Oxides of
Sulfur, and Particulate Matter—
Ecological Criteria
(Final)
Center for Public Health and Environmental Assessment
Office of Research and Development
U.S. Enviromnental Protection Agency
Research Triangle Park, NC
i
-------
DISCLAIMER
This document is an external review draft, for review purposes only. This information is
distributed solely for predissemination peer review under applicable information quality
guidelines. It has not been formally disseminated by EPA. It does not represent and
should not be construed to represent any Agency determination or policy. Mention of
trade names or commercial products does not constitute endorsement or recommendation
for use.
ii
-------
CONTENTS
AUTHORS, CONTRIBUTORS, AND REVIEWERS xxxii
CLEAN AIR SCIENTIFIC ADVISORY COMMITTEE xxxvii
Chair of the Charter Clean Air Scientific Advisory Committee xxxvii
Charter Clean Air Scientific Advisory Committee Members xxxvii
Chair, Oxides of Nitrogen, Oxides of Sulfur and Particulate Matter—Ecological Criteria Review
Panel xxxvii
Oxides of Nitrogen, Oxides of Sulfur and Particulate Matter—Ecological Criteria Review Panel
Members xxxvii
ACRONYMS AND ABBREVIATIONS xxxix
PREFACE xlix
Legislative Requirements for the Review of the National Ambient Air Quality Standards xlix
Overview and History of the Reviews of the Secondary National Ambient Air Quality Standards for
Nitrogen Dioxide, Sulfur Dioxide, and Particulate Matter I
Nitrogen Dioxide Secondary National Ambient Air Quality Standards I
Sulfur Dioxide Secondary National Ambient Air Quality Standards li
Particulate Matter Secondary National Ambient Air Quality Standards Iv
Combined Review of the Oxides of Nitrogen and Oxides of Sulfur National Ambient Air
Quality Standards Iviii
EXECUTIVE SUMMARY 1
ES.1 Purpose and Scope of the Integrated Science Assessment 1
Figure ES-1 Roadmap of the Integrated Science Assessment (ISA) linking
atmospheric concentrations and deposition, soil and aquatic
biogeochemistry, and biological effects. 4
ES.2 Emissions, Ambient Air Concentrations, and Deposition 4
Figure ES-2 Wet plus dry deposition of (A) oxidized nitrogen, (B) reduced
nitrogen, (C) total nitrogen, and (D) total sulfur over the 3-year
periods 2000-2002 and 2016-2018. 6
ES.3 Ecological Effects 7
Table ES-1 Causal determinations for relationships between criteria
pollutants and ecological effects from the 2008 NOx/SOx
Integrated Science Assessment (ISA) or the 2009 ISA for
Particulate Matter (PM), for other effects of PM, and the current
draft ISA. 8
Figure ES-3 Causal relationships between the criteria pollutants and
ecological effects. 12
ES.4 Direct Phytotoxic Effects of Gas-Phase Oxides of Nitrogen (NOy) and Oxides of Sulfur
(SOx) 13
ES.5 Ecological Effects of Nitrogen and Sulfur Deposition 13
ES.5.1 Acidification of Terrestrial and Freshwater Ecosystems 13
ES.5.2 Nitrogen Enrichment/Eutrophication of Terrestrial, Wetland, and Aquatic
Ecosystems 15
Figure ES-4 Summary of critical loads for nitrogen in the U.S. for shrubs and
herbaceous plants (yellow), trees (blue), lichens (green), and
mycorrhizae (gray). Values expressed by major U.S.
ecoregions. 17
ES.5.3 Sulfur (S) Enrichment of Wetland and Freshwater Ecosystems 19
iii
-------
ES.5.4 Ecological Effects of Particulate Matter Other Than Those Associated with
Nitrogen and Sulfur Deposition 20
ES.6 Ecosystem Services 20
ES.7 Integrating across Ecosystems 21
Figure ES-5 Causal relationships between the criteria pollutants and
ecological effects organized under ecosystem type. 22
INTEGRATED SYNTHESIS 1
15.1 Introduction to This Integrated Science Assessment (ISA) 1
IS.1.1 Purpose 1
IS. 1.2 Process and Development 4
Figure IS-1 Workflow for collecting relevant literature for the 2017 Integrated
Science Assessment for Oxides of Nitrogen, Oxides of Sulfur,
and Particulate Matter—Ecological Criteria. 5
15.2 Connections, Concepts, and Changes 7
15.2.1 Connections 7
Figure IS-2 Overview of atmospheric chemistry, deposition, and ecological
effects of emissions of oxides of nitrogen, oxides of sulfur, and
reduced nitrogen. 8
15.2.2 Concepts 8
Figure IS-3 An example of the matrix of information considered in defining
and calculating critical loads (see discussion in text). Note that
multiple alternative biological indicators, critical biological
responses, chemical indicators, and critical chemical limits could
be used. 13
15.2.3 Changes: New Evidence and Causal Determinations 20
Table IS-1 Causal determinations for relationships between criteria
pollutants and ecological effects from the 2008 NOx/SOx
Integrated Science Assessment (ISA) or the 2009 Particulate
Matter (PM) ISA, for other effects of PM, and the current draft
ISA. 22
Figure IS-4 Causal relationships between the criteria pollutants and
ecological effects. 25
15.3 Emissions and Atmospheric Chemistry 27
15.3.1 Sources and Atmospheric Transformations 28
15.3.2 Measurement and Modeling Techniques 29
15.3.3 Spatial and Temporal Variability in Deposition 29
Figure IS-5 Wet plus dry deposition of total nitrogen over 3-year periods.
Top: 2000-2002; Bottom: 2016-2018. 31
Figure IS-6 Wet plus dry deposition of oxidized nitrogen over 3-year
periods. Top: 2000-2002; Bottom: 2016-2018. 32
Figure IS-7 Wet plus dry deposition of reduced (inorganic) nitrogen over 3-
year periods. Top: 2000-2002; Bottom: 2016-2018. 33
Figure IS-8 Wet plus dry deposition of total sulfur over 3-year periods.
Top: 2000-2002; Bottom: 2016-2018. 35
Figure IS-9 Total acidifying deposition of total oxidized nitrogen, reduced
nitrogen, and oxidized sulfur expressed as H+ equivalents per
hectare per year over the contiguous U.S. 2016-2018. 36
15.4 Gas-Phase Direct Phytotoxic Effects 36
15.4.1 Sulfur Dioxide 37
15.4.2 Nitrogen Oxide, Nitrogen Dioxide, and Peroxyacetyl Nitrate 38
15.4.3 Nitric Acid 38
15.5 Terrestrial Ecosystem Nitrogen Enrichment and Acidification 39
IS.5.1 Soil Biogeochemistry 40
iv
-------
Table IS-2 Summary of key soil geochemical processes and indicators
associated with eutrophication and acidification. 41
15.5.2 Biological Effects of Terrestrial Nitrogen Enrichment 46
Figure IS-10 Summary of critical loads for nitrogen in the U.S. for shrubs and
herbaceous plants (yellow), trees (blue), lichens (green), and
mycorrhizae (gray). 56
15.5.3 Biological Effects of Acidification 57
Figure IS-11 Forest ecosystem critical loads for soil acidity related to base
cation soil indicators. 60
15.6 Freshwater Ecosystem Nitrogen Enrichment and Acidification 61
15.6.1 Freshwater Biogeochemistry 62
Table IS-3 Summary of key aquatic geochemical processes and indicators
associated with eutrophication and acidification. 64
15.6.2 Biological Effects of Freshwater Nitrogen Enrichment 69
15.6.3 Biological Effects of Freshwater Acidification 73
Figure IS-12 Surface water critical loads for acidity in the U.S. 10th percentile
aggregation for 36-km2 grids with sulfur (S) and nitrogen (N). 77
15.7 Estuarine and Near-Coastal Ecosystem Nitrogen Enrichment 79
15.7.1 Estuary and Near-Coastal Biogeochemistry 79
15.7.2 Biological Effects of Nitrogen Enrichment 82
15.7.3 National-Scale Sensitivity and Critical Loads 86
15.8 Wetland Ecosystem Nitrogen Enrichment and Acidification 88
15.8.1 Wetland Biogeochemistry 88
15.8.2 Biological Effects of Wetland Nitrogen Enrichment/Eutrophication 89
15.9 Freshwater and Wetland Ecosystem Sulfur Enrichment 93
15.9.1 Biogeochemistry 93
15.9.2 Biological Effects of Sulfur Enrichment 94
15.9.3 National-Scale Sensitivity and Critical Loads 97
IS. 10 Ecological Effects of Particulate Matter Other Than Nitrogen (N) and Sulfur (S) Deposition 98
IS. 11 Recovery of Ecosystems from Nitrogen (N) and Sulfur (S) Deposition in the U.S. 100
IS. 11.1 Overarching Concepts of Ecological Recovery from Acidification 100
15.11.2 Acidification Recovery in the U.S. 101
15.11.3 Nitrogen (N) Driven Nutrient Enrichment Recovery in the U.S. 103
IS. 12 Climate Modification of Ecosystem Response to Nitrogen (N) and Sulfur (S) Deposition 103
IS. 13 Ecosystem Services 104
IS. 14 Key Scientific Uncertainties 106
15.14.1 Atmospheric Science 107
15.14.2 Ecological Effects 110
IS. 14.3 Aquatic Acidification Index 116
APPENDIX 1 QUALITY ASSURANCE AND INTRODUCTION TO APPENDICES 1-1
APPENDIX 2 SOURCE TO DEPOSITION 2-1
2.1 Introduction 2-1
2.2 Sources of Nitrogen and Sulfur Compounds and Particulate Matter to the Atmosphere 2-5
2.2.1 National Emissions by Source 2-6
Table 2-1 Emissions of NOx (nitric oxide + nitrogen dioxide), sulfur
dioxide, and ammonia by source category for 2017 (Teragrams
N, S/yr). 2-7
2.2.2 Methods of Estimating Emissions 2-8
Figure 2-1 Modeling system used to compute 2014 Fertilizer Application
Emissions. 2-11
Figure 2-2 Process to produce specific location and practice specific daily
emission factors for livestock waste. 2-13
v
-------
2.2.3 Evaluation and Uncertainty 2-13
2.3 Atmospheric Chemistry of Nitrogen and Sulfur Species and Particulate Matter (PM) 2-16
2.3.1 Nitrogen Oxides 2-16
Figure 2-3 Schematic diagram showing pathways for reactive nitrogen
species in ambient air. 2-18
Table 2-2 Henry's law coefficients for selected reactive nitrogen species at
25°C in water. 2-19
2.3.2 Sulfur Oxides 2-21
Figure 2-4 Rate of conversion of sulfur (IV) to sulfur (VI) by different
oxidation paths as a function of pH. 2-22
2.3.3 Acid Neutralization by Ammonia 2-23
2.3.4 Organic Nitrogen and Sulfur 2-25
2.3.5 Organic Acids 2-26
2.3.6 Particulate Matter (PM) 2-27
Figure 2-5 Contributions of organic carbon (OC), elemental carbon (EC),
sulfate, nitrate, sea salt, and crustal components to PM2.5 at
selected sites (A) 2003-2005 (B) 2013-2015. 2-28
2.4 Concentration and Deposition Measurements 2-29
2.4.1 Monitoring Networks 2-29
Table 2-3 Summary of monitoring networks used by Schwede and Lear
(2014a). 2-31
2.4.2 NO2, NOx, and NOy 2-33
2.4.3 Ammonia 2-35
2.4.4 Sulfur Dioxide 2-38
Table 2-4 Sources of uncertainty for individual Ozone Monitoring
Instrument measurements in the study of Nowlan et al. (2014). 2-39
2.4.5 Filter-Based Concentration Measurements 2-39
Figure 2-6 Clean Air Status and Trends Network filter pack. 2-40
Figure 2-7 Comparison between weekly average measurements of sulfur
dioxide using the Clean Air Status and Trends Network filter
pack and the trace ultraviolet pulsed fluorescence monitor in
2014. 2-42
2.4.6 Deposition Measurements 2-43
2.5 Modeling Chemistry, Transport, and Deposition 2-46
2.5.1 Advances in Chemistry-Transport Model (CTM) Modeling 2-46
2.5.2 Modeling Deposition 2-48
Figure 2-8 Schematic diagram showing mechanisms for transferring
pollutants from the atmosphere to the surface. 2-49
Table 2-5 Average dry deposition velocities (cm/s) for several gases over
land surfaces. 2-50
Table 2-6 Deposition velocity (cm/s) for sulfur dioxide averaged over
different land use types for summer and winter. 2-50
Figure 2-9 Modeled and measured deposition velocities over grass (left
figure) and coniferous forest canopies (right figure) for particles
of density 1 gm/cm3 depositing under similar friction velocity (u*)
(35
-------
Table 2-7 Reported comparisons of chemical transport models and
observations of nitrogen and sulfur wet deposition 2-62
2.6 Geographic Distribution of Concentration and Deposition 2-62
2.6.1 pH and H+ Equivalents 2-65
Figure 2-12 (Left) pH of rainwater, 1989-1991; (Right) pH of rainwater,
2016-2018. 2-66
Figure 2-13 Difference in wet deposition of nitrate, ammonium, and sulfate
expressed as hydrogen ion equivalents (eq/ha/yr) over the
contiguous U.S. between 1989 to 1991 and 2016 to 2018. 2-67
2.6.2 Total Nitrogen 2-68
Figure 2-14 Total (wet + dry) deposition of nitrogen (kg N/ha/yr) over the
contiguous U.S. 2016-2018. 2-69
Figure 2-15 (A) Percentage of total nitrogen deposition as reduced inorganic
nitrogen over the contiguous U.S. 2016-2018. (B) Percentage
of total nitrogen deposition as oxidized nitrogen over the
contiguous U.S. 2016-2018. 2-70
Figure 2-16 Trends in U.S. total deposition flux of total nitrogen, oxidized
nitrogen, reduced nitrogen, and major nitrogen species
2000-2017. 2-71
Figure 2-17 Three-year average percentage of total nitrogen deposition by
species (i.e., those species that are not measured in the
networks) simulated by the Community Multiscale Air Quality
modeling system for 2016-2018. 2-72
Figure 2-18 Wet deposition of ammonium + nitrate (kg N/ha/yr) over the
contiguous U.S. in two, 3-year periods, 2016 to 2018 and 1989
to 1991. Also shown are active National Trends Network sites in
either period. 2-73
2.6.3 Oxidized Nitrogen 2-74
Figure 2-19 Geographic distribution of annual U.S. NOx emissions in 2017. 2-75
Figure 2-20 Distribution of annual average total oxidized nitrogen species
concentrations for 2011 simulated by Community Multiscale Air
Quality modeling system. 2-76
Figure 2-21 Seasonal average surface nitrogen dioxide mixing ratios in parts
per billion for winter (upper panel) and summer (lower panel)
derived by the Ozone Monitoring Instrument/GEOS-Chem
model for 2009-2011. The Ozone Monitoring Instrument has an
overpass at approximately 1:30 p.m. local standard time. 2-77
Figure 2-22 Three-year average (2016-2018) surface concentrations of
nitric acid based on monitoring data obtained at Clean Air
Status and Trends Network sites (black dots). 2-79
Figure 2-23 Three-year average (2016-2018) surface concentrations of
particulate nitrate based on monitoring data obtained at Clean
Air Status and Trends Network sites (black dots). 2-80
Figure 2-24 Trends in particulate nitrate concentration 1990-2017: (A)
average eastern U.S. concentration based on 34 sites; (B)
average western concentration based 16 sites. 2-81
Figure 2-25 Total oxidized nitrogen deposition over the contiguous U.S.
2016-2018. 2-82
Figure 2-26 (Left) nitrate wet deposition, 1989-1991; (Right) nitrate wet
deposition, 2016-2018. 2-83
Figure 2-27 Difference in wet deposition of nitrate (kg N/ha/yr) over the
contiguous U.S. between 1989 to 1991 and 2016 to 2018. The
range of positive values is smaller than that for negative values. 2-84
Figure 2-28 Trends in oxidized nitrogen emissions and deposition
2000-2017: (A) total national emissions; (B) national average
total deposition flux. 2-85
2.6.4 Reduced Nitrogen 2-85
vii
-------
Figure 2-29 Geographic distribution of annual U.S. ammonia (NH3)
emissions in 2017. 2-86
Figure 2-30 Average (2017) surface concentration of ammonia obtained by
the Ambient Ammonia Monitoring Network at select Clean Air
Status and Trends Network sites. Concentrations of ammonia
(|jg/m3) can be converted to mixing ratios (parts per billion) to
rough approximation at normal temperature and pressure by
multiplying by 1.4. 2-87
Figure 2-31 Three-year average (2016-2018) surface concentrations of
particulate ammonium (|jg/m3) based on monitoring data
obtained at Clean Air Status and Trends Network sites (black
dots). 2-88
Figure 2-32 Total reduced inorganic nitrogen deposition over the contiguous
U.S. 2016-2018. 2-89
Figure 2-33 (Left) ammonium wet deposition, 1989-1991; (Right)
ammonium wet deposition, 2016-2018. 2-90
Figure 2-34 Difference in wet deposition of ammonium (kg N/ha/yr) over the
contiguous U.S. between 1989 to 1991 and 2016 to 2018. 2-91
Figure 2-35 Trends in reduced nitrogen emissions and deposition
2000-2017: (A) total national NH3 emissions; (B) national
average reduced nitrogen deposition flux. 2-92
2.6.5 Sulfur Oxides 2-92
Figure 2-36 Geographic distribution of annual U.S. sulfur dioxide (SO2)
emissions by county from the 2017 National Emissions
Inventory. 2-93
Figure 2-37 Trends in total national sulfur dioxide emissions. 2-93
Figure 2-38 Three-year average (2016-2018) surface concentrations of
sulfur dioxide obtained by fusion of monitoring data obtained at
Clean Air Status and Trends Network sites (black dots) and
Community Multiscale Air Quality model system results.
Concentrations (|jg/m3) can be converted to mixing ratios (parts
per billion) at normal temperature and pressure) to rough
approximation by multiplying by 0.37. 2-95
Figure 2-39 Three-year average (2016-2018) surface concentrations of
particulate sulfate based on monitoring data obtained at Clean
Air Status and Trends Network sites (black dots). 2-96
Figure 2-40 Trends in oxides of sulfur oxides concentrations 1990-2017:
(A) average eastern U.S. SO2 concentration based on 34 sites;
(B) average western U.S. SO2 concentration based on 16 sites;
(C) average eastern U.S. sulfate concentration based on 34
sites; (D) average western U.S. sulfate concentration based on
16 sites. 2-97
Figure 2-41 Total deposition of sulfur (kg S/ha/yr) over the contiguous U.S.
2016-2018. 2-98
Figure 2-42 Percentage of deposition of total sulfur as dry deposition over
the contiguous U.S. 2016-2018. 2-99
Figure 2-43 (Left) sulfate wet deposition, 1989-1991; (Right) sulfate wet
deposition, 2016-2018. 2-100
Figure 2-44 Difference in wet deposition of sulfate (kg S/ha/yr) over the
contiguous U.S. between 1989 to 1991 and 2016 to 2018. The
range of positive values is much smaller than for negative
values. 2-100
Figure 2-45 Trends in average sulfur deposition flux for 34 monitoring sites
in the eastern U.S. 1989-2017. 2-101
2.6.6 Particulate Matter (PM) 2-101
Figure 2-46 Three-year average concentrations of particulate matter smaller
than 2.5 pm diameter (PM2.5) 2013-2015. 2-102
viii
-------
Figure 2-47 National monthly average concentrations of particulate matter
smaller than 2.5 |jm diameter (PM2.5; top) and sulfate in PM2.5
(bottom) from 2000-2016 (concentrations in |jg/m3). 2-103
Figure 2-48 98th percentile concentrations for PM10-2.5 between 2013-2015. 2-104
2.6.7 Distributions of Dry Deposition of Nitrogen Dioxide and Sulfur Dioxide Derived
Using Satellite-Based Measurements and Chemistry Transport Models 2-104
Figure 2-49 Top panel: modeled deposition velocities for nitrogen dioxide
and sulfur dioxide for 2005 to 2007; middle panel satellite-model
estimates of annual mean dry deposition fluxes of nitrogen
dioxide and sulfur dioxide; bottom panel: uncertainties in
estimates. 2-106
2.6.8 Background Concentrations and Deposition 2-107
Figure 2-50 Contributions to oxidized and reduced nitrogen deposition from
U.S.: anthropogenic (top), foreign anthropogenic (middle), and
natural sources (bottom). 2-108
2.7 Supplemental Material on Changes in Deposition since 2000 2-112
Figure 2-51 Wet plus dry deposition of total nitrogen over 3-year periods.
Top: 2000-2002; Bottom: 2016-2018. 2-114
Figure 2-52 Wet deposition of total nitrogen over 3-year periods.
Top: 2000-2002; Bottom: 2016-2018. 2-115
Figure 2-53 Dry deposition of total nitrogen over 3-year periods.
Top: 2000-2002; Bottom: 2016-2018. 2-116
Figure 2-54 Percent of total nitrogen as dry deposition over 3-year periods.
Top: 2000-2002; Bottom: 2016-2018. 2-117
Figure 2-55 Wet plus dry deposition of oxidized nitrogen over 3-year
periods. Top: 2000-2002; Bottom: 2016-2018. 2-118
Figure 2-56 Percent of total nitrogen as oxidized nitrogen over 3-year
periods. Top: 2000-2002; Bottom: 2016-2018. 2-119
Figure 2-57 Dry deposition of oxidized nitrogen over 3-year periods.
Top: 2000-2002; Bottom: 2016-2018. 2-120
Figure 2-58 Percent of total nitrogen dry deposited as oxidized nitrogen over
3-year periods. Top: 2000-2002; Bottom: 2016-2018. 2-121
Figure 2-59 Combined dry deposition of nitric acid and particulate nitrate
over 3-year periods. Top: 2000-2002; Bottom: 2016-2018. 2-122
Figure 2-60 Dry deposition of nitric acid over 3-year periods. Top:
2000-2002; Bottom: 2016-2018. 2-123
Figure 2-61 Dry deposition of particulate nitrate over 3-year periods.
Top: 2000-2002; Bottom: 2016-2018. 2-124
Figure 2-62 Dry deposition of modeled (unmeasured) nitrogen species over
3-year periods. Top: 2000-2002; Bottom: 2016-2018. 2-125
Figure 2-63 Percent of total nitrogen as modeled (unmeasured) species over
3-year periods. Top: 2000-2002; Bottom: 2016-2018. 2-126
Figure 2-64 Wet plus dry deposition of reduced (inorganic) nitrogen over 3-
year periods. Top: 2000-2002; Bottom: 2016-2018. 2-127
Figure 2-65 Percent of total nitrogen deposition by reduced (inorganic)
nitrogen over 3-year periods. Top: 2000-2002;
Bottom: 2016-2018. 2-128
Figure 2-66 Dry deposition of ammonia over 3-year periods. Top:
2000-2002; Bottom: 2016-2018. 2-129
Figure 2-67 Dry deposition of particulate ammonium over 3-year periods.
Top: 2000-2002; Bottom: 2016-2018. 2-130
Figure 2-68 Dry deposition of reduced (inorganic) nitrogen over 3-year
periods. Top: 2000-2002; Bottom: 2016-2018. 2-131
Figure 2-69 Percent of total nitrogen deposition by dry reduced (inorganic)
nitrogen over 3-year periods. Top: 2000-2002;
Bottom: 2016-2018. 2-132
ix
-------
Figure 2-70
Wet plus dry deposition of total sulfur over 3-year periods.
Top: 2000-2002; Bottom: 2016-2018.
2-133
Figure 2-71
Wet deposition of total sulfur over 3-year periods.
Top: 2000-2002; Bottom: 2016-2018.
2-134
Figure 2-72
Dry deposition of total sulfur over 3-year periods.
Top: 2000-2002; Bottom: 2016-2018.
2-135
Figure 2-73
Percent of total sulfur deposition by dry deposition over 3-year
periods. Top: 2000-2002; Bottom: 2016-2018.
2-136
Figure 2-74
Dry deposition of sulfur dioxide over 3-year periods.
Top: 2000-2002; Bottom: 2016-2018.
2-137
Figure 2-75
Dry deposition of particulate sulfate over 3-year periods.
Top: 2000-2002; Bottom: 2016-2018.
2-138
APPENDIX 3 DIRECT PHYTOTOXIC EFFECTS OF GASEOUS OXIDIZED NITROGEN AND
SULFUR ON VEGETATION 3-1
3.1 Introduction 3-1
Figure 3-1 The microarchitecture of a dicot leaf. While details among
species vary, the general overview remains the same. Light that
drives photosynthesis generally falls upon the upper (adaxial)
leaf surface. Carbon dioxide, oxides of sulfur, oxides of nitrogen,
and ozone gases generally enter by diffusion through the guard
cells (or stomata) on the lower (abaxial) leaf surface, while
water vapor exits through the stomata (transpiration). 3-3
3.2 Direct Phytotoxic Effects of Sulfur Dioxide on Vegetation 3-3
Figure 3-2 Map of maximum 3-hour daily max average sulfur dioxide
concentration reported at Air Quality System monitoring sites for
2016. 3-5
3.3 Direct Phytotoxic Effects of Nitric Oxide, Nitrogen Dioxide, and Peroxyacetyl Nitrate 3-7
Figure 3-3 Map of U.S. annual average nitrogen dioxide concentrations for
2013. 3-8
3.4 Direct Phytotoxic Effects of Nitric Acid 3-13
3.5 Direct Phytotoxic Effects of Reduced Nitrogen Gases 3-15
3.6 Summary 3-16
3.6.1 Sulfur Dioxide 3-16
3.6.2 Nitrogen Oxide, Nitrogen Dioxide, and Peroxyacetyl Nitrate 3-17
3.6.3 Nitric Acid 3-17
APPENDIX 4 SOIL BIOGEOCHEMISTRY 4-1
4.1 Introduction 4-1
4.2 Nitrogen and Sulfur Sources to Soil 4-2
Figure 4-1 Dominant sources of nitrogen across the U.S. at 8-digit
hydrologic unit codes. 4-3
Figure 4-2 Percentage of nitrogen input from nitrogen deposition at 8-digit
hydrologic unit codes. 4-3
4.3 Soil Pools and Processes 4-4
Table 4-1 Summary of key soil geochemical processes and indicators
associated with eutrophication and acidification. 4-6
4.3.1 Nitrogen Pathways and Pools 4-6
Table 4-2 Pathways and pools. 4-8
4.3.2 Nitrogen Accumulation, Saturation, and Leaching 4-11
Figure 4-3 A conceptual framework for the responses of the ecosystem
nitrogen (N) cycle to nitrogen (N) addition. 4-12
Figure 4-4 A hypothetical model to account for the effects of nitrogen
supply on plant nitrogen uptake and belowground carbon
x
-------
allocation, microbial production of inorganic nitrogen, and
nitrogen leaching. 4-15
Table 4-3 Nitrogen accumulation, saturation, and leaching. 4-16
4.3.3 Sulfate Accumulation, Adsorption, and Leaching 4-22
Table 4-4 Sulfate adsorption, accumulation, and leaching. 4-24
4.3.4 Base Cation Leaching and Exchange 4-27
Table 4-5 Base cation leaching and exchange. 4-30
4.3.5 Aluminum Mobilization 4-35
Table 4-6 Aluminum mobilization. 4-37
4.3.6 Nitrification and Denitrification 4-39
Table 4-7 Nitrification and denitrification. 4-41
Figure 4-5 Effects of nitrogen addition on biogenic nitrous oxide emission. 4-46
Figure 4-6 The weighted response ratio for the responses to nitrogen
addition for fluxes and pools related to the ecosystem nitrogen
cycle in agricultural (open bars) and nonagricultural (closed
bars) ecosystems. 4-47
4.3.7 Decomposition 4-47
Table 4-8 Decomposition. 4-49
4.3.8 Nitrogen Mineralization 4-56
Table 4-9 Nitrogen mineralization. 4-58
Figure 4-7 Effects of nitrogen inputs on soil carbon, carbon:nitrogen ratio,
and minimum nitrogen in forest floors (panels A-C) and mineral
soils (panels D-F). 4-60
4.3.9 Dissolved Organic Carbon Leaching 4-61
Table 4-10 Terrestrial dissolved organic carbon (DOC) leaching. 4-62
Figure 4-8 Conceptual diagram of positive (solid arrows) and negative
(dashed arrows) fluxes in nitrogen pools (squares) and carbon
pools (ovals) and the biological processes (no border) that are
affected by experimental nitrogen deposition. 4-67
4.3.10 Belowground Carbon Pools 4-67
Figure 4-9 Estimation of the changes in carbon budget of terrestrial
ecosystem under nitrogen addition. 4-69
Table 4-11 Belowground carbon pools. 4-70
4.3.11 New Biogeochemical Indicators 4-72
Table 4-12 New biogeochemistry indicators. 4-73
4.3.12 Differential Effects of Reduced and Oxidized Nitrogen 4-74
Table 4-13 The effects of different forms of inorganic nitrogen on
biogeochemical processes and indicators according to meta-
analyses. See Table 6-1 for the effects of different forms of
inorganic nitrogen on biological endpoints. 4-75
4.4 Soil Monitoring and Databases 4-76
Table 4-14 Biogeochemistry monitoring and databases. 4-77
4.5 Models 4-80
4.5.1 Updates to Key Previously Identified Models 4-81
Table 4-15 Photosynthesis and Evapotranspiration—Biogeochemical
(PnET-BGC) and DayCent. 4-87
4.5.2 New Models (Published since 2008) 4-89
4.5.3 Comparative Analyses 4-91
Table 4-16 Overview of properties of four dynamic soil chemistry models as
characterized in Bonten et al. (2015). 4-93
Table 4-17 Model comparison. 4-95
4.6 National-Scale Sensitivity 4-98
4.6.1 Acidification Recovery 4-98
Table 4-18 Recovery. 4-100
4.6.2 Critical Loads 4-103
xi
-------
Figure 4-10 Forest ecosystem critical loads for soil acidity related to base
cation soil indicators. 4-104
Figure 4-11 Map of critical loads for nitrate leaching by ecoregion in the U.S. 4-106
4.7 Modification of Terrestrial Soil Response to Nitrogen (N) 4-106
4.7.1 Disturbance and Stand Age Effects on Nitrogen Retention 4-107
4.7.2 Nitrogen and Phosphorus Interactions 4-107
4.7.3 Climate Modification of Acidification Effects on Soil 4-108
4.7.4 Climate Modification of Nitrogen-Driven Eutrophication in Soil 4-109
Figure 4-12 The potential mechanisms that regulate the responses of
carbon dioxide (CO2), methane (CH4), and nitrous oxide (N2O)
production and consumption to elevated nitrogen (N). 4-112
4.8 Summary 4-113
4.8.1 Sources 4-113
4.8.2 Soil Processes and Indicators 4-113
Table 4-19 Summary of key soil geochemical processes and indicators
associated with eutrophication and acidification. 4-116
4.8.3 Monitoring 4-119
4.8.4 Models 4-120
4.8.5 National-Scale Sensitivity 4-122
4.8.6 Climate Modification of Soil Response to Nitrogen Addition 4-123
APPENDIX 5 BIOLOGICAL EFFECTS OF TERRESTRIAL ACIDIFICATION 5-1
5.1 Introduction 5-1
Figure 5-1 Diagram based on Fenn et al. (2006) showing indicators of
forest physiological function, growth, and structure that are
linked to biogeochemical cycles through processes that control
rates of calcium supply. Calcium affects plant physiological
processes that influence growth rates and the capacity of plants
to resist environmental stresses, such as extremes of
temperature, drought, insects, and diseases. Therefore,
acidifying deposition, which can deplete soil calcium or interfere
with calcium uptake through mobilization of soil aluminum, can
affect forest health. 5-3
Table 5-1 Relationships between soil chemistry indicators and biological
endpoints that have been evaluated in the literature since the
2008 Integrated Science Assessment. 5-4
5.2 Effects on Terrestrial Organisms and Ecosystems 5-6
5.2.1 Trees and Forests 5-6
Table 5-2 Summary of calcium addition studies in North America. 5-9
Figure 5-2 Relationship between the proportion of seedlings that were
sugar maple and soil base saturation in the upper B-horizon. 5-13
5.2.2 Forest Understory and Grassland Species 5-20
5.2.3 Lichens 5-21
5.2.4 Soil Biota 5-22
5.2.5 Fauna 5-24
5.3 Characteristics, Distribution, and Extent of Sensitive Ecosystems 5-26
5.4 Application of Terrestrial Acidification Models 5-27
5.5 Levels of Deposition at Which Effects Are Manifested 5-27
5.5.1 Impacts of Elevated Nitrogen and Sulfur Deposition 5-28
Table 5-3 Impacts of acidifying nitrogen and sulfur deposition. 5-31
5.5.2 Impacts of Ambient Deposition 5-36
Table 5-4 Results of Spearman's rank correlation analysis comparing
growth versus critical load exceedance by species for the
forested plots in the northeastern U.S. [from Duarte et al.
(2013)]. Modeled sulfur and nitrogen deposition on the plots
xii
-------
ranged from 242 to 1,154 eq/ha/yr and 256 to 920 eq/ha/yr,
respectively. Correlations shown here are significant at a = 0.05. 5-37
Table 5-5 Results of Spearman's rank correlation analyses comparing tree
vigor, crown density and dieback, and canopy transparency
versus critical load exceedance by species for the forested plots
in the northeastern U.S. [from Duarte et al. (2013)]. Modeled
sulfur and nitrogen deposition on the plots ranged from 242 to
1,154 eq/ha/yr and 256 to 920 eq/ha/yr, respectively.
Correlations shown here are significant at a = 0.05. 5-38
5.5.3 Critical Loads and Exceedances 5-40
5.6 Climate Modification of Ecosystem Response 5-46
5.7 Summary 5-48
5.7.1 Physiology and Growth 5-48
5.7.2 Biodiversity 5-50
5.7.3 National-Scale Sensitivity and Critical Loads 5-50
Table 5-6 Mode of action for acidifying nitrogen and sulfur deposition. 5-52
APPENDIX 6 TERRESTRIAL ECOSYSTEMS: NITROGEN ENRICHMENT EFFECTS ON
ECOLOGICAL PROCESSES 6-1
6.1 Introduction 6-1
6.2 Linking Nitrogen Deposition to Changes in Physiology, Growth, and Productivity in
Terrestrial Ecosystems 6-2
6.2.1 Introduction 6-2
6.2.2 Mechanisms Operating across Terrestrial Ecosystems 6-3
Figure 6-1 Effects of nitrogen additions on plant growth and net primary
productivity. 6-8
Figure 6-2 Effects of added nitrogen on ecosystem carbon pools and
fluxes. 6-9
Table 6-1 The effects of different forms of inorganic nitrogen on biological
endpoints according to meta-analyses. See Table 4-13 for the
effects of different forms of inorganic nitrogen on
biogeochemical processes and indicators. 6-14
6.2.3 Forests 6-17
Table 6-2 Growth, productivity, and carbon cycle responses of
ectomycorrhizal fungi to nitrogen added via atmospheric
deposition or experimental treatments. 6-21
Table 6-3 Growth, productivity, and carbon cycle responses of arbuscular
mycorrhizal fungi to nitrogen added via atmospheric deposition
or experimental treatments. 6-31
Table 6-4 Abundance and carbon cycle responses of forest soil
microorganisms and soil invertebrates to nitrogen added in
experimental treatments. 6-33
Table 6-5 Growth and physiology responses of forest epiphytic lichens to
nitrogen added via atmospheric deposition or experimental
treatments. 6-41
Figure 6-3 Studies reporting the response of forest (A) aboveground
biomass carbon sequestration and (B) ecosystem carbon
sequestration to nitrogen deposition or long-term nitrogen
additions. 6-47
6.2.4 Arctic and Alpine Tundra and Grasslands 6-48
Table 6-6 Alpine and Arctic tundra plant productivity and physiology
responses to nitrogen added via atmospheric deposition or
experimental treatments. 6-50
Table 6-7 Alpine and Arctic tundra lichen growth and physiology
responses to nitrogen added via atmospheric deposition or
experimental treatments. 6-68
xiii
-------
6.3
6.4
6.5
Growth and biodiversity responses of ericoid mycorrhizal fungi
to nitrogen added in experimental treatments.
Alpine and Arctic tundra microbial biomass responses to
experimental nitrogen additions.
6.2.5
6.2.6
Table 6-8
Table 6-9
Grasslands
Table 6-10 Grassland microbial biomass responses to experimental
nitrogen additions.
Arid and Semiarid Ecosystems
Table 6-11 Arid and semiarid ecosystem plant productivity and physiology
responses to nitrogen added in experimental treatments.
Arid and semiarid microbial biomass responses to experimental
nitrogen additions.
Table 6-12
6-72
6-73
6-75
6-81
6-83
6-86
6-96
Relationships between Nitrogen Deposition and Terrestrial Species Composition, Species
Richness, and Biodiversity
6.3.1
6.3.2
6.3.3
Introduction
Mechanisms Operating across Terrestrial Ecosystems
Forests
Table 6-13
Forest plant diversity responses to nitrogen added via
atmospheric deposition or experimental treatments.
Table 6-14
Table 6-15
Table 6-16
Forest microbial biodiversity responses to nitrogen added via
atmospheric deposition or experimental treatments.
Ectomycorrhizal biodiversity responses to nitrogen added via
atmospheric deposition or experimental N additions.
Arbuscular mycorrhizal responses to nitrogen added via
atmospheric deposition or experimental treatments.
6.3.4
Table 6-17 Arthropod and other invertebrate responses to experimental
nitrogen additions.
Alpine and Arctic Tundra
Table 6-18 Alpine and Arctic tundra plant diversity responses to nitrogen
_ 6-98
_ 6-98
_ 6-99
6-103
6-106
6-113
6-117
6-122
6-126
6-130
6.3.5
6.3.6
Table 6-19
Grasslands
Table 6-20
added via atmospheric deposition or experimental treatments. 6-132
Alpine and Arctic tundra microbial diversity responses to
nitrogen added via experimental treatments. 6-137
6-139
Grassland microbial diversity responses to nitrogen added via
experimental treatments.
Arid and Semiarid Ecosystems
Table 6-21
Table 6-22
6.3.7 Lichens
Arid and semiarid ecosystem plant diversity responses to
nitrogen added via atmospheric deposition or experimental
treatments.
Arid and semiarid ecosystem microbial diversity responses to
nitrogen added via experimental treatments.
Table 6-23 Lichen biodiversity responses to nitrogen added via atmospheric
deposition or experimental treatments.
Most Sensitive Ecosystems
6.3.8
Climate Modification of Ecosystem Nitrogen Response
Critical Loads
6.5.1
6.5.2
Figure 6-4 Comparison of European and U.S. empirical critical loads for
nitrogen from Pardo et al. (2011a).
Mycorrhizal Fungi
Figure 6-5 Map of critical loads for mycorrhizal fungi by ecoregion in the
U.S.
Table 6-24 Mycorrhizal critical loads.
Lichens and Bryophytes
Figure 6-6 Map of critical loads for lichens by ecoregion in the U.S.
6-146
6-148
6-150
6-157
6-158
6-159
6-162
6-163
6-165
6-168
6-168
6-170
6-171
6-171
6-173
xiv
-------
Table 6-25 Lichen critical loads. 6-174
6.5.3 Herbaceous and Shrub Species 6-176
Table 6-26 Herbaceous and shrub critical loads. 6-178
Figure 6-7 Map of critical loads for herbaceous plants and shrubs by
ecoregion in the U.S. 6-181
Figure 6-8 Nitrogen deposition (gray scale) and critical loads for nitrogen
deposition based on total graminoid plus forb species richness
(colored symbols). 6-182
6.5.4 Trees 6-184
Table 6-27 Tree critical loads. 6-184
Figure 6-9 Map of critical loads for forest ecosystems by ecoregion in the
U.S. 6-185
6.5.5 Critical Loads Exceedance Studies 6-186
6.6 Summary 6-188
6.6.1 Physiology, Growth, and Productivity Summary 6-188
6.6.2 Biodiversity Summary 6-190
6.6.3 Critical Loads Summary 6-193
Table 6-28 Critical loads for nitrogen by Pardo et al. (2011c) and more
recent critical load information. 6-194
APPENDIX 7 AQUATIC BIOGEOCHEMISTRY 7-1
7.1 Biogeochemistry of Nitrogen and Sulfur in Freshwater Systems 7-1
Table 7-1 Summary of key freshwater indicators of eutrophication and
acidification. 7-4
7.1.1 Nitrogen and Sulfur Sources 7-4
Table 7-2 Summary of recent studies quantifying nitrogen deposition
contribution to total nitrogen loading in freshwater systems. 7-7
7.1.2 Ecosystem Processes, Effects, and Indicators 7-10
Figure 7-1 Nitrogen cycle in freshwater ecosystem. 7-11
Figure 7-2 Quarterly measured concentrations of a range of water
chemistry variables at Blue Lough from 1990 to 2014. 7-16
Figure 7-3 Dissolved organic carbon (DOC) concentrations from 1993 to
2013 for Bracey Pond (a), Salmon Pond (b), Jellison Pond (c),
Second Pond (d), Little Long Pond (e), and Tilden Pond (f) in
Maine. 7-30
7.1.3 Freshwater Monitoring and Databases 7-33
Table 7-3 Monitoring and resurvey results of aquatic acidification and/or
chemical recovery since the 2008 Integrated Science
Assessment for Oxides of Nitrogen and Sulfur—Ecological
Criteria. 7-35
7.1.4 Models 7-42
Table 7-4 Recent process-based model estimates of surface water
acidification and chemical recovery in the U.S. 7-44
7.1.5 National-Scale Sensitivity and Response 7-45
Table 7-5 Model projections of surface water sulfate and associated acid
neutralizing capacity, shown as changes between dates, for
Adirondack and Shenandoah streams. 7-47
7.1.6 Water Quality Criteria 7-57
Table 7-6 Numeric nutrient water quality criteria for rivers/streams by state
(all values in mg/L). 7-59
Table 7-7 U.S. EPA aggregate Level III ecoregion nutrient criteria (all
values in mg/L; U.S. EPA ecoregional nutrient criteria
documents for rivers and streams). 7-63
Figure 7-4 Total nitrogen criterion values by ecoregion. 7-64
Figure 7-5 Chlorophyll a criterion values by ecoregion. 7-64
xv
-------
7.1.7 Freshwater Biogeochemistry Summary 7-65
7.2 Biogeochemistry of Nitrogen in Estuarine and Near-Coastal Systems 7-68
Table 7-8 Summary of key indicators of nitrogen enrichment in estuaries. 7-70
7.2.1 Nitrogen Sources 7-70
Figure 7-6 Chemical nitrogen cascade in the Chesapeake Bay watershed
(metric tons/year). 7-72
Table 7-9 Summary of studies quantifying atmospheric nitrogen
contribution to total nitrogen in coastal areas via watersheds
and/or direct deposition to estuary surface waters. 7-73
7.2.2 The Estuary Environment 7-77
Figure 7-7 Schematic diagram illustrating sources, transformations, and
fate of nitrogen along the estuary-to-ocean continuum. Surface,
subsurface, and atmospheric pathways of externally supplied or
new nitrogen inputs attributable to anthropogenic activities are
shown as internal nitrogen cycling. The combined
anthropogenic nitrogen inputs are shown as a thick arrow
(upstream), which decreases in thickness downstream as a
portion of the nitrogen inputs settles to the bottom sediments
and is buried and/or denitrified. Nitrogen (N2) fixation is a
biologically mediated new nitrogen input. The linkage of
anthropogenically enhanced nitrogen inputs to accelerated
primary production or eutrophication and its trophic and
biogeochemical fate are also shown. 7-78
7.2.3 Dissolved Oxygen and Hypoxia 7-78
7.2.4 Estuarine and Near-Coastal pH 7-79
Figure 7-8 A conceptual model for a large river plume eutrophication and
subsurface water hypoxia and acidification. 7-80
7.2.5 Nitrogen in Surface Waters 7-82
7.2.6 Nitrogen Cycling 7-83
Figure 7-9 New complexities in nitrogen cycling have been detailed since
the 2008 ISA as shown in this illustration of the sedimentary N
cycle in the Lower St. Lawrence estuary. 7-84
7.2.7 Monitoring Data 7-95
Figure 7-10 Percentage of area in each coastal region scoring good, fair,
and poor based on the Water Quality Index for the NCCA 2010. 7-96
7.2.8 Modeling Estuaries and Near-Coastal Areas 7-99
7.2.9 National-Scale Sensitivity 7-104
7.2.10 Water Quality Criteria for Estuaries 7-105
Figure 7-11 State progress toward developing numeric nutrient criteria for
estuaries https://www.epa.gov/nutrient-policy-data/state-
progress-toward-developing-numeric-nutrient-water-quality. 7-106
7.2.11 Estuary and Near-Coastal Biogeochemistry Summary 7-107
APPENDIX 8 BIOLOGICAL EFFECTS OF FRESHWATER ACIDIFICATION 8-1
8.1 Introduction 8-1
8.2 Chronic versus Episodic Acidification 8-3
8.3 Aquatic Organisms Impacted by Acidifying Deposition 8-5
8.3.1 Plankton 8-5
8.3.2 Periphyton 8-9
8.3.3 Benthic Invertebrates 8-10
Figure 8-1 Total macroinvertebrate species (community) richness as a
function of median pH in 36 streams sampled in the western
Adirondack Mountains of New York, 2003-2005; the four
standard (New York State) impact categories for species
richness are defined. 8-12
xvi
-------
Table 8-1 Thresholds of biological response to changes in water acidity for
benthic invertebrates published since the 2008 Integrated
Science Assessment for Oxides of Nitrogen and Sulfur—
Ecological Criteria. 8-13
8.3.4 Bacteria, Macrophytes, and Bryophytes 8-14
8.3.5 Amphibians 8-15
8.3.6 Fish 8-16
Figure 8-2 Relationship between (a) inorganic monomeric aluminum and
gill aluminum for parr and smolt, and (b) acid neutralizing
capacity and gill aluminum. 8-20
Figure 8-3 Critical aquatic pH ranges for fish species. 8-22
Table 8-2 pH thresholds in fish published since the 2008 Integrated
Science Assessment for Oxides of Nitrogen and Sulfur—
Ecological Criteria. 8-24
Figure 8-4 Number of fish species per lake verses acidity status, expressed
as acid neutralizing capacity, for Adirondack lakes. 8-25
Table 8-3 Expected ecological effects and concern levels in freshwater
ecosystems at various levels of acid neutralizing capacity. 8-27
Table 8-4 Threshold values of aluminum for various fish species and
associated effects. 8-29
Figure 8-5 Relationship between (a) pH, (b) cationic aluminum, (c) acid
neutralizing capacity, and (d) gill aluminum as compared with
accumulated mortality of Atlantic salmon smolt. 8-33
8.3.7 Fish-Eating Birds 8-34
8.3.8 Aquatic Assemblages 8-34
Figure 8-6 Species richness of biotic groups in 30 Adirondack study lakes
relative to midsummer epilimnetic pH during sample years. 8-36
Figure 8-7 Structure of diatom assemblage in 20 streams across a pH
gradient of 4.5-8.5. (a) Species richness (log-io-transformed
numbers of diatom species per stream), (b) total abundance of
diatoms (log-io-transformed numbers of individuals per m2),
(c) chlorophyll a concentration (mg chlorophyll a per m2), as a
measure of algal biomass. 8-37
Figure 8-8 Macroinvertebrate community composition in 20 streams across
a pH gradient. Taxon richness (total number of primary
consumer taxa; left panel) and benthic density [number of
individuals logio(x+ 1)-transformed; right panel] plotted against
stream pH, all primary consumers (a, b), shredders (excluding
Leuctridae and Nemouridae) (c, d), herbivore-detritivores
(Leuctridae and Nemouridae) (e, f), collectors (g, h) and grazers
(i, j). 8-38
8.4 Documentation of Biological Recovery 8-39
8.4.1 Phytoplankton Recovery 8-42
Table 8-5 Paleolimnological responses to changing lake chemistry
published since the 2008 Integrated Science Assessment for
Oxides of Nitrogen and Sulfur—Ecological Criteria. 8-43
8.4.2 Zooplankton Recovery 8-44
Figure 8-9 (A) Midsummer sulfate concentration in the epilimnion of
Brooktrout Lake (o) and in annual wet deposition (•) at local
National Atmospheric Deposition Program/National Trends
Network Station NY52 from 1984-2012. (B) Midsummer
epilimnetic concentrations of inorganic monomeric aluminum (~)
and [hydrogen ion] (¦) in Brooktrout Lake from 1984-2012.
(C) Midsummer phytoplankton (A) and total plankton
(phytoplankton, rotifers, crustaceans) (A) species richness in
Brooktrout Lake from 1984-2012. 8-46
xvii
-------
Table 8-6 Midsummer values and productivity analytes in the epilimnion of
Brooktrout Lake in the Adirondack Park from the 1980s through
2010-2012. 8-47
8.4.3 Benthic Invertebrate Recovery 8-50
8.4.4 Fish and Amphibian Recovery 8-53
8.4.5 Bird Recovery 8-55
8.4.6 Mitigation 8-56
8.5 Levels of Deposition at Which Effects Are Manifested 8-58
8.5.1 Most Sensitive and Most Affected Ecosystems and Regions 8-58
Figure 8-10 Map of landscape sensitivity to acidic deposition for the
northeastern and mid-Atlantic U.S. Stippled areas were not
considered. 8-59
8.5.2 Extent and Distribution of Sensitive Ecosystems/Regions 8-60
Figure 8-11. Surface water Acid Neutralizing Capacity (ANC) map, based on
data compiled by Sullivan (2017) 8-62
Figure 8-12 (a) Minimum critical loads of surface water acidity for nitrogen
and sulfur. Grids represent the minimum calculated critical load
from all data within the 36 * 36-km grid cell (b) Mean critical
loads of surface water acidity. Grids represent the average
calculated critical load from all data within the 36 * 36-km grid
cell. The critical chemical criterion used was an acid neutralizing
capacity 50 peq/L. 8-63
8.5.3 Climate Modification of Ecosystem Response to Nitrogen and Sulfur 8-64
8.5.4 Critical Loads 8-64
Table 8-7 Recent empirical critical loads to protect against aquatic
acidification in U.S. ecosystems. 8-67
Table 8-8 Recent aquatic critical load and target load modeling studies in
the U.S. to protect against aquatic acidification. 8-68
Figure 8-13 Target loads for sulfur deposition in the Adirondack Park to
protect lake acid neutralizing capacity to 50 peq/L in the
year 2010 (left map) and their exceedance (right map). 8-74
Figure 8-14. Modeled target loads and median measured Acid Neutralizing
Capacity (ANC) in the period 1993 to 1996 from 30 Great
Smoky Mountain National Park (GRSM) streams to achieve an
ANC of 20 peq/L by the year 2150. 8-77
Figure 8-15. Exceedance level of current NO3" + SO42" atmospheric
deposition for 387 stream sites in the GRSM. Exceedances
were calculated for the years 2050 and 2150 using two targets
for modeled stream ANC recovery of 20 pmolc/L and
20 pmolc/L less than the simulated preindustrial ANC. 8-78
8.6 Aquatic Acidification Summary and Causal Determinations 8-80
Table 8-9 Ecological indicators for aquatic acidification. 8-82
8.6.1 Phytoplankton 8-82
8.6.2 Zooplankton 8-83
8.6.3 Benthic Invertebrates 8-83
8.6.4 Fish 8-84
8.6.5 Thresholds of Response 8-85
Table 8-10 Results of recent biological effects studies in surface waters
indicative of thresholds of biological response to changes in
water acidity. 8-86
8.6.6 Biological Recovery 8-86
8.6.7 Most Sensitive and Most Affected Regions 8-87
8.6.8 Critical Loads 8-88
APPENDIX 9 BIOLOGICAL EFFECTS OF FRESHWATER NITROGEN ENRICHMENT 9-1
9.1 Introduction to Nitrogen Enrichment and Eutrophication in Freshwater Systems 9-1
xviii
-------
Figure 9-1 Conceptual model of the influence of atmospheric nitrogen
deposition on freshwater nutrient enrichment. 9-2
9.1.1 Deposition to Freshwater Systems 9-5
Table 9-1 Summary of studies using diatoms as biological indicators of
nitrogen enrichment evaluated in the literature since the 2008
Integrated Science Assessment for Oxides of Nitrogen and
Sulfur—Ecological Criteria. 9-9
Figure 9-2 Phytoplankton responses to nitrogen and/or phosphorus
enrichment for Rocky Mountain lakes receiving low or high
atmospheric nitrogen deposition, given as the ratio of final
chlorophyll concentration in the enriched treatment
(+phosphorus, +nitrogen, or+nitrogen + phosphorus) to the
chlorophyll concentration in the unenriched control. 9-12
9.2 Biological Indicators 9-15
9.2.1 Diatoms 9-16
9.2.2 Ratios of Nitrogen and Phosphorus 9-17
9.2.3 Phytoplankton Biomass Nitrogen (N) to Phosphorus (P) Limitation Shift 9-18
9.2.4 Periphyton/Microbial Biomass 9-18
9.2.5 Chlorophyll a 9-20
Table 9-2 Summary of additional evidence for nitrogen limitation on
productivity of freshwater ecosystems that has been evaluated
since the 2008 Integrated Science Assessment for Oxides of
Nitrogen and Sulfur—Ecological Criteria. 9-24
9.2.6 Potential Biological Indicators 9-27
9.3 Community Composition, Species Richness, and Diversity 9-29
9.3.1 Archaea and Bacterial Diversity 9-30
9.3.2 Phytoplankton Diversity 9-30
Table 9-3 Summary of studies on nitrogen effects on species composition
and biodiversity that have been evaluated in the literature since
the 2008 Integrated Science Assessment for Oxides of Nitrogen
and Sulfur—Ecological Criteria. 9-31
9.3.3 Benthic Algal Diversity 9-39
9.3.4 Zooplankton Diversity 9-40
9.3.5 Macroinvertebrate Diversity 9-41
9.3.6 Macrophyte Diversity 9-42
9.3.7 Amphibian Diversity 9-43
9.3.8 Fish Diversity 9-43
9.4 Animal Behavior and Disease 9-43
9.4.1 Behavior 9-43
9.4.2 Disease 9-44
9.5 Summary of Thresholds, Levels of Deposition at Which Effects Are Manifested, and
Critical Loads 9-44
Table 9-4 Summary of critical loads for nitrogen eutrophication for surface
water in the U.S. [adapted from Pardo et al. (2011c) with newer
studies added], 9-45
Table 9-5 Summary of mean lake nitrate (NO3") concentrations, inorganic
nitrogen deposition amounts, and nutrient enrichment inflection
points where lake NO3" concentrations reflect increased
nitrogen deposition, [from Baron et al. (2011b)]. 9-48
9.6 Summary and Causal Determination 9-48
APPENDIX 10 BIOLOGICAL EFFECTS OF NITROGEN ENRICHMENT IN ESTUARIES AND
NEAR-COASTAL SYSTEMS 10-1
10.1 Introduction 10-1
Figure 10-1 Eutrophication can occur when the availability of nutrients
increases above normal levels. 10-3
xix
-------
Figure 10-2 Overall eutrophication condition on a national scale. 10-5
10.1.1 Nitrogen Sources to Estuaries and Coasts 10-5
10.1.2 Trends in Atmospheric Deposition of Nitrogen 10-6
10.1.3 Nitrogen Limitation 10-6
10.1.4 Characteristics of Coastal Systems Sensitive to Eutrophication 10-8
10.2 Indicators of Nutrient Enrichment 10-11
Table 10-1 Indicators of Estuarine Eutrophication. 10-12
Figure 10-3 Biological indicator responses to nutrient enrichment. 10-13
10.2.1 Chlorophyll a 10-13
Table 10-2 Chlorophyll a thresholds used in methods to evaluate the status
of phytoplankton in U.S. coastal and estuarine water bodies. 10-14
10.2.2 Harmful/Nuisance/Toxic Algal Blooms 10-16
Table 10-3 Levels and forms of nitrogen at which effects on phytoplankton
are manifest in U.S. coastal waters evaluated in the literature
since the 2008 Integrated Science Assessment for Oxides of
Nitrogen and Sulfur—Ecological Criteria. 10-20
10.2.3 Macroalgal Abundance 10-25
10.2.4 Dissolved Oxygen 10-26
Figure 10-4 The range of ecological impacts exhibited as dissolved oxygen
levels drop from saturation to anoxia. 10-27
Figure 10-5 Coastal eutrophic and hypoxic areas of North America and the
Caribbean. 10-29
10.2.5 Submerged Aquatic Vegetation 10-32
Table 10-4 Nitrogen loading thresholds from multiple watershed sources
versus eelgrass loss. 10-34
Figure 10-6 Extent of submerged aquatic vegetation in the Chesapeake Bay
1978-2016. 10-36
10.2.6 Indices of Estuarine Condition 10-37
10.3 Effects on Biodiversity 10-39
10.3.1 Paleontological Diversity 10-39
10.3.2 Phytoplankton Diversity 10-40
10.3.3 Diversity of Phytoplankton Under Different Forms of Nitrogen (Reduced vs.
Oxidized) 10-41
Figure 10-7 Summary conceptual schematic illustrating the effect of
changes in the proportion of NhV and NO3" in the loads of N
provided to a natural system. When NhV is the dominant form,
and when waters are warmer, flagellates, cyanobacteria, and
chlorophytes among other classes may proliferate, leading to
overall productivity dominated by the small size class of algae
(e.g., <5 |jm). In contrast, when NO3" is the dominant form
provided, especially under cooler water conditions, diatoms
more likely dominate, and their overall production will be more
likely dominated by cells of a larger size class (e.g., >5 |jm).
Moreover, chlorophyll a yield and total production may be higher
than under the NHV enrichment condition. 10-43
10.3.4 Diversity of Bacteria, Archaea, and Microzooplankton 10-44
10.3.5 Benthic Diversity 10-45
10.3.6 Fish Diversity 10-46
10.3.7 Trophic Interactions 10-46
10.3.8 Models Linking Indicators to Nitrogen Enrichment 10-47
Figure 10-8 Forecasting curves for effects on total nitrogen loadings on
(A) chlorophyll and (B) dissolved oxygen (mean and 95%
confidence interval) for selected estuaries demonstrating the full
range of sensitivity to relative total nitrogen loading. 10-49
10.4 Animal Behavior and Disease 10-50
10.4.1 Behavior 10-50
xx
-------
Figure 10-9 The pathway of effects of eutrophication on different
reproductive behaviors and selection forces in Gasterosteus
aculeatus. 10-51
10.4.2 Disease 10-52
10.5 Nitrogen Enrichment and Acidification Effects on Calcifying Organisms 10-53
Figure 10-10 Pathway from nitrogen loading to biological effects of nutrient-
enhanced coastal acidification. Both microbial respiration of
organic matter and increasing atmospheric carbon dioxide lower
pH of coastal waters. 10-55
10.6 Summary of Thresholds and Levels of Deposition at Which Effects Are Manifested 10-58
10.7 Summary and Causality Determination 10-59
APPENDIX 11 NITROGEN ENRICHMENT EFFECTS IN WETLANDS 11-1
11.1 Introduction 11-1
Table 11-1 Wetland classification used in the Integrated Science
Assessment. 11-2
11.2 Regional Sensitivity 11-4
Figure 11-1 Estimated extent of wetlands stressed by nonnative plants as
determined by Nonnative Plant Stressor Indicator, at a national
or regional level. 11-5
11.2.1 Climate Modification of Ecosystem Response to Nitrogen 11-6
11.3 Soil Biogeochemistry 11-6
11.3.1 Nitrogen Pools and Processes 11-7
Table 11-2 New studies on nitrogen addition effects on nitrogen cycling in
wetlands. 11-13
11.3.2 Soil Carbon Cycling 11-19
Table 11-3 Loading effects upon belowground carbon cycling. 11-25
Table 11-4 Nitrogen loading effects upon methane emissions. 11-34
11.4 Production and Aboveground Biomass 11-36
11.4.1 Salt Marsh 11-36
11.4.2 Mangrove 11-38
11.4.3 Freshwater Tidal Marsh 11-38
11.4.4 Intermittent Wetland 11-39
11.4.5 Bog and Fen 11-39
11.4.6 Summary Table 11-41
Table 11-5 Nitrogen loading effects upon production and biomass. 11-41
11.5 Plant Stoichiometry and Physiology 11-48
11.5.1 Salt Marsh 11-49
11.5.2 Mangrove 11-49
11.5.3 Freshwater Marsh 11-50
11.5.4 Riparian Wetland 11-51
11.5.5 Bog and Fen 11-51
11.5.6 Summary Table 11-56
Table 11-6 Nitrogen loading effects upon plant stoichiometry and
physiology. 11-56
11.6 Plant Architecture 11-68
11.6.1 Salt Marsh 11-68
11.6.2 Mangrove 11-69
11.6.3 Freshwater Tidal Marsh 11-69
11.6.4 Riparian Wetland 11-70
11.6.5 Summary Table 11-70
Table 11-7 Nitrogen loading effects upon architecture. 11-70
11.7 Demography 11-72
11.7.1 Mangrove 11-73
xxi
-------
11.7.2 Riparian Wetland 11-73
11.7.3 Summary Table 11-74
Table 11-8 Nitrogen loading effects upon demography. 11-74
11.8 Biodiversity/Community 11-74
11.8.1 Plants 11-75
Table 11-9 Nitrogen loading effects upon plant biodiversity and
communities. 11-79
11.8.2 Consumers 11-84
11.9 Critical Loads 11-85
11.9.1 Freshwater Wetland 11-85
11.9.2 Coastal Wetlands 11-85
11.9.3 Comparison to Critical Loads from Europe 11-86
11.10 Summary 11-87
11.10.1 Causality across Wetland Types 11-88
Figure 11-2 Summary of the levels of nitrogen addition where a change to
nitrogen cycling is observed. 11-89
Figure 11-3 Summary of new literature of nitrogen load effects on
belowground and aboveground carbon cycling. 11-91
Figure 11-4 Summary of the level of nitrogen load that caused a change in
the response variables of plant stoichiometry and physiology in
wetlands. 11-93
Figure 11-5 Summary of nitrogen addition studies on wetland biodiversity. 11-95
11.10.2 Coastal versus Freshwater Wetlands 11-96
Figure 11-6 Summary of field nitrogen addition studies for coastal wetlands
(blue borders) versus critical load (black border). 11-97
Figure 11-7 Summary of nitrogen load studies for freshwater wetlands as
well as current critical loads. 11-98
APPENDIX 12 NONACIDIFYING SULFUR ENRICHMENT EFFECTS 12-1
12.1 Introduction 12-2
Figure 12-1 Effects of sulfur oxide deposition on chemical (blue boxes),
biological (green boxes), and atmospheric (yellow boxes)
indicators of ecosystem change, as documented by the previous
Integrated Science Assessment and more recent research. 12-4
12.2 Ecosystem Effects of Altered Sulfur Cycling 12-4
12.2.1 The Sulfur Cycle 12-4
Figure 12-2 Sulfur cycle in terrestrial, forested ecosystems. 12-5
12.2.2 Deposition and Sulfur Stores 12-9
Table 12-1 New study on sulfur (S) deposition effects on sulfur (S) cycling. 12-10
12.2.3 Sulfide Toxicity 12-10
Table 12-2 Quantitative effects of sulfide on wetland and aquatic plant
species. 12-11
Figure 12-3 Schematic from Minnesota Pollution Control Agency that
illustrates the mitigating effect of iron on the toxicity of sulfide
and the stimulatory effect that organic carbon has on sulfide
production. 12-15
12.2.4 Internal Eutrophication 12-16
Figure 12-4 Mechanisms of linked sulfur, iron, and phosphorus cycling in
wetland waters and soils. 12-17
12.2.5 Sulfur Effects on Methane Emissions 12-18
Table 12-3 New studies on nonacidifying sulfur effects on methane
emissions. 12-20
12.3 Interactions between Sulfur Deposition and Mercury 12-22
xxii
-------
Figure 12-5. Sulfate, dissolved organic carbon, and inorganic mercury are all
important determinants of the rate and amount of methyl
mercury produced by sulfate-reducing prokaryotes. 12-23
12.3.1 Mercury Cycle and the Importance of Methylation 12-24
Figure 12-6 Cycling of mercury (Hg) and methylmercury (MeHg) in
ecosystems. 12-24
12.3.2 Biology of Sulfate-Reducing Prokaryotes 12-26
Table 12-4 New studies on the biology of sulfate-reducing prokaryotes. 12-27
Figure 12-7 The effect of different environmental factors on the relationship
between sulfate and mercury methylation. Methylmercury
(MeHg) accumulation is minimal at low and high sulfate
concentrations, with an optimum near 100 |jM sulfate (blue line).
High dissolved organic matter (DOM) will increase the
magnitude of MeHg production across the range of sulfate
concentrations (red line). Sulfide produced by methylators will
inhibit further methylation if it accumulates in the aqueous
environment where methylation occurs, shifting the MeHg
optimum left (purple line). However if ecosystem chemistry (iron
[Fe], reoxidation of sulfide, organic matter [OM]) allows for rapid
sequestration of sulfide to large particles and sediments, the
relaxation of the negative feedback of sulfide to sulfur-reducing
prokaryotes (SRPs) will shift the MeHg optimum right (green
line). 12-35
12.3.3 Environmental Drivers of Mercury Methylation Potential 12-36
Table 12-5 Environmental factors that affect mercury (Hg) methylation. 12-36
Figure 12-8 The relationship between surface water sulfate and total
mercury or methylmercury fraction in river/leaf litter mesocosms. _ 12-40
Table 12-6 New mesocosm or incubation studies on sulfur addition effects
on methylmercury. 12-41
Figure 12-9 Total methylmercury mass in water at Little Rock Lake, Wl,
annually (a), and in relationship to annual mercury (b) or
sulfur (c) deposition. 12-46
12.3.4 Mercury Methylation in Sulfur Addition Field Studies 12-53
Table 12-7 New studies on mercury methylation in sulfur-amended
ecosystems. 12-54
Table 12-8 New studies on mercury (Hg) and sulfur (S) cycling in the
Everglades Water Conservation Areas (WCA). 12-61
Figure 12-10 The relationship between surface water sulfate and
methylmercury concentrations in the Florida Everglades. 12-63
12.3.5 Drivers of Mercury Methylation under Ambient Conditions 12-66
Table 12-9 New studies on correlations between sulfur (S) and
methylmercury (MeHg) in ecosystems. 12-68
Figure 12-11 Concentrations of soil sulfur (S), elevated by sulfur oxides (SOx)
deposition, correlate with rice grain (methylmercury [MeHg]) and
rice grain %MeHg in rice paddies near industrial emitters in
China. 12-74
Figure 12-12 Interactions between mercury (Hg) and sulfur (S) cycles. 12-74
Table 12-10 New studies on deposition of sulfur (S) and mercury (Hg) and
their effect on methylmercury (MeHg). 12-75
Table 12-11 New studies of Interactions between organic matter and sulfur
(S) in Controlling methylmercury (MeHg). 12-78
12.4 Sulfur Impacts on Mercury in Wildlife 12-80
Figure 12-13 Bioconcentration and biomagnification result in methylmercury
concentrations about 1 million times higher in predator fish than
in stream water. 12-81
Table 12-12 New studies on sulfur (S) impacts upon mercury (Hg) in wildlife. 12-82
xxiii
-------
Figure 12-14 Tissue mercury concentrations as a function of surface water
sulfate concentrations (n = 2,360 surface water samples) in the
Everglades Protection Area. 12-86
12.5 Extent and Distribution of Sensitive Ecosystems 12-87
Figure 12-15 Fish mercury concentrations across the U.S. 12-88
12.6 Critical Loads 12-89
12.7 Summary of Nonacidifying Sulfur Effects 12-89
Table 12-13 Summary of quantitative effects of nonacidifying sulfur
enrichment. 12-90
12.7.1 Terrestrial Sulfur Cycling 12-92
12.7.2 Aquatic Sulfur Cycling 12-93
12.7.3 Sulfide Toxicity 12-93
Figure 12-16 Thresholds of sulfate or sulfide concentrations in water, which
cause biological and chemical effects in ecosystems. 12-94
12.7.4 Internal Eutrophication 12-95
12.7.5 Effects on Methane Production 12-95
12.7.6 The Role of Microbes in Mercury Methylation 12-95
Figure 12-17 Linear relationships between sulfate and methylmercury
concentrations in published studies. 12-97
12.7.7 Impacts of Sulfur upon Mercury Cycling 12-98
Figure 12-18 Thresholds of sulfate addition or deposition from published
studies which affect chemical or biological changes in
ecosystems. 12-99
12.7.8 Sensitive Ecosystems 12-100
12.7.9 Mercury Effects on Animal Species 12-100
12.8 Supplemental Materials: Mercury Cycling 12-102
12.8.1 Transfer of Mercury from the Atmosphere to Terrestrial Ecosystems 12-102
12.8.2 Transfer of Mercury from Terrestrial to Aquatic Ecosystems 12-103
12.8.3 Transfer of Mercury from Atmosphere to Aquatic Ecosystems 12-104
12.8.4 Methylmercury Cycling 12-104
12.8.5 Transfer of Methylmercury from Terrestrial to Aquatic Ecosystems 12-104
12.8.6 Transfer of Methylmercury from Wetlands to Aquatic Ecosystems 12-105
12.8.7 Methylmercury Cycling in Aquatic Ecosystems—Lake Onondaga, NY 12-105
Figure 12-19 Mercury and methylmercury mass balance cycle in Lake
Onondaga in 1992. The quantities of mercury in each flux are
indicated in parentheses by (kg total Hg/yr, kg MeHg/yr). 12-106
12.8.8 Methylmercury in Sediments and Water Column—Lakes 12-107
12.8.9 Methylmercury in Sediments and Water Column—Wetlands 12-108
12.8.10 Methylmercury in Sediments and Water Column—Estuarine and Marine
Ecosystems 12-109
APPENDIX 13 CLIMATE MODIFICATION OF ECOSYSTEM RESPONSE TO NITROGEN AND
SULFUR 13-1
13.1 Climate Modification of Soil Acidification and Nitrogen Enrichment 13-2
13.1.1 Nitrogen Transport and Transformation 13-2
Figure 13-1 Summary of key interactions between nitrogen, anthropogenic-
driven climate change, and hydrology. 13-3
Table 13-1 Summary on climate modification sulfur and nitrogen cycling in
Appendix 4 and Appendix 6 in addition to those in Appendix 13. 13-6
13.1.2 Nitrogen, Climate, and Carbon Cycling 13-7
Figure 13-2 The effects of increased nitrogen, temperature, and precipitation
on terrestrial carbon pools (left panel) and fluxes (right panel)
from published meta-analyses. 13-9
xxiv
-------
Table 13-2 Summary on climate modification of nitrogen (N) effects on
carbon (C) cycling in Appendix 4 and Appendix 6 in addition to
those in Appendix 13. 13-11
13.1.3 Climate and Acidification 13-12
Table 13-3 Summary on climate modification of acidification in Appendix 4
and Appendix 5 in addition to those in Appendix 13. 13-15
13.1.4 Nitrogen, Climate, and Biodiversity 13-16
Table 13-4 Summary on climate modification of nitrogen (N) effects on
biodiversity in Appendix 6 in addition to those in Appendix 13. 13-19
13.2 Estuaries 13-20
13.3 Wetlands 13-21
APPENDIX 14 ECOSYSTEM SERVICES 14-1
14.1 Introduction 14-1
14.2 Ecosystem Services Frameworks 14-2
14.3 United States Applications 14-4
14.3.1 Acidification 14-5
Table 14-1 Ecosystem services research related to ecosystem acidification. 14-7
14.3.2 Eutrophication 14-8
Figure 14-1 Economic nitrogen cascade in the Chesapeake Bay Watershed. 14-11
Table 14-2 Potential damage costs of nitrogen (N) ($/kg N; 2008 or as
reported) to air, land, and water resources in the conterminous
U.S. in the early 2000s as synthesized by Sobota et al. (2015).
Low, median, and high costs derive from the specific damage
cost reference. Negative values indicate an economic benefit. 14-12
Figure 14-2 Map of ecosystem services altered by nitrogen critical load
exceedance. 14-15
Table 14-3 Numbers of chains, Final Ecosystem Goods and Services
(FEGS), and beneficiaries (bens) associated with each initial
biological indicator (Clark et al., 2017). 14-16
Table 14-4 Ecosystem services research related to nitrogen-driven
eutrophication. 14-17
14.3.3 Nitrogen and Climate Modification 14-18
Table 14-5 Summary of recent literature examining economic impacts of
ocean acidification on U.S. fisheries. 14-19
14.4 European and Canadian Applications 14-20
Figure 14-3 Benefits and costs associated with the 25% decline in nitrogen
deposition in the U.K. since 1990. 14-21
14.5 Global Perspective 14-21
14.6 Summary 14-22
14.7 Supplemental Materials: Ecosystem Services Profiles of Select Species 14-24
14.7.1 Balsam Fir 14-24
14.7.2 Eel Grass 14-25
14.7.3 Green Turtle 14-27
14.7.4 White Ash 14-28
14.7.5 Lace Lichen 14-30
APPENDIX 15 OTHER ECOLOGICAL EFFECTS OF PARTICULATE MATTER 15-1
15.1 Introduction 15-1
15.2 Direct Effects of Particulate Matter on Radiative Flux 15-2
15.3 Particulate Matter Deposition to Ecosystems 15-3
15.3.1 Metals 15-3
15.3.2 Organics 15-4
15.4 Effects of Particulate Matter on Vegetation 15-5
XXV
-------
15.4.1 Vegetative Surfaces 15-5
15.4.2 Foliar Uptake of Particulate Matter 15-7
15.4.3 Particulate Matter Impacts on Gas Exchange Processes 15-8
15.4.4 Plant Physiology 15-8
15.4.5 Uptake of Particulate Matter by Plants from Soils 15-9
15.4.6 Effects on Plant Growth and Reproduction 15-10
15.5 Effects of Particulate Matter on the Soil Environment 15-11
15.5.1 Bioavailability in Soils 15-11
15.5.2 Soil Nutrient Cycling 15-12
15.5.3 Soil Community Effects 15-14
15.5.4 Soil Microbe Interactions with Plant Uptake of Particulate Matter 15-17
15.5.5 Effects of Particulate Matter on Physical Properties of Soils 15-18
15.6 Effects of Particulate Matter on Fauna 15-19
15.6.1 Laboratory Bioassays 15-19
15.6.2 Wildlife as Biomonitors of Particulate Matter 15-21
15.6.3 Biomagnification 15-22
15.7 Effects of Particulate Matter on Ecological Communities and Ecosystems 15-23
15.7.1 Gradient Effects near Smelters 15-24
15.7.2 Urban Environments 15-25
15.7.3 Aquatic Ecosystems 15-25
15.7.4 Experimental Microecosystems 15-26
15.8 Summary of Ecological Effects of Particulate Matter 15-26
APPENDIX 16 CASE STUDIES 16-1
Figure 16-1 Critical loads (CL) exceedance in Class I areas. 16-1
16.1 Northeastern U.S. Case Study: Acadia National Park, Hubbard Brook Experimental
Forest, and Bear Brook Watershed 16-2
16.1.1 Background 16-2
Figure 16-2 Locations of northeastern U.S. case study areas and nearby
human population centers. 16-3
Table 16-1 Selected characteristics of northeastern case study areas. 16-3
Figure 16-3 Site map of Hubbard Brook Experimental Forest in the White
Mountains of New Hampshire. 16-5
Table 16-2 Land use/land cover for northeastern case study areas. 16-7
Figure 16-4 Land cover in the Northeast case study region. 16-8
Table 16-3 Literature cited by Northeast U.S. case study area. 16-8
Table 16-4 Key recent research literature focused on the case study region. 16-12
16.1.2 Deposition 16-13
Figure 16-5 Total nitrogen deposition (A) and percentage of oxidized
nitrogen deposition (B) for the Northeast case study area
estimated by the National Atmospheric Deposition Program
Total Deposition Science committee. 16-14
Figure 16-6 Total sulfur deposition (A) for the Northeast case study area
estimated by the National Atmospheric Deposition Program
Total Deposition Science committee. Time series of wet
deposition (B) from the National Atmospheric Deposition
Program/National Trends Network in the Hubbard Brook
Experimental Forest, NH. 16-15
16.1.3 Critical Loads and Other Dose-Response Relationships 16-16
Table 16-5 Terrestrial empirical and modeling research on the response of
nitrogen and sulfur deposition for the northeastern U.S. 16-17
Table 16-6 Empirical critical loads for nitrogen in Acadia National Park, by
receptor, from Pardo et al. (2011c). 16-20
xxvi
-------
Figure 16-7
Table 16-7
Figure 16-8
Table 16-8
Table 16-9
Table 16-10
Annual stream calcium and magnesium export (paired bars),
and cumulative excess export in West Bear Brook compared
with East Bear Brook (line), over the study period 1989-2000 at
the Bear Brook Watershed experiment.
Critical loads of nutrient nitrogen for the Northern Forests
ecoregion.
Conceptual diagram showing the direct and indirect effects of
changes in temperature and precipitation on biogeochemical
processes in forests and on the services forests provide. Also
shown are feedbacks that further influence climatological
effects.
Aquatic empirical research on the response of nitrogen and
sulfur deposition for the northeastern U.S.
Critical and target load and exceedance modeling studies in the
northeastern U.S.
Empirical and modeled nitrogen critical loads applicable to the
northeastern U.S.
16.1.4 Long-Term Ecological Monitoring _
Table 16-11
Table 16-12
Table 16-13
Summary table of observed terrestrial and aquatic acidification
long-term trends in Hubbard Brook Experimental Forest and
Bear Brook Watershed.
Example surface water acidification chemistry studies in the
Northeast case study region.
Summary of observed terrestrial and aquatic nutrient
enrichment long-term responses in Acadia National Park,
Hubbard Brook Experimental Forest, Bear Brook Watershed,
and other northeastern regions.
16.1.5 Recovery _
16.2 Adirondack Case Study: Adirondack Region of New York
16.2.1
16.2.2
Background_
Figure 16-9
Deposition
Figure 16-10
Map of Adirondack park land cover.
Total nitrogen deposition (A) and percentage of oxidized
nitrogen deposition (B) for the Adirondack case study area
estimated by the National Atmospheric Deposition Program
Total Deposition Science committee.
16.2.3
Figure 16-11 Total sulfur deposition (A) for the Adirondack case study area
estimated by the National Atmospheric Deposition Program
Total Deposition Science committee. Time series of wet
deposition (B) from the National Atmospheric Deposition
Program/National Trends Network Whiteface Mountain, NY.
Critical Loads and Other Dose-Response Relationships
Table 16-14 Terrestrial empirical and modeling research on the response of
nitrogen and sulfur deposition for the Adirondack Mountains.
Number offish species per lake and acidity status, expressed
as acid neutralizing capacity, for Adirondack lakes.
Figure 16-12
Figure 16-13
Total macroinvertebrate species (community) richness as a
function of median pH in 36 streams sampled in the western
Adirondack Mountains of New York, 2003-2005; the four
standard (New York State) impact categories for species
richness are defined.
Figure 16-14
Table 16-15
Table 16-16
Species richness of biotic groups in 30 Adirondack study lakes
relative to midsummer epilimnetic pH during sample years.
Aquatic empirical research on the response of nitrogen and
sulfur deposition for the Adirondacks.
Critical and target load and exceedance modeling studies in
Adirondack Mountains.
16-22
16-23
16-25
16-26
16-30
16-32
16-33
16-34
16-37
16-40
16-45
16-46
16-46
16-48
16-49
16-50
16-51
16-52
16-52
16-56
16-58
16-59
16-60
16-62
xxvii
-------
16.3
16.2.4 Long-Term Monitoring
Figure 16-15 (A) Midsummer sulfate concentration in the epilimnion of
Brooktrout Lake (o) and in annual wet deposition (•) at local
National Atmospheric Deposition Program/National Trends
Network Station NY52 from 1984-2012. (B) Midsummer
epilimnetic concentrations of inorganic monomeric aluminum (~)
and hydrogen ion (¦) in Brooktrout Lake from 1984-2012.
(C) Midsummer phytoplankton (A) and total plankton
(phytoplankton, rotifers, crustaceans) (A) species richness in
Brooktrout Lake from 1984-2012.
Southeastern Appalachia Case Study
16.3.1
16-63
Background_
Table 16-17
Figure 16-16
Figure 16-17
16.3.2 Deposition
Figure 16-18
Figure 16-19
Figure 16-20
Species in the Southeast case study region that are listed as
threatened or endangered or as a species of concern.
Great Smoky Mountains National Park and nearby Class I
wilderness areas, with emphasis on water sampling locations
within GSMNP and critical loads for watersheds described in
Appendix 16.2.3.2.
Land cover in the southern Appalachian Mountains case study
region.
Deposition over Great Smoky Mountain National Park.
Trends in wet deposition of nitrogen and sulfur in Great Smoky
Mountain National Park, 1990-2014.
Total nitrogen deposition on left, total sulfur deposition on right,
for the 3-year average, 2011 -2013 in Great Smoky Mountain
National Park.
Modeled sulfur and nitrogen deposition to the Great Smoky
Mountain National Park for the year 2000.
16.3.3
16.3.4
Figure 16-21
Figure 16-22. Annual-averaged monthly footprint of reactive N deposition in
Great Smoky Mountain National Park (10.4 kg N/ha/yr), and pie
chart of fractional contribution from emission sectors, as
estimated by GEOS-Chem adjoint model.
Critical Loads and Other Dose-Response Relationships
Characterization and Long-Term Monitoring
Table 16-18 Example soil, terrestrial biota, and surface water acidification
characterization and long-term monitoring studies in the
southern Appalachian Mountains region.
16.4 Tampa Bay Case Study_
16.4.1 Background
Figure 16-23
Tampa Bay mainstem segments and watershed. Also shown
are the National Atmospheric Deposition (NADP) National
Trends Network wet deposition monitoring site FL41 at Verna
wellfield in Sarasota County and the NADP Atmospheric
Integrated Research Monitoring Network site FL18 in
Hillsborough County.
Figure 16-24
16.4.2 Deposition
Figure 16-25
Figure 16-26
Figure 16-27
Annual wet deposition of ammonium, nitrate, sulfate, and acidity
at the National Atmospheric Deposition National Trends
Network monitoring site closest to the Tampa Bay case study
area.
Estimated annual loads of total nitrogen from various sources to
Tampa Bay summarized from 1976 to 2011.
(A) Wet and dry nitrogen deposition in Tampa Bay and the
surrounding area. (B) Percentage of oxidized nitrogen
deposition in Tampa Bay and the surrounding area.
16-68
16-69
16-69
16-72
16-77
16-78
16-78
16-79
16-80
16-81
16-82
16-83
16-83
16-86
16-87
16-90
16-90
Tampa Bay overview map highlighting watershed development
and land use.
16-92
16-95
16-96
16-96
16-97
16-99
xxviii
-------
16.4.3
16.4.4
Long-Term Ecological Monitoring
Figure 16-28 Submerged aquatic vegetation cover loss in Tampa Bay.
Figure 16-29 Comparison of observed chlorophyll a and that predicted from
the total nitrogen load—chlorophyll a relationships for all four
mainstem Tampa Bay segments, for 1986-1998 and
1999-2007.
Figure 16-30 Total seagrass coverage in Tampa Bay circa 1950 through
2014.
Figure 16-31 Trend in hydrologically normalized total nitrogen load to Tampa
Bay relative to population increases in the Tampa Bay
metropolitan area.
Nitrogen Management
Table 16-19 Numeric nutrient criteria for chlorophyll a for the four mainstem
segments of Tampa Bay adopted by the Florida Department of
Environmental Protection.
Table 16-20 Numeric nutrient criteria for total nitrogen for the four mainstem
segments of Tampa Bay.
16.5 Rocky Mountain National Park Case Study
16.5.1 Background
Figure 16-32 Rocky Mountain National Park ecosystems.
Figure 16-33 Rocky Mountain National Park land coverage using the land
cover classifications as mapped by the National Land Cover
Dataset. Percentage of cover is shown for the four dominant
cover types.
Figure 16-34 Rocky Mountain National Park hydrologic unit code 12
watersheds.
16.5.2 Deposition
Figure 16-35
Total atmospheric nitrogen and sulfur deposition in the Rocky
Mountain National Park region based on TDEP calculations
averaged from 2011-2013 (see Appendix 2) and long-term
trends in wet atmospheric deposition from the Beaver Meadows
National Atmospheric Deposition Program Monitoring site within
Rocky Mountain National Park.
Figure 16-36
Figure 16-37
Figure 16-38
Figure 16-39
National Park from November 2008 to November 2009,
including organic nitrogen and particulate organic nitrogen.
Annual-averaged monthly footprint of reactive N deposition in
Rocky Mountain National Park(4.0 kg N/ha/yr), and pie chart of
fractional contribution from emission sectors, as estimated by
GEOS-Chem adjoint model.
16.5.3 Critical Loads and Other Dose-Response Relationships
Table 16-21
Table 16-22
Table 16-23
Table 16-24
Terrestrial empirical critical loads of nutrient nitrogen for the
Northwestern Forested Mountains ecoregion.
Hindcast absolute and percentage changes in species
abundance between 1900 and 2010 in response to historical
reconstructions of nitrogen deposition (Sullivan et al., 2005) and
historical climate change (IPCC, 2007b).
_ 16-99
16-101
16-102
16-104
16-105
16-106
16-106
16-107
16-109
16-109
16-110
16-113
16-114
16-115
Percentage of atmospheric nitrogen deposited in oxidized forms
and through wet deposition in the Rocky Mountain National Park
region based on TDEP calculations averaged from 2011-2013
(see Appendix 2).
Rocky Mountain National Park nitrogen cycle.
Montane forest and alpine ecosystem critical loads for nitrogen
deposition research published since the critical load assessment
by Pardo et al. (2011 c).
Critical loads for nitrogen for eutrophication for surface water
(high-elevation lakes) in the Rocky Mountains.
16-116
16-117
16-118
16-119
16-120
16-120
16-121
16-124
16-127
16-129
xxix
-------
Table 16-25
Table 16-26
Table 16-27
Lake water nitrate concentrations in nitrogen deposition studies
observing phytoplankton responses.
Paleolimnological biological responses in Rocky Mountain lakes
exposed to anthropogenic nitrogen deposition.
Critical loads of nitrogen or sulfur for surface water acidification
Rocky Mountain National Park and other high-elevation lakes in
the Rocky Mountains.
Figure 16-40
Figure 16-41
16.5.4
16.5.5
The continuum of ecological sensitivity to nitrogen deposition._
Critical load thresholds for current and possible future
biogeochemical and biological effects of nitrogen deposition.
Highlights of Additional Research Literature and Federal Reports since January
2008
Table 16-28
Summary of freshwater eutrophication studies in the Rocky
Mountains since 2008.
Rocky Mountain National Park Initiative
Figure 16-42 Rocky Mountain National Park Initiative glidepath and current
wet nitrogen deposition at Loch Vale in Rocky Mountain
National Park.
Figure 16-43
Rocky Mountain National Park Initiative accomplishment
timeline.
16.5.6 Interactions between Nitrogen Deposition, Precipitation, and Large-Scale
Ecological Disturbances
16.6 Southern and Central California
16.6.1 Background
Figure 16-44
Table 16-29
Figure 16-45
16.6.2 Deposition
Figure 16-46
Figure 16-47
Figure 16-48
Map of the distribution of vegetation types and land cover in
California.
Land coverages of Sequoia, Kings Canyons, and Joshua Tree
national parks.
Southern and central California case study region showing
locations of human population centers.
Patterns and temporal trends of nitrogen and sulfur deposition in
Joshua Tree National Park and surrounding region in California.
A and B show the 3-year average total deposition of nitrogen
and sulfur for 2011-2013
Patterns and temporal trends of nitrogen and sulfur deposition in
Joshua Tree National Park and surrounding region in California.
A shows the partitioning between oxidized and reduced
nitrogen; B and C show the 3-year average total percentage of
wet deposition of nitrogen and sulfur for 2011 -2013.
Figure 16-49
Figure 16-50
Patterns and temporal trends of nitrogen and sulfur deposition
of Sequoia and Kings Canyons national parks and surrounding
region in California. A and B show the 3-year average total
deposition of nitrogen and sulfur for 2011 -2013.
Patterns and temporal trends of nitrogen and sulfur deposition
of Sequoia and Kings Canyons national parks and surrounding
region in California. A shows the partitioning between oxidized
and reduced nitrogen, indicated as the fraction of total nitrogen
which is oxidized; B and C show the 3-year average total
percentage of wet deposition of nitrogen and sulfur for
2011-2013.
16-130
16-131
16-133
16-135
16-136
16-136
16-138
16-139
16-141
16-142
16-142
16-144
16-144
16-145
16-148
16-149
16-150
16-151
16-152
Patterns and temporal trends of nitrogen and sulfur deposition in
Joshua Tree National Park and surrounding region in California.
The 25-year time series for wet deposition of nitrate,
ammonium, sulfate, and hydrogen ion obtained from the
National Atmospheric Deposition Program/National Trends
Network monitoring site CA67.
16-153
16-154
16-155
XXX
-------
Figure 16-51 Patterns and temporal trends of nitrogen and sulfur deposition
of Sequoia and Kings Canyons and in Yosemite national parks
and surrounding regions in California. A and B show the 25-year
time series for wet deposition of nitrate, ammonium, sulfate, and
hydrogen obtained from the National Atmospheric Deposition
Program/National Trends Network monitoring sites CA99 and
CA75. 16-156
Figure 16-52 Annual-averaged monthly footprint of reactive N deposition in
Joshua Tree National Park (3.2 kg N/ha/yr) and Sequoia
National Park (5.7 kg N/ha/yr). Also shown for each park is a pie
chart of fractional contribution from emission sectors, as
estimated by GEOS-Chem adjoint model. 16-158
16.6.3 Critical Loads and Other Dose-Response Relationships 16-158
Figure 16-53 Composite critical load exceedance maps for all seven
vegetation types included in the study of Fenn et al. (2010)
showing the combined exceedance areas and the level of
exceedance (kg N/ha/yr). 16-161
Table 16-30 Summary of recent empirical dose-response and critical load
studies focused on the southern/central California case study
area and published since Pardo et al. (2011c). 16-162
Table 16-31 Terrestrial critical and target load and exceedance modeling
studies in southern/central California. 16-165
Table 16-32 Example surface water acidification studies in Sequoia and
Kings Canyons National Parks and other Sequoia and Kings
Canyons National Parks—relevant areas in the southern/central
California case study region. 16-168
16.6.4 Highlights of Additional Research Literature and Federal Reports since January
2008 16-173
Table 16-33 Key recent research literature focused on the case study region._ 16-173
16.6.5 Summary 16-175
Figure 16-54 Continuum of critical loads in southern/central California case
study area and relevant surrounding region. 16-176
REFERENCES R-1
xxxi
-------
AUTHORS, CONTRIBUTORS, AND REVIEWERS
Authors
Dr. Tara Greaver (Assessment Lead)—Center for Public Health and Environmental
Assessment, Office of Research and Development, U.S. Environmental Protection
Agency, Research Triangle Park, NC
Dr. Peter Byrley—Center for Public Health and Environmental Assessment, Office of
Research and Development, U.S. Environmental Protection Agency, Research Triangle
Park, NC
Ms. Marion Deerhake*—RTI International, Research Triangle Park, NC
Dr. Jean-Jacques Dubois—Center for Public Health and Environmental Assessment, Office
of Research and Development, U.S. Environmental Protection Agency, Research
Triangle Park, NC
Dr. Emmi Felker-Quinn—Air Resources Division, National Parks Service, Denver, CO
Dr. Jeffrey D. Herrick—Center for Public Health and Environmental Assessment, Office of
Research and Development, U.S. Environmental Protection Agency, Research Triangle
Park, NC
Dr. S. Douglas Kaylor—Center for Public Health and Environmental Assessment, Office of
Research and Development, U.S. Environmental Protection Agency, Research Triangle
Park, NC
Dr. Meredith Lassiter—Center for Public Health and Environmental Assessment, Office of
Research and Development, U.S. Environmental Protection Agency, Research Triangle
Park, NC
Dr. Stephen D LeDuc—Center for Public Health and Environmental Assessment, Office of
Research and Development, U.S. Environmental Protection Agency, Washington, DC
Dr. Stephen McDow—Center for Public Health and Environmental Assessment, Office of
Research and Development, U.S. Environmental Protection Agency, Research Triangle
Park, NC
Dr. Leigh Moorhead—Center for Public Health and Environmental Assessment, Office of
Research and Development, U.S. Environmental Protection Agency, Research Triangle
Park, NC
Dr. Jennifer Phelan*—RTI International, Research Triangle Park, NC
Dr. Robert Pinder—Health and Environmental Impacts Division, Office of Air and
Radiation, U.S. Environmental Protection Agency, Research Triangle Park, NC
Dr. Joseph P. Pinto—Center for Public Health and Environmental Assessment, Office of
Research and Development, U.S. Environmental Protection Agency, Research Triangle
Park, NC
Dr. Timothy J. Sullivan*—E&S Environmental Chemistry, Inc., Corvallis, OR
xxxii
-------
Dr. Alan Talhelm—Oak Ridge Institute for Science and Education, Center for Public Health
and Environmental Assessment, Office of Research and Development,
U.S. Environmental Protection Agency, Research Triangle Park, NC
* Under contract with RTI International
Contributors
Mr. Adam Benson—Oak Ridge Institute for Science and Education, Center for Public
Health and Environmental Assessment, Office of Research and Development,
U.S. Environmental Protection Agency, Research Triangle Park, NC
Dr. Mary C. Barber*—RTI International, Washington, DC
Ms. Tamara Blett—National Park Service, Lakewood, CO
Dr. Jana Compton—National Health and Environmental Effects Research Laboratory, Office
of Research and Development, U.S. Environmental Protection Agency, Corvallis, OR
Dr. Christopher M. Clark—Center for Public Health and Environmental Assessment, Office
of Research and Development, U.S. Environmental Protection Agency, Washington, DC
Dr. Linda Geiser—Air Resource Management National Program Leader, USDA Forest
Service, Washington, DC
Dr. Alan Knapp—Department of Biology, Colorado State University, Fort Collins CO
Dr. Jason Lynch—Office of Air and Radiation, Office of Atmospheric Programs, U.S.
Environmental Protection Agency, Washington, DC
Ms. April Maxwell—Oak Ridge Institute for Science and Education, Center for Public
Health and Environmental Assessment, Office of Research and Development,
U.S. Environmental Protection Agency, Research Triangle Park, NC
Ms. Margaret O'Neil*—RTI International, Research Triangle Park, NC
Ms. Jennifer Richkus*—RTI International, Research Triangle Park, NC
Ms. Elizabeth Sullivan*—RTI International, Research Triangle Park, NC
Dr. George L. Van Houtven*—RTI International, Research Triangle Park, NC
* Under contract with RTI International
Reviewers
Mr. Barry P. Baldigo, Water Science Center, U.S. Geological Survey, Troy, NY
Ms. Tamara Blett—National Park Service, Lakewood, CO
Dr. Michael Bell—National Park Service, Lakewood, CO
Dr. Micah Bennett—Center for Public Health and Environmental Assessment, Office of
Research and Development, U.S. Environmental Protection Agency, Washington, DC
Ms. Patricia Brewer—National Park Service, Lakewood, CO
Dr. Suzanne Bricker—National Oceanic and Atmospheric Administration, Washington, DC
Mr. Jim Cheatham—National Park Service, Lakewood, CO
xxxiii
-------
Dr. Christopher M. Clark—Center for Public Health and Environmental Assessment, Office
of Research and Development, U.S. Environmental Protection Agency, Washington, DC
Dr. Christopher B. Craft—Indiana University, Bloomington, IN
Dr. Eric Davidson—University of Maryland, Appalachian Laboratory, Frostburg, MD
Ms. Christine Davis—Office of Air and Radiation, Office of Air Quality Planning and
Standards, U.S. Environmental Protection Agency, Research Triangle Park, NC
Dr. Jean-Jacques Dubois—North Carolina State University, Raleigh, NC
Dr. Eric Edgerton—Atmospheric Research and Analysis, Cary, NC
Ms. Candace Edwards—Oak Ridge Institute for Science and Education, Center for Public
Health and Environmental Assessment, Office of Research and Development,
U.S. Environmental Protection Agency, Research Triangle Park, NC
Ms. Charlene Finley—Oak Ridge Institute for Science and Education, Center for Public
Health and Environmental Assessment, Office of Research and Development,
U.S. Environmental Protection Agency, Research Triangle Park, NC
Ms. Elizabeth Gatling—Oak Ridge Institute for Science and Education, Center for Public
Health and Environmental Assessment, Office of Research and Development,
U.S. Environmental Protection Agency, Research Triangle Park, NC
Dr. Cynthia Gilmour—Smithsonian Environmental Research Center, Edgewater, MD
Mr. William Grffin—Oak Ridge Institute for Science and Education, Center for Public
Health and Environmental Assessment, Office of Research and Development,
U.S. Environmental Protection Agency, Cincinnati, OH
Dr. David Grantz—College of Natural and Agricultural Sciences, Air Pollution Research
Center, University of California Riverside, Parlier, CA
Dr. Michael Griffith—Center for Public Health and Environmental Assessment, Office of
Research and Development, U.S. Environmental Protection Agency, Washington, DC
Dr. Scot Haggerthey—Center for Public Health and Environmental Assessment, Office of
Research and Development, U.S. Environmental Protection Agency, Washington, DC
Dr. Meredith K. Hastings—Earth, Environmental, and Planetary Sciences, Brown
University, Providence, RI
Dr. Laszlo Horvath—Plant Ecology Research Group, Szent Istvan University, Hungary
Dr. Robert Hetes—Office of Air and Radiation, Office of Air Quality Planning and
Standards, U.S. Environmental Protection Agency, Research Triangle Park, NC
Mr. James Hou—Office of the Regional Administrator, Region 8, U.S. Environmental
Protection Agency, Denver, CO
Dr. Bryan Hubbel—Office of Air Quality Planning and Standards, Office of Air and
Radiation, U.S. Environmental Protection Agency, Research Triangle Park, NC
Mr. Saturo Ito—Oak Ridge Institute for Science and Education, Center for Public Health and
Environmental Assessment, Office of Research and Development, U.S. Environmental
Protection Agency, Research Triangle Park, NC
xxxiv
-------
Mr. Ryan Jones—Center for Public Health and Environmental Assessment, Office of
Research and Development, U.S. Environmental Protection Agency, Research Triangle
Park, NC
Dr. James Kaldy—Office of Research and Development, U.S. Environmental Protection
Agency, Corvallis, OR
Mr. Matt Kulp—National Park Service, Gatlinburg, TN
Ms. Emily Lau—Oak Ridge Institute for Science and Education, Center for Public Health
and Environmental Assessment, Office of Research and Development,
U.S. Environmental Protection Agency, Research Triangle Park, NC
Dr. Sylvia Lee—Center for Public Health and Environmental Assessment, Office of
Research and Development, U.S. Environmental Protection Agency, Washington, DC
Dr. Jason Lynch—Office of Air and Radiation, Office of Atmospheric Programs, U.S.
Environmental Protection Agency, Washington, DC
Mr. Ihab Mikati—Oak Ridge Institute for Science and Education, Center for Public Health
and Environmental Assessment, Office of Research and Development,
U.S. Environmental Protection Agency, Research Triangle Park, NC
Ms. Danielle Moore—Oak Ridge Institute for Science and Education, Center for Public
Health and Environmental Assessment, Office of Research and Development,
U.S. Environmental Protection Agency, Research Triangle Park, NC
Ms. Kristi Morris—National Park Service, Lakewood, CO
Dr. Kristopher Novak—Center for Public Health and Environmental Assessment, Office of
Research and Development, U.S. Environmental Protection Agency, Research Triangle
Park, NC
Dr. Caroline Nowlan—Harvard-Smithsonian Center for Astrophysics, Harvard University,
Cambridge, MA
Mr. Kyle Painter—Oak Ridge Institute for Science and Education, Center for Public Health
and Environmental Assessment, Office of Research and Development,
U.S. Environmental Protection Agency, Research Triangle Park, NC
Ms. Rebecca Perrin—Office of the Regional Administrator, Region 8, U.S. Environmental
Protection Agency, Denver, CO
Mr. Jim Renfro—National Park Service, Gatlinburg, TN
Dr. Jennifer Richmond-Bryant—Center for Public Health and Environmental Assessment,
Office of Research and Development, U.S. Environmental Protection Agency, Research
Triangle Park, NC
Dr. Caroline Ridley—Center for Public Health and Environmental Assessment, Office of
Research and Development, U.S. Environmental Protection Agency, Washington, DC
Ms. Kristin Riha—Office of Air and Radiation, Office of Air Quality Planning and
Standards, U.S. Environmental Protection Agency, Research Triangle Park, NC
Ms. Alexandra Ross—Oak Ridge Institute for Science and Education, Center for Public
Health and Environmental Assessment, Office of Research and Development,
U.S. Environmental Protection Agency, Research Triangle Park, NC
Dr. Donald Scavia—University of Michigan, Ann Arbor, MI
xxxv
-------
Ms. Vicki Sandiford—Office of Air and Radiation, Office of Air Quality Planning and
Standards, U.S. Environmental Protection Agency, Research Triangle Park, NC
Dr. Erika Sasser—Office of Air and Radiation, Office of Air Quality Planning and
Standards, U.S. Environmental Protection Agency, Research Triangle Park, NC
Dr. Rich Scheffe—Office of Air Quality Planning and Standards, U.S. Environmental
Protection Agency, Research Triangle Park, NC
Dr. Bret Schichtel—National Park Service, Fort Collins, CO
Dr. Donna Schwede—Office of Research and Development, U.S. Environmental Protection
Agency, Washington, DC
Dr. J. Travis Smith—Office of Air and Radiation, Office of Air Quality Planning and
Standards, U.S. Environmental Protection Agency, Research Triangle Park, NC
Ms. Ginger Tennant—Office of Air and Radiation, Office of Air Quality Planning and
Standards, U.S. Environmental Protection Agency, Research Triangle Park, NC
Dr. Barbara J. Turpin—Gillings School of Global Environmental Health, University of
North Carolina, Chapel Hill, NC
Mr. Randy Waite—Office of Air and Radiation, Office of Air Quality Planning and
Standards, U.S. Environmental Protection Agency, Research Triangle Park, NC
Dr. John T. Walker—Office of Research and Development, U.S. Environmental Protection
Agency, Research Triangle Park, NC
Dr. Christopher Weaver—Center for Public Health and Environmental Assessment, Office
of Research and Development, U.S. Environmental Protection Agency, Washington, DC
Ms. Karen Wesson—Office of Air and Radiation, Office of Air Quality Planning and
Standards, U.S. Environmental Protection Agency, Research Triangle Park, NC
xxxvi
-------
CLEAN AIR SCIENTIFIC ADVISORY COMMITTEE
Chair of the Charter Clean Air Scientific Advisory Committee
Dr. Louis Anthony (Tony) Cox, Jr.—Cox Associates, Denver, CO
Charter Clean Air Scientific Advisory Committee Members
Dr. James Boylan—Georgia Department of Natural Resources, Atlanta, GA
Dr. Mark W. Frampton—University of Rochester Medical Center, Rochester, NY
Dr. Ronald J. Kendall—Texas Tech University, Lubbock, TX
Dr. Sabine Lange—Texas Commission on Environmental Quality, Austin, TX
Dr. Corey M. Masuca—Jefferson County Department of Health, Birmingham, AL
Dr. Steven C. Packham—Utah Department of Environmental Quality, Salt Lake City, UT
Chair, Oxides of Nitrogen, Oxides of Sulfur and Particulate
Matter—Ecological Criteria Review Panel
Dr. Ivan J. Fernandez**—Distinguished Maine Professor, School of Forest Resources and
Climate
Dr. Ronald J. Kendall***—Head of the Wildlife Toxicology Laboratory and Professor of
Environmental Toxicology, Texas Tech University
Oxides of Nitrogen, Oxides of Sulfur and Particulate
Matter—Ecological Criteria Review Panel Members
Dr. Edith Allen—Professor of Plant Ecology, Department of Botany and Plant Sciences,
University of California Riverside, Riverside, CA
Dr. Praveen Amar—Independent Consultant, Environment, Energy, and Climate Strategies,
Lexington, MA
Dr. James Boyd—Senior Fellow and Director, Center for the Management of Ecological
Wealth, Resources for the Future, Washington, DC
Dr. Elizabeth W. Boyer—Associate Professor of Water Resources, Department of
Ecosystem Science and Management, Pennsylvania State University, University Park, PA
Dr. Douglas Burns—Research Hydrologist, New York Water Science Center, U.S.
Geological Survey, Troy, NY
Ms. Lauraine Chestnut—Managing Economist, Stratus Consulting Inc., Boulder, CO
Dr. Charles T. Driscoll, Jr.—Distinguished Professor and University Professor of
Environmental Systems Engineering, Department of Civil and Environmental
Engineering, College of Engineering and Computer Science, Syracuse University,
Syracuse, NY
xxxvii
-------
Dr. Mark Fenn—Research Plant Pathologist, Pacific Southwest Research Station, USDA
Forest Service, Riverside, CA
Dr. James Galloway—Sidman P. Poole Professor of Environmental Sciences, Department of
Environmental Sciences, University of Virginia, Charlottesville, VA
Dr. Frank Gilliam, Professor—Department of Biological Sciences, Marshall University,
Huntington, WV
Dr. Robert A. Goldstein—Senior Technical Executive for Water and Ecosystems, Electric
Power Research Institute, Palo Alto, CA
Dr. Daven Henze—Assistant Professor and Charles C. Gates Faculty Fellow, Department of
Mechanical Engineering, University of Colorado, Boulder, CO
Dr. Robert W. Howarth—David R. Atkinson Professor of Ecology & Environmental
Biology, Department of Ecology and Evolutionary Biology, Cornell University, Ithaca,
NY
Dr. Donna Kenski—Data Analysis Director, Lake Michigan Air Directors Consortium,
Rosemont, IL
Dr. William McDowell—Professor of Environmental Science, Department of Natural
Resources and the Environment, University of New Hampshire, Durham, NH
Dr. Erik Nelson—Assistant Professor, Department of Economics, Bowdoin College,
Brunswick, ME
Dr. Hans Paerl—Kenan Professor of Marine and Environmental Sciences, Institute of
Marine Sciences, University of North Carolina—Chapel Hill, Morehead City, NC
Mr. Richard L. Poirot—Air Quality Planning Chief, Air Quality and Climate Division,
Vermont Department of Environmental Conservation, Montpelier, VT
Dr. Armistead (Ted) Russell—Howard T. Tellepsen Chair and Regents Professor of Civil
and Environmental Engineering, Department of Civil and Environmental Engineering,
Georgia Institute of Technology, Atlanta, GA
Dr. Stephen E. Schwartz—Senior Scientist, Environmental and Climate Sciences
Department, Brookhaven National Laboratory, Upton, NY
Dr. Kathleen Weathers—Senior Scientist, Cary Institute of Ecosystem Studies, Millbrook,
NY
Member of the statutory CASAC appointed by the U.S. EPA Administrator 2014-2018
*Member of the statutory CASAC appointed by the U.S. EPA Administrator 2019-2020
xxxviii
-------
ACRONYMS AND ABBREVIATIONS
Acronym/Abbreviation Meaning
ACCENT
Atmospheric Composition
Change: the European Network
of excellence
AIRMoN
Atmospheric Integrated
Research Monitoring Network
AIRS
Atmospheric Infrared Sounder
(instrument)
A1
aluminum
Al3+
aluminum ion
Ali
inorganic aluminum
Al2+
aluminum ion
Alo
organic aluminum
Al(OH)3
aluminum hydroxide
ALSC
Adirondack Lake Survey
Corporation
ALTM
Adirondack Long Term
Monitoring
AMD
acid mine drainage
ANC
acid neutralizing capacity
ANPP
aboveground net primary
production
AOD
aerosol optical depth
AOSR
Athabasca oil sands region
AQCD
Air Quality Criteria Document
AQEG
Air Quality Expert Group
AQI
Air Quality Index
AQS
Air Quality System (database)
Ar
argon
ARS
Agricultural Research Service
As
arsenic
ASI
Acid Stress Index
asl
above sea level
ATMOS
Atmospheric Trace Molecule
Spectroscopy
ATTILA
type of Lagrangian model
AUSPEX
Atmospheric Utility Signatures,
Predictions, and Experiments
AVIRIS
Airborne Visible and Infrared
Imaging Spectrometer
Acronym/Abbreviation Meaning
Ba barium
BBW Bear Brook Watershed
BBWM Bear Brook Watershed, Maine
Be or BC base cation
BCE exchangeable base cations
BCS base-cation surplus
BCw base cation weathering
BGC Biogeochemical (model)
B-IBI benthic index of biological
integrity
BMPs best management practices
BNF biological nitrogen fertilization
Br bromine
Br~ bromide ion
Bn molecular bromine
BrCl bromine chloride
BrO bromine monoxide
BUV Backscatter Ultraviolet
Spectrometer
BUVD Beneficial Use Values Database
C carbon; concentration
12C carbon-12, stable isotope of
carbon
13C carbon-13, stable isotope of
carbon
Ca ambient air concentration
Ca calcium
Ca2+ calcium ion
CAA Clean Air Act
CAAA Amendments to the Clean Air
Act
CAAAC Clean Air Act Advisory
Committee
CaCk calcium chloride
CaCC>3 calcium carbonate
CALIPSO Cloud-Aerosol Lidar and
Infrared Pathfinder Satellite
Observation (satellite)
Ca(NC>3)2 calcium nitrate
xxxix
-------
Acronym/Abbreviation
Meaning
Acronym/Abbreviation
Meaning
Ca(OH)2
calcium hydroxide
CINO2
nitryl chloride
CAPMoN
Canadian Air and Precipitation
Monitoring Network
CMAQ
Community Multiscale Air
Quality (modeling system)
CaS04-2H20
CASTNet
gypsum
Clean Air Status and Trends
Network
CMSA
CO
consolidated metropolitan
statistical area
carbon monoxide
CB4
Carbon Bond 4 (model)
CO2
carbon dioxide
Cd
cadmium
CO3-
carbonate
CEC
cation exchange capacity
CONUS
contiguous U.S.
CENTURY
model that simulates carbon,
nitrogen, phosphorus, sulfur, and
water dynamics in the soil-plant
system at monthly intervals over
time scales of centuries and
millennia
CPUE
CRREL
CS
catch per unit effort
U.S. Army Cold Regions
Research and Engineering
Laboratory
Consumer surplus
CFCs
chlorinated fluorocarbons
CS2
carbon disulfide
CG
cloud-to-ground (lightning flash)
CSS
coastal sage scrub (ecosystem)
Chi a
chlorophyll a
CTM
chemical transport model
CH4
methane
Cu
copper
C2H4
ethene
CV
contingent valuation
C2H6
ethane
CVM
contingent valuation method
C5Hs
isoprene
A, 5
delta, difference; change
CH3CHO
CH3C(0)
CH3C(0)00
CH2I2
acetaldehyde
acetyl radical
acetyl peroxy radical
diiodomethane
DayCent
DayCent-Chem
model for daily biogeochemistry
for forest, grassland, cropland,
and savanna systems
combination of DayCent-Chem
and PHREEQC models
CH-20
formaldehyde
DC
dichotomous choice
CH300H
methyl hydroperoxide
DDRP
Direct Delayed Response Project
CH3-S-CH3
dimethylsulfide, DMS
DDT
Damage Delay Time
CH3-S-H
(CH3)2SO
CH3S03H
CH3-S-S-CH3
Ci
CL
CI
methyl mercaptan
dimethyl sulfoxide, DMSO
methanesulfonic acid
dimethyl disulfide, DMDS
interstitial air concentration
critical load
chlorine
DECOMP
DEP
DIC
DIN
DMDS
decomposition model based on
soil-plant system dynamics
Department of Environmental
Protection
dissolved inorganic carbon
dissolved inorganic nitrogen
dimethyl disulfide,
CH3-S-S-CH3
cr
chloride ion
DMS
dimethyl sulfide, CH3-S-CH3
C12
molecular chlorine
DMSO
dimethylsulfoxide
CLaMS
type of Lagrangian model
DNDC
Denitrification-Decomposition
(model)
dissolved oxygen
CloudSat
NASA Earth observation
satellite
DO
-------
Acronym/Abbreviation Meaning
DOC dissolved organic carbon
DON dissolved organic nitrogen
EBB East Bear Brook
EC elemental carbon
EEAs Essential Ecological Attributes
ELA Experimental Lakes Area
ELS Eastern Lakes Survey
EMAP Environmental Monitoring and
Assessment Program
EMEFS Eulerian Model Evaluation Field
Study
EMEP Co-operative Programme for
Monitoring and Evaluation of
the Long-range Transmission of
Air Pollutants in Europe
EMF ectomycorrhizal fungi
EOS Earth Observation System
U.S. EPA U.S. Environmental Protection
Agency
eq equivalents
ecosystem respiration
EPT Ephemeroptera-Plecoptera-
Tricoptera (index)
ERP Episodic Response Project
ESA European Space Agency
EVRI Environmental Valuation
Reference Inventory
F flux
F~ fluoride ion
FAB First-order Acidity Balance
model
FACE free-air CO2 enrichment (studies)
Fe iron
FeP04 iron phosphate
FeS iron sulfide
F-factor fraction of the change in mineral
acid anions that is neutralized by
base cation release
FHM Forest Health Monitoring
FIA Forest Inventory and Analysis
(program)
FISH Fish in Sensitive Habitats
(project)
Acronym/Abbreviation
Meaning
FLEXPART
type of Lagrangian model
ForSAFE
three-component model using
nitrogen, carbon cycling, and
soil chemistry
FRM
Federal Reference Method
FTIR
Fourier Transform Infrared
Spectroscopy
FW2
black carbon soot
Fx
flux
YN2O5
uptake coefficient for N2O5 on
particles
GAW
Global Atmospheric Watch
(program)
GCE
Goddard Cumulus Ensemble
(model)
GDP
gross domestic product
GEOS
Goddard Earth Observing
System
GEOS-Chem
Goddard Earth Observing
System (with global chemistry
transport model)
GEOS-IDAS
Goddard Earth Observing
System Data Assimilation
System
GFED
Global Fire Emissions Database
GHG
greenhouse gas
GOES
Geostationary Operational
Environmental Satellites
GOME
Global Ozone Monitoring
Experiment
GPP
gross primary productivity
gs
stomatal conductance
GtC
gigaton carbon
Gton
gigaton
GWP
global warming potential
H
hydrogen; hydrogen atom
2H
hydrogen-2, deuterium, stable
isotope of hydrogen
H+
proton, hydrogen ion; relative
acidity
ha
hectare
HAPs
hazardous air pollutants
HBEF
Hubbard Brook Experimental
Forest
xli
-------
Acronym/Abbreviation Meaning
HBES
Hubbard Brook Ecosystem
Study
HBN
Hydrologic Benchmark Network
HC
hydrocarbon
HCHO
formaldehyde
HC1
hydrochloric acid
HC03~
bicarbonate
Hg
mercury
HNO2, HONO
nitrous acid
HNO3, HOONO
nitric acid
HNO4
pernitric acid
HO2
hydroperoxyl radical
H2O2
hydrogen peroxide
HO2NO2
peroxynitric acid
HOBr
hypobromous acid
HOC1
hypochlorous acid
HOX
hypohalous acid
HP
hedonic pricing
HSO3-
bisulfate ion
HSO4-
sulfuric acid ion
H2S
hydrogen sulfide
H2SO3
sulfurous acid
H2SO4
sulfuric acid
HUC-8s
8-digit Hydrologic Unit Codes
hv
energy of photon with
frequency v
I
iodine
I2
molecular iodine
IA
Integrated Assessment
IADN
Integrated Atmospheric
Monitoring Deposition Network
IC
intracloud (lightning flash)
ILWAS
Integrated Lake-Watershed
Acidification Study
IPC
International Cooperative
Programme
IEc
Industrial Economicsym
IIASA
International Institute for
Applied Systems Analysis
IMPROVE
Interagency Monitoring of
Protected Visual Environments
Acronym/Abbreviation Meaning
INOs
iodine nitrate
INTEX-NA
Intercontinental Chemical
Transport Experiment—North
America
IO
iodine oxide
IPCC
Intergovernmental Panel on
Climate Change
IPCC-AR4
Intergovernmental Panel on
Climate Change 4th Assessment
Report
IPCC-TAR
Intergovernmental Panel on
Climate Change 3rd Assessment
Report
IQR
interquartile range
IR
infrared
ISA
Integrated Science Assessment
J
flux from a leaf, deposition flux
(g/cm/second)
JK
Joyce Kilmer
JPL
Jet Propulsion Laboratory
JRGCE
Jasper Ridge Global Climate
Change Experiment
K
potassium
K+
potassium ion
Ka
dissociation constant
Kb
dissociation constant
KH
Henry's Law constant in M/atm
(M-atnT1)
kmol
kilomole
KNOs
potassium nitrate
Kw
ion product of water
LAF
Lake Acidification and Fisheries
LAR
leaf-area ratio
LB
laboratory bioassay
LC0.01
lethal concentration at which
0.01% of exposed animals die
LD33
lethal dose at which 33% of
exposed animals die
LDH
lactic acid dehydrogenase
LG
Linville Gorge
LIDAR
Light Detection and Ranging
(remote sensing system)
LIF
laser-induced fluorescence
-------
Acronym/Abbreviation Meaning Acronym/Abbreviation Meaning
LIMS
Limb Infrared Monitor of the
Stratosphere
MOZAIC
Measurement of Ozone and
Water Vapor by Airbus
LMCO
Lacasse-like multicopper
In-Service Aircraft
oxidase
MOZART
Model for Ozone and Related
LOD
limit of detection
Chemical Tracers
LP
long-path
MPAN
peroxymethacrylic nitrate
LRTAP
Long Range Transport of Air
Pollution
MPCA
Minnesota Pollution Control
Agency
LTER
Long-Term Ecological Research
MSA
metropolitan statistical area
(program)
Mt
million, or mega tons
LTM
Long-Term Monitoring (project)
N
nitrogen
M
air molecule
N, n
number of observations
MA
Millennium Ecosystem
Assessment
14N
nitrogen-14, stable isotope of
nitrogen
MAGIC
Model of Acidification of
15N
nitrogen-15, stable isotope of
Groundwater in Catchments
nitrogen
molecular nitrogen; nonreactive
(model)
N2
MAHA
Mid-Atlantic Highlands
nitrogen
Assessment of streams
N14C
plant soil N and C cycling model
MAQSIP
Multiscale Air Quality
Simulation Platform (model)
NA
not available; insufficient data
MAT
moist acidic tundra
Na
sodium
MAX-DOAS
multiple axis differential optical
Na+
sodium ion
absorption spectroscopy
NAAQS
National Ambient Air Quality
MBC
microbial biomass carbon
Standards
MBL
marine boundary layer
NaCl
sodium chloride
MDN
Mercury Deposition Network
NADP
National Atmospheric
MeHg
Methylmercury
Deposition Program
MEM
model ensemble mean
Na2MoC>4
sodium molybdate
Microequivalent
NAMS
National Air Monitoring Stations
Heq
Mg
Magnesium
NANI
Net anthropogenic nitrogen
inputs
Mg2+
magnesium ion
NAPAP
National Acid Precipitation
MIMS
membrane inlet mass
Assessment Program
spectrometry
NASQAN
National Stream Quality
MM5
National Center for Atmospheric
Accounting Network
Research/Penn State Mesoscale
NARSTO
program formerly known as
Model, version 5
North American Regional
Mn
Manganese
Strategy for Atmospheric Ozone
MOBILE6
Highway Vehicle Emission
NAS
National Academy of Sciences
Factor Model
NASA
National Aeronautics and Space
MODIS
Moderate Resolution Imaging
Administration
Spectroradiometer
Na2S04
sodium sulfate
MOPITT
Measurement of Pollution in the
NASQAN
National Stream Quality
Troposphere (satellite
Accounting Network
instrument)
xliii
-------
Acronym/Abbreviation Meaning
NATTS
National Air Toxics Trends
(network)
NAWQA
National Water Quality
Assessment (program)
NCore
National Core Monitoring
Network
N-dep
nitrogen deposition
NEE
net ecosystem exchange
NEG/ECP
New England Governors and
Eastern Canadian Premiers
NEI
National Emissions Inventory
NEON
National Ecological Observatory
Network
NEP
net ecosystem productivity
N-fert
nitrogen-fertilization
N-fix
nitrogen-fixing vegetation
NFI
net factor income
NH3
ammonia
nh2
amino (chemical group)
nh4+
ammonium ion
NH4CI
ammonium chloride
NH4NO3
ammonium nitrate
(NH4)2S04
ammonium sulfate
NHx
category label for NH3 plus
NH4+
NHy
total reduced nitrogen
Ni
nickel
NILU
Norwegian Institute for Air
Research
NITREX
Nitrogen saturation Experiments
nitro-PAH
nitro-polycyclic aromatic
hydrocarbon
NLCD
National Land Cover Data
Nmin
nitrogen mineralization
NMOC
nonmethane organic compound
NO
nitric oxide
NO2
nitrogen dioxide
no2~
nitrite
MV
nitrate
N2O
nitrous oxide
N2O5
dinitrogen pentoxide
Acronym/Abbreviation Meaning
NOAA
U.S. National Oceanic and
Atmospheric Administration
NOAA-ARL
U.S. National Oceanic and
Atmospheric Administration Air
Resources Laboratory
NOAEL
no-observed-adverse-effect level
NOEC
no-observed-effect concentration
NOx
sum of NO and NO2
NOy
sum of NOx and NOz; odd
nitrogen species; total oxidized
nitrogen
NOz
sum of all inorganic and organic
reaction products of NOx
(HONO, HNO3, HNO4, organic
nitrates, particulate nitrate,
nitro-PAHs, etc.)
NPOESS
National Polar-orbiting
Operational Environmental
Satellite System
NPP
net primary production
NPS
National Park Service
Nr
reactive nitrogen
NRC
National Research Council
NS orn.s.
nonsignificant
NSF
National Science Foundation
NSS
National Stream Survey
Nss
nonsea salt
NSTC
National Science and
Technology Council
NSWS
National Surface Water Survey
NTN
National Trends Network
NuCM
nutrient cycling model
O2
molecular oxygen
O3
ozone
16o
oxygen-16, stable isotope of
oxygen
18o
oxygen-18, stable isotope of
oxygen
19Q
oxygen-19, radioactive isotope
of oxygen
OC
organic carbon
O-CN
terrestrial biosphere model
OCO
Orbiting Carbon Observatory
OCS
carbonyl sulfide
xliv
-------
Acronym/Abbreviation
Meaning
O('D)
electronically excited oxygen
atom
OH
hydroxyl radical
OM
organic matter
OMI
Ozone Monitoring Instrument
0(3P)
ground-state oxygen atom
P
phosphorus
P,P
probability value
Pi
1st percentile
P5
5th percentile
P95
95th percentile
P99
99th percentile
PAHs
polycyclic aromatic
hydrocarbons
PAMS
Photochemical Assessment
Monitoring Stations
PAN
peroxyacetyl nitrate
PANs
peroxyacyl nitrates
PARASOL
Polarization and Anisotropy of
Reflectances for Atmospheric
Sciences coupled with
Observations from a Lidar
(satellite instrument)
Pb
lead
PBL
planetary boundary layer
PC
payment card
PCBs
polychlorinated biphenyl
compounds
PH
relative acidity
P(HN03)
production of nitric acid
PHREEQC
model for soil and water
geochemical equilibrium
PIRLA
Paleoecological Investigation ol
Recent Lake Acidification
(projects)
pKa
dissociation constant
PM
particulate matter
PM2.5
particulate matter with
aerodynamic diameter of #2.5
(im
PM10
particulate matter with
aerodynamic diameter #10 (im
Acronym/Abbreviation Meaning
PMlO-2.5
particulate matter with
aerodynamic diameter between
10 and 2.5 (im
PM-CAMx
Comprehensive Air Quality
Model with extensions and with
particulate matter chemistry
PnET
Photosynthesis and
EvapoTranspiration (model)
PnET-BGC
Photosynthesis and
E vapoT ranspiration-
Biogeochemical (model)
PnET-CN
Photosynthesis and
EvapoTranspiration model of C,
water, and N balances
PnET-N-DNDC
Photosynthesis and
EvapoTranspiration-
Denitrification-Decomposition
(model)
pN03"
particulate nitrate
P(03)
production of O3
PO,f, PO43-
phosphate
POPs
persistent organic pollutants
PPb
parts per billion
PPN
peroxypropionyl nitrate
ppt
parts per trillion
PRB
policy relevant background
PRE-STORM
Preliminary Regional
Experiment for STORM
PROFILE
model using soil mineralogy as
input
PS
producer surplus
pS042~
particulate sulfate
P(so42-)
production of sulfate
Q
flow rate; discharge
Q10
temperature coefficient
QAPP
Quality Assurance Project Plan
R
generic organic group attached
to a molecule
R2
coefficient of determination
r2
correlation coefficient
Ra
aerodynamic resistance
Raboveground
aboveground respiration
Rautotrophic
soil autotrophic respiration
Rb
boundary layer resistance
-------
Acronym/Abbreviation Meaning
Rc internal resistance
RADM Regional Acid Deposition Model
RAMS Regional Atmospheric Modeling
System
RAPS Regional Air Pollution Study
RCOO-s strongly acidic organic anions
RC(0)00 organic peroxy radical
RDT Recovery Delay Time
REMAP Regional Environmental
Monitoring and Assessment
Program
RH relative humidity
RLTM Regional Long-Term Monitoring
Rmicrobiai microbial respiration
RMCC Research and Monitoring
Coordinating Committee
RMSE root mean squared error
RO2 organic peroxyl; organic peroxy
RONO2 organic nitrate
RO2NO2 peroxynitrate
ROS rain on snow
RP revealed preferences
RRx lognormal-transformed response
ratio
response ratio
^soil total soil respiration
RuBisCO ribulose-1,5-bisphosphate
carboxylase/oxygenase
s second
S sulfur
32S sulfur-32, stable isotope of sulfur
34S sulfur-34, stable isotope of sulfur
35S sulfur-35, radioactive isotope of
sulfur
SAA sum of mineral acid anion
concentrations
SAFE Soil Acidification in Forest
Ecosystems (model)
SAMAB Southern Appalachian Man and
the Biosphere (program)
SAMI Southern Appalachian
Mountains Initiative
Acronym/Abbreviation Meaning
SAO Smithsonian Astrophysical
Observatory
SAPRAC Statewide Air Pollution Research
Center
SBC sum of base cation
concentrations
SBUV Solar Backscatter Ultraviolet
Spectrometer
SC safe concentration
SCAQS Southern California Air Quality
Study
SCIAMACE1Y Scanning Imaging Absorption
Spectrometer for Atmospheric
Cartography
Se selenium; standard error
SEARCH Southeastern Aerosol Research
and Characterization Study
(monitoring program)
Si silicon
SIP State Implementation Plan
SJAQS San Joaquin Valley Air Quality
Study
SLA specific leaf area
SLAMS State and Local Air Monitoring
Stations
SMART Simulation Model for
Acidification's Regional Trends
(model)
SMB Simple Mass Balance (model)
SMBE steady-state mass-balance
equations
SO sulfur monoxide
502 sulfur dioxide
503 sulfur trioxide
S032~ sulfite
S042~ sulfate ion
S2O disulfur monoxide
SOM soil organic matter
SONEX Subsonics Assessment Ozone
and Nitrogen Oxides Experiment
SOS Southern Oxidant Study
SOS/T State of Science/Technology
(report)
SOx sulfur oxides
xlvi
-------
Acronym/Abbreviation Meaning
SP
stated preferences
SPARROW
SPAtially Referenced
Regressions on Watershed
SR
Attributes (model)
Shining Rock
Sr
strontium
86 Sr
strontium-86, stable isotope of
strontium
87Sr
strontium-87, stable isotope of
strontium
SRB
sulfate-reducing bacteria
SRP
soluble reactive phosphorus
SSWC
Steady State Water Chemistry
(model)
STA
Soil Texture Approximations
(model)
STE
strata spheric-tropo spheric
exchange
STN
Speciation Trends Network
SUM06
seasonal sum of all hourly
average concentrations
>0.06 ppm
SVOC
semivolatile organic compound
SWAS
Shenandoah Watershed Study
T, T
tau, atmospheric lifetime
T
time; duration of exposure
TAF
Tracking and Analysis
Framework (model)
Tair
air temperature
TAMM
Timber Assessment Market
Model
TAR
Third Assessment Report
TC
total carbon; travel cost
TCM
travel cost method
TDLAS
Tunable Diode Laser Absorption
Spectrometer
Tg
teragram
TIME
Temporally Integrated
Monitoring of Ecosystems
(program)
TN
total nitrogen
TOC
total organic carbon
TOMS
Total Ozone Mapping
Spectrometer
Acronym/Abbreviation Meaning
TOR
tropospheric ozone residual
TP
total phosphorus
TRACE-P
Transport and Chemical
Evolution over the Pacific
TSI
timber-stand improvement
TSS
total suspended solids
T water
water temperature
UAN
urea and ammonium nitrate
fertilizer
UMD-CTM
University of Maryland
Chemical Transport Model
UNECE
United Nations Economic
Commission for Europe
USDA
U.S. Department of Agriculture
USFS
U.S. Forest Service
USGS
U.S. Geological Survey
UV
ultraviolet
UV-A
ultraviolet radiation of
wavelengths from 320 to 400 nm
UV-B
ultraviolet radiation of
wavelengths from 280 to 320 nm
Vd
deposition rate, deposition
velocity (cm/s)
voc
volatile organic compound
VSD
Very Simple Dynamic (soil
acidification model)
VTSSS
Virginia Trout Stream
Sensitivity Study
WARMS
Waterfowl Acidification
Response Modeling System
WATERSN
Watershed Assessment Tool for
Evaluating Reduction Scenarios
for Nitrogen
WBB
West Bear Brook
WEBB
Water, Energy, and
Biogeochemical Budgets
WFPS
water-filled pore space
WGE
Working Group on Effects
WLS
Western Lakes Survey
WMO
World Meteorological
Organization
WMP
Watershed Manipulation Project
WSA
Wadeable Stream Assessment
(survey)
xlvii
-------
Acronym/Abbreviation
Meaning
wt%
percent by weight
WTA
willingness-to-accept
WTP
willingness-to-pay
XNOs
nitrate halogen-X salt
XO
halogen-X oxide
yr
year
Zn
zinc
ZnO
zinc oxide
xlviii
-------
PREFACE
Legislative Requirements for the Review of the National Ambient
Air Quality Standards
Two sections of the Clean Air Act (CAA) govern the establishment, review, and revision
of the National Ambient Air Quality Standards (NAAQS). Section 108 [42 U.S. Code
(U.S.C.) 7408] directs the Administrator to identify and list certain air pollutants and then
to issue air quality criteria for those pollutants. The Administrator is to list those air
pollutants that in their "judgment, cause or contribute to air pollution which may
reasonably be anticipated to endanger public health or welfare," "the presence of which
in the ambient air results from numerous or diverse mobile or stationary sources," and
"for which ... [the Administrator] plans to issue air quality criteria ..." [42 U.S.C.
7408(a)(1); ]. Air quality criteria are intended to "accurately reflect the latest scientific
knowledge useful in indicating the kind and extent of all identifiable effects on public
health or welfare, which may be expected from the presence of [a] pollutant in the
ambient air ..." [42 U.S.C. 7408(b)], Section 109 [42 U.S.C. 7409; (CAA. 1990b)l
directs the Administrator to propose and promulgate "primary" and "secondary" NAAQS
for pollutants for which air quality criteria are issued. Section 109(b)( 1) defines a primary
standard as one "the attainment and maintenance of which in the judgment of the
Administrator, based on such criteria and allowing an adequate margin of safety, are
requisite to protect the public health."1 A secondary standard, as defined in
Section 109(b)(2), must "specify a level of air quality the attainment and maintenance of
which, in the judgment of the Administrator, based on such criteria, is requisite to protect
the public welfare from any known or anticipated adverse effects associated with the
presence of [the] air pollutant in the ambient air."2
In setting standards that are "requisite" to protect public health and welfare as provided in
Section 109(b), the U.S. EPA's task is to establish standards that are neither more nor less
stringent than necessary for these purposes. In so doing, the U.S. EPA may not consider
the costs of implementing the standards.3 Likewise, "[a]ttainability and technological
1 The legislative history of Section 109 indicates that a primary standard is to be set at"... the maximum permissible
ambient air level ... which will protect the health of any [sensitive] group of the population," and that for this
purpose "reference should be made to a representative sample of persons comprising the sensitive group rather than
to a single person in such a group" S. Rep. No. 91:1196, 91st Cong., 2d Sess. 10 (1970).
2 Section 302(h) of the Act [42 U.S.C. 7602(h)] provides that all language referring to effects on welfare includes,
but is not limited to, "effects on soils, water, crops, vegetation, man-made materials, animals, wildlife, weather,
visibility and climate, damage to and deterioration of property, and hazards to transportation, as well as effects on
economic values and on personal comfort and well-being ..." (CAA. 2005).
3 See generally. Whitman v. American Trucking Associations, 531 U.S. 457, 465-472, 475-476 (2001).
xlix
-------
feasibility are not relevant considerations in the promulgation of national ambient air
quality standards."1
Section 109(d)(1) requires that "not laterthan December 31, 1980, and at 5-year intervals
thereafter, the Administrator shall complete a thorough review of the criteria published
under Section 108 and the national ambient air quality standards ... and shall make such
revisions in such criteria and standards and promulgate such new standards as may be
appropriate Section 109(d)(2) requires that an independent scientific review
committee "shall complete a review of the criteria ... and the national primary and
secondary ambient air quality standards ... and shall recommend to the Administrator any
new ... standards and revisions of existing criteria and standards as may be
appropriate ...Since the early 1980s, this independent review function has been
performed by the Clean Air Scientific Advisory Committee (CASAC).2
Overview and History of the Reviews of the Secondary National
Ambient Air Quality Standards for Nitrogen Dioxide, Sulfur
Dioxide, and Particulate Matter
NAAQS are defined by four basic elements: indicator, averaging time, level, and form.
The indicator defines the pollutant to be measured in the ambient air for the purpose of
determining compliance with the standard. The averaging time defines the time period
over which air quality measurements are to be obtained and averaged or cumulated,
considering evidence of effects associated with various time periods of exposure. The
level of a standard defines the air quality concentration used (i.e., an ambient
concentration of the indicator pollutant) in determining whether the standard is achieved.
The form of the standard defines the air quality statistic that is compared to the level of
the standard in determining whether an area attains the standard. The Administrator
considers these four elements collectively in evaluating the protection to public health
provided by the primary NAAQS.
Nitrogen Dioxide Secondary National Ambient Air Quality Standards
The first air quality criteria and standards for oxides of nitrogen were issued in 1971
rOJ.S. EPA. 1971). 36 FR 8186], Both the primary and secondary standards were set at
0.053 parts per million (ppm), as an annual arithmetic mean (36 FR 8186). In 1982, the
U.S. EPA published Air Quality Criteria for Oxides of Nitrogen (U.S. EPA. 1982a).
1 See American Petroleum Institute v. Costle, 665 F. 2d at 1185.
2 Lists of chartered CASAC members and of members of the CASAC Panels are available at:
http://vosemite.epa.gov/sab/sabproduct.nsf/WebCASAC/CommitteesandMembersliip70penDocument.
1
-------
which updated the scientific criteria upon which the initial standards were based. On
February 23, 1984, the U.S. EPA proposed to retain these standards (49 FR 6866). After
taking into account public comments, the U.S. EPA published the final decision to retain
the existing standards on June 19, 1985 (50 FR 25532).
In November 1991, the U.S. EPA initiated another review and released an updated draft
air quality criteria document (AQCD) for review and comment by CASAC and the public
(56 FR 59285). The final AQCD was released later in 1993 (U.S. EPA. 1993). Staff of
the Office of Air Quality Planning and Standards (OAQPS) prepared a draft Staff Paper
that summarized and integrated the key studies and scientific evidence contained in the
revised air quality criteria document and identified the critical elements to be considered
in the review of the NO2 NAAQS. The Staff Paper was reviewed by the CASAC and the
public in December 1994, and in September 1995, the U.S. EPA finalized the Staff Paper
(U.S. EPA. 1995b). On October 2, 1995, the Administrator announced her proposed
decision not to revise either the primary or secondary NAAQS for NO2 based on the
information available in this review (60 FR 52874; October 11, 1995). After
consideration of public comments, the Administrator made a final determination that
revisions to neither the primary nor the secondary NAAQS for NO2 were appropriate at
that time (61 FR 52852; October 8, 1996).
The most recent review of the secondary NAAQS standards for oxides of nitrogen was
performed jointly with a review of the secondary NAAQS for oxides of sulfur beginning
in 2005 (described below).
Sulfur Dioxide Secondary National Ambient Air Quality Standards
Based on the 1969 sulfur oxides criteria document (HEW. 1969). the U.S. EPA
promulgated the initial primary and secondary NAAQS for SO2 on April 30, 1971 (36 FR
8186). The secondary standards were 0.02 ppm as an annual arithmetic mean and
0.5 ppm as a maximum 3-hour, not to be exceeded more than once per year. These
secondary standards were established on the basis of vegetation effects evidence
described in the 1970 criteria document. Based on additional data available in 1973,
revisions were made to Chapter 5 "Effects of Sulfur Oxide in the Atmosphere on
Vegetation" of the Air Quality Criteria for Sulfur Oxides (U.S. EPA. 1973). which led the
U.S. EPA to propose (38 FR 11355) and then finalize a revocation of the annual mean
secondary standard (38 FR 25678). At that time, the U.S. EPA additionally considered
welfare effects related to effects on materials, visibility, soils, and water. However, the
U.S. EPA concluded that either protection from such effects was afforded by the primary
li
-------
standard or that sufficient data were not then available to develop criteria for standards
based on these effects (38 FR 25680).
In 1980, the U.S. EPA released a combined AQCD for sulfur oxides and particulate
matter for CASAC review. Following its review of a draft revised criteria document in
August 1980, the CASAC concluded that acidic deposition was a topic of extreme
scientific complexity, noting that a fundamental problem of addressing acid deposition in
a criteria document is that acidic deposition is produced by several pollutants, including
oxides of sulfur, oxides of nitrogen, and the fine particulate fraction of suspended
particles RTJ.S. EPA. 1982b). pp. 125-126], Following CASAC closure on the criteria
document in December 1981, the U.S. EPA released a final AQCD (U.S. EPA. 1982b).
and the OAQPS prepared a Staff Paper that was released in November 1982 (U.S. EPA.
1982c). The issue of acidic deposition was not, however, assessed directly in the OAQPS
Staff Paper because the U.S. EPA followed the guidance given by CASAC.
In response to CASAC recommendations for a separate comprehensive discussion of
acidic deposition as part of the criteria documents, the U.S. EPA subsequently prepared
the following documents: The Acidic Deposition Phenomenon and Its Effects: Critical-
Assessment Review Papers, Volumes I andII (U.S. EPA. 1984a. b) and The Acidic
Deposition Phenomenon and Its Effects: Critical Assessment Document r(Bennett et al..
1985); 53 FR 14935-14936], Although these documents were not considered criteria
documents and had not undergone CASAC review, they represented the most
comprehensive summary of relevant scientific information completed by the U.S. EPA at
that point (58 FR 21355).
At about the same time in 1980 as the CASAC recommendation for a comprehensive
assessment of acidic deposition, Congress created the National Acid Precipitation
Assessment Program (NAPAP). During the 10-year course of this program, a series of
reports were issued and a final report was issued in 1990 (NAPAP. 1991).
On April 26, 1988, the U.S. EPA proposed not to revise the existing primary and
secondary standards. This proposal regarding the secondary SO2 NAAQS was due to the
Administrators conclusions that (1) based upon the then-current scientific understanding
of the acidic deposition problem, it would be premature and unwise to prescribe any
regulatory control program at that time and (2) when the fundamental scientific
uncertainties had been reduced through ongoing research efforts, the U.S. EPA would
draft and support an appropriate set of control measures (53 FR 14926). Subsequent to
proposal, Congress took up consideration of acidic deposition.
On November 15, 1990, Amendments to the CAA were passed by Congress and signed
into law by the President. In Title IV of these Amendments, Congress included a
lii
-------
statement of findings that had led them to take this action, including that: ""(1) the
presence of acidic compounds and their precursors in the atmosphere and in deposition
from the atmosphere represents a threat to natural resources, ecosystems, materials,
visibility, and public health; (2) the problem of acid deposition is of national and
international significance; and that (3) current and future generations of Americans will
be adversely affected by delaying measures to remedy the problem..." The goal of
Title IV was to reduce emissions of SO2 by 10 million tons and oxides of nitrogen
emissions by 2 million tons from 1980 emission levels in order to achieve reductions over
broad geographic regions/areas. In envisioning that further action might be necessary in
the long term, Congress included Section 404 of the 1990 Amendments. This section
requires the U.S. EPA to conduct a study on the feasibility and effectiveness of an acid
deposition standard or standards to protect "sensitive and critically sensitive aquatic and
terrestrial resources" and at the conclusion of the study, submit a report to Congress. Five
years later, the U.S. EPA submitted to Congress its report titled Acid Deposition Standard
Feasibility Study: Report to Congress (U.S. EPA. 1995a) in fulfillment of this
requirement. The Report to Congress concluded that establishing acid deposition
standards for sulfur and nitrogen deposition might at some point in the future be
technically feasible although appropriate deposition loads for these acidifying chemicals
could not be defined with reasonable certainty at that time.
The 1990 Amendments also added new language to sections of the CAA that pertain to
the scope or application of the secondary NAAQS designed to protect the public welfare.
Section 108 (g) specified that "the Administrator may assess the risks to ecosystems from
exposure to criteria air pollutants (as identified by the Administrator in the
Administrator's sole discretion)/' The definition of public welfare in Section 302 (h) was
expanded to state that the welfare effects identified should be protected from adverse
effects associated with criteria air pollutants ".. .whether caused by transformation,
conversion, or combination with other air pollutants."
In response to these legislative initiatives, the U.S. EPA and other federal agencies
continued research on the causes and effects of acidic deposition and related welfare
effects of SO2 and implemented an enhanced monitoring program to track progress
(58 FR 21357). In 1993, the U.S. EPA announced a decision not to revise the secondary
standard, concluding that revision to address acidic deposition and related SO2 welfare
effects was not appropriate at that time (58 FR 21351). In reaching this decision, the
U.S. EPA took into account the significant reductions in SO2 emissions, ambient SO2
concentrations and ultimately deposition expected to result from implementation of the
Title IV program, which was expected to significantly decrease the acidification of water
bodies and damage to forest ecosystems and to permit much of the existing damage to be
reversed with time (58 FR 21357). While recognizing that further action might be needed
liii
-------
to address acidic deposition in the longer term, the U.S. EPA judged it prudent to await
the results of the studies and research programs then underway, including those assessing
the comparative merits of secondary standards, acidic deposition standards, and other
approaches to control of acidic deposition and related effects, and then to determine
whether additional control measures should be adopted or recommended to Congress
(58 FR 21358).
In 2000, the U.S. EPA announced receipt of two items related to acidic deposition and the
NAAQS (65 FR 48699). The first was a petition submitted to the U.S. EPA in 1999 by
representatives of seven northeastern states for the promulgation of revised secondary
NAAQS for the criteria pollutants associated with the formation of acid rain (including
NO2, SO2, and fine particulate matter [PM2 5]). The petition states that the language in
Section 302(h) of the CAA "clearly references the transformation of pollutants resulting
in the inevitable formation of sulfate and nitrate aerosols and/or their ultimate
environmental impacts as wet and dry deposition, clearly signaling Congressional intent
that the welfare damage occasioned by sulfur and nitrogen oxides be addressed through
the secondary standard provisions of Section 109 of the Act." The petition further stated
that "recent federal studies, including the NAPAP Biennial Report to Congress: An
Integrated Assessment, document the continued—and increasing—damage being
inflicted by acid deposition to the lakes and forests of New York, New England, and
other parts of our nation, demonstrating that the Title IV program had proven
insufficient." The petition also listed other adverse welfare effects associated with the
transformation of these criteria pollutants, including visibility impairment, eutrophication
of coastal estuaries, global warming, tropospheric ozone, and stratospheric ozone
depletion.
The second item was a related request from the U.S. Department of Interior (DOI) that
the U.S. EPA address many of the same adverse environmental effects associated with
the same types of air pollutants and with ozone that the DOI asserted were occurring in
national parks and wilderness areas (65 FR 48699). Included among the effects of
concern identified in the request were acidification of streams, surface waters and/or
soils, eutrophication of coastal waters, visibility impairment, and foliar injury from ozone
(65 FR 48701). The U.S. EPA requested comment on the issues raised by these requests,
stating that it would consider any relevant comments and information submitted, along
with the information provided by the petitioners and DOI, before making any decision
concerning a response to these requests for rulemaking, which if commenced would
include opportunity for public review and comment (65 FR 48701).
liv
-------
Particulate Matter Secondary National Ambient Air Quality Standards
The U.S. EPA first established NAAQS for particulate matter (PM) in 1971 (36 FR 8186,
April 30, 1971) based on the original AQCD (NAPCA. 1969). The AQCD assessed the
evidence for a variety of PM-associated welfare effects, including visibility impairment
and materials damage (e.g., soiling, corrosion). Based on evidence for such effects, the
secondary standards were set at 150 (ig/m3 for the 24-hour average not to be exceeded
more than once per year and 60 (ig/m3 for the annual geometric mean. The federal
reference method (FRM) specified for determining attainment of the original standards
was the high-volume sampler, which collects PM up to a nominal size of 25 to
45 micrometers (|im: referred to as total suspended particulates or TSP).
In October 1979 (44 FR 56730; October 2, 1979), the U.S. EPA announced the first
periodic review of the air quality criteria and NAAQS for PM. Revised primary and
secondary standards were promulgated in 1987 (52 FR 24634; July 1, 1987). In the 1987
decision, the U.S. EPA changed the indicator for particles from TSP to PMi0 to focus on
the subset of inhalable particles small enough to penetrate to the thoracic region of the
respiratory tract (including the tracheobronchial and alveolar regions) referred to as
thoracic particles.1 The level of the 24-hour standards (primary and secondary) was set at
150 |ig/nr\ and the form was one expected exceedance per year, on average, over 3 years.
The level of the annual standards (primary and secondary) was set at 50 (ig/m3, and the
form was annual arithmetic mean averaged over 3 years.
In April 1994, the U.S. EPA announced its plans for the second periodic review of the air
quality criteria and NAAQS for PM, and in 1997, the U.S. EPA promulgated revisions to
the NAAQS (62 FR 38652, July 18, 1997). In the 1997 decision, the U.S. EPA
determined that the fine and coarse fractions of PMio should be considered separately.
This determination was based on evidence that serious health effects were associated with
short- and long-term exposures to fine particles in areas that met the existing PMio
standards. The U.S. EPA added new standards using PM2.5 as the indicator for fine
particles (with PM2.5 referring to particles with a nominal mean aerodynamic diameter
less than or equal to 2.5 (mi). These new standards were as follows: (1) an annual
standard with a level of 15.0 (ig/m3 based on the 3-year average of annual arithmetic
mean PM2.5 concentrations from single or multiple community-oriented monitors and
(2) a 24-hour standard with a level of 65 (.ig/rn3 based on the 3-year average of the 98th
percentile of 24-hour PM2.5 concentrations at each monitor within an area. Also, the
1 PMio refers to particles with a nominal mean aerodynamic diameter less than or equal to 10 |im. More specifically,
10 |im is the aerodynamic diameter for which the efficiency of particle collection is 50%. Larger particles are not
excluded altogether but are collected with substantially decreasing efficiency while smaller particles are collected
with increasing efficiency.
lv
-------
U.S. EPA established a new reference method for the measurement of PM2 5 in the
ambient air and adopted rules for determining attainment of the new standards. To
continue to address the coarse fraction of PM10 (referred to as thoracic coarse particles or
PM10-2.5; generally including particles with a nominal mean aerodynamic diameter
greater than 2.5 |im and less than or equal to 10 |im). the U.S. EPA retained the annual
PM10 standard and revised the form of the 24-hour PM10 standard to be based on the 99th
percentile of 24-hour PM10 concentrations at each monitor in an area. The U.S. EPA
revised the secondary standards by setting them equal in all respects to the primary
standards.
Following promulgation of the 1997 PM NAAQS, petitions for review were filed by a
large number of parties, addressing a broad range of issues. In May 1999, the U.S. Court
of Appeals for the District of Columbia Circuit (D.C. Circuit) upheld the U.S. EPA's
decision to establish fine particle standards, holding that "the growing empirical evidence
demonstrating a relationship between fine particle pollution and adverse health effects
amply justifies establishment of new fine particle standards." American Trucking
Associations v. EPA, 175 F. 3d 1027, 1055-56 (D.C. Cir., 1999). The D.C. Circuit also
found "ample support" for the U.S. EPA's decision to regulate coarse particle pollution,
but vacated the 1997 PM10 standards, concluding that the U.S. EPA had not provided a
reasonable explanation justifying use of PM10 as an indicator for coarse particles (175 F.
3d at 1054-55). Pursuant to the D.C. Circuit's decision, the U.S. EPA removed the
vacated 1997 PM10 standards, and the pre-existing 1987 PM10 standards remained in
place (65 FR 80776, December 22, 2000). The D.C. Circuit also upheld the U.S. EPA's
determination not to establish more stringent secondary standards for fine particles to
address effects on visibility (175 F. 3d at 1027).
The D.C. Circuit also addressed more general issues related to the NAAQS, including
issues related to the consideration of costs in setting NAAQS and the U.S. EPA's
approach to establishing the levels of NAAQS. Regarding the cost issue, the court
reaffirmed prior rulings holding that in setting NAAQS the U.S. EPA is "not permitted to
consider the cost of implementing those standards" (Id. at 1040-41). Regarding the levels
of NAAQS, the court held that the U.S. EPA's approach to establishing the level of the
standards in 1997 (i.e., both for PM and for the ozone NAAQS promulgated on the same
day) effected "an unconstitutional delegation of legislative authority" (Id. at 1034-40).
Although the court stated that "the factors U.S. EPA uses in determining the degree of
public health concern associated with different levels of ozone and PM are reasonable," it
remanded the rule to the U.S. EPA, stating that when the U.S. EPA considers these
factors for potential nonthreshold pollutants "what U.S. EPA lacks is any determinate
criterion for drawing lines" to determine where the standards should be set.
lvi
-------
The D.C. Circuit's holding on the cost and constitutional issues were appealed to the U.S.
Supreme Court. In February 2001, the Supreme Court issued a unanimous decision
upholding the U.S. EPA's position on both the cost and constitutional issues. Whitman v.
American Trucking Associations, 531 U.S. 457, 464, 475-76. On the constitutional issue,
the Court held that the statutory requirement that NAAQS be "requisite" to protect public
health with an adequate margin of safety sufficiently guided the U.S. EPA's discretion,
affirming the U.S. EPA's approach of setting standards that are neither more nor less
stringent than necessary.1
In October 1997, the U.S. EPA published its plans for the third periodic review of the air
quality criteria and NAAQS for PM (62 FR 55201; October 23, 1997). On September 21,
2006, the U.S. EPA announced its final decisions to revise the primary and secondary
NAAQS for PM to provide increased protection of public health and welfare,
respectively (71 FR 61144; October 17, 2006). With regard to the primary and secondary
standards for fine particles, the U.S. EPA revised the level of the 24-hour PM25 standards
to 35 |ig/m3. retained the level of the annual PM2 5 standards at 15.0 |ig/m3. and revised
the form of the annual PM2 5 standards by narrowing the constraints on the optional use of
spatial averaging. With regard to the primary and secondary standards for PM10, the
U.S. EPA retained the 24-hour standards, with levels at 150 |ig/m3. and revoked the
annual standards.2 The Administrator judged that the available evidence generally did not
suggest a link between long-term exposure to existing ambient levels of coarse particles
and health or welfare effects. In addition, a new reference method was added for the
measurement of PM10-2.5 in the ambient air, in order to provide a basis for approving
federal equivalent methods (FEMs) and to promote the gathering of scientific data to
support future reviews of the PM NAAQS.
Several parties filed petitions for review following promulgation of the revised PM
NAAQS in 2006. These petitions addressed the following issues: (1) selecting the level of
the primary annual PM2 5 standard; (2) retaining PM10 as the indicator of a standard for
1 The Supreme Court remanded the case to the Court of Appeals for resolution of any remaining issues that had not
been addressed in that court's earlier rulings (Id. at 475-76). In a March 2002 decision, the Court of Appeals
rejected all remaining challenges to the standards, holding that the U.S. EPA's PM2.5 standards were reasonably
supported by the administrative record and were not "arbitrary and capricious" American Trucking Associations v.
EPA, 283 F. 3d 355, 369-72 (D.C. Cir. 2002).
2 In the 2006 proposal, the U.S. EPA proposed to revise the 24-hour PM10 standard in part by establishing a new
PMi0-2.5 indicator for thoracic coarse particles (i.e., particles generally between 2.5 and 10 |im in diameter). The
U.S. EPA proposed to include any ambient mix of PM10-2.5 that was dominated by resuspended dust from high
density traffic on paved roads and by PM from industrial and construction sources. The U.S. EPA proposed to
exclude any ambient mix of PMi 0-2.5 that was dominated by rural windblown dust and soils and by PM generated
from agricultural and mining sources. In the final decision, the existing PM10 standard was retained, in part due to an
"inability.. .to effectively and precisely identify which ambient mixes are included in the [PMi0-2.5] indicator and
which are not" (71 FR 61197; October 17, 2006).
lvii
-------
thoracic coarse particles, retaining the level and form of the 24-hour PMio standard, and
revoking the PMio annual standard; and (3) setting the secondary PM2 5 standards
identical to the primary standards. On February 24, 2009, the U.S. Court of Appeals for
the District of Columbia Circuit issued its opinion in the case American Farm Bureau
Federation v. EPA, 559 F. 3d 512 (D.C. Cir. 2009). The court remanded the primary
annual PM2 5 NAAQS to the U.S. EPA because the U.S. EPA failed to adequately explain
why the standards provided the requisite protection from both short- and long-term
exposures to fine particles, including protection for at-risk populations (American Farm
Bureau Federation v. EPA, 559 F. 3d 512, 520-27; D.C. Cir. 2009). With regard to the
standards for PMio, the court upheld the U.S. EPA's decisions to retain the 24-hour PMio
standard to provide protection from thoracic coarse particle exposures and to revoke the
annual PMio standard (American Farm Bureau Federation, 559 F. 2d at 533-38). With
regard to the secondary PM2 5 standards, the court remanded the standards to the
U.S. EPA because the Agency failed to adequately explain why setting the secondary PM
standards identical to the primary standards provided the required protection for public
welfare, including protection from visibility impairment (American Farm Bureau
Federation, 559 F. 2d at 528-32). The U.S. EPA responded to the court's remands as
part of the next review of the PM NAAQS, which was initiated in 2007.
In June 2007, the U.S. EPA initiated the fourth periodic review of the air quality criteria
and the PM NAAQS by issuing a call for information in the Federal Register (72 FR
35462; June 28, 2007). In December 2012, the U.S. EPA announced its final decisions
with regard to the secondary PM standards, the U.S. EPA retained the 24-hour and annual
PM2 5 standards and the 24-hour PMio standard to address visibility and nonvisibility
welfare effects. On judicial review, the revised standards were upheld in all respects
(NAMv EPA, 750 F.3d 921; D.C. Cir. 2014).
Combined Review of the Oxides of Nitrogen and Oxides of Sulfur
National Ambient Air Quality Standards
In 2005, the U.S. EPA initiated a joint review of the air quality criteria for oxides of
nitrogen and sulfur and the secondary NAAQS for NO2 and SO2. In so doing, the
U.S. EPA assessed the scientific information, associated risks, and standards relevant to
protecting the public welfare from adverse effects associated jointly with oxides of
nitrogen and sulfur. Although the U.S. EPA has historically adopted separate secondary
standards for oxides of nitrogen and oxides of sulfur, the U.S. EPA conducted a joint
review of these standards because oxides of nitrogen and sulfur and their associated
transformation products are linked from an atmospheric chemistry perspective, as well as
from an environmental effects perspective. The joint review was also responsive to the
lviii
-------
National Research Council (NRC) recommendation for the U.S. EPA to consider
multiple pollutants, as appropriate, in forming the scientific basis for the NAAQS (NRC.
2004).
The review was initiated on December 13, 2005 with a call for information (70 FR
73236) for the development of a revised ISA. A draft Integrated Review Plan (IRP) was
released in October 2007, reviewed by CASAC; the final IRP was released in December
2007 (U.S. EPA. 2007). The first and second drafts of the ISA were released in
December 2007 and August 2008 (73 FR 10243), respectively, for CASAC and public
review. The final ISA(U.S. EPA. 2008a) was released in December 2008 (73 FR 75716).
Based on the scientific information in the ISA, the U.S. EPA developed a Risk and
Exposure Assessment (REA) to further assess the national impact of the effects
documented in the ISA. The Draft Scope and Methods Plan for Risk/Exposure
Assessment: Secondary NAAQS Review for Oxides of Nitrogen and Oxides of Sulfur
outlining the scope and design of the future REA was released in March 2008 (73 FR
10243). A first and second draft of the REA were released (August 2008 and June 2009)
for CASAC review and public comment. The final REA (U.S. EPA. 2009c) was released
in September 2009. Drawing on the information in the final REA and ISA, a first draft,
second draft, and final Policy Assessment (PA) were released in March 2010, September
2010, and January 2011, respectively (U.S. EPA. 2011a).
On August 1, 2011, based on consideration of the scientific information and quantitative
assessments, the U.S. EPA published a proposal to (1) retain the existing NO2 and SO2
secondary standards, (2) add secondary standards identical to the NO2 and SO2 primary
1-hour standards, and (3) not set a new multipollutant secondary standard in this review.
After consideration of public comments on the proposed standards and on design of a
new field pilot program to gather and analyze additional relevant data, the Administrator
signed a final decision in this rulemaking on March 20, 2012. The Administrator's
decision was that, while the current secondary standards were inadequate to protect
against adverse effects from deposition of oxides of nitrogen and sulfur, it was not
appropriate under Section 109(b) to set any new secondary standards at this time due to
the limitations in the available data and uncertainty as to the amount of protection the
metric developed in the review would provide against acidification effects across the
country (77 FR 20281). In addition, the Administrator decided that it was appropriate to
retain the current NO2 and SO2 secondary standards to address direct effects of gaseous
NO2 and SO2 on vegetation. Thus, taken together, the Administrator decided to retain and
not revise the current NO2 and SO2 secondary standards: an NO2 standard set at a level of
0.053 ppm as an annual arithmetic average, and an SO2 standard set at a level of 0.5 ppm
as a 3-hour average, not to be exceeded more than once per year (77 FR 20281).
lix
-------
The U.S. EPA's decision to not set a secondary NAAQS for oxides of nitrogen and sulfur
even though the Administrator had concluded that the existing standards are not adequate
to protect against the adverse impacts of aquatic acidification on sensitive ecosystems
was challenged by the Center for Biological Diversity and other environmental groups.
The petitioners argued that having decided that the existing standards were not adequate
to protect against adverse public welfare effects such as damage to sensitive ecosystems,
the Administrator was required to identify the requisite level of protection for the public
welfare and to issue a NAAQS to achieve and maintain that level of protection. The D.C.
Circuit disagreed, finding that the U.S. EPA acted appropriately in not setting a
secondary standard given the U.S. EPA's conclusions that "the available information was
insufficient to permit a reasoned judgment about whether any proposed standard would
be 'requisite to protect the public welfare ..."' (Center for Biological Diversity, etal. v.
EPA, 749 F.3d 1079, 1087; 2014). In reaching this decision, the court noted that the
U.S. EPA had "explained in great detail" the profound uncertainties associated with
setting a secondary NAAQS to protect against aquatic acidification.
lx
-------
EXECUTIVE SUMMARY
ES.1 Purpose and Scope of the Integrated Science Assessment
This Integrated Science Assessment (ISA) for Oxides of Nitrogen, Oxides of Sulfur, and
Particulate Matter—Ecological Criteria is a comprehensive evaluation and synthesis of
the most policy-relevant science aimed at characterizing the ecological effects caused by
these criteria pollutants.1 These criteria pollutants are reviewed here together because
they all contribute to nitrogen (N) and sulfur (S) deposition, which causes substantial
ecological effects. In this document, the term "oxides of nitrogen" refers to total oxidized
N (NOy), including nitric oxide (NO) and nitrogen dioxide (NO2) and all other gaseous
and particulate oxidized N containing compounds formed from NO and NO2.2 Total
sulfur oxides (SOx) includes gaseous chemical species (e.g., sulfur dioxide [SO2], sulfur
monoxide [SO], disulfiir monoxide [S2O], and sulfur trioxide [SO3]) as well as particulate
species, such as ammonium sulfate [(NH^SO-i] (U.S. EPA. 2011a). Particulate species
include SOx species like sulfites (SO;2 ) and sulfates (SO42 ). but among these two
species usually only SO42 make a major contribution to particulate mass. Throughout
this document SOx is defined as the sum of SO2 and particulate sulfate (SO42 ). which
together represent virtually all of the SOx mass in the atmosphere.3 Particulate matter
(PM) is composed of some or all of the following components: nitrate (NO;, ). SO42 .
ammonium (NH4+), metals, minerals (dust), and organic and elemental carbon.
This ISA serves as the scientific foundation for the review of the ecological effects
associated with the secondary (welfare-based) National Ambient Air Quality Standards
(NAAQS) for NOy, SOx, and PM. The health effects of these criteria pollutants are
considered in separate assessments for NOy (U.S. EPA. 2016f). SOx (U.S. EPA. 2016e).
and PM (U.S. EPA. 2019).4 The Clean Air Act definition of welfare effects includes, but
is not limited to, effects on soils, water, wildlife, vegetation, visibility, weather, and
1 The general process for developing an ISA, including the framework for evaluating weight of evidence and
drawing scientific conclusions and causal judgments, is described in a companion document. Preamble to the
Integrated Science Assessments (U.S. EPA. 2015e), www.epa.gov/isa.
2 This ISA reserves the abbreviation NOx strictly as the sum of NO and NO2—consistent with its use in the
atmospheric science community—and uses the term "oxides of nitrogen" to refer to the broader list of oxidized
nitrogen species. Oxides of nitrogen refers to NOy as the total oxidized nitrogen in both gaseous and particulate
forms. The major gaseous and particulate constituents of NOy include nitric oxide (NO), nitrogen dioxide (NO2),
nitric acid (HNO3), peroxyacetyl nitrate (PAN), nitrous acid (HONO), organic nitrates, and particulate nitrate
(NO;, ). This ISA uses the definitions adopted by the atmospheric sciences community.
3 The same definition of SOx used in the 2011 NOxSOx Policy Assessment (U.S. EPA. 2011a).
4 In this ISA, the blue electronic links can be used to navigate to cited materials as well as appendices, sections,
tables, figures, and studies from this ISA.
ES-1
-------
climate, as well as effects on man-made materials, economic values, and personal
comfort and well-being.
The current secondary NAAQS for NOy and SOx were set to protect against direct
damage to vegetation by NO2 or SO2. The secondary NAAQS for NO2 is identical to the
primary standard set in 1971: an annual average not to exceed 0.053 ppm N dioxide. The
secondary NAAQS for SO2, set in 1973, is a 3-hour average of 0.5 ppm SO2, not to be
exceeded more than once per year. The current secondary standards for PM are intended
to address PM-related visibility and nonvisibility welfare effects. These standards are a
3-year annual mean PM2.5 concentration of 15 |ig/nr\ with the 24-hour average PM2.5 and
PM10 set at concentrations of 35 (ig/m3 and 150 (ig/m3, respectively.
This ISA updates the 2008 ISA for Oxides of Nitrogen and Oxides of Sulfur—Ecological
Criteria [hereafter referred to as 2008 ISA (U.S. EPA. 2008a)l. as well as the ecological
portion of the 2009 ISA for Particulate Matter (U.S. EPA. 2009a) with studies and
reports published from January 2008 through May 2017. There are some studies included
that were published more recently than the May 2017 literature cutoff date; these studies
were added based on recommendations from the Clean Air Scientific Advisory
Committee (CASAC). The U.S. EPA conducted in-depth searches to identify
peer-reviewed literature on relevant topics. Subject-matter experts and the public were
also able to recommend studies and reports during a kick-off workshop held by the U.S.
EPA in March 2014 for NOy and SOx and in June 2016 for PM. CASAC recommended
the inclusion of additional studies during the review of the first draft. To fully describe
the state of available science, the U.S. EPA also carried over the most relevant studies
from previous assessments to include in this ISA.
This ISA determines whether NOy, SOx, and PM concentrations in the air or deposition
from the air cause ecological effects. The ecological effects of deposition are grouped
into three main categories: (1) acidification (caused by gaseous NOy, SOx, and
particulate NH/, NO;, . SO42 ). (2) N enrichment/N driven eutrophication (caused by
gaseous NOy and particulate NH44" and NO;, ). and (3) S enrichment (caused by SOx and
particulate forms of SO42 ). Ecological effects are further subdivided into terrestrial,
wetland, freshwater, and estuarine/near-coastal ecosystems. These ecosystems and effects
are linked by the connectivity of terrestrial and aquatic habitats through biogeochemical
pathways of N and S.
A schematic of the document organization is given by Figure ES-1. The Integrated
Synthesis (IS) brings together key information on specific subject matter found in the
appendices. Appendix 1 is an introduction to the purpose and organization of the material
covered in Appendix 2-Appendix 16. Appendix 2 characterizes the sources and
atmospheric processes involving NOy, SOx, and PM, as well as trends in ambient
ES-2
-------
concentrations and deposition. Appendix 3 describes direct effects of gas-phase NOy and
SOx on plants and lichens. Appendix 4 describes N and S deposition effects on terrestrial
biogeochemistry, and Appendix 5 and Appendix 6 describe the biological effects of
terrestrial acidification and terrestrial N enrichment, respectively. Appendix 7 describes
N and S deposition effects on aquatic biogeochemistry. Appendix 8 through Appendix 10
characterize the biological effects of freshwater acidification, freshwater N enrichment,
and marine eutrophication, respectively. Appendix 11 describes the effects ofN
deposition on wetlands. Appendix 12 describes the wetland and freshwater effects of S
enrichment. Appendix 13 discusses the climate modification of ecosystem response to N
and S deposition, and Appendix 14 presents information on N and S deposition effects on
ecosystem services. Information on the ecological effects of forms of PM beyond those
related to N or S deposition is presented in Appendix 15 (the nonecological welfare
effects associated with PM, such as visibility, climate, and material effects, are
considered as part of a separate review of PM [81 FR 87933, December 6, 2016]).
Appendix 16 includes six locations in the U.S. selected as case study areas that are
candidates for additional analysis of risk and exposure. These candidate sites were
selected because they have abundant data on ecological effects.
ES-3
-------
Integrative
Synthesis
Deposition of
N and S
(Appendix 2}
Exposure
Ambient Air Concentrations
(Appendix 2}
Direct exposure
Soil and aquatic biogeochemical pathways of
acidification (NOY+ NHX +SOx)
N enrichment/eutrophication (NOY +NHX)
S nutrient (SOx)
Terrestrial Ecosystems
Direct to organism/deposition
Directto soil, effects on soil biogeochemistry (Appendix 4)
Wetland Ecosystems
Directto soil and surface water, runoff from soil
Wetland biogeochemistry (Appendices 11&12)
Freshwater Ecosystems
Directto surface water, runoff from soil, effects on
freshwater biogeochemistry (Appendix 7)
Estuaries Ecosystems
Directto water, transport from watershed runoff, effects on
biogeochemistry along the freshwater to ocean continuum
(Appendix 7}
Climate Modification of Ecosystem
Response to N and S
(Appendix 13)
Ecosystem
Services
(Appendix 14)
Biological Effects
S02, N02, NO, PAN, HN03
(Appendix 3)
Plant foliar and lichen Injury
]
Biological effects of
acidification (NOY+ NHX +SOx)
N enrichment/eutrophication (NOy +NHX)
S nutrient (SQX)
Terrestrial Ecosystems
Acidification (Appendix 5)
N enrichment/eutrophication (Appendix 6)
Wetlands Ecosystems
N enrichment/eutrophication (Appendix 11)
S nutrient (Appendix 12)
Freshwater Ecosystems
Acidification (Appendix 8)
N enrichment/eutrophication (Appendix 9)
S nutrient (Appendix 12)
Estuarine Ecosystems
N nutrient/ eutrophication (Appendix 10)
N enhanced ocean acidification (Appendix 10)
Other Ecological
Effects of PM
(Appendix 15)
Case
Studies
(Appendix 16)
HN03 = nitric acid; N = nitrogen; NHX = reduced nitrogen; NO = nitric oxide; N02 = nitrogen dioxide; NOy = nitrogen oxides;
PAN = peroxyacetyl nitrate; PM = particulate matter; S = sulfur; S02 = sulfur dioxide; SOx = sulfur oxides.
Figure ES-1 Roadmap of the Integrated Science Assessment (ISA) linking
atmospheric concentrations and deposition, soil and aquatic
biogeochemistry, and biological effects.
ES.2 Emissions, Ambient Air Concentrations, and Deposition
The atmospheric chemistry from emission to deposition discussed in this ISA1 is for the
criteria pollutants NOy, SOx, and PM. NOy and SOx cause ecological effects in the gas
phase and/or after N and S deposition to surfaces. Particulate matter (PM) effects
discussed in this document focus on N and S containing species, which together usually
make up a large fraction of the PMgj mass in most areas of the U.S. NHx
(NHX = NH3 + NH_f) includes both gas-phase NH3 and the PM component NH4 . NH3 is
estimated to account for 19-63% of total observed inorganic N deposition, depending on
region (Appendix 2.1). Therefore, NH? is discussed in this ISA along with NOy and
relevant PM components to better understand and compare their contributions to both wet
and dry N deposition.
1 The term concentration is used throughout the ISA to denote either a mass per unit volume or a volume per unit
volume (mixing ratio). The use of concentration to denote abundance expressed as mixing ratio is firmly entrenched
in the literature; therefore, it is retained here.
ES-4
-------
Both gaseous and particulate forms of NOy, SOx, and NHX contribute to atmospheric wet
and dry deposition. The major components of PM in the U.S. are NO3 , SO42 . NFU+,
organic matter, elemental C, crustal material, and sea salt. Of these, NO;, . SO42 . and
NH4+ usually have a strong influence on acid deposition, and NO;, and NFU+, and in
some cases organic N (organic nitrates and reduced organic N), contribute substantially
to N deposition and eutrophication.
The sources and precursors to gaseous and particulate forms of NOy, SOx, and Mix vary.
The main contributors to acidifying precipitation are formed from precursor emissions of
the gases SO2, NOx, and NH3 (Appendix 2.2). Electricity-generating units (EGUs) are the
source of about half of national gaseous emissions of SO2, mainly from coal-fired power
plants. Notably, SO2 emissions from EGUs have been decreasing. NOx emissions have a
wider distribution of sources, with substantial contributions from highway and
off-highway vehicles, lightning, and EGUs. Fertilizer application and animal waste are
the main national-scale sources of NH3, with animal waste contributing the most. Primary
PM2.5 and PM10 emissions are dominated by dust and combustion products of fires, but
much of the PM2.5 mass in the U.S. is produced by reaction of gas-phase precursors to
form secondary PM2.5. In this process, particulate NFU+, NO; . and SO42 are primarily
derived from the gaseous precursors NH3, NOx, and SO2 (Appendix 2.3). Formation of
particulate N and S is described in the 2019 ISA for Particulate Matter (U.S. EPA. 2019).
An understanding of the sources, chemistry, and atmospheric processes for these
gas-phase and PM species is necessary to understand acidifying and N deposition.
Since the passage of the Clean Air Act Amendments in 1990, the emissions of NOx and
SO2 have declined dramatically. Total emissions of SO2 decreased by 89% from 1990 to
2017, resulting in a decrease in SO2 concentrations of 89% in the eastern U.S. and 45% in
the western U.S. Emissions of NOx in the U.S. declined by 61% between 1990 and 2017,
while nationwide annual average 98th percentile NO2 concentrations decreased by 53%
from 1990 to 2017. These reductions have in turn led to decreases in PM2.5 concentrations
because of declines in the amount of SO42 and NO3 produced, and a decrease in the
fraction of PM2.5 accounted for by SO42 . Between 1989 and 2017, average particulate
S042 concentration decreased by 75% in the eastern U.S. and 35% in the western U.S.,
and average particulate NO3 concentration decreased by 51% in the eastern U.S. and
37% in the western U.S.
Averaged across the contiguous U.S., deposition of total N (oxidized + reduced N, in kg
N/ha/yr) has changed only slightly since 2000 (Appendix 2.6.2). Figure ES-2 shows that
between 2000 and 2018 large decreases in oxidized nitrogen (Figure ES-2A) have
combined with large increases in reduced nitrogen deposition (Figure ES-2B) to produce
a small decrease in total nitrogen deposition (Figure ES-2C). There is large spatial
ES-5
-------
variability in N deposition over the contiguous U.S. (Figure ES-2C). According to
National Atmospheric Deposition Program Total Deposition Committee's (TDEP s)
estimates for 2016-2018 (Appendix 2.6.2). much of the eastern U.S. is estimated to
receive at least 10 kg N/ha/yr dry + wet deposition, with some areas receiving more than
15 kg N/ha/yr. Figure ES-2 A through C shows that between 2000 and 2018, large
decreases in oxidized nitrogen deposition occurred.
2000-2002
2016-2018
A)
Oxidized
Nitrogen •
* .
'< 4 —
\ |
B)
Reduced
Nitrogen
Mi
m
f*
C)
Total
Nitrogen
D)
Total
Sulfur
r/*"
li
tl
w
Ha = hectare; kg = kilogram; N = nitrogen; OxN = oxidized nitrogen; ReN = reduced nitrogen; S = sulfur.
Source: We acknowledge the Total Deposition (TDep) Science Committee of the National Atmospheric Deposition Program (NADP)
for their role in making the TDep data and maps available.
Figure ES-2 Wet plus dry deposition of (A) oxidized nitrogen, (B) reduced
nitrogen, (C) total nitrogen, and (D) total sulfur over the 3-year
periods 2000-2002 and 2016-2018.
ES-6
-------
For S, wet deposition tends to dominate over dry deposition in large areas of the
contiguous U.S. However, in some regions, mostly in the west, dry deposition of mainly
SO2 is greater than wet deposition. Anthropogenic emissions of S and subsequent
deposition have declined markedly since the 1990s, with the most pronounced declines in
the eastern U.S., as shown in Figure ES-2D. Currently, the highest values of total
(wet + dry) SOx deposition in the U.S. are in parts of the Ohio Valley region where they
range between 15 and 20 kg S/ha/yr.
Both N and S deposition contribute to acidification of ecosystems. The acidity of
rainwater has decreased, as indicated by the increase of rainwater pH across the U.S.
since 1990, coincident with decreases in the wet deposition of nitrate and sulfate.
However, widespread areas are still affected by acidifying precipitation, mainly in the
eastern U.S. (see Appendix 2.6.1). Total acidifying deposition (wet + dry N + S,
expressed as H+ equivalents) fluxes for 2016 to 2018 ranged from a few tenths of H+
keq/ha/yr over much of the western U.S. to over 1.5 H+ keq/ha/yr in parts of the Midwest
and the Mid-Atlantic regions, and in other isolated hotspots surrounding areas of
concentrated industrial or agricultural activity (Figure IS-6). Estimated deposition fluxes
greater than 1.5 keq/ha/yr covered a much smaller portion of the U.S. in 2016-2018 than
in 2000-2002.
ES.3 Ecological Effects
In this ISA, information on ecological effects from controlled exposure, field addition,
ambient deposition, and toxicological studies, among others, are integrated to form
conclusions about the causal nature of relationships between NOy, SOx, and PM and
ecological effects. Studies on the ecological effects are considered in relation to a range
of ambient concentration and deposition loads that are within two orders of magnitude
from current conditions [Preamble (U.S. EPA. 2015e). Section 5c]. A consistent and
transparent framework [Preamble (U.S. EPA. 2015e). Table II] is applied to classify the
ecological effect evidence according to a five-level hierarchy:
1. Causal relationship
2. Likely to be a causal relationship
3. Suggestive of, but not sufficient to infer, a causal relationship
4. Inadequate to infer a causal relationship
5. Not likely to be a causal relationship
The conclusions presented in Table ES-1 are based on recent findings integrated with
information from the 2008 ISA (U.S. EPA. 2008a). The conclusions of Table ES-1 are
based on careful consideration of errors and uncertainty in the supporting studies. We
ES-7
-------
also consider the coherence of findings integrated across studies of underlying
geochemical and biological mechanisms. There are 18 causality statements in this ISA
(Table ES-1). Fourteen are causal relationships repeated from the 2008 ISA or modified
from the 2008 ISA to include specific endpoints. For these causality statements, new
research strengthens the evidence base and is consistent with the 2008 ISA. There is one
likely causal relationship repeated from the 2009 ISA for Particulate Matter. Three causal
relationships are new endpoint categories not evaluated in the 2008 ISA. Although NOy
and SOx can cause phytotoxic injury, current monitored concentrations of gas-phase NOy
and SOx are not high enough to injure vegetation. For all other identified causal
relationships identified in this ISA, the evidence indicates a causal association from
current levels of S and/or N deposition.
Table ES-1 Causal determinations for relationships between criteria pollutants
and ecological effects from the 2008 NOx/SOx Integrated Science
Assessment (ISA) or the 2009 ISA for Particulate Matter (PM), for
other effects of PM, and the current draft ISA.
Causal Determination
Effect Category
2008 NOX/SOX ISA
Current Draft ISA
Gas-phase direct phytotoxic effects
Gas-phase SO2 and injury to vegetation
Causal relationship
Causal relationship
Section IS.3 and ADDendix 3.6.1
Gas-phase NO, NO2, and PAN and injury to vegetation
Causal relationship
Causal relationship
Section IS.3 and ADDendix 3.6.2
Gas-phase HNO3 and injury to vegetation3
Causal relationship
Causal relationship
Section IS.3 and ADDendix 3.6.3
N and acidifying deposition to terrestrial ecosystems
N and S deposition and alteration of soil biogeochemistry in
terrestrial ecosystems'5
Causal relationship
Causal relationship
Section IS.5.1 and ADDendix 4.1
N deposition and the alteration of the physiology and growth of
terrestrial organisms and the productivity of terrestrial
ecosystems0
Not included
Causal relationship
Section IS.5.2 and ADDendix 6.6.1
N deposition and the alteration of species richness, community
composition, and biodiversity in terrestrial ecosystems0
Causal relationship
Causal relationship
Section IS.5.2 and ADDendix 6.6.2
ES-8
-------
Table ES-1 (Continued): Causal determinations for relationships between criteria
pollutants and ecological effects from the 2008 NOx/SOx
Integrated Science Assessment (ISA) or the 2009 ISA for
Particulate Matter (PM), for other effects of PM, and the
current draft ISA.
Causal Determination
Effect Category
2008 NOX/SOX ISA
Current Draft ISA
Acidifying N and S deposition and the alteration of the physiology
and growth of terrestrial organisms and the productivity of
terrestrial ecosystemsd
Not included
Causal relationship
Section IS.5.3 and ADDendix 5.7.1
Acidifying N and S deposition and the alteration of species
richness, community composition, and biodiversity in terrestrial
ecosystemsd
Causal relationship
Causal relationship
Section IS.5.3 and ADDendix 5.7.2
N and acidifying deposition to freshwater ecosystems
N and S deposition and alteration of freshwater biogeochemistrye
Causal relationship
Causal relationship
Section IS.6.1 and ADDendix 7.1.7
Acidifying N and S deposition and changes in biota, including
physiological impairment and alteration of species richness,
community composition, and biodiversity in freshwater
ecosystems'
Causal relationship
Causal relationship
Section IS.6.3 and ADDendix 8.6
N deposition and changes in biota, including altered growth and
productivity, species richness, community composition, and
biodiversity due to N enrichment in freshwater ecosystems9
Causal relationship
Causal relationship
Section IS.6.2 and ADDendix 9.6
N deposition to estuarine ecosystems
N deposition and alteration of biogeochemistry in estuarine and
near-coastal marine systems
Causal relationship
Causal relationship
Section IS.7.1 and ADDendix 7.2.10
N deposition and changes in biota, including altered growth, total
primary production, total algal community biomass, species
richness, community composition, and biodiversity due to N
enrichment in estuarine environments11
Causal relationship
Causal relationship
Section IS.7.2 and ADDendix 10.7
N deposition to wetland ecosystems
N deposition and the alteration of biogeochemical cycling in
wetlands
Causal relationship
Causal relationship
Section IS.8.1 and ADDendix 11.10
ES-9
-------
Table ES-1 (Continued): Causal determinations for relationships between criteria
pollutants and ecological effects from the 2008 NOx/SOx
Integrated Science Assessment (ISA) or the 2009 ISA for
Particulate Matter (PM), for other effects of PM, and the
current draft ISA.
Causal Determination
Effect Category
2008 NOX/SOX ISA
Current Draft ISA
N deposition and the alteration of growth and productivity, species
physiology, species richness, community composition, and
biodiversity in wetlands
Causal relationship
Causal relationship
Section IS.8.2 and Appendix 11.10
S deposition to wetland and freshwater ecosystems
S deposition and the alteration of mercury methylation in surface
water, sediment, and soils in wetland and freshwater ecosystems'
Causal relationship
Causal relationship
Section IS.9.1 and Appendix 12.7
S deposition and changes in biota due to sulfide phytotoxicity,
including alteration of growth and productivity, species physiology,
species richness, community composition, and biodiversity in
wetland and freshwater ecosystems
Not included
Causal relationship
Section IS.9.2 and Appendix 12.7
2009 PM ISA
Current Draft ISA
Other ecological effects of PM
PM and a variety of effects on individual organisms and
ecosystems
Likely to be a causal
relationship
Likely to be a
causal relationship
Section IS.10 and Appendix 15.8
C = carbon; Hg = mercury; HN03 = nitric acid; ISA = Integrated Science Assessment; N = nitrogen; NO = nitric oxide;
N02 = nitrogen dioxide; PAN = peroxyacetyl nitrate; S = sulfur; S02 = sulfur dioxide.
aThe 2008 ISA causality statements for gas-phase HN03 was phrased as "changes in vegetation."
bThe 2008 ISA included two causality statements for terrestrial biogeochemistry phrased as "relationship between acidifying
deposition and changes in biogeochemistry" and "relationship between N deposition and the alteration of biogeochemical cycling
of N."
The 2008 ISA causality statement for biological effects of N enrichment in terrestrial ecosystems was phrased as "relationship
between N deposition and the alteration of species richness, species composition, and biodiversity."
dThe 2008 ISA causality statement for biological effects of acidifying deposition in terrestrial ecosystems was phrased as
"relationship between acidifying deposition and changes in terrestrial biota."
eThe 2008 ISA included three causality statements for freshwater biogeochemistry phrased as "relationship between acidifying
deposition and changes in biogeochemistry related to aquatic ecosystems," "relationship between N deposition and the alteration
of biogeochemical cycling of N," and "relationship between N deposition and the alteration of biogeochemical cycling of C."
'The 2008 ISA causality statement for biological effects of acidifying deposition in freshwater ecosystems was phrased as
"relationship between acidifying deposition and changes in aquatic biota."
9The 2008 ISA causality statement for biological effects of N deposition in freshwater ecosystems was phrased as "relationship
between N deposition and the alteration of species richness, species composition, and biodiversity in freshwater aquatic
ecosystems."
hThe 2008 ISA causality statement for biological effects of N deposition to estuaries was phrased as "relationship between N
deposition and the alteration of species richness, species composition, and biodiversity in estuarine ecosystems."
'The 2008 ISA causality statement for biological effects of S deposition effects on ecosystems was phrased as "relationship
between S deposition and increased methylation of Hg, in aquatic environments where the value of other factors is within
adequate range for methylation."
ES-10
-------
Figure ES-3 presents a visualization of the causality statements integrated into a single
diagram. There is not a one-to-one correspondence between the number of causality
statements, of which there are 18, and the cells indicated to have causal relationships in
the diagram because some causal statements include effects across more than one level of
biological organization. The main findings are that gaseous NOy and SOx cause
phytotoxic effects, while N and S deposition cause alteration in (1) biogeochemical
components of soil and water chemistry and (2) multiple levels of biological organization
ranging from physiological processes to shifts in biodiversity and ecological function.
ES-11
-------
NOx SOx PM Integrated Science Assessment for Ecological Effects*
Indicator
Gases * Nitrogen Deposition Sulfur Deposition ^'^Depo^on^
Class of Pollutant Effect
Direct
Phytotoxic N-enrichment/Eutrophication Sulfide Toxicity Mercury Methylation Acidification
Ecosystem
Terrestrial Terrestrial Wetland Fresh Water Estuary Wetland Fresh Water Wetland Fresh Water Terrestrial Fresh Water
| Scale of Ecological Response
Population
Geochemistry Individual Community Ecosystem
Individual
Productivity
Biodiversity
Growth rate
Physiological
alteration, stress
or injury
Soil or sediment
chemistry
Surface water
chemistry
u 1^1
* A causal relationship is likely to exist between deposition of PM and a variety of effects on individual organisms and ecosystems, based
on information from the previous review and limited new findings in this review
* Includes: NO, N02, HN03, S02, and PAN
Figure ES-3 Causal relationships between the criteria pollutants and ecological effects.
ES-12
-------
ES.4 Direct Phytotoxic Effects of Gas-Phase Oxides of Nitrogen
(NOy) and Oxides of Sulfur (SOx)
The current NO2 and SO2 secondary NAAQS are set to protect against direct damage to
vegetation by exposure to gas-phase oxides of NOy and SOx. NH3 can also have direct
phytotoxic effects, but reduced N gases such as NH3 are not criteria air pollutants.
Research continues to support causal relationships between SO2, NO2, NO, peroxyacetyl
nitrate (PAN), HNO3, and injury to vegetation (e.g., visible foliar injury, damage to
photosynthesis, decline of growth and abundance; (Table IS-1. Section IS.4. Appendix 3).
but research that tests plant response to the lower exposure levels representative of
current atmospheric NOy and SOx concentrations is limited. Consequently, few studies
are available to help determine whether current monitored concentrations of gas-phase
NOy and SOx are high enough to injure vegetation. It is also known that these can be
gases taken up by plants and alter the N cycle in some ecosystems.
ES.5 Ecological Effects of Nitrogen and Sulfur Deposition
It is clear from the body of evidence that NOy, SOx, and PM contribute to total N and S
deposition. In turn, N and S deposition cause alteration of the biogeochemistry and the
physiology of organisms, resulting in harmful declines in biodiversity in terrestrial,
freshwater, wetland, and estuarine ecosystems in the U.S. Decreases in biodiversity mean
that some species become relatively less abundant and may be locally extirpated. In
addition to the loss of unique living species, the decline in total biodiversity can be
harmful because biodiversity is an important determinant of the stability of ecosystems
and their ability to provide socially valuable ecosystem services (see more on biodiversity
in Section IS.2.2.4).
ES.5.1 Acidification of Terrestrial and Freshwater Ecosystems
Several decades of research have documented that N and S deposition cause freshwater
and terrestrial ecosystem acidification in the U.S. New evidence strengthens the causal
relationships for ecosystem acidification determined in the 2008 ISA (Table IS-1).
Many of the terrestrial and freshwater ecosystems most sensitive to acidification in the
U.S. are found in the Northeast and Southeast. In the West, freshwater and terrestrial
ecosystems acidified from deposition are now limited in extent and occur mostly in
high-elevation sites. Watershed sensitivity to acid inputs depends on characteristics such
as underlying geology (Appendix 4 and Appendix 7) and the sensitivity of species in the
local biological community (Appendix 5 and Appendix 8). Regional heterogeneity of
deposition levels that cause ecological effects are in part due to historic exposure and
ES-13
-------
climate. In the East, especially the southern Appalachian Mountains, and the Northeast,
the effects of acidifying deposition have been studied for several decades.
Acidified aquatic habitats have a lower number of species (species richness) of fishes,
macroinvertebrates, and phytoplankton. The effects of acidifying deposition on aquatic
ecosystems also include physiological impairment or mortality of sensitive species and
shifts in biodiversity of both flora and fauna. Organisms at all trophic levels are affected
by acidification, with clear linkages to chemical indicators for effects on algae,
zooplankton, benthic invertebrates, and fish. Acid-neutralizing capacity (ANC) is a
measure of the buffering capacity of natural waters against acidification. Even though
ANC does not directly alter the health of biota, it is a key metric of acidification that
relates to pH and aluminum levels. Biological effects are primarily attributable to low pH
and high inorganic aluminum concentration. Characterization of ANC and its levels of
concern have not changed appreciably with the newly available information since the
2008 ISA. Few or no fish species are found in lakes and streams that have very low ANC
(near zero) and low pH (near 5.0), and the number of fish species generally increases
with higher ANC and pH (Appendix 8.3). The fish lost to acidification include culturally
and recreationally important species.
Acidified terrestrial habitats are characterized by the detrimental physiological effects
seen on vegetation, including inhibited growth and decreased plant health. Acidifying
deposition can decrease membrane stability and freezing tolerance in young red spruce
needles. For many species, calcium (Ca) depletion from the soil and aluminum
mobilization cause decreased root uptake of Ca and disrupt fine root physiological
functions. Reduced availability of (base) cations in the soil can also make trees more
vulnerable to other stresses, such as damage from insects and other pathogens. Within the
eastern U.S., the physiological effects of acidifying deposition have been well
documented for the several culturally and commercially important tree species with
known ecosystem services, particularly sugar maple (Acer saccharum) and red spruce
(Piceci rubens). Consistent and coherent evidence available before and since the 2008
ISA suggests acidifying deposition among these species can decrease foliar cold
tolerance, increase rates of crown dieback, decrease tree growth, suppress seedling
regeneration, and increase mortality rates. (Section IS.5.3; Appendix 5). Since the 2008
ISA, studies from the northeastern U.S. have shown that Ca addition can alleviate many
of these effects, demonstrating that acidification effects can be ameliorated in the short
term by soil amendments, suggesting the potential for recovery. However, Ca additions
have been studied in only a few areas. Acidifying deposition has also been linked to
changes in forest understory plant community composition in the northeastern U.S., grass
and forb biodiversity in eight ecoregions across the U.S., and decreased grassland plant
species richness in Europe.
ES-14
-------
Examples of improvement in acidification have been documented in some aquatic
ecosystems in the regions most affected. Along with those improvements in acidification,
chemical recovery has been observed in the Northeast, as seen by trends in water quality
indicators (NO3 , SO42 . pH, ANC, inorganic monomeric Al, MeHg) towards inferred
preindustrial values or, in the case of inorganic Al and MeHg, below water quality
threshold values protective of biota and human health. Chemical recovery has not been
observed in studies of the southern Appalachians. In a few examples in the Northeast,
chemical recovery co-occurs with the movement of biological indicators toward
recovery. However, biological recovery has been highly variable among ecosystems and
taxonomic groups. Biological recovery lags behind, sometimes by decades, chemical
recovery. In addition, the biological recovery trajectory may exhibit hysteresis, in which
a system does not follow the same path from acidification to recovery. Most biological
communities studied to date where signs of reversal are found have not returned to
preacidification conditions and are unlikely to do so, given the extirpation of some
species, fundamental alterations in function and structure, decade-long depletion of base
cations, and changes in other interacting influences such as climate and land use.
ES.5.2 Nitrogen Enrichment/Eutrophication of Terrestrial, Wetland, and
Aquatic Ecosystems
Terrestrial, wetland, freshwater, and estuarine ecosystems in the U.S. are affected by N
enrichment/eutrophication caused by N deposition. N enrichment/eutrophication refers to
N nutrient-driven changes in growth, physiology, and biodiversity. These effects have
been consistently documented across the U.S. for hundreds of species. New evidence
strengthens the causal relationships for ecosystem N enrichment/eutrophication
determined in the 2008 ISA (Table IS-1).
The 2008 ISA documented that the N enrichment effect in sensitive terrestrial and
wetland ecosystems starts with the accumulation of N in the soil. This increases the
availability of N, a nutrient that increases the growth of some species of soil microbes
and vascular plants at the expense of other species, which may decrease biodiversity.
Since the 2008 ISA, the largest increase in ecological evidence is for terrestrial N driven
enrichment/eutrophication effects (Section IS.5.1. Section IS.5.2; Appendix 4. and
Appendix 6).
This new research confirms the causal relationship between N deposition and ecological
effects documented in the 2008 ISA and improves our understanding of the mechanistic
links that inform causal determinations between N deposition, biogeochemistry, and biota
in terrestrial ecosystems (Table IS-1). A new causal determination has been added to
reflect more specific categories of effects to include physiology, growth, and ecosystem
productivity. Further, there is now stronger empirical evidence from across most regions
ES-15
-------
of the U.S. to quantify critical loads (CLs) for N deposition. The figure below provides
estimates of CLs across broad ecoregions and shows the ranges for different functional
groups within these systems (Figure ES-4). Under the CLs, significant harmful effects
from N deposition do not occur according to present knowledge, while at or above CLs,
N deposition can cause a myriad of ecological effects, including decreased tree growth
and increased mortality, and declines in grasses/forbs, lichens, and mycorrhizal fungi.
Since the 2008 ISA, studies have strengthened evidence of species-specific effects of N
deposition on tree growth and mortality in the U.S. Although overall tree growth has
generally been enhanced by N deposition over the last several decades, there is wide
variation among species in growth and mortality responses. Moreover, within some
individual species, N deposition can increase growth and/or survival at low levels, while
decrease growth and/or survival at higher levels. Species with varying responses have
also been shown to co-occur in places in the U.S., suggesting overstory tree community
composition shifts with N deposition.
Since the 2008 ISA, studies have also strengthened the findings of N effects on
decreasing lichen and mycorrhizal fungi biodiversity and provided additional CL
estimates. In terrestrial ecosystems, new evidence provides support that epiphytic lichens
(an algal- and/or cyanobacteria-fungal symbiont) and mycorrhizae (a plant-fungal
symbiosis at the tips of plant roots) are the organisms most sensitive to atmospheric N
deposition and acidifying deposition. Although lichens typically are only a small portion
of terrestrial biomass, these changes in lichen communities are meaningful because
lichens provide food and habitat for insects, birds, and mammals; contribute to nutrient
and hydrologic cycling; have many traditional human uses; and have considerable
potential for pharmaceutical use. Changes in the community composition of mycorrhizal
fungi and declines in mycorrhizal abundance have been observed in the U.S. These fungi
are important for supplying nutrients and water to plants, influencing soil C sequestration,
and producing fruiting bodies (mushrooms) used by humans and wildlife.
ES-16
-------
Southern Semi-Arid Highlands
Temperate Sierras
Mediterranean California
North American Deserts
C
o
§? Great Plains
o
<_>
LlJ
Northwest Forested Mountains
Marine West Coast Forests
Eastern Temperate Forests
Northern Forests
0 5 10 15 20 25 30 35 40 45
kg N/ha/yr
No Data
i
4-7
O
•
o
6-33
5-6
7,8-9.2
O
CO-8.4 # 1
3
G
o
5-2
sO
8
r
4-10 O
•
2.5- 7.1
n
HI o
CO 07- 9-2
•i
<3
©
o
4-80 G
5-12
G
O
wO
5-7 m
• i
17.5
<3
0-21
Omernik Ecoregion I
Ecoregion I
| FASTERN TEMPERATE FORESTS
[ 1 GREAT PLAINS
| MARINE WEST COAST FOREST
¦ MEDITERRANEAN CALIFORNIA
| NORTH AMERICAN DESERTS
| NORTHERN FORESTS
B NORTHWESTERN FORESTED MOUNTAINS
I I SOUTHERN SEMI-ARID HIGHLANDS
| TEMPERATE SIERRAS
I TROPICAL WET FORESTS
CL = critical load; ha = hectare; kg = kilogram; N = nitrogen; yr = year.
The rectangles indicate the range of CLs designated by Pardo et al. (2011a); the circles indicate new papers that have specified
CLs; data from Table 6-28.
Figure ES-4 Summary of critical loads for nitrogen in the U.S. for shrubs and
herbaceous plants (yellow), trees (blue), lichens (green), and
mycorrhizae (gray). Values expressed by major U.S. ecoregions.
ES-17
-------
For wetland ecosystems, the 2008 ISA documented that wetlands receiving a larger
fraction of their total water budget in the form of precipitation are more sensitive to the
effects of N deposition. For example, bogs and fens (55-100% of hydro logical input
from rainfall) are more sensitive to N deposition than coastal wetlands (10-20% as
rainfall). Since the 2008 ISA, CLs for U.S. coastal and freshwater wetlands have been
established. The CL for freshwater wetlands is based on C cycling, as well as biodiversity
represented by the morphology and population dynamics of the purple pitcher plant
(Sarraceniapurpurea). The CL for coastal wetlands is based on several different
ecological endpoints, including plant community composition, microbial activity, and
biogeochemistry.
The 2008 ISA documented that the process ofN eutrophication is similar in freshwater
and estuarine ecosystems and typically begins with a nutrient-stimulated algal bloom that
is followed by anoxic conditions. The lack of oxygen in the water due to the respiration
and decomposition of the algae affects higher tropic species. The contribution of N
deposition to total N loading varies among freshwater lakes and stream ecosystems.
Atmospheric deposition is the main source of new N inputs to most headwater stream,
high-elevation lake, and low-order stream watersheds far from the influence of other N
sources like agricultural runoff and wastewater effluent. N deposition was known at the
time of the 2008 ISA to alter biogeochemical processes, nutrient ratios, and
concentrations in recipient freshwater ecosystems. New CLs published since the 2008
ISA support previous observations of increased productivity of phytoplankton and algae,
species changes, and reductions in diversity in atmospherically N enriched lakes and
streams. The productivity of many freshwater ecosystems is N limited. Thus, even small
amounts of N can shift nutrient ratios and affect the trophic status of lakes and streams.
As reported in the 2008 ISA and newer studies, a shift from N limitation to either
colimitation by N and P or limitation by P has been observed in some alpine lakes in the
U.S. and other countries, with these shifts correlated with elevated N deposition.
Estuaries support a large biodiversity of flora and fauna and play a role in nutrient
cycling. At the time of the 2008 ISA, N was recognized as the major cause of harm to the
majority of estuaries in the U.S. Elevated N inputs to coastal areas can alter key processes
that influence N and C cycling in near-coastal environments. Data evaluating sources of
N to estuaries, from the 2008 ISA and newer studies reviewed in this ISA, indicate that N
from atmospheric sources ranges from <10% to approximately 70% of total estuary N
inputs; the atmospheric input for most estuaries is between 15 to 40% of total N inputs. N
from atmospheric and other sources contributes to increased primary productivity,
leading to eutrophication. In some coastal areas eutrophication from N loading may affect
carbonate chemistry under certain circumstances, potentially contributing to acidifying
conditions along with atmospheric anthropogenic CO2 inputs and other factors. Since
ES-18
-------
2008, new paleontological studies, observational studies, and experiments have further
characterized the effects of N on phytoplankton growth and community dynamics,
macroinvertebrate response, and other indices of biodiversity in streams, rivers, lakes and
estuaries. For this ISA, new information is consistent with the 2008 ISA, and the causal
determinations for N enrichment in aquatic systems have been updated to reflect more
specific categories of effects, including measures of productivity and altered growth of
biota (Table ES-1).
ES.5.3 Sulfur (S) Enrichment of Wetland and Freshwater Ecosystems
SOx deposition increases SO42 concentration in surface waters. New evidence supports
links between aqueous S concentrations in freshwater ecosystems and both mercury (Hg)
methylation and sulfide toxicity (Table ES-1); however, quantitatively linking these
outcomes to atmospheric deposition remains a challenge.
Increasing SO42 concentration in surface waters can stimulate the microbial
transformation of inorganic Hg into methylmercury (MeHg; Appendix 12). MeHg is the
most persistent and toxic form of Hg affecting animals in the natural environment.
Indicators of S deposition effects upon Hg methylation include increases in MeHg
concentrations or fraction of total Hg in water, sediments, and peat, as well as increases
in MeHg concentrations in periphyton, submerged aquatic plants, invertebrates, and fish.
New evidence confirms the relationship between aqueous concentrations of SO42 and
MeHg and broadens our understanding of where methylation occurs from the wetlands
and lakes reported in the 2008 ISA to include rivers, reservoirs, streams, and saturated
forest soils. Hg methylation occurs at anoxic-oxic boundaries in peat moss and
periphyton, as well as in wetland, lake, estuarine, and marine sediments. There are
published quantitative relationships between surface water SO42 concentrations and
MeHg concentrations, MeHg and total Hg in water, and Hg load in larval mosquitoes and
fish. There is also evidence that decreasing S deposition loads over time (observational
studies of SOx deposition, experimental studies of simulated SOx wet deposition) result
in lower concentrations of MeHg in water, invertebrates, and fish.
There is new evidence since the 2008 ISA to infer a causal relationship between S
deposition and sulfide phytotoxicity, which alters growth and productivity, species
physiology, species richness, community composition, and biodiversity in wetland and
freshwater ecosystems (Appendix 12). This new causal statement reflects new research
on sulfide phytotoxicity in North American wetlands, as the 2008 ISA described sulfide
phytotoxicity only in European ecosystems. Current levels of S deposition cause sulfide
toxicity in wetland and aquatic plants. Indicators of sulfide phytotoxicity caused by S
deposition include increases in water or sediment sulfide concentrations. Sulfide
ES-19
-------
negatively effects growth, competitive ability, and persistence in several wetland species,
including the economically important species of wild rice and the keystone sawgrass
species in the Everglades marshes. To date, no published studies have established
regional sensitivities to sulfide phytotoxicity, although studies have observed its effects in
New York, Minnesota, and Florida freshwater marshes. There are no S deposition-based
critical loads for Hg methylation or sulfide phytotoxicity, although researchers have
proposed water quality values to protect biota against these effects in several ecosystems
(Appendix 12).
ES.5.4 Ecological Effects of Particulate Matter Other Than Those Associated
with Nitrogen and Sulfur Deposition
There is a likely causal relationship between PM and ecological effects on biota other
than those associated with N and S deposition (Table ES-1; Appendix 15). Since
publication of the 2009 PM ISA, new literature has built upon the existing knowledge of
ecological effects associated with PM components, especially metals and organics. In
some instances, new techniques have enabled further characterization of the mechanisms
of PM on soil processes, vegetation, and effects on fauna. New studies provide additional
evidence for community-level responses to PM deposition, especially in soil microbial
communities. However, uncertainties remain due to the difficulty in quantifying
relationships between ambient concentrations of PM and ecosystem response.
ES.6 Ecosystem Services
"Ecosystem services" refers to the concept that ecosystems provide benefits to people,
directly or indirectly (Costanza et al.. 2017) and produce socially valuable goods and
services deserving of protection, restoration, and enhancement.
The ecosystem services literature has expanded since the 2008 ISA to include studies that
better characterize ecosystem service valuation and quantification related to acidification
and N enrichment/eutrophication.
Several new studies have paired biogeochemical modeling and benefit transfer equations
informed by willingness-to-pay surveys to estimate the monetary damage done to
ecosystems and the services they provide in the Adirondacks and Shenandoah regions
due to ecosystem acidification (Appendix 14). Despite this progress, for many regions
and specific services, poorly quantified relationships between deposition, ecological
effects, and services are the greatest challenge in developing specific data on the
economic benefits of emission reductions.
ES-20
-------
In the 2008 ISA, there were no publications that had specifically evaluated the effects of
N deposition on ecosystem services associated with N driven enrichment/eutrophication.
Since the 2008 ISA, several comprehensive studies have been published on the
ecosystem services related to excessive N in U.S. water bodies. These include an
evaluation of services affected by multiple N inputs (including N deposition) to the
Chesapeake Bay, a synthesis of the cost-benefits on N loading across the nation, an
estimation of the social cost of nitrogen when applied as fertilizer, and an analysis of how
N lost from its intended area of application (e.g., agricultural fields) affects ecosystem
services of adjacent ecosystems. Most notably, new work identifies over 1,000 links
between N deposition and human beneficiaries.
Considering the full body of literature on ecosystem services related to N and S, the
following conclusions are offered: (1) there is evidence that N and S emissions/deposition
have a range of effects on U.S. ecosystem services and their social value; (2) some
economic studies demonstrate such effects in broad terms, but it remains
methodologically difficult to derive economic costs and benefits associated with specific
regulatory decisions/standards; and (3) numerous, but still inadequately quantified,
relationships are now documented between N and S air pollution and changes in final
ecosystem goods and services.
ES.7 Integrating across Ecosystems
Overall, new evidence since the 2008 ISA increases the weight of evidence for ecological
effects, confirming concepts previously identified and improving quantification of
dose-response (or deposition-ecological indicator) relationships, particularly for N and S
deposition. The ecological effects are described by the causality determinations in
Figure ES-5. which reorganizes the information in Figure ES-3 to show a visualization of
the effects of NOy, SOx, and PM by ecosystem type (e.g., terrestrial, wetland, freshwater,
and estuarine). With this organization, the multiple effects occurring in each ecosystem
due to various pollution combinations of NOy, SOx, and PM are emphasized. Between
two and four different classes of pollutant effects may occur in each ecosystem type in
the U.S. For more information on key messages, see the expanded discussion in the
Integrated Synthesis; detailed information on specific ecosystem types and specific
classes of pollutant effects included in the ISA may be found in the appendices.
ES-21
-------
Ecosystem
Terrestrial Wetland Fresh Water Estuary
Class of Pcfutarvt Effect
Direct . . M-enrichment/ M-enrichment/ . Mercury . . M-enrichment/ , ^Mercury H-ennchment/
.. - Acidification . Suffide Toxicity . AcKfrfcaton Sulfide Toxicity .
Phytotoxe Eutropneation Eutropneaton Methylaton Eutropneaton Metnylaton Eutroph cation
imficator
Gases 4 N+S dep M dep N dep S dep S dep N+S dep M dep s dep 5 dep N dep
01
in
c
8.
ut
&
V
t
0
V
LU
*5
_0j
V
tn
Population
Individual
Productivity
Biodiversity
Growth rate
Physiological
alteration, stress or
injury
Soil or sediment
chemistry
Surface water
chemistry
1 1 1 1 1 —^ 1 l 1
| Causality framework
u
Causal
Suggestive
Inadequate
Not likely
Not evaluated in causal framework
u
* Includes: NO, N02, HNOj, S02, and PAN
Figure ES-5 Causal relationships between the criteria pollutants and ecological effects organized under
ecosystem type.
ES-22
-------
INTEGRATED SYNTHESIS
IS.1 Introduction to This Integrated Science Assessment (ISA)
IS.1.1 Purpose
The Integrated Science Assessment (ISA) is a comprehensive evaluation and synthesis of
the policy-relevant science. Policy-relevant science is that which is "useful in indicating
the kind and extent of all identifiable effects on public health or welfare which may be
expected from the presence of [a] pollutant in the ambient air," as described in
Section 108 of the Clean Air Act (CAA. 1990a).1 This ISA communicates critical science
judgments on the ecological criteria for oxides of nitrogen, oxides of sulfur, and
particulate matter (PM). Accordingly, this ISA is the scientific foundation for the review
of the ecological effects of the current secondary (welfare-based) National Ambient Air
Quality Standards (NAAQS) for oxides of nitrogen, oxides of sulfur, and particulate
matter. The Clean Air Act definition of welfare effects includes, but is not limited to,
effects on soils, water, wildlife, vegetation, visibility, weather, and climate, as well as
effects on man-made materials, economic values, and personal comfort and well-being.
The nonecological welfare effects associated with particulate matter, such as climate and
visibility, are considered part of a separate, ongoing review of PM that is outlined in the
Integrated Review Plan (IRP) for the National Ambient Air Quality Standards for
Particulate Matter (U.S. EPA. 2016d). The human health effects are evaluated in
separate assessments conducted as part of the review of the primary (human
health-based) NAAQS for oxides of nitrogen (U.S. EPA. 2016f). oxides of sulfur (U.S.
EPA. 2016e). and as noted above, particulate matter (U.S. EPA. 2019).
Oxides of nitrogen, oxides of sulfur, and particulate matter are reviewed here together
because they are interrelated through complex chemical and physical atmospheric
processes and because they all contribute to nitrogen (N) and sulfur (S) deposition, which
in turn contributes to well-documented ecological effects. In this document, the term
"oxides of nitrogen" refers to all forms of oxidized nitrogen (NOy) compounds, including
1 The general process for developing an ISA, including the framework for evaluating weight of evidence and
drawing scientific conclusions and causal judgments, is described in a companion document. Preamble to the
Integrated Science Assessments (U.S. EPA. 2015e).
IS-1
-------
NO, NO2, and all other oxidized N containing compounds formed from NO and NO2.1
Oxides of sulfur2 are defined here to include sulfur monoxide (SO), sulfur dioxide (SO2),
sulfur trioxide (SO3), disulfiir monoxide (S2O), and sulfate (S042 ). However, SO, SO3,
and S2O are present at much lower ambient levels than SO2 and SO42 and are therefore
not discussed further. Particulate matter is composed of some or all of the following
components: nitrate (NO;, ). SO42 . ammonium (NH4+), metals, minerals (dust), and
organic and elemental carbon (C).
This ISA updates the 2008 Integrated Science Assessment for Oxides of Nitrogen and
Sulfur—Ecological Criteria [hereafter referred to as the 2008 ISA (U.S. EPA. 2008a)l. as
well as the ecological portion of the Integrated Science Assessment for Particulate Matter
(U.S. EPA. 2009a). with studies and reports published from January 2008 through May
2017. Thus, this ISA updates the state of the science that was available for the 2008 ISA,
which informed decisions on the secondary oxides of nitrogen and oxides of sulfur
NAAQS in the review completed on March 20, 2012. In the final rulemaking, the
Administrator's decision was that, while the current secondary standards were inadequate
to protect against adverse effects from deposition of oxides of nitrogen and oxides of
sulfur, it was not appropriate under Section 109(b) to set any new secondary standards at
this time due to the limitations in the available data and uncertainty as to the amount of
protection the metric (Aquatic Acidification Index—see Section IS.2.2.6) developed in
the Policy Assessment (U.S. EPA. 2011a) would provide against acidification effects
across the country (77 FR 20281). In addition, the Administrator decided that it was
appropriate to retain the current nitrogen dioxide (NO2) and sulfur dioxide (SO2)
secondary standards to address direct effects of gaseous NO2 and SO2 on vegetation.
Thus, taken together, the Administrator decided to retain and not revise the current NO2
and SO2 secondary standards: an NO2 standard set at a level of 0.053 ppm, as an annual
arithmetic average, and an SO2 standard set at a level of 0.5 ppm, as a 3-hour average, not
to be exceeded more than once per year (77 FR 20281). The current secondary standards
for PM are intended to address PM-related welfare effects, including visibility
impairment, ecological effects, and effects on materials and climate. These standards are
a 3-year annual mean PM2.5 concentration of 15 (ig/m3, with the 24-hour average PM2.5
and PM10 set at concentrations of 35 (ig/m3 and 150 |ig/nr\ respectively.
1 This ISA reserves the abbreviation NOx strictly as the sum of NO and NO2—consistent with that used in the
atmospheric science community—and uses the term "oxides of nitrogen" to refer to the broader list of oxidized
nitrogen species. Oxides of nitrogen refers to NOy as the total oxidized nitrogen in both gaseous and particulate
forms. The major gaseous and particulate constituents of NOy include nitric oxide (NO), nitrogen dioxide (NO2),
nitric acid (HNO3). peroxyacetyl nitrate (PAN), nitrous acid (HONO), organic nitrates, and particulate nitrate (NO3).
This ISA uses the definitions adopted by the atmospheric sciences community.
2 Oxides of sulfur refers to the criteria pollutant category.
IS-2
-------
This new review of the secondary oxides of nitrogen, oxides of sulfur, and particulate
matter NAAQS is guided by several policy-relevant questions that were identified in The
Integrated Review Plan for the Secondary National Ambient Air Quality Standard for
Nitrogen Oxides, Sulfur Oxides, and Particulate Matter [hereafter referred to as the 2017
IRP (U.S. EPA. 2017c)I.
To address these questions, this ISA aims to characterize the evidence available in the
peer-reviewed literature for ecological effects associated with:
• the major gaseous and particulate constituents of total oxidized N (NOy), which
include NO, NO2, HNO3, PAN, HONO, organic nitrates, and NO;, :
• the major gaseous and particulate constituents of SOx, which include SO2 and
SO42 ; and
• PM composed of some or all of the following components: particulate NO3 ,
particulate SO42 , ammonium (NIL+), metals, minerals (dust), and organic and
elemental carbon (C).
The assessment activities include:
• Identifying policy-relevant literature.
• Evaluating strength, limitations, and consistency of findings.
• Integrating findings across scientific disciplines and across related ecological
outcomes.
• Considering important uncertainties identified in the interpretation of the scientific
evidence.
• Assessing policy-relevant issues related to quantifying ecological risks, such as
ambient air concentrations, deposition, durations, and patterns associated with
ecological effects; the relationship between ambient air concentrations, deposition,
and ecological response and the existence of thresholds below which effects do
not occur; and species and populations potentially at increased risk of ecological
effects.
New analyses with the goal of quantifying risk, such as new model runs, Critical Loads
(CLs) exceedance maps, and quantified uncertainties regarding modeled scenarios are not
conducted in the ISA. These types of analyses, if pursued, require the selection of
chemical or biological limits that define CLs and represent adversity. These analyses
would also require choosing a time period over which to average deposition. Such
scope-of-analysis decisions are more appropriate for the Risk and Exposure Assessment,
as described in the 2017 IRP (U.S. EPA. 2017c). The information summarized in this ISA
will serve as the scientific foundation of the Risk and Exposure and Policy Assessments
during the current review of the secondary oxides of nitrogen, oxides of sulfur, and
particulate matter NAAQS.
IS-3
-------
IS.1.2 Process and Development
The U.S. EPA uses a structured and transparent process to evaluate scientific information
and determine the causality of relationships between air pollution and ecological effects
[see Preamble (U.S. EPA. 2015e)l. The ISA development includes approaches for
literature searches, criteria for selecting and evaluating relevant studies, and a framework
for evaluating the weight of evidence and forming causal determinations. As part of this
process, the ISA is reviewed by the public and by the Clean Air Scientific Advisory
Committee (CASAC), which is a formal independent panel of scientific experts. This ISA
informs the review of the secondary oxides of nitrogen, oxides of sulfur, and particulate
matter NAAQS and therefore integrates and synthesizes information characterizing NOy,
SOx, and PM air concentrations. It also examines deposition of these substances and their
ecological effects. Relevant studies include those examining atmospheric chemistry,
spatial and temporal trends, and deposition, as well as U.S. EPA analyses of air quality
and emissions data. Relevant ecological research includes geochemistry, microbiology,
physiology, toxicology, population biology, and community ecology. The research
includes experimental laboratory and field additions of the pollutants, as well as gradient
studies.
The U.S. EPA conducted literature searches to identify relevant peer-reviewed studies
published since the previous ISA (i.e., from January 2008 through May 2017;
Figure IS-1).
IS-4
-------
HERO = Health and Environmental Research Online; ISA = Integrated Science Assessment.
Figure IS-1 Workflow for collecting relevant literature for the 2017 Integrated
Science Assessment for Oxides of Nitrogen, Oxides of Sulfur, and
Particulate Matter—Ecological Criteria.
Multiple search methods were used in the Web of Science database [Preamble (U.S.
EPA. 2015e). Appendix 21. including searches by keyword and by citations of 2008 ISA
references. Subject-matter experts and the public were also permitted to recommend
studies and reports during kick-off workshops held by the U.S. EPA in March 2014 for
oxides of nitrogen and oxides of sulfur and in February 2015 for particulate matter. The
new references were sorted by automated methods into topic areas based on wording in
the publication's abstract or numbers of citations of 2008 ISA references, and the
resultant relevant literature was reviewed by the ISA authors. Studies were screened first
based on the title and then by the abstract; studies that did not address a relevant research
topic based on this screening were excluded. The U.S. EPA also identified studies from
previous assessments as definitive works on particular topics to include in this ISA. The
HERO project page for this ISA
(https://heronet.epa.gov/heronet/index.cfin/proiect/page/proiect id/2965) contains the
references that are cited in the ISA and electronic links to bibliographic information and
abstracts.
IS-5
-------
The Preamble to the Integrated Science Assessments (U.S. EPA. 2015e) describes the
general framework for evaluating scientific information, including criteria for assessing
study quality and developing scientific conclusions. For ecological studies, emphasis is
placed on studies that characterize quantitative relationships between criteria pollutants
and ecological effects that occur at concentration and deposition levels relevant to current
ambient levels in the U.S. However, experimental studies with higher exposure
concentrations are included if they contribute to an understanding of mechanisms.
This ISA draws conclusions about relationships between NOy, SOx, and PM and
ecological effects by integrating information across scientific disciplines and related
ecological outcomes and synthesizing evidence from previous and recent studies.
Determinations are made about causation, not just association, and are based on
judgments of consistency, coherence, and scientific plausibility of observed effects, as
well as related uncertainties. The ISA uses a formal causal framework [Table II of the
Preamble (U.S. EPA. 2015e)l. which is based largely on the aspects for causality
proposed by Sir Austin Bradford Hill to classify the weight of evidence according to the
five-level hierarchy summarized below.
• Causal relationship
• Likely to be a causal relationship
• Suggestive of, but not sufficient to infer, a causal relationship
• Inadequate to infer the presence or absence of a causal relationship
• Not likely to be a causal relationship
APPENDIX 7This ISA includes the Preface (legislative requirements and history of the
secondary oxides of nitrogen, oxides of sulfur, and particulate matter NAAQS), an
Executive Summary, an Integrated Synthesis, and 16 appendices. The general process for
developing an ISA is described in a companion document, Preamble to the Integrated
Science Assessments (U.S. EPA. 2015e). The Integrated Synthesis summarizes the
scientific evidence that best informs policy-relevant questions that frame this review.
Appendix 1 is an introduction to the appendices. Appendix 2 characterizes the sources,
atmospheric processes, and the trends in ambient concentrations and deposition of NOy,
SOx, and PM. Appendix 3 describes direct effects of NOy and SOx gases on plants and
lichens. Appendix 4-Appendix 6 describe N and S deposition effects on terrestrial
biogeochemistry and the terrestrial biological effects of terrestrial acidification and N
enrichment. Appendix 7 describes the effects of N and S deposition on aquatic
biogeochemistry. Appendix 8-Appendix 10 characterize the biological effects of
freshwater acidification, freshwater N enrichment, and N enrichment in estuaries and
near-coastal systems. Appendix 11 describes the effects of N deposition on wetlands, and
Appendix 12 characterizes the ecological effects of S as a nutrient. Appendix 13 presents
IS-6
-------
information on climate modification of ecosystem response to N and S, and Appendix 14
discusses ecosystem services. Appendix 15 is a review of the ecological effects of forms
of PM that are not related to N or S deposition. Finally, Appendix 16 presents case
studies for six locations in the U.S. (southern/central California, northeastern U.S., Rocky
Mountain National Park, southeastern Appalachia, Tampa Bay, and the Adirondacks)
where data are sufficient to well characterize the ecological effects of N and S deposition.
These sites would therefore make good candidates for further study to better understand
the linkages across various effects and ecosystems and to better assess risk and exposure.
IS.2 Connections, Concepts, and Changes
IS.2.1 Connections
Although scientific material in this ISA is divided into separate appendices for
atmospheric science and the multiple ecological effects, the strong links between the
atmosphere and terrestrial and aquatic ecosystems are acknowledged (Figure IS-2).
Emissions of NOy, SOx, and PM contribute to an accumulation of N and S in the
environment that creates a multitude of effects on terrestrial, wetland, and aquatic
ecosystems. Nitrogen is a vital component of all biological systems, serving as an
essential element to molecules such as amino acids and nucleic acids, which are among
the biochemical building blocks of life. As an organizing concept to understand the
effects of N within the environment, the sequence of transfers, transformations, and
environmental effects has been described as the ""N cascade" (Galloway and Cowling.
2002). The concept of cascading effects also applies to S, which is also an essential
macronutrient. Specifics of biogeochemical cycling and biological effects of N are
discussed in Section IS.5 and for S are discussed in Section IS.9.
IS-7
-------
Ecosystem Services
Appendix 14
Biological Effects
of S-riutrient
Appendix 12
Gas-phase
Ecological Effects
Appendix 3
Oxidation
so2—»-h2so4
NOx—>¦ HNOj
Dissolution
2H++ S042"
->-H++N0r
Other Ecological
Effects of PM
Appendix 15
Dry deposition
NO ,NH , SO
Biological Effects of
Terrestrial Acidification
Appendix 5
Biological Effects of
Terrestrial N Enrichment
Appendix 6
n2o no,
Wetlands
Appendix 11
Soil BGC
Appendix 4
Aquatic BGC
Appendix 7
Wet deposition
H+, NH4+, N03", SQ,2"
Biological Effects of
Freshwater Acidification
Appendix 8
Biological Effects
of Freshwater j
N Enrichment
Appendix 9
Biological Effects of
Estuarine N Enrichment
Appendix 10
Deposition
Ecological
Effect
Atmospheric Sciences
Appendix 2
Climate Modification
of Ecological Effects
Appendix 13
Ambient Air
Concentration
Ca2+ = calcium ion; GHG = greenhouse gas; H+ = hydrogen ion; H2S04 = sulfuric acid; HN03 = nitric acid; Mg2+ = magnesium ion;
N20 = nitrous oxide; N = nitrogen; NH3 = ammonia; NH4+ = ammonium; NHX = NH3 + NH4* + reduced organic nitrogen compounds;
NO = nitric oxide; N02 = nitrogen dioxide; N03" = nitrate; NOx = NO + N02; PAN = peroxyacetyl nitrate; PM = particulate matter;
S02 = sulfur dioxide; S042" = sulfate; SOx = S02 + S042"; VOC = volatile organic compounds.
The sum of reactive oxidized nitrogen species is referred to as NOY (NOY = NO + N02 + HN03 + 2N205 + HONO + N03" + N20
PAN + other organic nitrates).
Although not explicitly indicated, wet and dry deposition of PM components (e.g., metals, minerals, and secondary organic aerosols)
also occur and contribute to ecological effects.
Source: Modified from U.S. EPA (2008a).
Figure IS-2 Overview of atmospheric chemistry, deposition, and ecological
effects of emissions of oxides of nitrogen, oxides of sulfur, and
reduced nitrogen.
IS.2.2 Concepts
This ISA draws on many methodological approaches and disciplines within the larger
scientific fields of ecology and atmospheric sciences. The studies discussed herein are
best understood in the context of some general concepts within these fields, such as
ecosystem scale, structure, and function (Section IS.2.2.1); deposition and source
apportionment to ecosystems (Section IS.2.2.2); critical loads (Section IS.2.2.3);
biodiversity (Section IS.2.2.4); the effects of reduced versus oxidized forms of N
IS-8
-------
(Section IS.2.2.5); and the metric developed in the previous secondary NAAQS review,
the Aquatic Acidification Index (AAI; Section IS.2.2.6). The topics discussed in this
"Concepts" section do not have separate sections dedicated to them in the Integrated
Synthesis. The topics of ecosystem recovery, ecosystem services, and uncertainty, while
conceptual in nature, are not discussed here because they are the focus of more detailed
discussions in Section IS. 11. Section IS. 13. and Section IS. 14. respectively.
Ecosystem structure comprises both biodiversity and geography. Biodiversity
encompasses many quantitative measures of the abundance and distribution of organisms
within a defined geographical area (for a more explicit definition, see Section IS.2.2.1
and Section IS.2.2.4). Ecosystem function refers to processes that control fluxes and
pools of matter and energy in the ecosystem (Section IS.2.2.1). The loss of biodiversity is
a key consequence of the air pollutants discussed in this ISA. The importance of
preserving biodiversity and ecosystem function contributes to the sustainability of
ecosystem services that benefit human welfare and society in general (Section IS.2.2.4
and Appendix 14).
In human health assessments, dose-response relationships are used to identify
quantitative relationships between chemical exposure (dose) and health outcomes
(response), with emphasis on identifying thresholds, or the lowest doses at which
negative health outcomes are observed. In ecology, CLs provide a similar quantitative
relationship between chemical dose (e.g., deposition) and specific, quantitative changes
in ecological properties or processes (Section IS.2.2.3). For CLs to be used in evaluating
the effects of deposition upon ecosystems that receive N or S from multiple sources,
those other sources must be considered in comparison to deposition level
(Section IS .2.2.2). as well as the heterogeneous sensitivities of organisms and ecosystems
to different chemical forms of deposition (Section IS.2.2.5).
IS.2.2.1 Ecosystem Scale, Structure, and Function
For this assessment, an ecosystem is defined as the interactive system formed from all
living organisms (biota) and their abiotic (chemical and physical) environment within a
given area (IPCC. 2007a). Ecosystem spatial boundaries are somewhat arbitrary,
depending on the focus of interest or study. Thus, the spatial extent of an ecosystem may
range from very small, well-circumscribed systems such as a small pond, to biomes at the
continental scale, or the entire globe (U.S. EPA. 2008a). Ecosystem spatial scale does not
always correlate with complexity. A small pond may be a complex system with multiple
trophic levels ranging from phytoplankton to invertebrates to several feeding guilds of
fish. A large lake, on the other hand, may be a very simple ecosystem, such as the Great
IS-9
-------
Salt Lake in Utah that covers approximately 1,700 square miles but contains only
bacteria, algae, diatoms, and two invertebrate species (U.S. EPA. 2013b). All ecosystems,
regardless of size or complexity, have multiple interactions between biota and abiotic
factors. Ecosystems include both structural (geography and biodiversity [e.g., soil type
and food web trophic levels]) and functional (flow of energy and matter
[e.g., decomposition, nitrification]) attributes. Ecosystem changes are often considered
undesirable if important structural or functional components of the ecosystems are altered
following pollutant exposure (U.S. EPA. 2013b. 1998a).
Biotic or abiotic structure may define an ecosystem. Abiotic structure includes climatic
and edaphic components. Biotic structure includes species abundance, richness,
distribution, evenness, and composition, measured at the population, species, community,
ecosystem, or global scale. A species (for eukaryotic organisms) is defined by a common
morphology, genetic history, geographic range of origin, and ability to interbreed and
produce fertile offspring. A population consists of interbreeding groups of individuals of
the same species that occupy a defined geographic space. Interacting populations of
different species occupying a common spatial area form a community (Barnthousc ct al..
2008). Community composition may also define an ecosystem type, such as a pine forest
or a tall grass prairie. Pollutants can affect the ecosystem structure at any of these levels
of biological organization (Suter et al.. 2005).
Individual plants or animals may exhibit changes in metabolism, enzyme activities,
hormone function, or may suffer gross lesions, tumors, deformities, or other pathologies.
However, only some organism-level endpoints affected by pollution, such as growth,
survival, and reproductive output, have been definitively linked to effects at the
population level and above (U.S. EPA. 2013b). Population-level effects of pollutants
include changes over time in abundance or density (number of individuals in a defined
area), age or sex structure, and production or sustainable rates of harvest (Barnthouse et
al.. 2008). Community-level attributes affected by pollutants include species richness,
species abundance, composition, evenness, dominance of one species over another, or
size (area) of the community (U.S. EPA. 2013b). Pollutants may affect communities in
ways that are not observable in organisms or populations (Bartell. 2007). including
(1) effects resulting from interactions between species, such as altered predation rates or
competitive advantage; (2) indirect effects, such as reducing or removing one species
from the assemblage and allowing another to emerge (Petraitis and Latham. 1999); and
(3) alterations in trophic structure.
Alternatively, ecosystems may be defined on a functional basis. "Function" refers to the
suite of processes and interactions among the ecosystem components that involve energy
or matter. Examples include water dynamics and the flux of trace gases such as rates of
IS-10
-------
photosynthesis, decomposition, nitrification, or carbon cycling. Pollutants may affect
biotic structure indirectly. For example, a pollutant may first alter abiotic conditions
(e.g., soil chemistry), which in turn influences biotic structure and function (Bartell.
2007).
Some ecosystems, and some aspects of particular ecosystems, are less vulnerable to
long-term consequences of pollutant exposure. Other ecosystems may be profoundly
altered if a single attribute is affected by pollution. Thus, spatial and temporal definitions
of ecosystem structure and function become essential factors in defining affected
ecosystem services and in determining CLs for certain pollutants, either as single
pollutants or in combination with other stressors.
The main causal determinations of this ISA (Section IS.2.3) are that N and S deposition
affect ecosystem structure, with effects ranging from biogeochemical alterations in soil
and water chemistry to multiple levels of biological organization, including species-level
alterations of physiological processes and shifts in biodiversity and ecological function.
IS.2.2.2 Deposition and Source Contribution of Nitrogen (N) and Sulfur (S) to
Ecosystems
Deposition of N and S results from a variety of human activities and atmospheric
processes. Emissions from stationary, mobile, and agricultural sources undergo
atmospheric transformation (Section IS.3.1) to form products that are eventually
deposited out of the air onto the land or waterscape (Section IS.3.3). The contribution of
atmospheric deposition to total loading for N and S varies within and among terrestrial,
wetland, freshwater, and estuarine ecosystems.
In the 2008 ISA, atmospheric deposition was identified as the main source of
anthropogenic N to unmanaged terrestrial ecosystems. This conclusion has been
confirmed by new studies on N sources to lands and waterways (Appendix 4.2). Across
all watersheds, atmospheric N deposition is the second largest overall human-mediated N
source; agriculture is the largest, and the largest N source to 33% of watersheds. Current
deposition levels in the U.S. are discussed in Appendix 2 and Section IS.3.3. No new
information has been published on nonatmospheric sources of S in terrestrial ecosystems
(Appendix 4.2); S inputs from the atmosphere are discussed in Appendix 2 and
Section IS.3.3.
In the 2008 ISA, atmospheric deposition was also identified as the main source of N to
some freshwater ecosystems, including headwater streams, high-elevation lakes, lower
order streams in undisturbed areas, and freshwater wetlands (e.g., bogs and fens).
Evidence for the influence of N deposition on water chemistry has been further supported
IS-11
-------
by new studies that quantify the contribution of N deposition to total N loading in
freshwater lakes and streams, and which quantify atmospheric contributions during storm
events (Table 7-1). As shown in these studies, deposition can represent a substantial
portion of total N loading to surface waters. However, other nonpoint and point sources
of N dominate N inputs to high-order streams.
In fresh surface waters and wetlands, S that contributes to enrichment induces acidifying
effects. Sources of S include weathering of minerals in sediments and rocks, leaching
from terrestrial S cycling, internal cycling, and direct atmospheric deposition. The 2008
ISA showed that drought can release S stored in wetlands or lake sediments because
bound sulfide (S2 ) is exposed to atmospheric oxygen and oxidized to SO42 .Increases in
waterborne SO42 concentration through various concurrent processes has been observed
as a result of drought in whole-lake observational research (93% increase in Little Rock
Lake, WI, from 1.5 to 2.9 mg/L), and in response to variation in water levels from
climate change-induced droughts in modelling using Model of Acidification of
Groundwater in Catchments (MAGIC). New evidence confirms that fluctuating water
levels in wetlands increase SO42 concentration in pulses following water level recovery.
The importance of atmospheric deposition as a cause of estuarine eutrophication is
determined by the relative contribution of the atmospheric versus nonatmospheric sources
of N input. Sources of N in coastal areas may include direct deposition to the water
surface, coastal upwelling from oceanic waters, and transport from watersheds.
Freshwater inflows to estuaries often transport N from agriculture, urban, wastewater,
and atmospheric deposition sources. Atmospheric deposition constitutes less than half of
the total N supply in most, but not all, estuaries (Table 7-9). Both point sources and
nonpoint sources (including runoff, as well as atmospheric deposition) have been
identified as targets for mitigation of N loading in coastal areas. Seawater contains high
concentrations of SO42 , so atmospheric inputs of S are unlikely to contribute
substantially to biogeochemical or biological effects in coastal areas.
IS.2.2.3 Critical Loads Concept and General Approaches
The following section provides a discussion of important concepts regarding Critical
Loads (CLs). The definition of a CL is, "a quantitative estimate of an exposure to one or
more pollutants below which significant harmful effects on specified sensitive elements
of the environment do not occur according to present knowledge" (Nilsson and Grennfelt.
1988). This definition is intended as background material to support a better
understanding of the CL calculations presented throughout the ISA. The main concepts
presented here include CLs as an organizing principle, CL heterogeneity across the
IS-12
-------
landscape, more than one CL for a given location, the pros and cons of methods used to
calculate CLs (e.g., empirical, steady state, and dynamic), and a comparison of CLs
versus target loads. Uncertainty in calculating CLs is discussed in Section IS. 13.
Throughout this ISA, the CL concept is used as an organizing principle to relate
atmospheric deposition to ecological endpoints that indicate impairment. The
development of a quantitative CL estimate requires a number of steps. An illustrative
example of the eight general steps is shown in Figure IS-3.
*
Al = aluminum; ANC = acid-neutralizing capacity; C = carbon; Ca = calcium; L = liter; |jeq = microequivalents; N = nitrogen;
NH4 = ammonium; N03 = nitrate; S04 = sulfate.
Source: U.S. EPA (2008al
Figure IS-3 An example of the matrix of information considered in defining
and calculating critical loads (see discussion in text). Note that
multiple alternative biological indicators, critical biological
responses, chemical indicators, and critical chemical limits could
be used.
1) Disturbance
Acidification
Eutrophication
2) Receptor
Forest
Lake
Grassland
Lake
3) Biological
indicator
Sugar
Maple
Norway
Spruce
Brook trout
Fish species
richness
Species
diversity
Primary
productivity
4) Critical
biological
response
Failure to
reproduce
Seedling
death
Presence
absence
Species
loss
Species
loss
Excess
productivity
5) Chemical
indicator
Soil % Base
Saturation
Soil Ca/AI
ratio
Lake water
ANC
Lakewater
ANC
Soil C/N
ratio
Lakewater
N03
6) Critical
chemical
limit
10%
1.0
0 peq/L
50 peq/L
20
10 peq/L
7) Atmospheric
pollutant
C/)
o
z ^
Ca>
S04, N03,
nh4
S04, N03,
nh4
S04, N03,
nh4
no3, nh4
no3, nh4
8) Critical
pollutant load
???
???
???
???
???
???
It is important to recognize that there is no single "definitive" CL for an ecological effect.
CL estimates reflect the current state of knowledge and the selected limits, indicators, and
responses. Changes in scientific understanding may include, for example, new
dose-response relationships, better resource maps and inventories, larger survey data sets,
continuing time-series monitoring, and improved numerical models.
IS-13
-------
Calculating multiple CLs for a given pollutant at a single location is not uncommon
because of the nested sequence of disturbances, receptors, and biological indicators
considered for a given pollutant. Multiple CL values may also arise from an inability to
agree on a single definition of "harm/' Calculation of CLs for multiple definitions of
"harm" may be deemed useful in subsequent discussions of the analysis and in the
decision-making steps that may follow CL calculation.
The heterogeneity of natural environments can affect responsiveness of ecosystems to
deposition load. For example, the high spatial variability of soils almost guarantees that
for any reasonably sized soil-based "receptor" that might be defined in a CL analysis,
there will be a continuum of CL values for any indicator chosen. Although the range of
this continuum of values might be narrow, there is nevertheless an a priori expectation in
any CL analysis that multiple values (or a range of values) will result from the analysis.
Given the heterogeneity of ecosystems affected by N and S deposition, published CL
values for locations in the U.S. vary depending on both biological and physical factors.
The three approaches to developing CLs (i.e., empirical observation, steady-state
modeling, and dynamic modeling) each have strengths and limitations. It is suggested
that the combined approach of calculating CLs from biogeochemical simulation models
in conjunction with empirical analyses is the most effective way to characterize the
effects of deposition to a given environment (Fenn et al.. 2015). For all three types of
models, spatial boundaries of where to apply a CL are important. For example, a CL may
apply to a watershed, ecoregion, or species range, depending on how the CL is defined.
An important advantage of empirical CLs is that they are based on measured
(vs. modeled) changes in ecological variables in response to inputs. Consequently, the
links between deposition and the measured response variable are direct; full process-level
knowledge is not required. Empirical CLs are important for validating CL values
determined with models (Fenn et al.. 2015).
Fenn et al. (2015) discussed that the advantages of models, "are that ecosystem responses
to alternative scenarios can be tested. These might include changes in atmospheric
deposition, disturbance or climatic conditions, and responses to silvicultural treatments,
grazing, fire, and other disturbances. Simulation modeling allows temporal aspects of
ecosystem response in relation to CLs and CL exceedances to be evaluated, including
evaluation of historical and future conditions."
Two key ways that steady-state and dynamic models differ in their modeling of CLs is by
how they assume ecosystem equilibrium and by the amount of input data they need for
parameterization.
IS-14
-------
Steady-state models assume that the ecosystem is in equilibrium with the CL of
deposition; therefore, the long-term sustainable deposition, that is the CL, is indicated.
This is the relevant information needed to provide protection from deposition in
perpetuity as the system comes into equilibrium with the pollutant CL. In the U.S., few
(if any) ecosystems qualify as steady-state systems. Therefore, the assumption of
equilibrium in the steady-state model is often false. The steady-state models give no
information concerning the time to achieve the equilibrium or what may happen to the
receptor along the path to equilibrium. The recovery of an ecosystem based on a CL from
a steady-state model may take several hundred years. In other words, the assumption that
attainment of a deposition value below the steady-state CL will result in biological
recovery within a specified time period may not be valid. Dynamic models calculate
time-dependent CLs and, therefore, do not assume an ecosystem that is in equilibrium.
The time-dependent calculation is relevant information to provide protection from
damage by the pollutant within a specific time frame. Generally, the shorter the time
frame selected, the lower the CL.
Data requirements for steady-state models tend to be much lower than for dynamic
models. Therefore, the data required to conduct dynamic modeling are not available for
as many places as the data required to conduct steady-state modeling. The few
national-scale modeling efforts for both terrestrial and aquatic acidification are both done
with steady-state models for this reason.
The results of all three CL approaches are difficult to extrapolate across geographic
space. Spatially, variation in biological and biogeochemical processes imposed by
climate, geology, biota, and other environmental factors may alter the
deposition-response relationship. Empirical CLs may only be applied with confidence to
sites with highly similar biotic and environmental conditions (Pardo et al.. 2011a). This is
particularly problematic in areas where deposition has received sparse research
attention—as is sometimes the case for CLs of N deposition related to N driven
eutrophication (Appendix 6.4). Models may be run at different locations, but the data
needed to parameterize them is not always available.
CLs are different from target loads. Fenn et al. (2011b) defined the "target load" as
follows: "The acceptable pollution load that is agreed upon by policy makers or land
managers. The target load is set below the CL to provide a reasonable margin of safety,
but could be set higher than the CL at least temporarily/' Target loads are selected based
on the level of ecosystem protection desired, economic considerations, and stakeholder
input at a given location.
IS-15
-------
IS.2.2.4 The Importance of Biodiversity
There are causal relationships between additions of N and/or S to an ecosystem and
biodiversity loss in terrestrial, freshwater, wetland, and estuarine ecosystems in the U.S.
(Table IS-1). What does it mean to lose biodiversity? Biodiversity loss not only means
the extirpation of unique living species, it represents the potential loss of ecosystem
function and ecosystem services, as shown by several decades of research in a wide
variety of natural svstcms(Hoopcr et al.. 2012; Balvanera et al.. 2006; Tilman. 2000).
Numerous studies demonstrate that the number and diversity of organisms in a system
control the abundance of habitat for other species, the biogeochemical cycling of
nutrients and carbon, and the efficiency at which biotic systems are able to transform
limited resources into biomass (Cardinale et al.. 2011). Among plant communities, higher
biodiversity leads to higher overall plant productivity and greater retention of soil
nutrients (Reich et al.. 2012; Tilman. 2000). In multitrophic systems, higher prey
diversity leads to both higher predator growth rates and a smaller impact of predation on
prey abundance (Duffy et al.. 2007). Positive impacts of biodiversity on ecosystem
services have been documented in forests (Gamfeldt et al.. 2013; Zhang et al.. 2012b).
grasslands (Tilman et al.. 2012). arid and semiarid ecosystems (Maestre et al.. 2012). and
marine systems (Gamfeldt et al.. 2015; Worm et al.. 2006) and include effects such as
greater carbon storage, fruit production, wood production, and nutrient cycling. In marine
ecosystems, biodiversity loss has been linked to increased rates of exponential decreases
in water quality through metrics such as higher numbers of beach closures and harmful
algal blooms [HABs; Worm et al. (2006)1. Notably, HABs are linked to increased disease
prevalence among humans, domestic animals/pets, and aquatic organisms (Johnson et al..
2010). In addition to the relationship between HABs and disease, there is now empirical
evidence from many ecosystems of a broader link between declines in biodiversity and
increased transmission and severity of disease (Johnson et al.. 2015b) caused by plant,
wildlife, and human pathogens. As a whole, these decades of research have produced an
overwhelming body of evidence indicating that the loss of biodiversity risks a
deterioration of the ecosystem goods and services on which humanity depends on
(Gamfeldt et al.. 2015; Cardinale et al.. 2012).
One of the most important consensus observations in biodiversity research is that
ecosystem processes are more stable (have less temporal variability) at higher levels of
diversity (Cardinale et al.. 2012; McCann. 2000; Naeem and Li. 1997; Tilman and
Downing. 1994). This stability occurs because species respond differently to
environmental variation. In diverse communities, it is more likely that declines in the
growth of one species caused by an environmental change will provide more resources
for competing species (Cardinale et al.. 2012; Tilman. 2000). This property was predicted
by economists and is similar to how more diversified investment portfolios provide
IS-16
-------
enhanced stability under fluctuating market conditions (Doak et al.. 1998; Tilman ct al..
1998). Notably, there is also consensus that the impact of biodiversity on ecosystem
processes is nonlinear, wherein declines in ecosystem processes accelerate as the number
of species in a system declines (Card in ale et al.. 2012). Accelerating ecosystem service
declines in response to species loss may be because different ecosystem functions require
the presence of different sets of species (Isbell et al.. 2015; Reich et al.. 2012; Zavaleta et
al.. 2010). The increased stability of diverse ecosystems makes these systems less
vulnerable to environmental change or collapse caused by external forces such as drought
or human disturbance (Isbell et al.. 2015; Tilman et al.. 2012; Isbell et al.. 2011; Worm et
al.. 2006). For example, coastal systems with higher species diversity had lower rates of
fishery collapse and extinction for commercially important fish and invertebrate species,
and large marine ecosystems with higher fish diversity recovered more quickly from
collapse (Worm et al.. 2006). Thus, there is strong evidence that high biodiversity helps
sustain ecosystem services and makes these ecosystem services more resilient to
environmental change.
IS.2.2.5 Reduced versus Oxidized Nitrogen Effects across Ecosystems
Individual biochemical and geochemical processes involve specific chemical forms of N,
suggesting that there may be consequences in many ecosystems from the ongoing trend
of decreasing NOy deposition and increasing NHx deposition in many parts of the U.S.
(Section IS.3). The largest body of evidence that the effects of reduced versus oxidized N
may have different consequences for ecological structure and function is for estuaries
where the form of N delivered to some coastal areas of the U.S. is shifting from primarily
NO;, to an increase in reduced forms of N. Although unlikely to be attributed solely to
atmospheric sources due to the large contribution of N from wastewater, agriculture, and
other sources, inputs of ammonia (NH3) and NH4+ selectively favor specific
phytoplankton functional groups (e.g., cyanobacteria, dinoflagellates) including harmful
species (Figure 10-7). Shifts in phytoplankton community composition to species that
respond strongly to reduced N have been observed in some coastal regions
(Appendix 10.3.2). Growth of some species of phytoplankton (Appendix 10.2.2) and
macroalgae (seaweed; Appendix 10.2.3) appear to be related to the form of N. There is
also increasing evidence in freshwater systems for the importance ofN in harmful algal
blooms (HABs), and several studies have shown that the form of N influences freshwater
algal species composition (Appendix 9.2.6.1). In terrestrial systems, oxidation-reduction
status of inorganic N seems to have little influence on the biological responses to N
deposition (Appendix 4.3.12).
IS-17
-------
Because some soil biogeochemical processes involve specific chemical forms of N
(e.g., denitrification, ammonium toxicity), there is the potential that biological responses
to N deposition (or N addition) could depend on whether the dominant form of deposited
N is oxidized (NOy) or reduced (NHx). Different responses to individual forms of N have
been observed for some soil biogeochemical processes (Table 4-13) and terrestrial
biological responses (Table 6-1). Moreover, a number of individual studies have
observed differential effects of NH44" versus NO;, additions on plant community diversity
[e.g., Kleiin et al. (2008); Dias et al. (2014)1. In general, however, meta-analyses in the
literature have tended to find no difference in the effects of individual forms of N on
terrestrial biological endpoints like plant productivity or microbial biomass (Table 6-1).
This result suggests that terrestrial community diversity is also generally not
differentially affected by the form of N, possibly because plant uptake of N is mediated
by soil biogeochemical cycles that often rapidly transform N between oxidized and
reduced forms.
Evidence of wetland responses to different chemical forms of N come primarily from N
addition experiments conducted outside of the U.S. In European bogs and fens, both
forms of N addition decreased ecosystem N retention, but oxidized N addition caused
dissolved organic nitrogen (DON) leaching, while reduced N caused dissolved inorganic
nitrogen (DIN) leaching as well as cation leaching (Appendix 11.3.1.6). Reduced N
caused greater physiological stress or injury than equivalent loads of oxidized N in moss
species (Appendix 11.4.5 and Appendix 11.5.5).
IS.2.2.6 Aquatic Acidification Index (AAI)
The 2017 IRP (U.S. EPA. 2017c) described the Aquatic Acidification Index (AAI) to be
a novel approach for a multipollutant standard intended to address deposition-related
effects. Scientifically, the AAI represented an advancement in ecological methodology to
(1) calculate CLs for aquatic acidification on a national scale, when previously CLs had
been calculated on the spatial scale of a watershed and (2) provide a uniform level of
ecological protection at the national scale. These advancements were accomplished by
first aggregating CLs calculated for the same chemical limit within a defined spatial
region. Next, the distribution of the "population" of CL values was evaluated, and the
percentage of water bodies to protect was selected as a potential method to evaluate
different conservation targets. The AAI also presented novel advancements in
atmospheric sciences, including (1) using transference ratios to relate atmospheric
concentrations of criteria pollutants to deposition levels and (2) allowing quantification of
criteria pollutants (NOy and SOx) and noncriteria pollutant (e.g., NHX) contributions to
total acidifying deposition. As a scientific publication, the AAI is documented in Scheffe
IS-18
-------
et al. (2014). The AAI was originally developed in the 2011 NOxSOx Policy Assessment
(U.S. EPA. 2011a). and the equation is described with terms that traditionally define a
NAAQS [the indicator,1 averaging time,2 form,3 and level4—further described in the
2017 IRP (U.S. EPA. 2017c)l.
Key scientific aspects of the AAI equation, as the form of a potential standard, are
described in the following excerpt from 2017 IRP (U.S. EPA. 2017c):
"The AAI, as described in the PA (U.S. EPA. 2011a). was constructed
from steady-state ecosystem modeling, and included atmospheric
transference ratios and deposition of reduced forms of nitrogen
(ammonia gas and ammonium ion, expressed as NHx). These
nonoxidized forms of nitrogen were included since ecosystems respond
to total nitrogen deposition, whether from oxidized or reduced forms.
More specifically, the AAI equation was defined in terms of four
ecological and atmospheric factors and the ambient air indicators NOy
and SOx:
AAI = F1-F2- /3[NOy] - /3[S0x]
Equation IS-1
where Fl5 represents the ecosystems natural ability to provide
acid-neutralizing capacity (e.g., geology, plant uptake of nitrogen
deposition) and other processes; F26 represents acidifying deposition
associated with reduced forms of nitrogen, NHx; and F31 and F4S are the
1 The "indicator" of a standard defines the chemical species or mixture that is measured in determining whether an
area attains the standard.
2 The "averaging time" defines the time period over which ambient measurements are averaged (e.g., 1-hour, 8-hour,
24-hour, annual).
3 The "form" of a standard defines the air quality statistic that is compared to the level of the standard in determining
whether an area attains the standard.
4 The "level" defines the allowable concentration of the criteria pollutant in the ambient air.
5 Fl is defined as: . I.Y<"illn + CLJOr, with . I \'<"|IITI representing a target ANC level. With regard to (the PA
developed distributions of calculated critical loads for a specific ecoregion; in setting an AAI-based standard, a
percentile would need to be specified to reference the value of CLr to be used in the AAI equation [U.S. EPA
(2011a). p. 7-37], The PA described the percentile as an aspect of the form for the standard [U.S. EPA (2011a).
Section 7.7],
6 F2 is defined as: NHv/6>r. where NHX is the deposition divided by O, [U.S. EPA (2011a). p. 7-37],
7 F3 is defined as: 7NOy/6>i. where 7NOy is the transference ratio that converts deposition of NOy to ambient air
concentrations of NOy [U.S. EPA (2011a). p. 7-37],
8 F4 is defined as: 7:SOv/6>r. where 7SOv is the transference ratio that converts deposition of SOx to ambient air
concentrations of SOx [(U.S. EPA. 2011a). p. 7-37],
IS-19
-------
transference ratios that convert concentrations ofNOy and SOxto related
deposition of nitrogen and sulfur ITJ.S. EPA (2011a). Section 7.7]."
Several other key scientific considerations are included in the AAI that were discussed in
the 2011 NOxSOx Policy Assessment (U.S. EPA. 2011a).
• Spatial heterogeneity of factors in the AAI equation: The value of factors in the
AAI equation vary across the U.S. Factors could be calculated for a spatial
boundary based on an ecologically similar landscape (e.g., Omernick ecoregion).
• Temporal heterogeneity: There is a relatively high degree of interannual
variability expected in the AAI because it is so strongly influenced by the amount
and pattern of precipitation that occurs within a region from year to year;
therefore, averaging calculated annual AAI values over 3 to 5 years would provide
reasonable stability.
• Level: With regard to a level for the AAI, the 2011 NOxSOx Policy Assessment
(U.S. EPA. 2011a) concluded that consideration should be given to a level within
the range of 20 to 75 (j,eq/L, noting that a target Acid Neutralizing Capacity
(ANC) value of 20 |icq/L would be a reasonable lower end of this range, so as to
protect against chronic acidification-related adverse impacts on fish populations
which have been characterized as severe at ANC values below this level.
IS.2.3 Changes: New Evidence and Causal Determinations
Since the 2008 ISA, several conceptual changes have occurred in our understanding of
the atmospheric sciences and ecological effects of NOx, SOx and PM. They include our
understanding of the sources of N deposition and in the relationship between atmospheric
concentration and deposition (Section IS.3 and Appendix 2). Models of N deposition rely
on accurate emissions data. Since the 2008 ISA, deposition of oxidized nitrogen has been
decreasing but deposition of reduced nitrogen has been increasing. As a result, the
uncertainty in total reactive N emissions (NOx + NHx) has increased because emissions
estimates that have the lowest levels of uncertainty are from stationary and mobile
sources, which contribute more to NOx than NHx emissions, and higher levels of
uncertainty are associated with agricultural emissions, which contribute more to NHx
than NOx emissions.
A better understanding of the relationship between atmospheric concentration and
deposition has resulted from advances in understanding bidirectional exchange of NH3
and NOy chemistry within canopies. These advances have led to the first efforts to
provide a detailed characterization of N and S deposition on a national scale, by using
both measured and modeled values to provide estimates of total sulfur and nitrogen
deposition across the U.S.
IS-20
-------
New evidence since the 2008 ISA increases the weight of evidence for ecological effects,
confirming concepts previously identified and improving quantification of dose
(deposition)-response relationships, particularly for N deposition. The ecological effects
are described by the causality determinations. There are 18 causality statements in this
ISA (Table IS-1). Fourteen are causal relationships repeated from the 2008 ISA or
modified from the 2008 ISA to include specific endpoints. One is a likely causal
relationship repeated from the 2009 PM ISA. Three are new endpoint categories not
evaluated in the 2008 ISA. Table IS-3 shows that N and S deposition cause alteration of
(1) biogeochemical components of soil and water chemistry and (2) multiple levels of
biological organization ranging from physiological processes to shifts in biodiversity and
ecological function (Figure IS-4).
The current NO2 and SO2 secondary NAAQS are set to protect against direct damage to
vegetation by exposure to gas-phase oxides of nitrogen and oxides of sulfur. Research
continues to support causal relationships between SO2, NO2, NO, peroxyacetyl nitrate
(PAN), HNO3, and injury to vegetation (Table IS-1). but research that tests plant response
to the lower exposure levels that represent current atmospheric NOy and SOx
concentrations is limited. Therefore, little evidence is available to help determine whether
current monitored concentrations of gas-phase NOy and SOx are high enough to injure
vegetation.
It is clear that the criteria pollutants NOy, SOx, and PM, in addition to the noncriteria
pollutant NH3, contribute to total N and S deposition, which alters the biogeochemistry
and the physiology of organisms, resulting in harmful declines in biodiversity. Decreases
in biodiversity mean that some species become relatively less abundant and may be
locally extirpated. The current period in Earth's history is the Anthropocene. In addition
to a spike in soil radiocarbon from nuclear bomb testing (Turnev et al.. 2018). a defining
attribute of the Anthropocene is global, human-driven mass extinctions of many species.
The biodiversity loss reported in this assessment contributes to the Anthropocene loss of
biodiversity (Rockstrom et al.. 2009). In addition to the loss of unique living species, the
decline in total biodiversity is harmful because biodiversity is an important determinant
of the stability of ecosystems and the ability of ecosystems to provide services to
humanity (see more on biodiversity in Section IS.2.2.4).
IS-21
-------
Table IS-1 Causal determinations for relationships between criteria pollutants
and ecological effects from the 2008 NOx/SOx Integrated Science
Assessment (ISA) or the 2009 Particulate Matter (PM) ISA, for other
effects of PM, and the current draft ISA.
Causal Determination
Effect Category
2008 NOx/SOx ISA
Current ISA
Gas-phase direct phytotoxic effects
Gas-phase SO2 and injury to vegetation
Causal relationship
Causal relationship
Section IS.3 and ADDendix 3.6.1
Gas-phase NO, NO2, and PAN and injury to vegetation
Causal relationship
Causal relationship
Section IS.3 and ADDendix 3.6.2
Gas-phase HNO3 and injury to vegetation3
Causal relationship
Causal relationship
Section IS.3 and ADDendix 3.6.3
N and acidifying deposition to terrestrial ecosystems
N and S deposition and alteration of soil biogeochemistry
in terrestrial ecosystems'5
Causal relationship
Causal relationship
Section IS.5.1 and ADDendix 4.1
N deposition and the alteration of the physiology and
growth of terrestrial organisms and the productivity of
terrestrial ecosystems0
Not included
Causal relationship
Section IS.5.2 and ADDendix 6.6.1
N deposition and the alteration of species richness,
community composition, and biodiversity in terrestrial
ecosystems0
Causal relationship
Causal relationship
Section IS.5.2 and ADDendix 6.6.2
Acidifying N and S deposition and the alteration of the
physiology and growth of terrestrial organisms and the
productivity of terrestrial ecosystemsd
Not included
Causal relationship
Section IS.5.3 and ADDendix 5.7.1
Acidifying N and S deposition and the alteration of
species richness, community composition, and
biodiversity in terrestrial ecosystemsd
Causal relationship
Causal relationship
Section IS.5.3 and ADDendix 5.7.2
N and acidifying deposition to freshwater ecosystems
N and S deposition and alteration of freshwater
biogeochemistrye
Causal relationship
Causal relationship
Section IS.6.1 and ADDendix 7.1.7
IS-22
-------
Table IS-1 (Continued): Causal determinations for relationships between criteria
pollutants and ecological effects from the 2008 NOx/SOx
Integrated Science Assessment (ISA) or the 2009
Particulate Matter (PM) ISA, for other effects of PM, and
the current draft ISA.
Causal Determination
Effect Category
2008 NOx/SOx ISA
Current ISA
Acidifying N and S deposition and changes in biota,
including physiological impairment and alteration of
species richness, community composition, and
biodiversity in freshwater ecosystems'
Causal relationship
Causal relationship
Section IS.6.3 and ADDendix 8.6
N deposition and changes in biota, including altered
growth and productivity, species richness, community
composition, and biodiversity due to N enrichment in
freshwater ecosystems9
Causal relationship
Causal relationship
Section IS.6.2 and Appendix 9.6
N deposition to estuarine ecosystems
N deposition and alteration of biogeochemistry in
estuarine and near-coastal marine systems
Causal relationship
Causal relationship
Section IS.7.1 and Appendix 7.2.10
N deposition and changes in biota, including altered
growth, total primary production, total algal community
biomass, species richness, community composition, and
biodiversity due to N enrichment in estuarine
environments11
Causal relationship
Causal relationship
Section IS.7.2 and Appendix 10.7
N deposition to wetland ecosystems
N deposition and the alteration of biogeochemical cycling
in wetlands
Causal relationship
Causal relationship
Section IS.8.1 and Appendix 11.10
N deposition and the alteration of growth and productivity,
species physiology, species richness, community
composition, and biodiversity in wetlands
Causal relationship
Causal relationship
Section IS.8.2 and Appendix 11.10
S deposition to wetland and freshwater ecosystems
S deposition and the alteration of mercury methylation in
surface water, sediment, and soils in wetland and
freshwater ecosystems'
Causal relationship
Causal relationship
Section IS.9.1 and Appendix 12.7
IS-23
-------
Table IS-1 (Continued): Causal determinations for relationships between criteria
pollutants and ecological effects from the 2008 NOx/SOx
Integrated Science Assessment (ISA) or the 2009
Particulate Matter (PM) ISA, for other effects of PM, and
the current draft ISA.
Causal Determination
Effect Category 2008 NOx/SOx ISA Current ISA
S deposition and changes in biota due to sulfide Not included Causal relationship
phytotoxicity, including alteration of growth and
productivity, species physiology, species richness,
community composition, and biodiversity in wetland and
freshwater ecosystems
Section IS.9.2 and Appendix 12.7
2009 PM ISA Current Draft ISA
Other ecological effects of PM (course and fine particles, without regard to chemical speciation)
PM and a variety of effects on individual organisms and Likely to be a causal Likely to be a causal
ecosystems relationship relationship
Section IS. 10 and Appendix 15.7
C = carbon; Hg = mercury; HN03 = nitric acid; ISA = Integrated Science Assessment; N = nitrogen; NO = nitric oxide;
N02 = nitrogen dioxide; PAN = peroxyacetyl nitrate; S = sulfur; S02 = sulfur dioxide.
aThe 2008 ISA causality statements for gas-phase HN03 was phrased as "changes in vegetation."
bThe 2008 ISA included two causality statements for terrestrial biogeochemistry which were phrased as "relationship between
acidifying deposition and changes in biogeochemistry" and "relationship between N deposition and the alteration of
biogeochemical cycling of N."
The 2008 ISA causality statement for biological effects of N enrichment in terrestrial ecosystems was phrased as "relationship
between N deposition and the alteration of species richness, species composition, and biodiversity."
dThe 2008 ISA causality statement for biological effects of acidifying deposition in terrestrial ecosystems was phrased as
"relationship between acidifying deposition and changes in terrestrial biota."
eThe 2008 ISA included three causality statements for freshwater biogeochemistry phrased as "relationship between acidifying
deposition and changes in biogeochemistry related to aquatic ecosystems," "relationship between N deposition and the alteration
of biogeochemical cycling of N," and "relationship between N deposition and the alteration of biogeochemical cycling of C."
'The 2008 ISA causality statement for biological effects of acidifying deposition in freshwater ecosystems was phrased as,
"relationship between acidifying deposition and changes in aquatic biota."
9The 2008 ISA causality statement for biological effects of N deposition in freshwater ecosystems was phrased as "relationship
between N deposition and the alteration of species richness, species composition, and biodiversity in freshwater aquatic
ecosystems."
hThe 2008 ISA causality statement for biological effects of N deposition to estuaries was phrased as "relationship between N
deposition and the alteration of species richness, species composition, and biodiversity in estuarine ecosystems."
'The 2008 ISA causality statement for biological effects of S deposition effects on ecosystems was phrased as "relationship
between S deposition and increased methylation of Hg, in aquatic environments where the value of other factors is within
adequate range for methylation."
IS-24
-------
NOx SOx PM Integrated Science Assessment for Ecological Effects*
Indicator
Gases * Nitrogen Deposition Sulfur Deposition ^'^epositoon^
Class of Pollutant Effect
Direct
Phytotoxic N-enrichment/Eutrophication Sulfide Toxicity Mercury Methylation Acidification
Scale of Ecological Response
Population
Geochemistry Individual Community Ecosystem
Individual
Ecosystem
Productivity
Biodiversity
Growth rate
Physiological
alteration, stress
or injury
Soil or sediment
chemistry
Surface water
chemistiy
Terrestrial Terrestrial Wetland Fresh Water Estuaiy Wetland Fresh Water Wetland Fresh Water Terrestrial Fresh Water
| Causality framework
Causal
Likely causal
Suggestive
Inadequate
Not likely
Not evaluated in causal framework
*A causal relationship is likely to exist between deposition of PM and a variety of effects on individual organisms and ecosystems, based
on information from the previous review and limited new findings in this review
* Includes: NO, N02, HN03, S02, and PAN
Figure IS-4 Causal relationships between the criteria pollutants and ecological effects.
IS-25
-------
Since the 2008 ISA, there is more evidence to support the direct effects of gaseous SOx
and NOy on vegetation. This causality determination is uniquely modified by the
observation that there is little or no evidence that such effects are continuing at current,
lower levels of exposure now occurring in the U.S.
Since the 2008 ISA, the largest increase in ecological evidence is for terrestrial N driven
eutrophication effects (Section IS.5.1. Section IS.5.2. Appendix 4. and Appendix 6). This
new research confirms the causal relationship between N deposition and ecological
effects documented in the 2008 ISA. Further, this new research improves our
understanding of the mechanistic links that inform causal determinations between N
additions via atmospheric deposition, biogeochemistry, and biota in terrestrial ecosystems
(Table IS-1). There is now stronger empirical evidence from across most regions of the
U.S. to quantify the levels of N deposition (empirical CLs) that cause biodiversity
declines of lichens and grasses/forbs. There is new evidence to quantify empirical CLs
across much of the U.S. for nitrate leaching, tree survivorship, and mycorrhizal
biodiversity. Many of the N deposition effects are due to historical and continuing N
deposition.
New research confirms that N + S deposition causes terrestrial ecosystem acidification, as
documented in the 2008 ISA (Table IS-1). New evidence to characterize terrestrial
acidification (soil biogeochemistry changes and biological effects) across large regions of
the U.S. is available; in particular, new modeling work has improved calculation of CLs
for soil acidification (Section IS. 5.3; Appendix 4 and Appendix 5). Many of the
acidification effects are due to historical and continuing N and S deposition
(Section IS. 11).
New evidence for freshwater acidification CLs builds on several decades of research
documenting freshwater acidification effects on aquatic biota in the U.S. and confirms the
causal relationships determined in the 2008 ISA (Table IS-1). Many of the acidification
effects are due to historical and continuing N and S deposition (Section IS. 11).
The sources of N driven eutrophication of fresh waters, estuaries, and wetlands include
atmospheric N deposition and N from agricultural and other wastewaters. New research
has helped show how these respective sources contribute to total loading. In freshwater
ecosystems where atmospheric deposition is the primary source of N, such as in high
alpine watersheds, new CLs since the 2008 ISA support previous observations of
increased algal productivity, species changes, and reductions in diversity. New evidence
also supports clear links between aqueous S concentrations in aquatic systems and both
mercury methylation and sulfide toxicity; however, quantitatively linking these outcomes
to atmospheric deposition remains a challenge.
IS-26
-------
IS.3 Emissions and Atmospheric Chemistry
The atmospheric chemistry from emission to deposition discussed in this ISA is for the
criteria pollutants NOy, SOx, and PM. In addition to gas-phase indicators like SO2 and
NO2 used to monitor criteria pollutant trends, deposition of total N, total S, and total N +
S that accounts for a wider range of species is also a main focus.
A wide variety of N containing compounds (oxidized + reduced, and organic + inorganic)
contribute to wet and dry N deposition (Appendix 2.1). NHX (NHX = NH3 + NH4+)
includes both the PM component NH44" and gas-phase NH3. The contribution of NH3 to
total observed inorganic N deposition may range from 19% in northwestern U.S.
locations to 63% in locations in the southwestern U.S. and is generally greater in the
summer than in the winter. Therefore, NH3 is discussed in the ISA along with NOy and
relevant PM components to better understand and compare their contributions to both wet
and dry N deposition. In addition, PM impacts discussed in this document are also mainly
focused on N and S containing species, which together usually make up a large fraction
of PM25 mass in most areas of the U.S. and have greater and better understood ecological
impacts than other PM components.
Gaseous, particulate, and dissolved forms of NOy, SOx, and NHx all contribute to
atmospheric wet and dry deposition. The major components of particulate matter in the
U.S. are NO3 , SO42 . NH44", particulate organic matter, elemental carbon, crustal
material, and sea salt. While organic matter usually accounts for a large fraction of PM2 5,
only a small portion can be identified at a molecular level. As a result, there is little
information on organic PM impacts, except for individual compounds that make minor
contributions to mass. Assessment of ecological impacts of major PM species is largely
limited to NO3 , SO42 . and NH/. Of these, SO42 and NO3 are also components of total
oxides of sulfur and nitrogen, respectively. NO3 , SO42 . and NH44" usually have a strong
influence on acid deposition. NO3 and NH44", and in some cases organic nitrogen
(organic nitrates and reduced organic N), make a substantial contribution to N deposition.
Since the 2008 ISA, there have been several new developments including:
• Expansion of ambient monitoring networks to include NH3 and NOy at selected
sites, and comparisons of monitoring methods with research grade instruments
(Appendix 2.4);
• Adoption of new methods, such as data-model fusion, to integrate deposition
information across the U.S. (Appendix 2.5);
• Incorporation of bidirectional exchange into models of dry deposition
(Appendix 2.5.2); and
IS-27
-------
• Improvements in techniques using satellite-based measurements and chemical
transport model simulations to estimate emissions, concentrations, and dry
deposition ofNC>2, SO2, andNHa (Appendix 2.6).
IS.3.1 Sources and Atmospheric Transformations
Both gaseous and particulate forms of N and S contribute to atmospheric deposition. The
main contributors to acidifying precipitation are H2SO4, HNO3, and NH44", which are
formed from precursor emissions of SO2, NOx (NO + NO2), and NH3 (Appendix 2.2).
Gaseous emissions of NH3 are dominated by agricultural fertilizer application and animal
waste from intensive animal feeding operations, with important local contributions from
motor vehicles and episodic contributions from wildland and agricultural fires. Roughly
half of SO2 emissions are from by electricity-generating units (EGUs), mainly coal-fired
power plants. Notably, SO2 emissions from EGUs have been decreasing. NOx emissions
have a wider distribution of sources, with substantial contributions from highway and
off-highway vehicles, lightning, and EGUs. Primary PM2.5 and PM10 emissions are
dominated by dust and fires, but much of the PM2.5 mass in the U.S. is produced by
reactions that form secondary PM2.5 from gas phase precursor N and S species. Because
of these processes, a sharp decrease in SO2 emissions and smaller, but substantial
decreases in NOx emissions have occurred since the passage of the Clean Air Act
Amendments in 1990. Emissions of NOx in the U.S. declined 61% between 1990 and
2017 (U.S. EPA. 2020a). while nationwide annual average 98th percentile NO2
concentrations decreased by 53% from 1990 to 2017 (U.S. EPA. 2016f). Total emissions
of SO2 decreased by 89% from 1990 to 2017 (U.S. EPA. 2020a). resulting in a decrease
in SO2 concentrations of 89% in the eastern U.S. and 45% in the western U.S.
(Appendix 2.6.5). National annual NH3 emissions have fluctuated as a result of changes
in both emissions and methods of estimating emissions. However, no clear trend is
evident for national NH3 emissions, with estimates for 1990 and 2017 differing by less
than 1% (U.S. EPA. 2020a). National NH3 monitoring is too recent for evaluating
long-term concentration trends, although more limited studies of NH3 emissions,
concentrations and deposition each suggest slight increases may have occurred
(Appendix 2.6.4).
Major components of particulate N and S include NH44", NO3 . and SO42 . which are
primarily derived from gaseous precursors NH3, NOx, and SO2 (Appendix 2.3). Together,
NO3 , SO42 . and NH4+ make up a large fraction of PM2.5 mass in most areas of the U.S.
Formation of particulate N and S is described in the 2019 ISA for Particulate Matter
(U.S. EPA. 2019). An understanding of the sources, chemistry, and atmospheric
processes for these gas-phase and PM species provides a background for understanding
acidifying and N deposition.
IS-28
-------
IS.3.2 Measurement and Modeling Techniques
Monitoring networks across the U.S. measure NOy, SOx, and NHx species involved in
deposition (Appendix 2.4.1). The National Atmospheric Deposition Program/National
Trends Network (NADP/NTN) has monitored precipitation chemistry for several decades
at many U.S. sites. The Clean Air Status and Trends Network (CASTNET) has monitored
concentrations of inorganic gas and particulate-phase N and S species since 1990.
Monitoring of NH3 (Appendix 2.5.3) in the Ammonia Monitoring Network (AMoN), part
of the NADP network, was initiated at a subset of CASTNET sites in 2007. NH3 was also
measured as a part of the Southern Aerosol Characterization (SEARCH) network from
2004 until its termination in 2016. The Interagency Monitoring of Protected Visual
Environments (IMPROVE) network and the Chemical Speciation Network (CSN)
measure PM and PM components, including NO;, and SO42 . although these data are not
routinely used to estimate deposition rates (Appendix 2.4.1).
Atmospheric N deposition rates are calculated from measurements and models. Direct
measurement of NO2 concentration has limited utility for quantifying NOy deposition
rates in areas with less urban influence. Because NOy is composed of diverse chemical
species with a wide range of deposition velocities and physical properties, concentrations
of unmeasured component species of NOy in general and of all NOy species in
data-sparse regions must be provided by regional models. For NO2 and NH3 this can be
done in conjunction with satellite-based remote sensing data (Appendix 2.4.2).
Estimates of dry deposition (Appendix 2.5.2) over the contiguous U.S. are inferred by
atmospheric models, used with monitoring network data. When combined with accurate
estimates of historical trends in emissions and meteorology, these models are able to
capture the historical long-term changes in PM2.5 SO42 . NO;, . and NH44", but are subject
to uncertainties in their treatment of turbulence, surface interactions, and in particular,
seasonal variability in NO3 deposition, mainly because of uncertainties in NH3
emissions. Consequently, dry deposition rates (and ratios of wet-to-dry deposition)
continue to be uncertain.
IS.3.3 Spatial and Temporal Variability in Deposition
Overall deposition of total N (oxidized + reduced N) has decreased slightly over the past
since 2000 (Appendix 2.6.2). This is because although NOy deposition has declined
considerably in the contiguous U.S., deposition of NHx has increased. The large spatial
variability in N deposition and changes in geographic distribution of 3-year average N
deposition between 2000-2002 and 2016-2018 are evident in the maps (Figure IS-5) of
IS-29
-------
3-year average annual dry + wet deposition of NOy and NHX over the contiguous U.S.
estimated using the TDEP (Total Deposition) modeling approach (Appendix 2.6). which
combines output from the Community Multiscale Air Quality (CMAQ) system with wet
deposition from the NADP/NTN (Schwede and Lear. 2014b) and air concentrations from
CASTNET.
According to TDEP estimates for 2016-2018 (Appendix 2.6). much of the eastern
contiguous U.S. is estimated to receive at least 10 kg N/ha/yr dry + wet deposition, with
some areas receiving more than 15 kg N/ha/yr. Estimates for the spatial extent of the
areas receiving at least 10 kg N/ha/yr of deposition and the overall amount of N deposited
could be low because reduced organic N species are not routinely monitored.
In general, wet deposition of reduced N exceeds that of oxidized N across the contiguous
U.S. According to estimates based on CASTNET and NADP data and CMAQ modeling
results (Figure 2-16). deposition of N nationwide occurs mainly by dry deposition of
HNO3 and NH3 (with NH3 dominant) and wet deposition of NH4 and NO;, (with NH4
dominant). Hybrid satellite/modeling and CMAQ results indicate that dry deposition of
NO2 is also a nontrivial source of deposited N in many areas (Appendix 2.6.6). Over the
past 30 years, NADP/NTN data show that wet deposition of inorganic N
(oxidized + reduced) decreased in areas such as the Northeast but remained constant or
increased in areas such as the central U.S. (see Figure 2-18 in Appendix 2.6). Wet
deposition of total inorganic N has remained fairly constant over the past 30 years,
despite declines in NOx emissions, indicating that most of the increases in N wet
deposition seen today is of reduced inorganic N. Data for total (wet + dry) deposition are
available for a shorter time series than wet deposition, but show a similar increase in the
share of reduced N relative to oxidized N. Figure IS-6 shows reductions in TDEP 3-year
average oxidized N deposition over the contiguous U.S. between 2000-2002 and
2016-2018, while Figure IS-8 shows the decrease in reduced N deposition compared
between the same periods.
IS-30
-------
Source: C ASTNTTT/CM AQ/N A DP
Tola] deposition of nitrogen 0002
USHPA02/19/19
Total N
(kg-N/ha)
1
-0
-2
-4
-6
-B
-10
-12
-14
-16
-18
I
->20
Source: CASTTnET/CMAQ/NADP
Total deposition of nitrogen 1618
USEPA 10/21/19
Total N
(kg-N/ha)
Ha = hectare; kg = kilogram; N = nitrogen.
Source; CASTNET/NTN/AMON/SEARCH. We acknowledge the Total Deposition (TDep) Science Committee of the National
Atmospheric Deposition Program (NADP) for their role in making the TDep data and maps available.
Figure IS-5 Wet plus dry deposition of total nitrogen over 3-year periods.
Top: 2000-2002; Bottom: 2016-2018.
IS-31
-------
Total deposition of oxidized N 0002
USEPA 09/12/18
Source: C ASTNHT/CM AQ/N A DP
Total oxN
(kg-N/ha)
Souicft: CASTNtnVCMAQ/NADf'
Total deposition of oxidized N 1618
USEPA 10/21/19
OxN = oxidized nitrogen.
Source: CASTNET/CMAQ/NTN/AMON/SEARCH. We acknowledge the Total Deposition (TDep) Science Committee of the National
Atmospheric Deposition (NADP) Program for their role in making the TDep data and maps available.
Figure IS-6 Wet plus dry deposition of oxidized nitrogen over 3-year periods.
Top: 2000-2002; Bottom: 2016-2018.
IS-32
-------
Source: CASTNET/CMAQ/NADP
Total reN
(kg-N/ha)
-0
-1
-2
-3
-4
-5
c
— D
-7
-8
-9
->10
Total deposition of reduced N 0002
USEPA 09/12/18
Total deposition of reduced N 1618
USEPA 10/21/19
Source: CASTNET/CMAQ/NADP
Total reN
(kg-N/ha)
-0
"1
-2
-3
-4
-5
-6
-7
-8
i
-9
I
->10
reN = reduced nitrogen.
Source: CASTNET/CMAQ/NTN/AMON/SEARCH. We acknowledge the Total Deposition (TDep) Science Committee of the National
Atmospheric Deposition Program (NADP) for their role in making the TDep data and maps available.
Figure IS-7 Wet plus dry deposition of reduced (inorganic) nitrogen over
3-year periods. Top: 2000-2002; Bottom: 2016-2018.
IS-33
-------
For S deposition, wet deposition tends to dominate over dry deposition in large areas of
the contiguous U.S. However, in some regions mainly in the West, dry deposition of
mainly SO2 makes a greater contribution than wet deposition. Anthropogenic emissions
of S and subsequent deposition have declined markedly since the 1990s, with the most
pronounced declines in the eastern U.S. Currently, some of the highest values of total
(wet + dry) SOx deposition in the U.S. are in parts of the Ohio Valley region
(Figure 2-41). However, Figure IS-8 shows that TDEP 3-year average total S deposition
has decreased substantially between 2000-2002 and 2016-2018, especially in this
region.
Both N and S deposition contribute to acidification of ecosystems. The pH of rainwater
has increased markedly across the U.S. since 1990, coincident with decreases in the wet
deposition of nitrate and SO42 . However, there are still widespread areas affected by
acidifying precipitation, mainly in the eastern U.S. (see Appendix 2.6). Total acidifying
deposition (wet + dry N + S, expressed as H+ equivalents) fluxes for 2016 to 2018 ranged
from a few tenths of H+ keq/ha/yr overmuch of the western U.S. to over 1.5 H+ keq/ha/yr
in parts of the Midwest and the Mid-Atlantic regions, and in other isolated hotspots
surrounding areas of concentrated industrial or agricultural activity (Figure IS-9).
IS-34
-------
Source: CASTNET/CMAQ/NADP
Total deposition of sulfur 0002
USEPA 09/12/18
Total S
(kg-S/ha)
i
-0
-2
-4
-6
-8
-10
-12
-14
-16
-18
1
->20
Total S
(kg-S/ha)
[i
-8
-10
r12
¦->20
of sulfur 1618
USEPA 10/21/19
Total deposition
Source: CASTNET/CMAQ/NADP
S = sulfur.
Source; CASTNET/CMAQ/NTN/AMON/SEARCH. We acknowledge the Total Deposition (TDep) Science Committee of the National
Atmospheric Deposition Program (NADP) for their role in making the TDep data and maps available.
Figure IS-8 Wet plus dry deposition of total sulfur over 3-year periods.
Top: 2000-2002; Bottom: 2016-2018.
IS-35
-------
Total N+S
(keq/ha)
-0.0
-0.2
-0.4
-0.6
-0.8
-1.0
-1.2
-1.4
-1.6
-1.8
1
->2.0
Total N+S deposition 1618
USEPA 10/21/19
Source: CASTNET/CMAQ/NADP
eq. = equivalents; H+ = hydrogen ion; ha = hectare; N = nitrogen; S = sulfur; yr = year.
Source: NADP. Note: We acknowledge the Total Deposition (TDep) Science Committee of the National Atmospheric Deposition
Program (NADP) for their role in making the TDep data and maps available.
Figure IS-9 Total acidifying deposition of total oxidized nitrogen, reduced
nitrogen, and oxidized sulfur expressed as H+ equivalents per
hectare per year over the contiguous U.S. 2016-2018.
Dry deposition rates are a strong function of surface characteristics, which modify the
structure of surface layer turbulence and the resistance to uptake by vegetation
(Appendix 2.5.2). As a result, spatially aggregated estimates of dry deposition fluxes are
subject to uncertainty, in addition to uncertainties that are inherent in the measurement of
species concentrations and in the inference of dry fluxes (see Section IS. 13). Wet fluxes
are not directly influenced by surface characteristics (although orography affects
transport and precipitation) but are subject to smaller uncertainties in the measurement of
rainfall and chemistry.
IS.4 Gas-Phase Direct Phytotoxic Effects
New evidence supports the causal determinations made in the 2008 ISA regarding
gas-phase effects on vegetation, and there are no new causal statements for gas-phase
IS-36
-------
effects. As in the 2008 ISA, the current ISA concludes that there are causal relationships
between SO2, NO2, NO, PAN, HNO3, and injury to vegetation. This determination is
based on consistent, coherent, and biologically plausible evidence (Appendix 3.2.
Appendix 3.3. and Appendix 3.4; Table IS-1). The clearest evidence for these
conclusions comes from studies available at the time of the 2008 ISA, but there have
been some additional studies since then. Most evidence on the direct effects of gaseous
NOy and SOx comes from controlled exposure studies across many species of vegetation.
Most controlled exposure studies over the past several decades have used concentrations
of gas-phase NOy and SOx above current ambient conditions observed in the U.S.
Relevant information is lacking on exposures and effects reflecting the more recent lower
pollutant conditions. Therefore, there is little evidence available to inform whether
current monitored concentrations of gas-phase NOy and SOx are high enough to injure
vegetation.
NH3 can also have direct phytotoxic effects if the uptake exceeds the ability of a plant to
detoxify and assimilate it. However, reduced N gases such as NH3 are not criteria air
pollutants or oxides of N and, therefore, are not the focus of this review of the gas-phase
effects. Direct damage from NH3 to foliage can occur on higher plants and effect
bryophytes and lichens. Declines in shrubs and lichens and changes in peat bogs have
been reported with NH3 exposure. Besides being potentially phytotoxic to vegetation,
NH3 exposure can lead to more N inputs into plants and ecosystems through foliage
uptake. Ammonia deposition that leads to N enrichment is an important consideration
when evaluating total N deposition. These N nutrient effects to vegetation are discussed
in Appendix 6.
IS.4.1 Sulfur Dioxide
In the 2008 ISA, evidence was sufficient to infer a causal relationship between exposure
to SO2 and injury to vegetation. The current secondary standard for SO2 is a 3-hour
average of 0.50 ppm, which is designed to protect against acute foliar injury in
vegetation. There has been limited research on acute foliar injury since the 1982 PM-SOx
Air Quality Criteria Document (AQCD), and there is no clear evidence of acute foliar
injury below the level of the current standard. The limited new research since 2008 adds
more evidence that SO2 can have acute negative effects on vegetation but does not
change conclusions from the 2008 ISA regarding the causal relationship between SO2
exposure and vegetation damage or the SO2 levels producing these effects (see
Appendix 3.1). Consistent with the 2008 ISA, the body of evidence is sufficient to infer
a causal relationship between gas-phase SO2 and injury to vegetation.
IS-37
-------
Increased SO2 exposure concentrations and longer exposure times are associated with
decreases in plant growth and yield. The 1982 PM-SOx AQCD concluded that more
definitive concentration-response studies were needed before useable exposure metrics
could be identified. However, very few studies of the effects of SO2 on the growth of
vegetation in the U.S. have been conducted since 1982. Recent studies from eastern
Europe indicate recovery of tree growth in response to decreases in SO2 concentrations
since the 1980s and that annual SO2 concentrations of 4 ppb decreased silver fir (Abies
alba) growth. In West Virginia, the growth of eastern red cedar (Junipenis virginiana)
trees increased with declines in SO2 emissions since the 1980s.
IS.4.2 Nitrogen Oxide, Nitrogen Dioxide, and Peroxyacetyl Nitrate
In the 2008 ISA, evidence was sufficient to infer a causal relationship between exposure
to NO, NO2, and PAN and injury to vegetation. It is well known that in sufficient
concentrations, NO, NO2, and PAN can have phytotoxic effects on plants by decreasing
photosynthesis and inducing visible foliar injury. However, the 1993 Oxides of Nitrogen
AQCD concluded that concentrations of NO, NO2, and PAN in the atmosphere are rarely
high enough to have phytotoxic effects on vegetation (U.S. EPA. 1993). and very little
new research has been performed at concentrations currently observed in the U.S. (see
Appendix 3.3). It is also known that these gases alter the N cycle in some ecosystems,
and nutrient effects of N are discussed in Section IS.5. Thus, consistent with the previous
2008 ISA, the body of evidence is sufficient to infer a causal relationship between
gas-phase NO, NO2, and PAN and injury to vegetation.
IS.4.3 Nitric Acid
In the 2008 ISA, evidence was sufficient to infer a causal relationship between exposure
to HNO3 and changes to vegetation. The 2008 ISA reported experimental exposure to
HNO3 resulted in damage to the leaf cuticle of pine and oak seedlings, which may
predispose those plants to other stressors such as drought, pathogens, and other air
pollutants. Since the 2008 ISA, Padgett et al. (2009b) investigated dry deposition of
HNO3 on the foliage in a fumigation study and confirmed the earlier research. Nitric acid
can also add to N nutrient enrichment of ecosystems and is discussed in Section IS.5. The
2008 ISA also reported several lines of evidence that past and current HNO3
concentrations may be contributing to the decline in lichen species in the Los Angeles
basin. Subsequent studies conducted in the Los Angeles basin since the 2008 ISA provide
further evidence of the impacts (see Appendix 3.4). These new studies continue to
IS-38
-------
support the causal findings of the 2008 ISA, such that the body of evidence is sufficient
to infer a causal relationship between gas-phase HNO3 and changes to vegetation.
IS.5 Terrestrial Ecosystem Nitrogen Enrichment and Acidification
For terrestrial ecosystems, new evidence reinforces causal findings from the 2008 ISA
and provides the basis for two new causal statements that reflect a more comprehensive
understanding of how N and acidifying deposition alter terrestrial ecosystem biota
(Table IS-1). In general, N deposition may cause soil N enrichment and stimulate the
growth of opportunistic species. However, in sensitive soils, deposition of N and/or S can
cause soil acidification, which may decrease growth and cause mortality among sensitive
plant species. Atmospheric deposition ofN and S alter the species composition of
terrestrial systems by one of four mechanisms: (1) nutrient enrichment (eutrophication;
Appendix 4 and Appendix 6). (2) acidification (Appendix 4 and Appendix 5). (3) direct
damage (Appendix 3). and (4) secondary effects (e.g., wildfire; Appendix 6). Ecosystems
and communities may be simultaneously affected by one or more mechanisms depending
on the sensitivity of environmental and biological properties to each mechanism.
Despite the abundance of N in the environment, plants are unable to directly access the
large pools of N contained in the atmosphere as N2 gas and in the soil as large organic
molecules. Consequently, the limited availability of reactive N often constrains biological
activity in terrestrial ecosystems. N deposition is therefore considered nutrient
enrichment because N additions generally stimulate plant growth and productivity
(cumulative growth of all vegetation within a community), which has been recognized
since the second half of the 19th century. In comparison, the biological effects of
acidifying deposition are less common and largely constrained to ecosystems with
historically high rates of deposition and that are vulnerable because of factors such as
geology and climate. While S is also an essential macronutrient, less S is required for
growth than N, and areas affected by acidifying deposition typically receive S at rates
that greatly exceed biotic demand. Instead, the impact of acidifying deposition stems
from the disruptions to biochemical processes caused by decreased pH and shifts in soil
physiochemical processes that decrease the supply of other essential nutrients (e.g., Ca,
Mg) and from increased mobilization of toxic forms of Al.
Current knowledge of soil biogeochemistry indicates soil N enrichment and soil
acidification occur in sensitive ecosystems across the U.S. at present levels of deposition.
Newly published work indicates decreasing SO2 emissions and S deposition have led to
early signs of recovery from acidification in some northeastern watersheds, but areas in
the Southeast do not show recovery (Appendix 4). There are many well-defined soil
IS-39
-------
indicators related to the biological effects of acidifying (N + S) deposition. New evidence
uses these indicators to describe the status of ecosystems, either by empirical observation
or models. Soil indicators for acidification are more typically modeled than those for
eutrophication effects. There is an abundance of new information on biogeochemical
pools and processes, including a new conceptual framework for the N saturation of
terrestrial ecosystems.
The enrichment of terrestrial ecosystems by N deposition often increases plant
productivity and causes changes in physiology and growth rates that vary among species.
This has been observed for herbaceous plants and trees across ecoregions. The changing
growth rates transform competitive interactions between species, and consequently, lower
species diversity is often observed with increasing N deposition within terrestrial
communities. The level of N deposition negatively affecting community composition is
often expressed as a Critical Load (CL). There are many new CLs available since the
2008 ISA, including those for lichens, herbaceous plants, and mycorrhizae.
The process of terrestrial acidification has been well understood and documented for
decades. Recent research, since the 2008 ISA, has confirmed and strengthened this
understanding and provided more quantitative information, especially across
regional-scale landscapes. Several studies have evaluated the relationships between soil
chemical indicators of acidification and ecosystem biological endpoints (see Table 5-6).
and some biogeochemical models are well established. There have been new advances in
the parameterization of acidification models to U.S. soils since the 2008 ISA
(Appendix 4.5) resulting in better certainty of CLs. Biological endpoints included in the
evaluations include physiological and community responses of trees and other vegetation
(such as lichens), soil biota, and fauna.
The following section summarizes the main effects of N and S deposition on terrestrial
ecosystems.
IS.5.1 Soil Biogeochemistry
In the 2008 ISA, evidence was sufficient to infer causal relationships between
(1) acidifying deposition and changes in terrestrial biogeochemistry and (2) between N
deposition and terrestrial biogeochemical cycling of N. There is new evidence of how
deposition contributes to total loading in ecosystems, as well as new information from
addition, gradient, and time-series studies characterizing how deposition affects soil pools
and processes. Much of the new work focuses on the effects of N deposition, with
relatively little work focusing on S deposition. Soil N enrichment and soil acidification
occur in sensitive ecosystems across the U.S. at present levels of deposition. Decreasing
IS-40
-------
S emissions have led to early signs of recovery from acidification in some northeastern
watersheds, but areas in the Southeast do not show recovery (for additional discussion on
recovery see Section IS. 11). Deposition rates of total N (NOy + NHX) are relatively
unchanged across much of the contiguous US (Appendix 2.7). Accordingly, there are no
signs of recovery from N enrichment effects. CL determinations have been made at the
ecoregion scale for NO, leaching. CLs for biological effects are summarized below
(Section IS.5.1.2. Section IS.5.2.2. and Section IS.5.3.3). The body of evidence is
sufficient to infer a causal relationship between N and S deposition and alteration of
soil biogeochemistry in terrestrial ecosystems, which is consistent with the conclusions
of the 2008 ISA.
IS.5.1.1 Soil Processes and Indicators
Deposition ofN or N + S alters soil chemistry, which can have cascading effects on
aquatic ecosystems (for effects on aquatic biology and chemistry see
Appendix 7-Appendix 10). Soil acidification is a natural process that can be accelerated
by N or S deposition. Deposition in the forms of HNO3 and H2SO4 can directly acidify
soils; however, deposition of reduced forms of N (e.g., NHx) can also cause soil
acidification by releasing hydrogen ions (H+) during the microbial oxidation of NH44" to
NO3 . There are a number of soil biogeochemical processes associated with acidification
(Table IS-2). Base cations can counterbalance acid anions. Base cations are added to the
soil by weathering and atmospheric deposition and are removed by leaching and
biological uptake. Where acidifying deposition rates are high relative to base cation input,
deposition can deplete exchangeable base cation pools in soils. There are several useful
indicators of soil acidification (Table IS-2) that have quantitative relationships to
biological responses (Appendix 5).
Table IS-2 Summary of key soil geochemical processes and indicators
associated with eutrophication and acidification.
N Driven
Nutrient
Endpoint Enrichment Acidification The Effect of Deposition
Process
N saturation X X New empirical evidence suggests revising the N saturation
concept; specifically, it is now observed that NO3" leaching
can occur even if the ecosystem N capacity to retain N has
not yet been saturated.
IS-41
-------
Table IS-2 (Continued): Summary of key soil geochemical processes and
indicators associated with eutrophication and
acidification.
N Driven
Nutrient
Endpoint Enrichment Acidification The Effect of Deposition
Soil N accumulation X X New meta-analyses across ecosystem types confirm
inorganic soil NO3" concentration increases with N addition.
A new gradient study confirms that N concentration
increases with N deposition. A new addition study confirms
increased soil N accumulation. New studies on Soil N
accumulation are summarized in Table 4-3.
NO3" leaching X X New meta-analyses confirm leaching increases with N
additions. Regional-scale gradient analyses: <8 kg N/ha/yr
onset of leaching; <1 kg N/ha/yr in European forests; in the
NE U.S., 90% retention for sites receiving 7 kg N/ha/yr to
60% retention for sites receiving 11 kg N/ha/yr.
New USFS CLs for the onset of leaching: 8-10 kg N/ha/yr in
eastern and western U.S., 17 kg N/ha/yr in the Sierra
Nevada and San Bernardino Mountains. New studies on
Soil N accumulation are summarized in Table 4-3.
S accumulation and X Some soils (notably in many watersheds in the SE U.S.)
adsorption have the capacity to adsorb substantial quantities of S, with
essentially no acidification of drainage water. Nevertheless,
S adsorption capacity is finite, and under continual high S
deposition loading, the adsorptive capacity of soil will
eventually be exceeded.
New studies of 27 watersheds in the SE indicate most will
begin releasing SO42" in the next two decades; NE
watersheds show a net loss of S from soils now in response
to decreased levels of atmospheric S deposition. New
studies on soil S accumulation are summarized in Table 4-4.
SO42" leaching X Atmospheric S deposition generally increases leaching of
SO42" to surface waters. The amount of deposition that
causes the onset of leaching varies across the landscape.
New studies on soil SO42" leaching are summarized in
Table 4-4.
Base cation leaching X Base cation (Ca, Mg, K, Na) release from soil particles to the
and exchange soil solution occurs in response to the input of acid anions
(SO42" and NO3") from deposition.
New studies confirm base cation depletion continues to
occur in the Rocky Mountains (threshold 28 kg N/ha/yr) and
in U.K. grasslands, while in a NE forest, 17 yr of N addition
did not cause further depletion. A meta-analysis suggests
cation depletion soon after increased deposition of acid
anions, but this depletion tapers off with time. New studies
on base cation leaching and exchange are summarized in
Table 4-5.
IS-42
-------
Table IS-2 (Continued): Summary of key soil geochemical processes and
indicators associated with eutrophication and
acidification.
Endpoint
N Driven
Nutrient
Enrichment
Acidification
The Effect of Deposition
Al mobilization
X
The threshold for inorganic Al mobilization from soil is
<15-20% soil base saturation. This is an extremely
important effect of acidifying deposition because inorganic
monomeric Al is toxic to biota (Appendix 5 and Appendix 8).
Inorganic Al is minimally soluble at pH 6.0, but solubility
increases steeply at pH below 5.5.
New studies on Al in soils are summarized in Table 4-6.
Nitrification
X
X
Nitrification releases 2 mol hydrogen ion (H+) per mol NHV
converted to NO3", acidifying soils. As soil inorganic N
accumulates, net nitrification rates often increase, and NO3"
can leach from the ecosystem.
New N gradient and meta-analysis studies confirm N
addition increases nitrification. New studies on nitrification
are summarized in Table 4-6.
Denitrification
X
Denitrification is the microbial reduction of NO3" to NO2",
NO, the greenhouse gas N2O, and N2, which occurs under
anaerobic conditions. In Europe, soil switched from a source
to a sink after two decades of N deposition exclusion. New
meta-analysis confirms N addition increases denitrification
rates. New studies on denitrification are summarized in
Table 4-6.
DOC leaching
X
X
In recent years, the DOC of many lakes and streams has
risen, with the source likely from the soils in the adjacent
terrestrial watershed. However, the mechanism causing the
observed increase is unclear and may be due to a
combination of soil recovery from acidification, changes in
climate (e.g., temperature and precipitation), and N
deposition among other mechanisms. New studies are
summarized in Table 4-10.
Decomposition
X
X
The addition of N can stimulate the breakdown of labile
compounds that degrade during the initial stages of
decomposition, but added N can suppress the
decomposition of more recalcitrant material. There are new
addition studies and meta-analyses on mechanisms and
response trends.
New studies are summarized in Table 4-8.
Indicator
Soil [N]
X
X
Increases in soil [N] indicate soil N accumulation and the
size of the soil N pool that may be assimilated by organisms
or mobilized via leaching.
Soil C:N ratio
X
X
Decreasing soil C:N linked to changes in decomposition and
increases in nitrification and NO3" leaching.
<20-25 causes increased nitrification and elevated risk of
NO3" leaching in the U.S. and <25-30 for increased NO3"
leaching in Europe.
IS-43
-------
Table IS-2 (Continued): Summary of key soil geochemical processes and
indicators associated with eutrophication and
acidification.
N Driven
Nutrient
Endpoint Enrichment Acidification The Effect of Deposition
Soil base saturation X Increasing N + S deposition decreases the soil pool of
exchangeable base cations.
<15-20% exchange ion chemistry is dominated by inorganic
Al and may cause injury to vegetation (see Appendix 5).
Soil Bc:AI ratio X Increasing N + S deposition decreases the soil pool of
exchangeable base cations, often decreasing the Ca:AI
ratio.
Ca:AI <1.0 causes physiological stress, decreased growth,
and mortality of sensitive plant species (see Appendix 5).
Fungi-to-bacteria ratio X New indicator: increasing N deposition decreases the
fungi-to-bacteria ratio and causes a transition from N to C
limitation among soil food webs.
Al = aluminum; Al3+ = aluminum(lll); Be = base cations; C = carbon; Ca = calcium; DOC = dissolved organic carbon; H+ = hydrogen
ion; ha = hectare; K= potassium; kg = kilogram; Mg = magnesium; N = nitrogen; N2 = molecular (atmospheric) nitrogen;
N20 = nitrous oxide; Na = sodium; NE = northeastern; NH4+ = ammonium; NO = nitric oxide; N02" = nitrite; N03" = nitrate;
S = sulfur; SE = southeastern; S042" = sulfate; U.K. = United Kingdom; U.S. = United States; USFS = U.S. Forest Service;
yr = year.
IS-44
-------
Some of the same processes and indicators associated with acidification are also
associated with the N enrichment of soils in response to the input of exogenous N
(Appendix 4.3). The 2008 ISA documented that the increase in global reactive N (defined
as NOy + NHx + organic N) that occurred over the previous century was largely due to
three main causes: (1) widespread cultivation of crops that promote conversion ofN2 gas
to organic N through biological N fixation, (2) fossil fuel combustion converting
atmospheric N2 and fossil N to NOx, and (3) the Haber-Bosch process, which converts
nonreactive N2 to reactive N to sustain food production and some industrial activities
(Galloway et al.. 2003; Galloway and Cowling. 2002).
The 2008 ISA documented that atmospheric deposition of N can increase soil N, the
accumulation of which is linked to increased N leaching and decreased retention of N.
CLs for the onset of elevated NO;, leaching are given in Appendix 4.6.2.2.
The 2008 ISA described the conceptual model of N saturation, which occurs when N
input rates to terrestrial ecosystems exceed the uptake capacity of the soils and biota and
is indicated by the onset of increased soil N leaching. However, more recent work has
revised the N saturation model in response to observations in which N leaching resulted
from N input rates that are faster than vegetation and soil uptake rates, thus distinguishing
capacity N saturation from kinetic N saturation. Budgets from 83 forested watersheds in
the northeastern U.S. show that N retention averages 76% of the incoming atmospheric N
deposition and decreases from 90% retention at 7 kg N/ha/yr of deposition to 60%
retention at 11 kg N/ha/yr of deposition.
The 2008 ISA documented that N enrichment is associated with changes in microbially
mediated biogeochemical processes, including nitrification, denitrification, and
decomposition (Appendix 4.3). The addition of N can increase nitrification (the microbial
conversion of NH44" to NO, ). which contributes to soil acidification. N deposition to soils
can decrease surface soil C:N ratio, which can stimulate nitrification when C:N ratios fall
below 20 to 25. The NO;, created by nitrification may be leached, biologically
immobilized, or denitrified. Denitrification is the microbial reduction of NOs" to NO; .
NO, the greenhouse gas N2O, and N2, which occurs under anaerobic conditions. Several
syntheses have been published since 2008 evaluating N addition effects on denitrification
and nitrification in terrestrial ecosystems. A new meta-analysis shows N addition
substantially increases denitrification from many types of ecosystems (e.g., coniferous
forest, deciduous forest, tropical forest, wetland, grassland), but not heathlands. Among
five chemical forms of N studied, NO3 addition showed the strongest stimulation of N2O
emission. Using data extracted from 206 peer-reviewed papers, a second meta-analysis
observed that the largest changes in the ecosystem N cycle caused by N addition were
increased nitrification (+154%), N2O emissions (+134%), and denitrification (+84%).
IS-45
-------
IS.5.1.2 National-Scale Sensitivity and Critical Loads
As of the 2008 ISA, the regions of the U.S. with abundant acid-sensitive soils had been
well delineated. These acid-sensitive ecosystems are generally located in mountainous
terrain in the eastern U.S. and are underlain by bedrock resistant to weathering. However,
a similar delineation of the areas sensitive to the eutrophication effects of N had not yet
been completed. There is strong evidence demonstrating that biogeochemical sensitivity
to deposition-driven eutrophication and acidification is the result of historical loading,
geologic/soil conditions (e.g., mineral weathering and S adsorption), and/or natural
sources of N and S loading to the system.
Since the 2008 ISA, several new publications have advanced our understanding of soil
recovery from acidification and CLs. New publications report the results of field
observations and modeling studies on soil recovery from acidification, specifically in the
northeastern U.S., and the lack of recovery in the southern Appalachian Mountains
(Table 4-18). New ecoregion-scale terrestrial CLs for NOs" leaching were published in
2011 and have been updated by more recent published work. Finally, Clark et al. (2018)
estimated areas exceeding CLs for terrestrial acidification and NO;, leaching for the
contiguous U.S. for 1800 to 2025. For terrestrial acidification, area exceeding the
minimum CL peaked at almost 2.8 million km2 by 1975 before declining; whereas, for
NO;, leaching, the area exceeding the minimum CL peaked at roughly 3.4 million km2
around 1995.
IS.5.2 Biological Effects of Terrestrial Nitrogen Enrichment
The enrichment of terrestrial ecosystems by N deposition often increases plant
productivity and causes changes in physiology and growth rates that vary among species.
This combination of effects can alter the composition and decrease diversity of terrestrial
communities by transforming competitive interactions between species and changing the
availability of other essential resources, including light, water, and nutrients. Because N
deposition can cause both eutrophication and acidification and these processes can occur
simultaneously, the relationship between N deposition and community composition has
often been derived empirically. Many of the effects ofN deposition are similar across
ecosystems and life forms because N is an essential macronutrient, but the composition
and magnitude of how these effects are expressed within an ecosystem can differ as a
result of biotic and abiotic influences. Consequently, as with the 2008 ISA, we have
grouped the effects of N deposition on physiology and biodiversity by biome (e.g., forest,
tundra, grassland, and arid lands), with further framing by life form (e.g., plants,
microorganisms) and functional groups (e.g., trees, herbaceous plants). In comparison,
IS-46
-------
the broadest CLs created by the scientific community are at the ecoregion level, in which
spatial boundaries across the landscape are typically defined based on ecological,
climatological, and geological differences.
The 2008 ISA documented consistent evidence that N additions increased plant
productivity broadly across a wide range of terrestrial ecosystems. Since 2008, a large
body of new research on the biological effects of added N in terrestrial ecosystems has
been published from investigations of plant and microbial physiology, long-term
ecosystem-scale N addition experiments, regional and continental-scale monitoring
studies, and syntheses. These studies have been conducted in ecosystems representing
biomes in the U.S., including tundra, grasslands, arid and semiarid lands, and tropical,
temperate, and boreal forests. Because of the breadth of this research, there is a strong
mechanistic and empirical understanding for many of the biological effects of added N.
This body of evidence is sufficient to infer a causal relationship between N
deposition and the alteration of the physiology and growth of terrestrial organisms
and the productivity of terrestrial ecosystems.
The varying effects of N deposition on the growth and physiology of individual species
have consequence(s) for biodiversity. In the 2008 ISA, evidence was sufficient to infer a
causal relationship between N deposition and the alteration of species richness, species
composition, and biodiversity in terrestrial ecosystems. The 2008 ISA documented
consistent evidence of reduced species richness and altered community composition from
N addition studies in the U.S. and N deposition gradient studies in Europe for grassland
plant diversity, forest understory plants, and forest mycorrhizal fungi. There was also
consistent evidence of altered plant and mycorrhizal community composition from N
addition studies in arid and semiarid ecosystems, particularly in southern/central
California. There was little evidence of changes in forest overstory tree composition.
Since the 2008 ISA, new research techniques have been developed to understand
community composition, a larger number of communities have been surveyed, and new
regional and continental-scale studies have made it possible to isolate the influence of N
deposition from other environmental factors. This new research has provided more
extensive and mechanistic evidence, and combined with the findings of the 2008 ISA,
this body of evidence is sufficient to infer a causal relationship between N deposition
and the alteration of species richness, community composition, and biodiversity in
terrestrial ecosystems.
IS-47
-------
IS.5.2.1 Physiology and Biodiversity
At the time of the 2008 ISA, terrestrial ecologists had used meta-analyses to broadly
quantify the effects that N deposition can have on the growth of terrestrial plants,
concluding that N additions stimulate plant productivity by 20-30% in grasslands,
forests, tundra, and wetlands, increase aboveground productivity in herbaceous plant
communities, alter plant tissue chemistry, decrease biomass of mycorrhizal fungi, and
alter litter decomposition (Appendix 6.6.1). Recent research has provided further
coherent and consistent evidence that N additions stimulate plant growth and
productivity, but this research is still dominated by studies of temperate ecosystems and
aboveground plant responses (Figure 6-1 and Figure 6-2).
In the 2008 ISA, the positive plant growth response to N deposition was attributed to
higher rates of photosynthesis. However, evidence for this is mixed: increases in
photosynthesis following N additions have been observed across a variety of plant
functional types, but higher rates of photosynthesis have not been consistently observed
in response to chronic N additions meant to simulate atmospheric deposition. There is
new support for another mechanism that would increase aboveground growth: decreases
in the quantity of C allocated by plants to roots and mycorrhizae. There was evidence in
the 2008 ISA that N additions increase aboveground biomass more than belowground
biomass, raising the shoot-to-root ratio among plants, but evidence is now more
consistent and widespread. Plants also invest substantial amounts of C to support
mycorrhizal fungi, but there is evidence this investment declines when N is added to
terrestrial ecosystems.
Evidence that biodiversity change can be a consequence of N deposition has accumulated
since 2008 and includes new information for major taxonomic groups, including
herbaceous plants, overstory trees, and two groups of symbionts (lichens and
mycorrhizae). Evidence is now more widespread for decreases in lichen species richness
as the result of N deposition in the U.S. There are direct observations that N deposition in
the U.S. is changing mycorrhizal community composition and altering herbaceous plant
species richness across a broad range of ecosystems, including forests, grasslands, arid
and semiarid ecosystems, and alpine tundra. In addition, based on variation in mortality
and growth rates of co-occurring tree species, there is also indirect evidence that N
deposition is altering overstory tree community composition.
A substantial body of research linking changes in biodiversity to shifts in N availability
has been developed. Two hypotheses for species loss are (1) the random-loss hypothesis
and (2) the functional trait hypothesis. The random-loss hypothesis suggests rare species
are most likely to disappear as increased competition for resources, such as light,
eliminates less successful individuals; whereas the functional trait hypothesis predicts
IS-48
-------
that organisms with disadvantageous traits (e.g., shorter plants) will be outcompeted
when N is added. Both hypothesized mechanisms can operate simultaneously, and both
tie the changes in physiology, growth, and productivity caused by increased N
availability to declines in biodiversity.
As noted in Appendix 4. soil microorganisms have important roles in regulating N and C
cycling. There are several mechanisms to alter soil microbial biomass and physiology,
including changes in soil pH, increases in inorganic N availability, shifts in soil food
webs, and changes in the quantity and quality of available C. There were some
observations in the 2008 ISA that added N decreases microbial biomass, but there is now
more evidence that added N generally negatively or neutrally affects microbial biomass C
and microbial biomass N (Table 6-4).
IS.5.2.1.1 Forests
Forests occur within every U.S. state, but are most abundant in the eastern U.S., montane
and coastal portions of the western U.S., and Alaska. The distribution of forests is bound
by water availability, cold temperatures, and land management. In the 2008 ISA, there
was consistent evidence that N additions stimulated forest productivity, but these
responses varied widely and included both neutral and negative effects of N additions on
tree growth. However, there had been no empirical analyses of how atmospheric N
deposition altered forest productivity in the U.S. at broad scales. The 2008 ISA lacked
information on whether N deposition had any impact on the diversity and composition of
forest overstory trees, but it did present evidence for changes in the composition of
herbaceous vegetation, epiphytic lichens, and microbial communities. The addition of
new research since the 2008 ISA provides coherent evidence that N deposition alters the
physiology and growth of overstory trees and provides indirect evidence that N
deposition changes the community composition of overstory trees. Further, new research
supports N deposition altering the physiology, growth, and community composition of
understory plants, lichens, mycorrhizal fungi, soil microorganisms, and arthropods
(Appendix 6.2.3 and Appendix 6.3.3).
As of the 2008 ISA, most long-term N addition experiments were conducted in temperate
forests in the northeastern U.S. or in temperate or boreal forests in Europe. In these
studies, conifer species were less likely than broadleaf species to exhibit positive growth
responses to added N and more frequently exhibited increased mortality and decreased
growth. Since the 2008 ISA, new observations from experiments, forest inventory
studies, model simulations, and data synthesis efforts have been published, quantifying
increases in forest net primary productivity (NPP), net ecosystem productivity (NEP),
and ecosystem C storage (Figure 6-3). Overall, evidence is consistent that N deposition
IS-49
-------
broadly increases tree growth and forest productivity, including specific evidence
indicating that current rates of N deposition in the contiguous U.S. broadly stimulate
aboveground forest productivity (Appendix 6.2.3.1).
Despite these broad effects, it is also clear that with N addition growth and mortality
responses vary by tree species. Many of the observations in the 2008 ISA have been
reinforced by more recent research, including long-term forest inventory data collected
from across the U.S. and Europe. Recent analyses of U.S. forest inventory data by Horn
et al. (2018) found that tree species vary in their growth and mortality responses to N
deposition (Appendix 6.2.3.1). Responses of individual tree species ranged from
consistently increasing growth with greater N deposition; to increasing growth at lower N
deposition but decreasing growth at higher levels; to consistently decreasing growth with
greater N deposition. Mortality responses showed a similar pattern between species.
Notably, species with varying responses in growth and mortality co-occurred in places in
the U.S. Thus, this indirect evidence suggests that changes in tree community
composition are occurring due to N deposition (Appendix 6.3.3.1). These analyses
represent an advancement in our understanding from the time of the 2008 ISA.
In comparison, there is direct evidence that N deposition is altering the composition of
forest understory plant communities (Appendix 6.3.3.2). The evidence for altered forest
understory plant communities (also known as herbaceous layer or groundcover
vegetation) comes from both the 2008 ISA and from the more recent literature. Changes
in understory plant communities have been observed in monitoring plots along
atmospheric N deposition gradients in the U.S. and in Europe. In Europe, forest
understory plant communities have shifted with increasing N toward more
nutrient-demanding and shade-tolerant plant species.
Higher rates of aboveground tree growth in response to N deposition might be due to
shifts in C allocation away from belowground processes. Changes in C allocation in
response to additional N have been accompanied by decreases in the abundance of
mycorrhizal fungi and changes in mycorrhizal community composition (Table 6-2.
Table 6-14). Evidence for composition change is particularly abundant in
ectomycorrhizal fungal communities (Table 6-14); there are fewer observations of how
arbuscular mycorrhizal fungal communities change in response to N additions (Table 6-3;
Table 6-16). There are also numerous observations of altered total microbial (including
bacterial) biomass and community composition. For microbial biomass, most studies
identified since 2008 showed either negative or neutral effects of N additions, consistent
with the results of syntheses published before the 2008 ISA (Table 6-4). Changes in soil
microbial community composition were identified along an N deposition gradient in
Europe and in all three N addition studies (Table 6-14). The effects of N additions on
IS-50
-------
individual microbial taxonomic groups (bacteria, archaea, fungi, etc.) have been less
consistent (Table 6-15). Overall, there is evidence that N additions can decrease total
microbial biomass and alter microbial communities in forest soils.
Within soil food webs, soil microorganisms have both direct and indirect links to
arthropods. Because arthropods feed upon both microorganisms and litter, they can be
important regulators of decomposition, nutrient cycling, and forest productivity. Several
studies have examined the response of forest arthropod communities to added N,
including a group of studies on insect herbivores conducted in southern California
(Table 6-17). There is coherent evidence that N additions can alter forest arthropod
communities.
Epiphytic lichens have long been recognized as sensitive to air pollution. Although these
organisms often make up a small portion of forest biomass, they have important roles in
hydrologic cycling, nutrient cycling, and as sources of food and habitat for other species.
New research on lichen community composition identified since the 2008 ISA has further
added to the consistent and coherent evidence that lichen communities in the U.S. and
Europe are sensitive to current levels of atmospheric N deposition (Appendix 6.2.6;
Table 6-23). In particular, the U.S. Forest Service's Forest Inventory and Analysis
Program has ample data on the abundance of lichens throughout the U.S., and shifts in
lichen community composition clearly attributable to atmospheric N pollution have been
observed in forests throughout the West Coast, in the Rocky Mountains, and in
southeastern Alaska. Shifts in epiphytic lichen growth or physiology have been observed
along atmospheric N deposition gradients in the highly impacted area of southern
California, but also in more remote locations such as Wyoming and southeastern Alaska
(Table 6-5). Experimental N studies have also created more detailed insight into changes
in lichen physiology processes.
Overall, there is widespread evidence from forests that N deposition alters the growth and
physiology of trees, and indirectly suggests N deposition affects tree community
composition. Nitrogen deposition in forests also alter the growth, physiology, and
biodiversity of herbaceous plants, lichens, soil microorganisms, and arthropods.
IS.5.2.1.2 Tundra
Within the U.S., tundra ecosystems are limited to Arctic ecosystems in Alaska and to
relatively isolated, high elevation sites. Although these ecosystems tend to be remote, the
influence of atmospheric N deposition is distinct and there was evidence in the 2008 ISA
indicating that alpine tundra plant communities were sensitive to atmospheric N
deposition. Alpine organisms may be more sensitive to N deposition because of the
IS-51
-------
unique nature of N cycling in these ecosystems, which tend to have limited inorganic N
availability. Since the 2008 ISA, numerous studies of tundra physiological, productivity,
and community composition responses to added N have been published, providing further
evidence that N deposition alters the growth and physiology of alpine plant communities,
including vascular plants (herbaceous and woody), bryophytes, and lichens
(Appendix 6.2.4). as well as evidence of altered soil microbial communities (Table 6-8;
Table 6-19).
As in forests, increases in N content in response to additional N are widespread in tundra
plant communities (Table 6-6). Higher tissue N concentrations in response to added N
have been observed in multiple studies for vascular plants, bryophytes, and lichens. The
2008 ISA noted that plant growth and biomass responses tended to be species specific.
Subsequent studies have confirmed this result (Table 6-6). showing varying responses to
added N among ecosystem types, plant functional groups, and species. Whereas vascular
plants tend to show a positive response to added N, both bryophytes and lichens tend to
decrease in biomass and cover (Table 6-5; Table 6-6).
Given the varying effects of N addition on species physiology and growth, the numerous
observations of N addition impacts on species richness, species diversity, and community
composition among vascular plants, bryophytes, and lichens in alpine and Arctic tundra
ecosystems are unsurprising (Appendix 6.3.4; Table 6-18). Within the U.S., these
observations have included effects of N additions on plant community composition in
Colorado and Washington. In northern Europe, decreases in plant species richness along
atmospheric N deposition gradients have been documented. Overall, this new research
has provided further evidence that experimental N additions can alter plant biodiversity in
alpine and Arctic tundra ecosystems and has provided new evidence that current rates of
atmospheric N deposition in Europe are associated with a loss of plant species richness in
these ecosystems.
There are relatively few observations regarding the effect of N additions on total
microbial biomass or the biomass response of individual microbial taxonomic groups in
tundra ecosystems, and these results have also been largely inconsistent. However, new
research has provided evidence that N additions can alter microbial community
composition in alpine tundra ecosystems (Table 6-8; Table 6-19).
IS.5.2.1.3 Grasslands
Grasslands are most prevalent in the central U.S., yet also are widely distributed across
the U.S. in areas where woody vegetation is excluded by environmental factors or
management. There was widespread evidence at the time of the 2008 ISA that the
IS-52
-------
growth, physiology, and productivity of grassland plants could be altered by N
deposition. In addition, there were multiple lines of evidence in the 2008 ISA that
grassland plant, mycorrhizal, and microbial communities were sensitive to N inputs.
Combined with subsequent research, the evidence is clear that physiology, growth, and
community composition of plants, mycorrhizae, soil microorganisms, and arthropods are
sensitive to N inputs in grasslands.
Although NPP can be limited by multiple factors (e.g., water, herbivores, other nutrients)
in all ecosystems, limitations other than N tend to be more marked in grasslands than
forests, making it harder to understand and predict the effects of increased N availability.
However, the general response is similar (Appendix 6.2.5): N additions stimulate NPP,
increase foliar N, and increase allocation to aboveground biomass (increased ratio of
shoot:root mass).
Evidence from the U.S. of grassland plant community composition change in the 2008
ISA was based on N addition studies in Mediterranean grasslands in California and
northern prairie ecosystems. However, large-scale assessments of biodiversity across
observed atmospheric N deposition gradients were restricted to Europe. Recent research
provides further evidence that N deposition reduces grassland biodiversity in the U.S. and
Europe (Appendix 6.3.5). Since 2008, there have been direct observations of reduced
species richness along atmospheric N deposition gradients for grasslands in the U.S. and
Europe. These gradient studies have documented an interaction with soil pH, noting that
N deposition causes a greater loss of species richness and a shift in community
composition at sites with lower pH. Together, these findings from deposition gradients in
the U.S. and Europe provide coherent evidence that N deposition causes shifts in plant
community composition and the loss of plant species richness through mechanisms of
both acidification and eutrophication. Experimental studies published since 2008 have
provided more insight into the mechanisms linking changes in plant and microbial
community composition to increased N availability, providing evidence that declines in
species richness increase with time and that competition for resources such as water may
exacerbate the effects of N addition on diversity.
Overall, the additional studies in grassland ecosystems have confirmed that many of the
responses observed in the older N addition studies also occur at present rates of
atmospheric N deposition. These changes include losses in forb species richness (which
make up the majority of grassland biodiversity), greater growth of grass species (which
make up the majority of grassland biomass), changes in reproductive rates, as well as
shifts in mycorrhizal (Table 6-16). soil microbial (Table 6-20). and arthropod
populations. In total, because of the prevalence of N limitation in grasslands and the
dominance by fast-growing species that can shift in abundance rapidly (in contrast to
IS-53
-------
forest trees), grasslands appear especially sensitive to N input rates comparable to N
deposition across much of the contiguous U.S.
IS.5.2.1.4 Arid and Semiarid
Arid and semiarid ecosystems are abundant in areas of the western U.S. where climate or
orography create annually or seasonally dry conditions. At the time of the 2008 ISA, a
large amount of information was available on how N deposition affected the growth and
physiology of plants and microorganisms in arid and semiarid ecosystems, and there was
coherent evidence that plant communities in these ecosystems could be altered by the
added N. Evidence for these effects was particularly strong in coastal sage scrub (CSS),
chaparral, and Mojave Desert ecosystems in southern California. Within the CSS
ecosystems, N deposition has been linked to increased mortality in native shrubs,
decreased abundance of arbuscular mycorrhizal fungi, higher cover of invasive annual
plants, and increased wildfire activity. Similar increases in invasive annual plant cover
and fire frequency have also been attributed to N deposition in areas of the Mojave
Desert downwind of urban centers in southern California. Research since 2008 has
further documented these effects, with consistent evidence that N deposition can affect
the physiology, growth, and community composition of plants and soil microorganisms
in arid and semiarid systems.
The effects of N deposition on physiological and biogeochemical processes in arid and
semiarid ecosystems are even more clearly dependent on moisture availability than in
grasslands (Appendix 6.2.6). In these ecosystems, inorganic N often accumulates in the
soil during dry periods, and growth and physiological responses to additional N are only
observed when and where sufficient moisture is available. Two additional important
effects of aridity include (1) higher soil base saturation and pH that buffer these systems
from the acidification effects of N deposition and (2) spatially patchy nutrient availability
that develops beneath isolated shrub canopies. One important effect of N deposition on
arid and semiarid ecosystems is to decrease the patchiness of nutrient availability, which
promotes the growth of invasive annual plants in the spaces between the isolated shrubs.
The growth of these annual plants creates a more continuous fuel bed for wildfires,
increasing the prevalence of fire, and shifting plant community composition toward more
fire-adapted plant species.
Since 2008, increases in aboveground plant biomass or plant cover have been observed in
the U.S. in the Mojave and Sonoran Deserts, and in southern California chaparral, and
internationally in China and Spain (Appendix 6.2.6). Given the linkage to fire, it is
notable that there have been multiple observations of increased annual plant growth in the
Mojave Desert in response to added N.
IS-54
-------
New research has also provided further evidence that N deposition alters plant
communities in arid and semiarid ecosystems, particularly in southern California, but also
in other locations (Appendix 6.3.6). Many of these studies documented changes in plant
community composition, with fewer observations of plant species loss or changes in plant
diversity. Overall, this body of research has provided consistent and coherent evidence
that N deposition is altering the growth, physiology, and community composition of
plants in arid and semiarid ecosystems. Relative to plants, there are fewer studies of
microbial communities (Table 6-12; Table 6-22). but these studies provided evidence that
N additions can alter the abundance, physiology, and community composition of soil
microorganisms in arid and semiarid ecosystems.
IS.5.2.2 National-Scale Sensitivity and Critical Loads
At the time of the 2008 ISA, there had been little quantification of the extent and
distribution of N sensitivity in terrestrial ecosystems in the U.S. In the 2008 ISA, there
was no published U.S. national CL assessment. Since then, substantial work has been
done on quantifying N CLs for U.S. ecoregions. The most notable new work is the U.S.
Department of Agriculture—Forest Service (USDA-FS)^ ssessmeni of Nitrogen
Deposition Effects and Empirical Critical Loads (Pardo et al.. 2011a). That assessment
was organized by Level 1 ecoregions, and where data were available, CL calculations
were made for individual ecosystem types (e.g., forests within the Mediterranean
California ecoregion) and life forms (i.e., lichens, mycorrhizal fungi). This ISA largely
follows that structure, reporting terrestrial N CLs for life forms (e.g., herbaceous plants)
within each ecoregion, which is a geographically defined area within a broader biome
(e.g., forests) based on distinct physical and biological features (e.g., Northwest Forested
Mountains, Eastern Temperate Forests, etc.).
Newer CL studies are presented in tandem with the CLs reported by Pardo et al. (2011a)
in Table 6-28 and Figure IS-10. The majority of values for new CLs are within the range
of CLs identified by Pardo et al. (2011a). Notably, however, Simkin et al. (2016)
identified a new average CL for herbaceous plants in open canopy (7.9 kg N/ha/yr)
forests in the Eastern Temperate Forest ecoregion, and new lower CLs were derived for
alpine ecosystems in the Northwest Forested Mountains ecoregion. There are also new
CLs for herbaceous species in two ecoregions previously lacking a CL for herbaceous
plants rTable 6-28. Simkin et al. (2016)1.
Recently, Clark et al. (2018) estimated CL exceedance areas for the contiguous U.S. over
a more than 200 year period. Overall, this analysis showed that terrestrial N CLs have
been exceeded for many decades in areas across the U.S. Exceedance areas peaked in
IS-55
-------
1995 for changes in lichen communities and plant community composition at 3.47 and
2.87 million km2, respectively, before declining marginally by 2006. The minimum forest
tree health CL was exceeded in 2.41 million km2 by 1855 and did not change much over
time, primarily because the relatively low CL compared to deposition values in the
Eastern Temperate and Northern forest ecoregions.
Southern Semi-Arid Highlands
No Data
1
Temperate Sierras
4-7
O
o
Mediterranean California
O
6-33
17-39
3-6
7.8-9.2
•
North American Deserts
CO-8.4 O
3
O
o
5-25 O
Great Plains
M •
4-10 O
1 •
Northwest Forested Mountains
O *- ¦
IS- 7.1(J
5-10
Marine West Coast Forests
5 O
CD 0-7-9-2
1
•
Eastern Temperate Forests
<3
o
O
4-dO O
5-12
o
o
mO
5-7
• |
17.5
Northern Forests
<3
0-21
0 5 10 15 20 25 30 35 40 45
kg N/ha/yr
CL = critical load; ha = hectare; kg = kilogram; N = nitrogen; yr = year.
The rectangles indicate the range of CLs designated by Pardo et al. (2011a1: the circles indicate new papers that have specified
CLs; data from Table 6-28.
Figure IS-10 Summary of critical loads for nitrogen in the U.S. for shrubs and
herbaceous plants (yellow), trees (blue), lichens (green), and
mycorrhizae (gray).
IS-56
-------
IS.5.3 Biological Effects of Acidification
Since publication of the 2008 ISA, the overarching understanding of terrestrial
acidification has not appreciably changed. Recent research has confirmed and
strengthened this understanding that acidification can be caused by acidifying deposition
(N + S) and provided more quantitative information, especially across regional-scale
landscapes. Several studies have evaluated the relationships between soil chemistry
indicators of acidification and ecosystem biological endpoints (see Table 5-6). Soil
chemistry indicators examined in recent literature include exchangeable base cations
(Be), soil pH, exchangeable acidity (H+ and Al), exchangeable Bc:Al ratio, base
saturation, and Al concentrations. The most common indicator used in determining CLs
is the soil solution Be Al ratio. Appendix 5.2.1 discusses the uncertainty considerations
when using this indicator. Biological endpoints included in the evaluations consisted of
physiological and community responses of trees and other vegetation, lichens, soil biota,
and fauna.
IS.5.3.1 Physiology and Growth
In the 2008 ISA, evidence was sufficient to infer a causal relationship between acidifying
deposition and changes in terrestrial biota; the evidence included changes in plant
physiology, plant growth, and terrestrial biodiversity. The physiological effects of
acidification on terrestrial ecosystems in the U.S. were well characterized at the time of
the 2008 ISA and included slower growth and increased mortality among sensitive plant
species. Consistent and coherent evidence from multiple species and studies in 2008
showed that the biological effects of acidification on terrestrial ecosystems were
generally attributable to physiological impairment caused by Al toxicity and decreased
ability of plant roots to take up base cations (Appendix 3.2.2.3 of the 2008 ISA). Much of
the new evidence for the negative effects of acidifying deposition comes from Ca
addition experiments, in which the addition of Ca has alleviated many of the negative
plant physiological and growth effects. Consistent with the findings of the 2008 ISA, the
body of evidence is sufficient to infer a causal relationship between acidifying N and
S deposition and the alteration of the physiology and growth of terrestrial organisms
and the productivity of terrestrial ecosystems.
In the 2008 ISA, acidifying deposition, in combination with other stressors, was found to
be a likely contributor to physiological effects that led to the decline of sugar maple (Acer
sacchctriim) trees occurring in portions of the eastern U.S. with base-poor soils. Studies
since the 2008 ISA support these findings (see Appendix 5.2.1.1). For example, recent
field studies have shown relationships between soil chemical indicator threshold values
IS-57
-------
and tree responses. Substantial declines in sugar maple regeneration have been found at
soil base saturation levels <20%, which is consistent with the range reported in the 2008
ISA.
In new studies, sugar maple grew more rapidly and showed increased regeneration
responses with increasing exchangeable base cations, base saturation, and soil pH,
however, growth was stunted and regeneration reduced with increasing exchangeable Al.
In other studies, the growth, regeneration, and physiological responses of sugar maple to
the soil conditions created by acidifying deposition were reversed or ameliorated by Ca
additions. Similarly, the 2008 ISA reported that processes associated with soil
acidification contributed to physiological stress, high mortality rates, and decreasing
growth trends of red spruce (Picea rubens) trees. New evidence from Ca addition studies
provides further support for these mechanisms (see Appendix 5.2.1.2). Added Ca
reversed or ameliorated many of the physiological responses to acidification.
In the 2008 ISA, there was limited information on the effects of acidification on other
tree species. Since the 2008 ISA, research has observed varying physiological sensitivity
to soil acidification among eight eastern U.S. tree species. New studies since the 2008
ISA have also added new information about the effects of acidifying deposition on forest
understory vegetation, grasslands, lichens, and higher trophic level organisms (snails and
salamanders) that support the terrestrial acidification conclusions of the 2008 ISA.
IS.5.3.2 Biodiversity
The 2008 ISA noted strong evidence that acidifying deposition could alter terrestrial
community composition and cause a loss of terrestrial biodiversity. The physiological and
growth effects of acidifying deposition are not uniform across species, resulting in altered
species composition and decreased biodiversity whereby sensitive species are replaced by
more tolerant species. For example, increasing soil base cation availability was tied to
greater sugar maple growth and seedling colonization, whereas American beech (Fagns
grcmdifolia) was relatively more dominant on soils with lower base cation availability
(see Appendix 5.2.1.3.1). Measurements of soil acid-base chemistry have been used as a
predictor of understory species composition, with 50 understory species associated with
high soil base cation status. In another set of studies, soil acid-base chemistry was
correlated with soil biodiversity and community composition. For example, addition of
Ca resulted in changes in soil bacterial community composition and bacterial community
structure that were correlated with soil exchangeable Ca, pH, and P (see Appendix 5.2.4).
Based on research included in the 2008 ISA and these new studies, the body of evidence
is sufficient to infer a causal relationship between acidifying N and S deposition and
IS-58
-------
the alteration of species richness, community composition, and biodiversity in
terrestrial ecosystems.
IS.5.3.3 National-Scale Sensitivity and Critical Loads
The sensitivity of soils to acidifying deposition is discussed in detail in Appendix 4. In
general, surficial geology is the principal factor governing the sensitivity of terrestrial
ecosystems soil to acidification from S and N deposition. Other factors that contribute to
the sensitivity of soils to acidifying deposition include topography, soil chemistry, and
land use. Several widely accepted models are currently used in the U.S. to assess the
terrestrial effects of S and N deposition (Appendix 4.5). These models are typically used
to evaluate acidification effects on biota by assigning a value of a soil parameter that
relates to the onset of a harmful biological effect. Since the 2008 ISA, estimates of base
cation weathering (BCw), which are input to soil acidification models have improved and
are being applied for deriving new CLs in the U.S. Forests of the Adirondack Mountains
of New York, Green Mountains of Vermont, White Mountains of New Hampshire, the
Allegheny Plateau of Pennsylvania, and mountain tops and ridges forest ecosystems in
the southern Appalachians are the regions that are most sensitive to terrestrial
acidification from atmospheric deposition (Appendix 3.2.4.2 of the 2008 ISA).
Models used to determine CLs of acidifying deposition included SMB, STA, MAGIC,
ForSAFE-VEG, and empirical models. Several models and extrapolation methods to
estimate BCw rates were also investigated. The PROFILE model was evaluated as a
model to estimate soil BCw rates to support estimates of SMB CLs in the U.S. (see
Appendix 4.5). In general, recently published models used soil solution Bc:Al ranging
from 1.0 to 10.0 as an indicator to estimate CLs in North America.
Ecosystem sensitivities to ambient N and S deposition were also characterized by
developing CLs and exceedances (see Appendix 4.6; Figure IS-11. and Appendix 5.5).
Calculated CLs for forest plots based on the soil solution Bc:Al of 10.0 in the
northeastern U.S. ranged from 11 to 6,540 eq/ha/yr (eq quantifies the supply of available
H+ ions, combining the acidifying effects of N and S deposition), and 15-98% (calculated
using maximum and minimum weathering rates) of these plot-level CLs were exceeded
by N and S deposition. In this region, correlation analyses showed that the growth of
17 tree species were negatively correlated with CL exceedance. In Pennsylvania, CLs
based on the soil solution Bc:Al of 10.0 for hardwood forests ranged from 4 to
10,503 eq/ha/yr and were exceeded by estimated N and S deposition in the year 2002 in
53% of the plots. Several studies found that CL and exceedance determinations could be
influenced by BCw rates, soil chemical indicators, N retention, tree species-specific base
IS-59
-------
cation uptake, and the type and accuracy of deposition estimates (i.e., wet, bulk, total,
measured or modeled).
Forest Ecosystems Critical Loads for Acidity
| 6,001-8,800
States
No Data
eq = equivalent; ha = hectare; yr = year.
(A) McNultv et al. (2007): CLs are mapped at 1 -km2 grids (center map). For uncertainty, see Li and McNultv (2007). (B) Duarte et ai
(2013); CLs are mapped at 4-km2 grids. (C and D) Pheian et ai. (2014): CLs are mapped for each sampling site (Pennsylvania),
McDonnell et al. (2014b); Sullivan et al. (2011b); Sullivan et al. (2011a); CLs are mapped as a single point at the center point of the
watershed (New York and North Carolina).
Source; http://nadp.slh.wisc.edu/committees/clad.
Figure IS-11 Forest ecosystem critical loads for soil acidity related to base
cation soil indicators.
IS-60
-------
IS.6 Freshwater Ecosystem Nitrogen Enrichment and
Acidification
For freshwater systems, new evidence reinforces causal findings from the 2008 ISA
(Table IS-1). It also expands the scope of existing causal findings to include additional
biota affected by N enrichment and acidifying deposition and supports quantification of
these effects with new CLs (see Section IS.6.3.2). Freshwater systems include lakes
(lentic systems) and rivers and streams (lotic systems). In freshwater ecosystems, N may
cause N enrichment/eutrophication. Aquatic eutrophication results in increased
productivity of algae and aquatic plants, altered nutrient ratios, and sometimes decreased
oxygen levels. Deposition of N, S, or N + S can cause acidification, which affects
watershed biogeochemical processes and surface water chemistry. Freshwater N
enrichment and acidification take place in sensitive ecosystems across the U.S. at present
levels of deposition and may occur simultaneously in some water bodies.
New studies have added to the body of evidence in the 2008 ISA that N nutrient
enrichment and acidifying deposition alter freshwater biogeochemistry with subsequent
biological effects. There is new information on biogeochemical processes including
cycling of N and S. Both N enrichment/eutrophication and acidification can impact
physiology, survival, and biodiversity of sensitive aquatic biota. The 2008 ISA and new
studies provide examples of lakes and streams that show signs of eutrophication,
especially increased algal growth and shifts in algal biodiversity, in response to N
addition. The current causal statement for nutrient enrichment effects of N deposition
now includes altered algal growth and productivity as well as the endpoints of species
richness, community composition, and biodiversity reported in the 2008 ISA
(Table IS-1). For biological effects of aquatic acidification, the current causal statement
has been expanded from the 2008 ISA to include the specific endpoints of physiological
impairment, alteration of species richness, community composition, and biodiversity
(Table IS-1). New studies also show that despite reductions in acidifying deposition,
many aquatic ecosystems across the U.S. are still experiencing changes in ecological
structure and functioning at multiple trophic levels. Although there is evidence for
chemical recovery in many previously acidified ecosystems, biological recovery has been
limited (Appendix 8.4).
A number of freshwater monitoring efforts have facilitated the analysis of long-term
trends in surface water chemistry and ecological response in areas affected by acidifying
(N + S) deposition (Appendix 7.1.3). Many of these studies have been conducted in the
U.S., especially in the Northeast and the Appalachian Mountains. Although many of
these monitoring programs were in existence at the time of the 2008 ISA and were
considered in that analysis, more recent publications reflect the longer period of
IS-61
-------
monitoring and strengthen previous conclusions. Surface water chemistry data from
long-term monitoring by federal, state, and local agencies, as well as university research
groups and nonprofits has been combined into several publicly available metadatabases
(Appendix 7.1.3.2) enabling further regional trend analysis. Since the early 2000s,
U.S. EPA, together with the states, tribes, and other entities and individuals, have
collaborated on a series of statistically representative surveys (National Aquatic Resource
Surveys [NARS]) of the nation's waters, including surveys of lakes (U.S. EPA. 2016h.
2009b). streams (U.S. EPA. 2016i). wetlands (U.S. EPA. 2016i). and coastal waters (U.S.
EPA. 2016g). These periodic surveys, which are based on standard sampling and analysis
protocols and consistent quality assurance, include chemical and biological indicators of
nutrient enrichment and acidification (Appendix 7.1.3).
IS.6.1 Freshwater Biogeochemistry
In the 2008 ISA, evidence was sufficient to infer a causal relationship between N and S
deposition and the alteration of biogeochemical cycling of N and C in freshwater
ecosystems, and between acidifying deposition and changes in biogeochemistry of fresh
waters. As documented in the 2008 ISA and by newer studies, biogeochemical processes
and surface water chemistry are influenced by characteristics of the catchment and the
receiving waters. A number of studies since 2008 have focused on improving
understanding of aquatic acidification and eutrophication processes mediated by N. Many
of these studies have focused on pathways of pollutant and other constituent movement
within ecosystems, including monitoring studies of various kinds. Chemical indicators of
N deposition identified by the 2008 ISA were NO;, and DIN concentrations in surface
waters. Increased N deposition to freshwater systems via runoff or direct atmospheric
deposition, especially to N limited and N and phosphorus (P) colimited systems, can alter
N cycling (Appendix 7) and stimulate primary production (Appendix 9). Data from
long-term monitoring, experimental manipulations, and modeling studies provide
consistent and coherent evidence for biogeochemical changes associated with acidifying
N and S deposition. The strongest evidence for a causal relationship between acidifying
deposition and aquatic biogeochemistry comes from studies of changes in surface water
chemistry. Surface water chemistry indicators of acidic conditions and acidification
effects include concentrations of SO42 , NO;, . inorganic aluminum (Al), calcium (Ca),
sum and surplus of base cations, acid-neutralizing capacity (ANC), and surface water pH.
New information on biogeochemical cycling of N and S, acidifying deposition effects on
biogeochemical processes and changes to chemical indicators of surface water chemistry
associated with acidification and N nutrient enrichment is consistent with the conclusions
IS-62
-------
of the 2008 ISA, and the body of evidence is sufficient to infer a causal relationship
between N and S deposition and the alteration of freshwater biogeochemistry.
IS.6.1.1 Freshwater Processes and Indicators
Key processes and geochemical indicators of freshwater acidification and N enrichment
(Table IS-3) link to biological effects (Appendix 8 and Appendix 9). Surface water
chemistry integrates the sum of soil and water processes that occur upstream within a
watershed. Several key biogeochemical processes cause or contribute to surface water
eutrophication and acidification, and these processes have been the focus of substantial
research over the last three decades. Since the 2008 ISA, experimental studies, isotopic
analyses, and monitoring and observational studies have further investigated the cycling
of S, N, C, and base cations; these studies substantiate and further quantify earlier
findings.
Spatial and temporal patterns of NO, in lakes and streams have typically been used as
indicators that a freshwater system is receiving excess N which will cause acidification or
eutrophication. Qualitatively, northeastern U.S. spatial patterns in surface water NO,
concentrations suggest an influence by atmospheric N deposition. However, considerable
variation in the relationship between stream chemistry and deposition was associated
with land use and watershed attributes. It was well known at the time of the 2008 ISA
that key processes such as nitrification and denitrification are quantitatively important
portions of the N cycle and that they can be influenced by atmospheric inputs. More
recent research has further substantiated these earlier findings and provided additional
quantitative context (Appendix 7.1.2.3).
Deposition is a source of S to watersheds that, along with geologic sources of S such as
sulfide minerals, contribute S042 to surface waters (Appendix 4). The 2008 ISA found
that S deposition alters soil and drainage water chemistry through sustained leaching of
SO42 , associated changes in soil chemistry, and accumulation of S in the soil through
adsorption and biological assimilation. Declines in lake SO42 concentrations have been
observed in locations where S deposition has decreased significantly such as in the
Adirondack Mountains (Appendix 7.1.5.1). In addition, internal watershed sources of S,
which were earlier believed to be relatively minor in the northeastern U.S., have and will
likely continue to become proportionately more important as S deposition continues to
decline. Reductions in SOx deposition have not consistently resulted in increases of ANC
in surface water.
IS-63
-------
Table IS-3 Summary of key aquatic geochemical processes and indicators
associated with eutrophication and acidification.
Endpoint
N Driven
Nutrient
Enrichment
Acidification
The Effect of Deposition
Process
NOs"
leaching into
water bodies
X
X
Leaching from terrestrial ecosystems is an important source of NO3 in
freshwater ecosystems. See NO3" leachinq in Table IS-2.
so42~
leaching into
water bodies
X
Leaching from terrestrial ecosystems is an important source of SC>42~
in freshwater ecosystems. See SO42" leachina in Table IS-2.
Nitrification
X
X
Nitrification is an acidifying process, releasing 2 mol hydrogen ion (H+)
per mol NhV converted to NO3". As the N cycle becomes enriched
through cumulative N addition, net nitrification rates often increase,
and NO3" concentrations increase.
Denitrification
X
Denitrification is the microbial process that transforms NO3" by
anaerobically reducing it to NO2", NO, N2O, and N2.
DOC
leaching into
water bodies
X
X
DOC contributes to acidity of freshwater ecosystems. See DOC
leachina in Table IS-2.
Indicator
Surface
water [NO3"]
X
X
Increased N deposition (to surface waters or to terrestrial watershed;
see Table IS-2) increases the water NOs" concentration.
High concentrations of NO3" in lakes and streams, indicative of
terrestrial ecosystem N saturation, have been found at a variety of
locations throuahout the U.S. (U.S. EPA. 2006c: Stoddard. 1994).
Comparison of preindustrial estimates to modern measurements
suggested elevated concentrations in water bodies as a result of N
deposition (Fenn et al.. 2011b).
Surface
water DIN
X
Increased N deposition increases DIN in most freshwater aquatic
environments, largely as NO3".
Surface
water N:P
ratios
X
Increased N deposition can alter the ratio of N to P in freshwater
systems. Freshwater biota have different nutrient requirements and
changes in nutrient ratios may alter species richness, community
structure, and biodiversity, especially primary producers.
Surface
water [SO42"]
X
Increased S deposition (to surface waters or to terrestrial watershed,
see Table IS-2) increases the water SO42" concentration.
Comparison of preindustrial estimates to modern measurements
suggested elevated concentrations in water bodies are a result of S
deposition.
IS-64
-------
Table IS-3 (Continued): Summary of key aquatic geochemical processes and
indicators associated with eutrophication and
acidification.
N Driven
Nutrient
Endpoint Enrichment Acidification The Effect of Deposition
Surface X Several studies in the eastern U.S. suggested that base cation
water (base concentrations in surface waters increased during the initial phases of
cation) acidification into the 1970s. This trend reversed, and base cations
have decreased primarily in response to decreasing SO42" and NO3"
concentrations. Many base cations (especially Ca2+) are important
nutrients for aquatic biota.
Surface X Increased N and S deposition decrease ANC. Surface water ANC
water ANC correlates with other biologically influential chemical metrics, including
pH, inorganic Al concentration, Ca concentration, and organic acidity.
ANC <50-100 peq/L typically poses a risk for species survival, species
richness, and biodiversity.
Surface X Surface water pH is a common alternative to ANC as an indicator of
water pH acidification, but ANC is a better indicator at pH >6.0 and is less
sensitive to dissolved CO2. N and S deposition are associated with
decreasing pH in surface waters.
Surface X Acidifying N and S deposition increase mobilization of inorganic Al
water from terrestrial ecosystems into surface water, increasing surface
Inorganic Al water concentrations. Inorganic Al in surface waters is (1) widely toxic
and (2) leaches from terrestrial ecosystems only in response to acidic
conditions. Earlier studies demonstrated reduced growth and survival
of various species offish at inorganic Al concentrations between
approximately 2 and 7.5 pmol/L. Most recently, 20% mortality of
young-of-the year brook trout was documented in situ during a 30-day
period with a median inorganic Al concentration of 2 pmol/L.
Al = aluminum; ANC = acid-neutralizing capacity; Ca = calcium; C02 = carbon dioxide; DIN = dissolved inorganic nitrogen;
DOC = dissolved organic carbon; H+ = hydrogen ion; ha = hectare; kg = kilogram; L = liter; |jeq = microequivalents;
|jmol = micromole; N = nitrogen; N2 = molecular (atmospheric) nitrogen; N20 = nitrous oxide; NE = northeast; NH4+ = ammonium;
NO = nitric oxide; N02 = nitrogen oxide; N03" = nitrate; P = phosphorus; S = sulfur; S042" = sulfate; U.S. = United States;
USFS = U.S. Forest Service; yr = year.
IS.6.1.1.1 Acidification
The acidifying effects of N and S deposition in U.S. waters have been well characterized
for several decades. Traditionally, acidification involves both chronic and episodic
processes. Driscoll et al. (2001b) characterized chronically acidic lakes and streams as
having ANC of <0 |icq/L throughout the year, while episodic acidification occurs when
ANC falls below 0 (j,eq/L only for hours to weeks. Chronic acidification refers to average
conditions and is often measured as summer and fall chemistry for lakes and as spring
baseflow chemistry for streams. Chronic acidification is no longer prevalent in regions of
the U.S. affected by acidic deposition (Fakhraei et al.. 2016; Fakhraei et al.. 2014).
Episodic acidification is associated with precipitation or snowmelt events when high
IS-65
-------
volumes of water flow through watersheds. Episodes generally cause changes in the
following chemical parameters: ANC, pH, base cations, SO42 concentration, NO;,
concentration, inorganic A1 concentration, organic acid anions, or DOC. New studies
show that both N and S contributed to episodic acidification over a 20-year period at Bear
Brook, ME (see Appendix 7.1.5.1.2). It is known that the biota in many streams/lakes are
impacted when the ANC is consistently below 50 j^icq/L. For example, the U.S. EPA
National Lakes Assessment used an ANC threshold of >50 j^icq/L as indicative of
nonacidified water bodies (U.S. EPA. 2009b).
The most widely used measure of surface-water acidification is ANC. As reported in the
2008 ISA and newer studies, ANC is the primary chemical indicator of historic
acidification and for predicting the recovery expected from decreasing atmospheric
deposition. ANC correlates with the surface water constituents (pH, Ca2+, and inorganic
A1 concentration) that contribute to or ameliorate acidity effects in biota. As reported in
the 2008 ISA, lake and stream ANC values decreased throughout much of the 20th
century in a large number of acid-sensitive lakes and streams throughout the eastern U.S.
This effect has been well documented in monitoring programs, paleolimnological studies,
and model simulations (Appendix 7.1.5.1). Biological indicators of acidification, such as
decreased fish species richness, are discussed in Appendix 8.3.
Surface water pH is another indicator of acidification. It also correlates with surface
water chemical constituents that have biotic effects (inorganic Al, Ca2+, and organic
acids). The 2008 ISA included the scientific consensus that low pH can have direct toxic
effects on aquatic species (U.S. EPA. 2008a; Driscoll et al.. 2001b). A pH value of 6.0 is
the level below which biota are at increased risk from acidification (Appendix 8.3). The
2008 ISA noted that increasing trends in pH (decreasing acidification) were common in
surface waters in the northeastern U.S. through the 1990s and up to 2004. This trend has
continued in more recent times at many locations (Appendix 7.1.2.5). Rates of change
have generally been relatively small.
As stated in the 2008 ISA, the concentration of dissolved inorganic monomeric Al in
surface waters is an especially useful indicator of acidifying deposition because (1) it is
toxic to many aquatic species and (2) it leaches from soils only under acidic conditions
including acidifying deposition, acid mine drainage, or from rare geologic deposits.
Inorganic Al has well-documented effects on aquatic biota at specific thresholds
(Appendix 8.3) and is often the greatest threat to aquatic biota below pH 5.5. The 2008
ISA noted that concentrations of inorganic Al decreased slightly in some surface waters
in the northeastern U.S. in response to decreased levels of acidifying deposition,
suggesting chemical recovery in some of these surface waters (U.S. EPA. 2008a). and
IS-66
-------
this trend has generally continued (Appendix 7.1.5; see discussion on recovery
Section IS. 11).
Assessments of acidifying deposition effects dating from the 1980s and reported in the
2008 ISA showed S042 to be the primary acidifying ion in most acid-sensitive waters in
the U.S. The 2008 ISA presented temporal data that showed a trend of increasing
concentrations of S042 in surface waters before the period of peak S emissions in the
early 1970s. After the peak, SO42 surface water concentrations decreased in a
widespread trend. The rate of recovery varied by ecosystem, and new studies indicate that
as atmospheric S deposition has declined, soils with large stores of historically deposited
S (e.g., the Blue Ridge Mountain region) have begun releasing this adsorbed S to
drainage water (Appendix 4). preventing or slowing aquatic recovery.
As stated in the 2008 ISA, the quantitatively most important component of the overall
surface water acidification and chemical recovery responses has been the change in base
cation supply. Decreases in base cation concentrations in surface waters in the eastern
U.S. have been ubiquitous over the past two to three decades and closely tied to trends in
S042 concentrations in surface waters. Change in base cation supply with surface water
acidification was highlighted in Charles and Christie (1991) and in the 2008 ISA. Base
cations are added to watershed soils by weathering of minerals and atmospheric
deposition, and are removed by uptake into growing vegetation or by leaching. Acidic
deposition increased leaching of base cations, because SO42 anions percolating through
the soil tend to carry base cations along with them to maintain the charge balance. In
watersheds that received high levels of historical acidic deposition, current exchangeable
concentrations of Ca2+ and other base cations are substantially reduced from likely
preindustrial levels, having been depleted by many years of acidic deposition. This base
cation depletion in watersheds constrains ANC and pH recovery of surface waters, as
described in the 2008 ISA. New studies of base cations, which include experiments,
modeling, and gradient studies, have further corroborated these earlier findings.
Changes in DOC concentration or properties can affect the acid-base chemistry of surface
waters and perhaps the composition of aquatic biota. In soils and water, DOC constitutes
only a portion of dissolved organic matter (DOM), which also includes other constituents
such as organic nitrogen, phosphorus, and sulfur. However, the very large majority of
studies that include DOC do not explicitly include all of DOM. It has been recognized
that surface water DOC concentrations decreased to some extent as a result of
acidification, and that these concentrations would likely increase with recovery.
However, the strength of this response and the magnitude of DOC changes have
exceeded scientific predictions. Recent research on this topic has been diverse and has
included experiments, observation, isotope studies, and synthesis and integration work.
IS-67
-------
Overall, these studies illustrate large increases in DOC with acidification recovery in
some aquatic systems. Increases in DOC constrain the extent of ANC and pH recovery,
but decrease the toxicity of dissolved A1 by converting some of it from inorganic to
organic forms (Lawrence et al.. 2013). However, DOC is not an indicator of recovery
everywhere; some recovering sites have not shown increasing trends in DOC.
IS.6.1.1.2 Nitrogen Enrichment/Eutrophication
In aquatic systems, N is a nutrient that stimulates growth of primary producers (algae
and/or aquatic plants). Atmospheric deposition of N to freshwater systems can increase
the absolute supply of nutrients and alter N and P ratios. The freshwater ecosystems in
the U.S. most likely to be sensitive to nutrient enrichment from N deposition are
headwater streams, lower order streams, and alpine lakes, which have very low nutrients
and productivity and are far from local pollution sources ITJ.S. EPA (2008a);
Appendix 9.1.1.41. These nutrient shifts alter stoichiometric composition of water
chemistry, thereby shifting the nutrient status of lakes. Even small inputs of N in low
nutrient water bodies can affect biogeochemical processing of N and increase the
productivity of photosynthesizing organisms, resulting in a larger pool of fixed carbon
(C). Nutrient enrichment leads to changes in aquatic assemblages and biodiversity in
freshwater (Appendix 9) and coastal regions (Appendix 10).
Indicators of altered N cycling include changes in the concentrations of NO3 in surface
waters. The concentration of NO3 in water provides an index of the balance between
removal and addition of N to terrestrial ecosystems. Studies of several types have been
conducted in recent years to elucidate these processes and include experimental studies,
isotopic analyses, and monitoring and observational studies. Both water column and
sediment N transformations have been further characterized (Appendix 7.1.2.3). New
research suggests that denitrification may, in some situations, play a larger role than was
previously recognized in the 2008 ISA in removing oxidized N from the watershed.
As reported in the 2008 ISA and in newer studies, atmospheric N has been positively
correlated to total N in lakes along gradients of atmospheric deposition. N deposition in
some high-deposition lakes has changed the nutrient status of these lakes from a
more-or-less balanced (mainly N deficient) state to more consistently P limited
conditions (Appendix 9.2.4). Since the 2008 ISA, several studies have reported increases
in P deposition to water bodies in the U.S., possibly affecting shifts in lake trophic status
from P to N limitation or colimitation, as well as prolonging N limitation
(Appendix 9.1.1.2). In higher order streams, N deposition typically mixes with N derived
from other nonatmospheric sources, including urban/suburban point and nonpoint
IS-68
-------
sources, industrial sources, and agricultural sources, with atmospheric sources typically
being most pronounced during high flow conditions (Table 7-2).
IS.6.1.2 Models
Models used to assess the effects of N and S deposition on U.S. ecosystems were
reviewed in the 2008 ISA (Annex A). Several of the models used for terrestrial
ecosystems (see Section IS.5.3.3) such as MAGIC and PnET/BGC are also applicable to
aquatic systems. Both models have been widely applied, mainly to relatively small,
upland watersheds. Three other models, Spatially Referenced Regressions on Watershed
Attributes (SPARROW), Watershed Assessment Tool for Evaluating Reduction
Scenarios for Nitrogen (WATERS-N), and Surface Water Assessment Tool (SWAT)
have been used to evaluate N loading to mixed-use watersheds in larger river systems.
Another model that has been applied to the analysis of nutrient enrichment in aquatic
systems is AQUATOX, which simulates nutrient dynamics and effects on aquatic biota.
Few new freshwater acidification or eutrophication models have been developed and
published since 2008. A new national water quality modeling system (Hydrologic and
Water Quality System, HAWQS) is under development by Texas A&M University and
the USDA for the U.S. EPA's Office of Water (https: //epahawqs .tamu.edu/). The model
is intended to assist resource managers and policy makers in evaluating the effectiveness
of water pollution control efforts. Freshwater eutrophication and acidification models are
described in greater detail in Appendix 7.1.4.2.
IS.6.1.3 National-Scale Sensitivity
Sensitivity of lakes, streams, and rivers to biogeochemical changes associated with N and
S deposition varies across the U.S. The biogeochemical sensitivity to acidifying
deposition will be discussed together with biological sensitivity in Section IS.6.2.2.
Sensitivity to N enrichment will be discussed with biological sensitivity in
Section IS.6.3.2.
IS.6.2 Biological Effects of Freshwater Nitrogen Enrichment
In the 2008 ISA, evidence was sufficient to infer a causal relationship between N
deposition and the alteration of species richness, community composition, and
biodiversity in freshwater ecosystems. The freshwater systems most affected by nutrient
enrichment due to atmospheric deposition of N were remote oligotrophic high-elevation
IS-69
-------
lakes with low N retention capacity. In these ecosystems, N changes the biota, especially
by increasing algal growth and shifting algal communities. Freshwater organism
responses to N enrichment can be assessed through biological indicators, including
chlorophyll a, phytoplankton and periphyton (algae attached to a substrate) biomass,
diatoms, and trophic status. The current causal statement has been expanded to include
effects on algal growth and productivity (Table IS-1). New evidence since 2008 of N
enrichment includes paleolimnology, phytoplankton community dynamics,
macroinvertebrate response, and indices of biodiversity. This new evidence is consistent
with the conclusions and strengthens the evidence base of the 2008 ISA, and together, the
body of evidence is sufficient to infer a causal relationship between N deposition and
changes in biota, including altered growth and productivity, species richness,
community composition, and biodiversity due to N enrichment in freshwater
ecosystems.
IS.6.2.1 Physiology and Biodiversity Effects
Inputs of N to freshwater systems stimulate algal growth, which leads to a cascade of
effects on algal community composition and biodiversity. Algal species have differential
responses to N loading and shifts in nutrient ratios, so dominant species may change in
response to N enrichment. As reported in the 2008 ISA and in newer studies, shifts in
nutrient limitation from N limitation to colimitation by N and P, or to P limitation, have
been observed in some alpine lakes. New biodiversity studies are summarized in
Table 9-3. Since the 2008 ISA, several meta-analyses have reported an increase in P
atmospheric deposition to water bodies, highlighting the need to account for how
sustained P deposition can modify the effects of anthropogenically emitted N deposition
on productivity (Appendix 9.1.1.4). P addition delays the shift to P limitation (prolonged
N limitation) for phytoplankton.
IS.6.2.1.1 Primary Producers
The body of evidence for biological effects of N enrichment in remote freshwater
systems (where atmospheric deposition is the predominant source of N) is greatest for
phytoplankton, the base of the freshwater food web. Most studies focused on
phytoplankton, although several new studies indicate that both benthic and pelagic
primary producers respond to N inputs, and at least some studies have shown that
periphyton outcompeted phytoplankton for limiting nutrients (Appendix 9.3.3). The 2008
ISA and new studies include lake surveys, fertilization experiments, and nutrient
bioassays that show a relationship between increased N concentrations in the water
IS-70
-------
column and increased pelagic and benthic algal productivity (measured by chlorophyll a
concentration). An increase in lake phytoplankton biomass with increasing N deposition
was reported in the Snowy Range in Wyoming and in Europe. New studies in the
Colorado Rocky Mountains, where atmospheric deposition ranged from 2 to 7 kg
N/ha/yr, found correlations between higher chlorophyll a and higher rates of deposition
(Appendix 9.2.1).
The 2008 ISA and newer studies (Table 9-1 and Appendix 9.3.2) show a general shift in
algal dominance from chrysophytes that dominate low N lakes to cyanophytes and
chlorophytes in higher N lakes. Two nitrophilous species of diatom, Asterionellct formosct
and Fragilaria crotonensis, serve as indicators of N enrichment in lakes; however,
increased relative abundance of A. formosct has also been attributed to lake warming in
some regions where N deposition is decreasing (Appendix 9.3.2). New studies show that
glacial meltwater has higher NO;, relative to snow meltwater with different influences on
algal community composition in some regions of the U.S. (Appendix 9.3.2). In a
comparison of lakes in the Rockies with different meltwater sources, fossil diatom
richness in snowpack-fed lakes was at least double the richness of lakes with both glacial
and snow meltwater inputs; however, alterations in phytoplankton community structure
were not observed in lakes in the northern Cascade Mountains, WA. Some shifts in algal
biodiversity observed in high-elevation waters are attributed to climate change or nutrient
effects and climate as costressors (Appendix 13).
The role of N in freshwater harmful algal bloom formation has been further researched
since the 2008 ISA. Additional evidence continues to show that availability and form of
N influences algal bloom composition and toxicity, and inputs of inorganic N selectively
favor some HAB species, including those that produce microcystin. Microcystin is
prevalent in U.S. waters as reported in recent regional and national surveys. The risk of
HAB formation is low in high-elevation oligotrophic water bodies where N deposition is
the dominant source of N, but transport of atmospheric inputs can exacerbate eutrophic
conditions in downstream water bodies. Increased understanding of the role of N as a
limiting nutrient in many freshwater systems has led to recommendations to consider
both N and P in nutrient-reduction strategies.
Few studies in the U.S. have considered the effects of atmospheric deposition on aquatic
macrophytes, although declines in macrophyte occurrence were noted in a new survey of
Lake Tahoe that compared the lake's biota with that from a survey conducted in the
1960s (Caires et al.. 2013). Atmospheric N contributions are a substantial portion
(approximately 57%) of the total N loading to Lake Tahoe.
IS-71
-------
IS.6.2.1.2
Zooplankton
Compared to changes in primary producers, biological responses to N deposition at
higher trophic levels are not well characterized, but atmospheric N can alter food web
interactions (see Appendix 9.3.4). A few studies in the 2008 ISA and newer studies
showed zooplankton responses to N related shifts in phytoplankton biomass potentially
altering food web interactions.
IS.6.2.1.3 Macroinvertebrates
Few studies published since the 2008 ISA have linked atmospheric N deposition to
taxonomic shifts and declines in invertebrates (Appendix 9.3.5). These studies do not
attribute shifts in the abundance of higher invertebrates to N deposition alone, because
their abundance is also determined by additional factors such as climate and the presence
of invasive species. New studies provide additional evidence that trophic interactions
may moderate algal growth following nutrient loading. In Lake Tahoe, which receives
57% of N inputs from atmospheric sources, endemic invertebrate taxa have declined 80
to 100% since the 1960s due to nutrient inputs and invasive species.
IS.6.2.2 National-Scale Sensitivity and Critical Loads
New data have not appreciably changed the identification of sensitive lakes and streams
in the U.S. since the 2008 ISA. Nutrient enrichment effects from N most likely occur in
undisturbed, low-nutrient headwater and lower order streams and lakes at higher
elevations in the western U.S. (Appendix 9.1). including the Snowy Range in Wyoming,
the Sierra Nevada, and the Colorado Front Range. A portion of these lakes and streams
where effects are observed are in Class I wilderness areas which are afforded special
Clean Air Act protections. The responses of high-elevation lakes vary with catchment
characteristics (Appendix 9.1) and N deposition estimates at these high elevation sites are
associated with considerable uncertainty, especially dry deposition (Appendix 2). In these
systems, even low inputs of atmospheric N can shift N limitation to colimitation by N and
P, or to P limitation (Appendix 9.2.4). altering algal species composition and
productivity.
In the 2008 ISA, diatom assemblage shifts were reported at N deposition rates as low as
1.5 kg/N/ha/yr. Additionally, a hindcasting exercise in remote alpine Rocky Mountain
National Park lakes associated algal changes between 1850 and 1964 with an increase in
wet N deposition of 1.5 kg N/ha/yr. Since the 2008 ISA, empirical and modeled CLs for
the U.S. have been estimated based on surface water NO, concentration, diatom
IS-72
-------
community shifts, and phytoplankton biomass nutrient limitation shifts indicative of a
shift from N limitation to P limitation. A CL ranging from 3.5 to 6.0 kg N/ha/yr was
identified for high-elevation lakes in the eastern U.S. based on the nutrient enrichment
inflection point [where NO;, concentrations increase in response to increasing N
deposition; Baron et al. (201 lb)l. Another CL of 8.0 kg N/ha/yr for eastern lakes based
on the value ofN deposition at which significant increases in surface water NO,
concentrations occur was estimated by Pardo et al. (2011c). In both Grand Teton and
Yellowstone National Parks, CLs for total N deposition ranged from <1.5 ± 1.0 kg
N/ha/yr to >4.0 ± 1.0 kg N/ha/yr (Nanus et al.. 2017). Exceedance estimates were as high
as 48% of the Greater Yellowstone area study region, depending on the threshold value
ofNCV concentration in lake water selected as indicative of biological harm. An
empirical CL of 4.1 kg N/ha/yr above which phytoplankton biomass P limitation is more
likely than N limitation was identified by Williams et al. (2017b) for the western U.S.
Modeled CLs ranged from 2.8 to 5.2 kg/N/ha/yr.
IS.6.3 Biological Effects of Freshwater Acidification
The 2008 ISA found evidence sufficient to infer a causal relationship between acidifying
deposition and changes in aquatic biota, including strong evidence that acidified aquatic
habitats had lower species richness of fishes, macroinvertebrates, and phytoplankton. The
effects of acidifying deposition on aquatic ecosystems also include physiological
impairment or mortality of sensitive species and shifts in biodiversity of both flora and
fauna. Organisms at all trophic levels are affected by acidification, with clear linkages to
chemical indicators for effects on algae, benthic invertebrates, and fish (Table 8-9).
Biological effects are primarily attributable to low pH and high inorganic Al
concentration. ANC integrates chemical components of acidification (Table IS-2) but
does not directly alter the health of biota.
Effects of acidification on fish species are especially well characterized and many species
are harmed. Both in situ and lifestage experiments in fish support thresholds of chemical
indicators for biological effects. Most of these effects were documented in a rigorous
review of acidification effects on aquatic biota that was included in the 2008 ISA.
Overall, the updated research synthesized in this ISA reflects incremental improvements
in scientific knowledge of aquatic biological effects and indicators of acidification as
compared with knowledge summarized in the 2008 ISA. The fundamental understanding
of mechanisms has not changed, and the causal relationships between acidifying
deposition and biological effects on aquatic ecosystems are now, and were in 2008, well
supported. New studies also show that despite reductions in acidifying deposition,
alterations in aquatic biodiversity and ecosystem functioning caused by acidification
IS-73
-------
persist. Although there is evidence for chemical recovery in many ecosystems, biological
recovery has been limited (Section IS.6.2.2). New research is consistent with the causal
determination in the 2008 ISA and has strengthened the evidence base for these effects.
The current causal statement has been expanded to include specific endpoints of
physiological impairment, as well as effects at higher levels of biological organization
(Table IS-1). The body of evidence is sufficient to infer a causal relationship between
acidifying deposition and changes in biota, including physiological impairment and
alteration of species richness, community composition, and biodiversity in
freshwater ecosystems.
IS.6.3.1 Physiology and Biodiversity Effects
The deterioration in water quality caused by acidification affects the physiology,
survivorship, and biodiversity of many species from several taxonomic groups and at
multiple trophic levels. As stated in the 2008 ISA, biological effects are primarily
attributable to low pH (or ANC) and high inorganic A1 concentrations under chronic or
episodic acidic conditions. During acidification episodes, water chemistry may exceed
the acid tolerance of resident aquatic biota, with effects that include fish mortalities,
changes in species composition, and declines in species richness across multiple taxa.
Studies reviewed in the 2008 ISA showed that the earlier aquatic lifestages were
particularly sensitive to acidification. New effects thresholds have been identified for
aquatic organisms consistent with observations in the 2008 ISA (Table 8-10). New
evidence is congruent with findings in the 2008 ISA that high levels of acidification (to
pH values below 5 and ANC lower than the range of 50 to 100 (j,eq/L) eliminate sensitive
species from freshwater streams. This information is reviewed below.
IS.6.3.1.1 Primary Producers
Phytoplankton are primary producers at the base of the aquatic food web. These
photosynthetic organisms vary in tolerance of acidic conditions and include diatoms,
cyanobacteria, dinoflagellates, and other algal groups. The 2008 ISA reported reduced
species richness of freshwater plankton in response to acidification-related decreases in
pH and increases in inorganic Al. Effects were most prevalent when pH decreased to the
5 to 6 range. Effects on productivity are uncertain. Since the 2008 ISA, several
paleolimnological and field studies have further linked phytoplankton community shifts
to chemical indicators of acidification (Appendix 8.3). For example, Lacoul et al. (2011)
reviewed information on the effects of acidification and observed that the largest declines
in phytoplankton species richness occur over a pH range of 4.7 to 5.6 in Atlantic Canada.
IS-74
-------
IS.6.3.1.2 Zooplankton
Zooplankton comprise many groups of freshwater unicellular and multicellular organisms
including protozoans, rotifers, cladocerans, and copepods. Zooplankton feed on
phytoplankton or other zooplankton. Decreases in ANC and pH and increases in
inorganic A1 concentration have been shown to contribute to the loss of zooplankton
species or abundance in lakes. In the 2008 ISA, thresholds for zooplankton community
alteration were between pH 5 and 6. In the Adirondacks, a decrease in pH from 6 to 5
decreased zooplankton richness in lakes, and at ANC <0, zooplankton richness was only
45% of the richness in unacidified lakes. Newer studies support effects in a similar pH
range (see Appendix 8.3.1.2). Zooplankton have also been used as indicators of
biological recovery (Appendix 8.4.2).
15.6.3.1.3 Benthic Invertebrates
Acidification has strong impacts on aquatic invertebrates because H+ and A1 are directly
toxic to sediment-associated invertebrates like bivalves, worms, gastropods, and insect
larvae. In the 2008 ISA and in new studies in Appendix 8.3.3. decreases in ANC and pH
and increases in inorganic A1 concentration contribute to declines in abundance or
extirpation of benthic invertebrate species in streams. Acidification to pH values below
5 eliminates mayflies (Ephemeroptera), a taxa indicative of stream water quality, along
with other aquatic organisms. Since the 2008 ISA, a survey of benthic macroinvertebrates
by Baldigo et al. (2009) in the Adirondack Mountains indicated that macroinvertebrate
communities were intact at apH above 6.4, with moderate acidification effects at pH 5.1
to 5.7, and severe acidification effects at a pH less than 5.1. Similarly, thresholds of
pH 5.2 to 6.1 were identified for sensitive invertebrates from Atlantic Canada
(Appendix 8.3.3).
15.6.3.1.4 Fish
The effects of low pH and ANC and of high inorganic Al concentrations have been well
characterized in fish for many decades (Appendix 8.3.6). The 2008 ISA reported that
acidification impairs gill function and can cause respiratory and circulatory failure in fish.
Sensitivity to pH and inorganic Al varies among fish species, and among lifestages within
species, with early lifestages more sensitive to acidification. The most commonly studied
species were brown trout (Sctlmo trutta), brook trout (Salvelinus fontinalis), and Atlantic
salmon (Sctlmo salar). Studies published since the 2008 ISA, especially in Atlantic
salmon, add to the existing information on sublethal effects and confirm variation in
IS-75
-------
sensitivity among lifestages (Appendix 8.3.6.1). Since 2008, new studies include
acidification effects on migratory activities and behavior. New studies on fish show
behavioral effects at pH <6.6 (Appendix 8.3.6.5).
As summarized in Baker et al. (1990a) and the 2008 ISA, fish populations in acidified
streams and lakes of Europe and North America have declined, and some have been
eliminated as a result of atmospheric deposition of N and S and the resulting changes in
pH, ANC, and inorganic Al concentrations in surface waters. There is often a positive
relationship between pH and the number of fish species, particularly between pH 5.0 and
6.5. Additional pH thresholds published since the 2008 ISA (Table 8-2) support this
range, and several new studies consider the role of DOC in controlling pH and
subsequent effects on biota. In the 2008 ISA and in new research, few or no fish species
are found in lakes and streams that have very low ANC (near zero; Figure 8-4 and
Table 8-3) and low pH (near 5.0). The number of fish species generally increases at
higher ANC and pH values. Al is very toxic to fish, and thresholds to elevated
concentrations of this metal in acidified waters are summarized in Table 8-4.
IS.6.3.2 National-Scale Sensitivity, Biological Recovery, and Critical Loads
The extent and distribution of acid-sensitive freshwater ecosystems and sensitive regions
in the U.S. were well known at the time of the 2008 ISA. Measured data on lake and
stream ANC across the U.S. exhibit clear spatial patterns (Figure 8-11). Surface waters in
the U.S. that are most sensitive to acidification are largely found in the Northeast,
southern Appalachian Mountains, Florida, the upper Midwest, and the mountainous West
(Figure IS-12). Levels of acidifying deposition in the West are low in most areas and rare
in acidic surface waters, and the extent of chronic surface water acidification to date has
been very limited. However, episodic acidification occurs in both the East and West at
sensitive locations, and this is partly natural and partly caused by humans. Geographic
patterns in acidification sensitivity vary in response to spatial differences in geology,
hydrologic flow paths, presence and depth of glacial till, climate, and other factors
(Appendix 8.5.1). In the eastern U.S., acid-sensitive ecosystems are generally located in
upland, mountainous terrain underlain by weathering-resistant bedrock. Some of the most
in-depth studies of the effects of acid stress on fish were conducted in streams in
Shenandoah National Park in Virginia and in lakes in the Adirondack Mountains of New
York. Effects on fish have also been documented in acid-sensitive streams of the Catskill
Mountains of southeastern New York, and the Appalachian Mountains from
Pennsylvania to Tennessee and South Carolina.
IS-76
-------
Surface Water Critical Loads for Acidity
S+N meq/m2/yr - 10th Percentile
12-50
51 - 100
101 -200
201-4,215 Conditions:
Nn Data (1) ANC threshold: East = 50 peq/L, West = 20 peq/L
(2) Negative Critical Loads = 0.1 meq/m2/yr
States (3) NCLD v2.5 - 3/17/2015
ANC = acid-neutralizing capacity; meq = milliequivalent; yr = year.
Source: http://nadp.slh.wisc.edu/committees/clad.
Figure IS-12 Surface water critical loads for acidity in the U.S. 10th percentile
aggregation for 36-km2 grids with sulfur (S) and nitrogen (N).
Biological recovery in acid-affected areas is discussed in Section IS.ll. Typically,
biological recovery occurs only if chemical recovery (Appendix 7.1.5.1) is sufficient to
allow growth, survival, and reproduction of acid-sensitive plants and animals. Surface
water chemistry recovery varies by region, with the strongest evidence for improvement
in the Northeast and little or no recovery in central Appalachian streams. Acidification
and recovery of fresh waters will also be affected by the physical, chemical, and
biological modifications to acid inputs projected to occur with changes in annual mean
temperature and magnitude of precipitation (Appendix 8.5.3). As reported in the 2008
ISA and in new studies, biological recovery is slower than chemical recovery in many
IS-77
-------
systems (see Appendix 8.4). The time required for biological recovery is unknown and
only partial biological recovery may be possible.
Since the 2008 ISA, considerable CL research has focused on aquatic acidification in the
U.S. The CLs for deposition for aquatic acidification are expressed in eq/ha/yr of S, N, or
S + N because one or both pollutants can contribute to the observed effects. New
empirical CLs include 571 eq N/ha/yr in the Northeast and 286 eq N/ha/yr in the West to
prevent episodic acidification in high-elevation lakes (Table 8-7). Steady-state CLs have
been derived at many locations since the 2008 ISA (Table 8-8). Steady-state CLs of
acidifying deposition for lakes in the Adirondack Mountains (1,620 eq/ha/yr) and for the
central Appalachian streams (3,700 eq/ha/yr) were calculated to maintain a surface water
ANC of 50 (j,eq/L on an annual basis (NAPAP. 2011). CL values of less than
500 eq/ha/yr were calculated for one-third of streams in the Blue Ridge ecoregion, to
maintain stream ANC at 50 (ieq/L. For lakes in Class I and II wilderness areas in the
Sierra Nevada, CLs for acidifying deposition in 2008 were estimated at ANC values of 0,
5, 10, and 20 |icq/L. which span the range of minimum ANC values observed in Sierra
Nevada lakes. The median CL for granitic watersheds based on a critical ANC limit of
10 (ieq/L was 149 eq/ha/yr. Slightly more than one-third of these lakes had estimated
rates of acidifying deposition higher than their CL.
In addition to the steady-state and empirical loads described above, CL estimates have
been derived from dynamic modeling (Appendix 8.5.4). For example, there is new work
on simulated past and future effects of N and S on stream chemistry in the Appalachians
and Adirondack Mountain lakes. In 12 watersheds in the Great Smoky Mountain
National Park, target levels of ANC to protect aquatic life were used and ranged from
minimal (0 j^ieq/L) to considerable protection (50 j^icq/L). For the 12 study streams, target
levels ofNCV + SO42 deposition ranged from 270 to 3,370 eq/ha/yr to reach an ANC of
0 (j,eq/L by 2050 and 0 to 1,400 eq/ha/yr to reach an ANC of 50 |icq/L by 2050. However,
the majority of streams could not achieve the ANC target of 50 (ieq/L. Modeling also
suggests that complete recovery from acidification may not be possible by the year 2100
at all sites in the southern Blue Ridge region (Sullivan et al.. 201 lb) even if S emissions
cease entirely. In Shenandoah National Park, MAGIC modeling based on simulations of
14 streams identified a target load of about 188 eq S/ha/yr to achieve an ANC = 50 j^ieq/L
(preindustrial level based on hindcast simulations) in 2100 in sensitive streams. In a
dynamic modeling simulation in the Adirondack Mountains, about 30% of the lakes in
the region had a target load <500 eq/ha/yr to protect lake ANC to 50 j^ieq/L (Sullivan et
al.. 2012a). Future decreases in SO42 deposition are suggested to be more effective in
that region in increasing Adirondack lake water ANC than equivalent decreases in NO;,
deposition. In another modeling study of 20 Adirondack watersheds, estimates of
IS-78
-------
preindustrial ANC for the study lakes ranged from 18 to 190 (ieq/L, and simulations
estimate that lake ANC has decreased by 26 to 100 j^ieq/L as a legacy of acidification.
IS.7 Estuarine and Near-Coastal Ecosystem Nitrogen Enrichment
For estuaries (areas where fresh water from rivers meets the salt water of oceans) and
near-coastal systems, causality determinations from the 2008 ISA are further supported
and strengthened by additional studies (Table IS-1). Estuaries support a large biodiversity
of flora and fauna and play a role in nutrient cycling. N from the atmosphere and other
sources contributes to increased primary productivity, leading to eutrophication
(Table 10-1). and N pollution is the major cause of harm to most estuaries in the U.S.
(Appendix 10). Source apportionment data in the 2008 ISA and newer studies indicate
that atmospheric contributions to estuarine N are heterogeneous across the U.S., ranging
from <10% to approximately 70% of total estuary N inputs (Table 7-9). In estuaries,
increasing nutrient over-enrichment leading to eutrophication is indicated by water
quality deterioration, resulting in numerous harmful effects, including areas of low
dissolved oxygen (DO) concentration (hypoxic zones), species mortality, and HABs.
New studies support the 2008 ISA's causal findings that increased N loading to coastal
areas can alter biogeochemical processes and lead to shifts in community composition,
reduced biodiversity, and mortality of biota. The current causal statement of biological
effects of N enrichment in estuarine ecosystems has been expanded to include total
primary production, altered growth, and total algal community biomass (Table IS-1).
IS.7.1 Estuary and Near-Coastal Biogeochemistry
In the 2008 ISA, the evidence was sufficient to infer a causal relationship between
reactive N deposition and biogeochemical cycling of N and C in estuarine and
near-coastal marine systems. Evidence reviewed in the 2008 ISA, along with new studies,
indicates elevated N inputs to coastal areas can alter key processes that influence N and C
cycling in near-coastal environments. As external organic matter loading to coastal areas
has increased in recent decades in many parts of the U.S., the varying rates of different N
cycling processes within estuaries themselves can also affect the magnitude of
eutrophication experienced as a result of external N enrichment. Nitrogen additions not
only cause the total pool of N to be larger but may also perturb N cycling in such a way
that the system may exacerbate eutrophication to a greater extent than expected based on
N additions alone. Research conducted since the 2008 ISA has shown that many of these
N cycling processes are more important in the estuarine environment than previously
understood. The removal of N through denitrification is a valuable ecosystem service in
IS-79
-------
terms of constraining the extent and magnitude of eutrophication. Additional research has
established dissimilatory NO;, reduction to NH4+ (DNRA) as a more important N
reduction pathway in some estuaries. Ammonium produced via DNRA can lead to
enhanced productivity and respiration, which may exacerbate hypoxia. Recent studies
indicate that DNRA rates are higher in warmer months and can also take up a larger
percentage of total N reduction activity when temperatures are higher. The roles of
sedimentary microbial processes of denitrification and N2 production via anaerobic
ammonium oxidation (anammox) have been further characterized. New research has
shown that the community of N fixing microorganisms is more diverse in estuarine and
coastal waters than previously thought, and that N fixation occurs more widely than
previously assumed. Influence of benthic macrofauna on N cycling has received
increased research attention in part due to the potential for these organisms to mitigate
external N enrichment. Along with atmospheric anthropogenic CO2 inputs and other
factors, eutrophication from N loading may affect carbonate chemistry in coastal areas,
contributing to acidifying conditions in some circumstances such as where there is spatial
or temporal decoupling of production and respiration processes. Monitoring of coastal
areas shows that excess nutrient inputs continues to be a widespread problem in many
parts of the U.S. New research further supports conclusions of the 2008 ISA, and the
body of evidence is sufficient to infer a causal relationship between N deposition and
the alteration of biogeochemistry in estuarine and near-coastal marine systems.
IS.7.1.1 Nitrogen Enrichment
Estuarine biogeochemistry is complicated because it directly controls more than just the
N cycle; the response to N loading resulting in eutrophication affects the chemical
cycling of metals and DO (Appendix 7.2.3). redox conditions, pH (Appendix 7.2.4). and
ultimately energy transfer (e.g., food webs from microbes to humans). The response to N
loading is also tightly controlled by the availability of organic matter (i.e., C) and its
lability and reactivity. External organic matter loading to estuarine and coastal waters
appears to be increasing and these excess nutrient inputs are occurring within the context
of other stressors such as climate change (Appendix 7.2.6.12) and rising atmospheric
CO2, which further modify coastal biogeochemistry (Doncv. 2010). As reported in the
2008 ISA, estuaries are generally N limited, and have received sufficiently high levels of
N input from human activities (including deposition, agricultural runoff, and wastewater)
to cause eutrophication. Highly variable environments within estuaries are characterized
by a gradient of increasing salinity toward the ocean. As N moves downstream, some
fraction is taken up by phytoplankton or removed by microbial denitrification. Key
processes that influence N cycling include hypoxia, nitrification, denitrification, and
IS-80
-------
decomposition. Until recently, it was generally believed that NH3 oxidation was
accomplished only by Proteobacteria in marine environments. New research has
discovered that some archaea can also oxidize NH3. These ammonia-oxidizing archaea
are dominant in some estuaries, while ammonia-oxidizing bacteria are more important in
others.
In the complex environment of the freshwater-to-ocean continuum, there are many
chemical and biological indicators of eutrophic condition. One approach is to measure
total nutrient loading and concentrations; however, these data need to be interpreted in
the context of the physical and hydrological characteristics that determine ecosystem
response. Water quality measures such as pH and DO, along with key biological
indicators such as chlorophyll a, phytoplankton abundance, HABs, macroalgal
abundance, and submerged aquatic vegetation (SAV; rooted vascular plants that do not
emerge above the water), can all be used to assess responses to nutrient loading
(Table 10-1). Nitrogen removal from the estuary is also influenced by faunal as well as
microbial communities.
Organic particles in coastal regions sink to the sediment-water interface where they
accumulate and decompose. Decomposition of these organic particles transforms
nutrients and depletes O2 in the water. Decreasing DO can create hypoxic (<2 mg/L of
dissolved O2) or anoxic zones inimical to fish and other aerobic life forms. Oxygen
depletion largely occurs only in bottom waters under stratified conditions, not throughout
the entire water column. This can result in seasonal hypoxia in shallow coastal regions,
particularly those that are receiving high inputs of nutrients from coastal rivers.
Development of hypoxia is increasingly a concern in estuaries across the U.S.
(Appendix 10.2.4).
Since the 2008 ISA, a number of papers have identified links between nutrient
enrichment and effects on estuarine carbonate chemistry, resulting in coastal acidification
or basification (Appendix 7.2.4). Eutrophication and acidification/basification are
complex biogeochemical processes that are driven by the same hydrological
(stratification) and biological (production/respiration) processes that can result in hypoxia
and enhanced organic matter loading. Acidification can occur by direct atmospheric
anthropogenic CO2 dissolution into the ocean. But under certain conditions N enrichment
can contribute to acidifying/basifying conditions, such as in systems with strong thermal
stratification or with spatial or temporal decoupling of production and respiration
processes. With increasing N inputs to coastal waters, CO2 in the water column is
produced from degradation of excess organic matter from changing land use, as well as
respiration of living algae and seagrasses, which in turn can make the water more acidic.
IS-81
-------
Estuarine carbonate chemistry is complex, responding to a wide variety of natural,
anthropogenic, physical (mixing), chemical and biological drivers.
IS.7.1.2 Models
Since the 2008 ISA, several new applications of existing models have quantified
eutrophication processes in estuaries and near-coastal marine ecosystems. These have
included studies that focused primarily on N cycling or hypoxia. Other models of
estuarine eutrophication focus on N load apportionment, or on relationships between N
loads and ecological endpoints. Since the 2008 ISA, SPARROW has been used to
estimate total N loads within watersheds to determine sources of N to streams and rivers;
it has also been applied at regional and national scales. Additional models and tools that
include the contribution of N directly from the atmosphere have been applied to U.S.
estuaries, including the Watershed N Loading Model (NLM) and the Watershed
Deposition Tool (WDT). The latter was developed by the U.S. EPA to map atmospheric
deposition estimates to watersheds using wet and dry deposition data from CMAQ
(Schwede et al.. 2009). This tool links air and water quality modeling data for use in total
maximum daily load (TMDL) determinations and analysis of nonpoint-source impacts.
New model applications include studies that focused primarily on endpoints of N cycling,
hypoxia, and HABs. Models of coastal eutrophication are described in greater detail in
Appendix 7.2.8.
IS.7.1.3 National-Scale Sensitivity
Sensitivity of estuaries to biogeochemical changes associated with N enrichment varies
across the U.S. The biogeochemical sensitivity of estuaries and near coastal areas will be
discussed together with national-scale biological sensitivity to N enrichment in
Section IS.7.3.
IS.7.2 Biological Effects of Nitrogen Enrichment
In the 2008 ISA, evidence was sufficient to infer a causal relationship between N
deposition and the alteration of species richness, community composition, and
biodiversity in estuarine ecosystems. The strongest evidence for a causal relationship was
from changes in biological indicators of nutrient enrichment (chlorophyll a, macroalgal
[seaweed] abundance, HABs, DO, and changes in SAV; Table 10-1). Some indicators,
such as chlorophyll a, are directly linked to nutrient enrichment and provide evidence of
IS-82
-------
early ecosystem response; other indicators, such as low DO and decreases in SAV,
indicate more advanced eutrophication. Phytoplankton are the base of the coastal food
web and increases in primary producer biomass and altered community composition
associated with increased N can lead to a cascade of direct and indirect effects at higher
trophic levels. At the time of the 2008 ISA, N was recognized as the major cause of harm
to the most estuaries in the U.S. Since 2008, new paleontological studies, observational
studies, and experiments have further characterized the effects of N on phytoplankton
growth and community dynamics, macroinvertebrate response, and other indices of
biodiversity. For this ISA, new information is consistent with the 2008 ISA and the
causal determination has been updated to reflect more specific categories of effects to
include total primary production, altered growth, and total algal community biomass.
This new research strengthens the evidence base and is consistent with the 2008 ISA
(Table IS-1) that the body of evidence is sufficient to infer a causal relationship
between N deposition and changes in biota including total primary production,
altered growth, total algal community biomass, species richness, community
composition, and biodiversity due to N enrichment in estuarine environments.
Since the 2008 ISA, additional evidence has shown that reduced forms of atmospheric N
play an increasingly important role in estuarine and coastal eutrophication and HAB
dynamics. New studies emphasize that N inputs interact with physical and hydrologic
factors to increase primary productivity and eutrophication in coastal areas.
Climate-related changes in temperature, precipitation, and wind patterns, as well as
extreme weather events, stronger estuary stratification, increased metabolism and organic
production, and rising sea-levels are all expected to modify coastal habitats
(Appendix 10.1.4.1).
IS.7.2.1 Primary Producers
Algae are the base of the coastal food web, and the 2008 ISA showed that changes in
chemical composition of N inputs can shift the algal community and cascade up the food
web. Chlorophyll a is a broadly recognized indicator of phytoplankton biomass and is
used as a proxy for assessing effects of estuarine nutrient enrichment. It can signal an
early stage of water quality degradation related to nutrient loading and is incorporated
into water quality monitoring programs and national-scale assessments including U.S.
EPA's National Coastal Condition Assessment (Appendix 7.2.7). Phytoplankton
sampling, microcosms studies, and sediment core analysis have shown changes in
phytoplankton community structure in estuaries with elevated N inputs (Appendix 10.3).
These shifts at the base of the food web to species that are not as readily grazed
(e.g., cyanobacteria, dinoflagellates) have a cascade of effects including poor trophic
IS-83
-------
transfer and an increase in unconsumed algal biomass, which could stimulate
decomposition and O2 consumption, and thus increase the potential for hypoxia.
There is consistent and coherent evidence that the incidence of HAB outbreaks is
increasing in both freshwater and coastal areas, a problem that has been recognized for
several decades (Appendix 10.2.2). Of the 81 estuary systems for which data were
available for the National Estuarine Eutrophication Assessment (NEEA), 26 exhibited a
moderate or high symptom expression for nuisance or toxic algae (Bricker et al.. 2007).
Since the 2008 ISA, HAB conditions and effects of HAB toxins on wildlife have been
further characterized (Appendix 10.2.2). Toxins released during HABs can be harmful to
fish and shellfish and may be transferred to higher trophic levels. The form of N affects
phytoplankton growth and toxin production of some HAB species. Increasing loads of
NH3+/NH4+ have been linked to the expansion of HABs and altered phytoplankton
community dynamics (Appendix 10.3.3). Cyanobacteria, and many chlorophytes and
dinoflagellates, may be better adapted to NH44", while diatoms generally thrive in the
presence of oxidized forms of N such as NO;, (Figure 10-7).
Macroalgal (seaweed) growth is also stimulated by increased N inputs, which increase
the dominance of faster growing benthic or pelagic macroalgae to the exclusion of other
species (Appendix 10.2.3). Studies published since the 2008 ISA provide further
evidence that macroalgae respond to the form of N, with some species showing greater
assimilation and growth rates with NH4 than with NO;, . Increased abundance of
macroalgae, which block light, and increased epiphyte loads on the surface of SAV may
reduce the growth and biomass of SAV. SAV, including the eelgrass Zostera marina, are
important ecological communities found within some coastal bays and estuaries that are
sensitive to elevated nutrient loading, and the loss of this habitat can lead to a cascade of
ecological effects because many organisms are dependent upon seagrasses for cover,
breeding, and as nursery grounds. Recently, the presence of seagrass beds was linked to
decreased bacterial pathogens of humans, fishes, and invertebrates in the water column
and lower incidence of disease in adjacent coral reefs (Appendix 10.2.5). The 2008 ISA
reported correlations between increased N loading and declines in SAV abundance, and
newer studies have further characterized this relationship. In a survey of southern New
England estuaries, reduced eelgrass extent was observed at increased watershed N
loading. New studies have characterized the role of invertebrate mesograzers, such as
small crustaceans and gastropods, in controlling algal growth, potentially buffering
eutrophication effects on seagrass communities (Appendix 10.3.7). Macroalgae may not
be a good indicator of eutrophication in some upwelling-influenced estuaries in the
Pacific Northwest because an increase in macroalgal biomass in these systems does not
appear to be associated with temporal declines in eelgrass (Appendix 10.2.3).
IS-84
-------
IS.7.2.2 Bacteria and Archaea
Ammonia-oxidizing prokaryotes carry out nitrification in estuarine waters.
Ammonia-oxidizing archaea have been described relatively recently, and several studies
since the 2008 ISA have considered community responses of ammonia-oxidizing bacteria
and archaea. Community structure of ammonia-oxidizers is related to nutrient inputs and
affected by the form of available N (Appendix 10.3.4).
15.7.2.3 Invertebrates
In coastal areas with severe seasonal hypoxia, the community of benthic organisms shifts
toward shorter life spans and smaller body size (Appendix 10.2.4). Reduced species
density and diversity in the northern Gulf of Mexico are linked to persistent hypoxic
events. The form of N present has been shown to affect molluscan taxonomic
assemblages (Appendix 10.3.5). Shifts in algal composition and productivity can affect
growth of shellfish that feed on phytoplankton. Shellfish contribute to N and C cycling
and can improve water quality, and recent research has explored the use of these
organisms for coastal N remediation (Appendix 7.2.6.11). Harvest of shellfish for human
consumption removes nutrients from estuaries.
N enrichment is one of several factors linked to increased disease susceptibility,
bleaching, and reduced calcification rate in corals (Appendix 10.4.2). Several studies
have isolated effects of N, which affects corals via pathways that are distinct from P. The
threatened status of staghorn coral (Acropora cervicornis) and elkhorn coral (Acropora
palmata) under the U.S. Endangered Species Act has been linked to indirect N pollution
effects, specifically low DO, algal blooms that alter habitat, and other non-nutrient
stressors (Hernandez et al.. 2016). Increasing acidification of coastal waters, which may
be exacerbated by elevated N inputs under certain circumstances (Appendix 7.2.4). is
projected to alter marine habitat, have a wide range of effects at the population and
community level and affect food web processes. Although the interactions between
elevated CO2, decreasing pH, and nutrient inputs are complex, calcareous plankton,
oysters, clams, sea urchins, and coral that produce calcium carbonate shells may be
affected by long-term decreases in pH (Appendix 10.5).
15.7.2.4 Fish
Fish biodiversity is altered by increased N inputs and resulting changes in biological and
chemical indicators (Appendix 10.3.6). Many fish are unable to persist at DO levels
IS-85
-------
below 2 mg/L (Figure 10-4). Recent studies in the southern Gulf of Saint Lawrence have
linked SAV loss to declines in fish biodiversity, although organisms did not change
positions within food webs. In laboratory conditions, turbidity associated with
eutrophication alters fish reproductive behaviors. Hypoxia has also recently been shown
to affect reproduction in fish. For example, hypoxia acts as an endocrine disruptor in
Atlantic croaker (Micropogonias iincliilaiiis: Appendix 10.2.4).
IS.7.3 National-Scale Sensitivity and Critical Loads
The NEEA, the most recent comprehensive survey of eutrophic conditions in U.S.
estuaries conducted by the National Oceanic and Atmospheric Administration, defined
eutrophication susceptibility as the natural tendency of an estuary to retain or flush
nutrients (Bricker et al.. 2007). In estuaries that have longer water residence times,
nutrients are more likely to lead to eutrophic conditions (Appendix 10.1.4). As reported
in the 2008 ISA and newer studies, nutrient loading accelerates hypoxia, which is more
likely in marine waters with limited water exchange, water column stratification, and
high production and settling of C to bottom waters. Other factors identified in the
2008 ISA that increase estuary sensitivity to eutrophication include human population,
agricultural production, and the size of the estuary relative to its drainage basin. The
NEEA reported that the most eutrophic estuaries in the U.S. occur in the mid-Atlantic
region, and the estuaries with the lowest degree of eutrophication are in the North
Atlantic (Figure 10-2). Estuaries identified in the 2008 ISA as susceptible to
eutrophication include the Chesapeake Bay, Pamlico Estuary in North Carolina, Long
Island Sound, as well as along the continental shelf adjacent to the Mississippi and the
Atchafalaya River discharges to the Gulf of Mexico. New research at the regional scale
includes long-term studies of several coastal systems that are looking at trends in coastal
water quality and chemistry. A 23-year study of the Chesapeake Bay concluded that
water quality has decreased and chlorophyll a levels have increased since 1986, in part
due to long-term climate trends (see Appendix 10.2.5).
Since the 2008 ISA, there is additional information on the extent and severity of
eutrophication and hypoxia in sensitive regions. Areas of eutrophication-related hypoxia
are found on the U.S. eastern and western coasts and the Gulf of Mexico (Figure 10-5).
The 2008 ISA reported that the largest zone of hypoxic coastal water in the U.S. was the
northern Gulf of Mexico on the Louisiana-Texas continental shelf. In the summer of
2017, the hypoxic zone in the Gulf was the largest ever measured at 14,123 km2
[8,776 mi2; U.S. EPA (2017f)l. Atmospheric deposition to watersheds in the
Mississippi/Atchafalaya River Basin contributes approximately 16 to 26% of the total N
load to the Gulf of Mexico (Appendix 10.2.4). Long Island Sound also experiences
IS-86
-------
periods of anoxia. In other U.S. coastal systems, hypoxia incidence is increasing, but DO
impacts are relatively limited temporally and spatially. In the Pacific Northwest, coastal
upwelling not related to anthropogenic sources can be a large source of nutrient loads,
and the advection of this upwelled water can introduce hypoxic water into estuaries.
The NEEA suggested that only a small fraction of the estuary systems evaluated showed
moderate to high SAV loss (Bricker et al.. 2007). mostly in the mid-Atlantic region.
While seagrass coverage is improving in some estuaries, such as Tampa Bay (Tampa Bay
Case Study, Appendix 16). many estuaries continue to see declines in seagrass extent.
SAV is often at a competitive disadvantage under N enriched conditions because of the
fast growth of opportunistic macroalgae that preferentially take up NH44" and can block
light from seagrass beds.
There are thresholds of response identified for some biological and chemical indicators of
N enrichment in estuaries (Appendix 10). The amount of chlorophyll a is an indicator of
phytoplankton biomass, and thus, a proxy for assessing estuarine nutrient enrichment. In
general, 0-5 (ig/L chlorophyll a is considered a good condition, concentrations between 5
and 20 (ig/L are classified as fair condition, and concentrations of >20 j^ig/L indicate poor
conditions (Table 10-2). A new response threshold of tidal-averaged total N
concentration of <0.34 mg/L has been identified for healthy eelgrass in Massachusetts
waters. Markedly decreased eelgrass coverage is observed at N loading rates
>100 kg N/ha/yr, and levels above 50 kg N/ha/yr are likely to impact SAV habitat extent
in shallow New England estuaries (Table 10-4). Greaver et al. (2011) identified the range
of 50-100 kg N/ha/yr total N loading as the empirical CL for loss of eelgrass based on
Latimer and Rego (2010). In terms of DO, concentrations of 0 mg/L are anoxic, 0-2 are
indicative of hypoxic conditions, and 2-5 mg/L are biologically stressful conditions
(Figure 10-4). Oxygen depletion largely occurs only in bottom waters under stratified
conditions, not throughout the entire water column.
The indicators of nutrient enrichment in coastal areas (chlorophyll a, HABs, macroalgal
abundance, DO, SAV, and benthic diversity) have been incorporated into indices of
coastal eutrophication. In the 2008 ISA, the Assessment of Estuarine Tropic Status
(ASSETS) categorical Eutrophication Condition index (ECI) developed for the NEEA
was used as an assessment framework for coastal U.S. estuaries (Bricker et al.. 2007).
Additional indices of estuarine functioning that incorporate biological indicators have
since been developed both in the U.S. and internationally (Appendix 10.2.6).
Comparisons of these frameworks have identified robust methods to measure estuarine
response, such as incorporation of annual data, frequency of occurrence, spatial coverage,
secondary biological indicators, and a multicategory rating scale.
IS-87
-------
Since the 2008 ISA, N enrichment has been linked to coral bleaching and reduced
calcification rates (Appendix 10.4.2). Near-coastal coral reefs in the U.S. occur off south
Florida, Texas, Hawaii, and U.S. territories in the Caribbean and Pacific.
IS.8 Wetland Ecosystem Nitrogen Enrichment and Acidification
New evidence, including new CLs, supports and strengthens the causal findings from the
2008 ISA regarding N enrichment effects in wetlands (Table IS-1). In freshwater
wetlands and coastal wetland ecosystems, deposition of N and S does not tend to cause
acidification-related effects at levels currently common in the U.S. However, the 2008
ISA documented that wetlands can be sensitive to N enrichment and eutrophication
effects. Newer studies have characterized N effects on biogeochemistry, physiology,
biodiversity, national sensitivity, and CLs for freshwater and coastal wetlands; coastal
wetlands are typically tolerant of higher N loading than freshwater wetlands.
IS.8.1 Wetland Biogeochemistry
In the 2008 ISA, evidence was sufficient to infer a causal relationship between N
deposition and the alteration of wetland biogeochemical cycling. Although sources and
rates of N inputs vary widely among wetlands, N deposition contributes substantially to
total loading in many wetlands. This additional N alters C cycling, N cycling, and the
release of nutrients to hydrologically connected surface waters. New research together
with the information included in the 2008 ISA shows that the body of evidence is
sufficient to infer a causal relationship between N deposition and the alteration of
biogeochemical cycling in wetlands.
The 2008 ISA reported that N enrichment altered N cycling in wetland ecosystems.
Chemical indicators of N deposition in wetlands include NO;, and NH4 leaching, DON
leaching, N mineralization, denitrification rates, and N2O emissions. A wetland can act as
a source, sink, or transformer of atmospherically deposited N, and these functions vary
with season and hydrological conditions. Vegetation type, physiography, local hydrology,
and climate all influence source/sink N dynamics in wetlands. A new synthesis of global
wetland data showed that a wetland "s reactive N removal and water quality improvement
is proportional to its reactive N load, and removal efficiency is 26% higher in nontidal
than tidal wetlands. Further, a new meta-analysis showed that N enrichment increases
wetland N2O emissions by 207%. New studies have also evaluated the effects of N
loading/N addition on other endpoints related to N cycling in peat bog, riparian,
mangrove, and salt marsh wetlands (see Appendix 11.3.1). The endpoints evaluated
IS-88
-------
include ecosystem N retention, wetland export of N to surface waters, N fixation, N
mineralization, denitrification, emission of N2O, and bacterial abundance, activity, and
composition in wetland soils. The results of North American studies are summarized in
Figure 11-2. Across studies, N enrichment decreases the ability of wetlands to retain and
store N, which may diminish the wetland ecosystem service of improving water quality.
In the 2008 ISA, evidence from Canadian and European peatlands showed that N
deposition had negative effects on Sphagnum (moss) bulk density and mixed effects on
Sphagnum productivity depending on the history of deposition. There is new information
on how N deposition alters biogeochemical cycling of C in wetlands. Chemical indicators
of N deposition in wetlands include soil organic matter, total soil C or peat C, CO2
emissions, and CH4 emissions. Long-term C storage is an important ecosystem service of
wetlands for which measures of physical marsh stability can serve as a proxy, and
physical indicators of N deposition can include temperature, bulk density, physical
resistance, and soil water content. In addition, changes to plant growth rates and
productivity indicate altered C cycling in wetlands, and are summarized in Section IS.8.2.
The literature evaluates the effects of N deposition, N loading, or experimental N
addition on C cycling in bogs, fens, riparian or intermittent marshes, freshwater tidal
marshes, mangroves, and salt marshes (see Appendix 11.3.2). Significant effects of N
loading upon biogeochemical cycling of C in North American wetlands (in which the N
addition was 500 kg N/ha/yr or lower) are summarized in Figure 11-3. N enrichment
decreases wetland retention of C, as indicated by new studies and a new meta-analysis
that show that N enrichment increases methane production in salt marshes. New studies
of marshes along the Gulf Coast and East Coast find that N enrichment also decreases the
bulk density of salt marshes, making marshes less resilient to physical stresses from tidal
or storm flooding, and may accelerate coastal marsh loss.
IS.8.2 Biological Effects of Wetland Nitrogen Enrichment/Eutrophication
In the 2008 ISA, evidence was sufficient to infer a causal relationship between N
deposition and the alteration of species richness, species composition, and biodiversity in
wetlands. New evidence is presented in the following sections regarding the effects of N
upon wetland plant physiology, architecture, demography, and biodiversity. The body of
evidence is sufficient to infer a causal relationship between N deposition and the
alteration of growth and productivity, species physiology, species richness,
community composition, and biodiversity in wetlands.
IS-89
-------
IS.8.2.1 Growth, Productivity, and Physiology
In the 2008 ISA, evidence from Canadian and European bogs and fens showed that N
deposition had negative or mixed effects on Sphagnum (moss) productivity, depending on
history of deposition. In Canadian ombrotrophic peatlands experiencing deposition of
2.7-8.1 kg N/ha/yr, peat accumulation increased with N deposition, but accumulation
rates had slowed by 2004, indicating a degree of N saturation. Coastal wetlands
responded to N enrichment with increased primary production, which shifted microbial
and plant communities and altered pore water chemistry, although many of the studies in
coastal wetlands used N enrichment levels more like those of wastewater than
atmospheric deposition. New research on N enrichment effects on growth and
productivity was conducted in ombrotrophic bogs, intermittent wetlands, freshwater tidal
marsh, mangroves, and coastal salt marshes (see Appendix 11.4). Ecological endpoints
evaluated to assess N loading effects on growth and productivity include plant
aboveground biomass and productivity, plant belowground biomass of roots and
rhizomes, and growth rates, and are summarized along with N effects on C cycling in
Figure 11-3. The effects of N additions on plant physiology were not addressed in the
2008 ISA, but information regarding these effects is available for bogs and fens, riparian
wetlands, freshwater tidal marshes, mangroves, and salt marshes (see Appendix 11.5).
Ecological endpoints evaluated to assess N loading effects on plant physiology include
stoichiometry (i.e., nutrient concentrations and ratios of multiple nutrients in plant tissue),
nutrient acquisition efficiency (including insectivory rates in carnivorous plants), nutrient
use efficiency, and nutrient reabsorption efficiency. These endpoints are summarized in
Figure 11-4.
In general, across types of wetlands, nitrogen loading increases aboveground growth and
productivity while decreasing or not affecting belowground growth and productivity. In
bogs and fens, N deposition decreases growth of state-listed Sarmcenia purpurea (purple
pitcher plant), and N enrichment increases aboveground productivity of emergent sedges
more than of peat-building moss species. These changes cascade up to affect biodiversity
in bogs and fens (see below, Section IS.8.2.2). In freshwater and tidal marshes, N
enrichment increases aboveground productivity while decreasing belowground
productivity, and this shift from belowground to aboveground plant productivity may
account for changes in wetland C storage (see Section IS.8.1).
Changes to plant physiology and stoichiometry vary by species tolerance to N and N
acquisition strategies. In bogs, N enrichment typically causes increased plant tissue N
concentrations, decreased N use efficiency, and decreased N resorption efficiency during
senescence. After several years of exposure to high rates of N loading, bog plants may
experience leafN saturation and limitation by other nutrients (e.g., P, K, and Ca,
IS-90
-------
indicated by increasing reabsorption efficiencies), resulting in leaf damage in sensitive
species. S. purpurea (purple pitcher plant) decreases its dependence upon insectivory for
nutrition at N deposition rates of 4.4 kg N/ha/yr. In freshwater marshes, N enrichment
also increases plant tissue N concentrations while increasing P limitation and altering
resorption efficiencies.
Plant architecture was not addressed in the 2008 ISA, and demography was addressed
only for bogs and fens. Aboveground, plant architecture includes branching patterns, as
well as the size, shape, and position of leaves and flower organs. New studies find N
enrichment affects plant architecture in a salt marsh, in mangroves, in freshwater tidal
marshes, and in a riparian wetland (Appendix 11.6). In terms of plant demography, the
2008 ISA found positive population growth rates for S. purpurea at 0 or 1.4 kg N/ha/yr,
but population losses at 14 kg N/ha/yr. N deposition above 6.8 kg N/ha/yr increases
population extinction risk of S. purpurea. New studies show that N addition has
species-specific effects on reproduction of West Coast salt marsh plant species and that it
increases mortality across the global distribution of mangrove species (Appendix 11.7).
IS.8.2.2 Biodiversity
In the 2008 ISA, evidence was sufficient to infer a causal relationship between N
deposition and the alteration of species richness, species composition, and biodiversity in
wetlands. Notably, the 2008 ISA cited 4,200 native plant species in U.S. wetlands, 121 of
which are federally endangered. Given their relative area, wetlands provide habitat to a
disproportionally high number of rare plants. Many wetland species have adapted to N
limited conditions, including endangered species in the genera Isoetes (3 endangered
species) and Sphagnum (15 endangered species), as well as insectivorous plants such as
pitcher plants (Sarracenia spp.) and sundews (Drosera rotundifolici).
Coastal wetlands responded to N enrichment with increased primary production,
changing microbial and plant communities, and altered pore water chemistry, although
many of the studies available in 2008 used high N enrichment levels more similar to N
loading from wastewater than from atmospheric deposition. New research since 2008
across environmentally relevant N levels including N deposition gradient studies,
experimental N addition studies, and observational studies show that N enrichment
altered biodiversity in bogs and fens, intermittent wetlands, freshwater wetlands,
freshwater tidal wetlands, and coastal salt marshes (see Appendix 11.8).
New research from wetland ecosystems strengthens the 2008 causal statement. New
research confirms that, as in terrestrial systems, N addition can decrease the abundance
and richness of sensitive species while increasing the abundance and richness of tolerant
IS-91
-------
species. In bogs and fens, N enrichment decreases the survival of insectivorous plants and
the cover of mosses, while increasing the cover of shrub species. In freshwater marshes,
N enrichment changes plant community composition, increases the abundance of and
stresses caused by invasive plant species, promotes the harmful algal species that produce
the toxin microcystin, and increases mosquito larvae that are vectors for zoonotic
diseases (see Figure 11-1). In freshwater tidal and coastal marshes, N enrichment changes
plant community composition, increases cover of invasive plant species, increases
herbivory by invertebrates, and increases herbivory by the invasive mammal Mvocastor
coy pus (nutria).
IS.8.2.3 National Sensitivity and Critical Loads for Wetlands.
Freshwater and coastal wetlands tend to have different sensitivity to added N. Broadly,
wetlands that receive a larger fraction of their total water budget in the form of
precipitation are more sensitive to the effects of N deposition. For example, bogs
(70-100% of hydrological input from rainfall) are more sensitive to N deposition than
fens (55-83% as rainfall), which are more sensitive than coastal wetlands (10-20% as
rainfall).
Since the 2008 ISA, an N CL for U.S. coastal wetlands has been established. The CL is
based on several different ecological endpoints, including plant community composition,
microbial activity, and biogeochemistry (63-400 kg N/ha/yr) and that this CL includes
total N loading values not just N deposition. Figure 11-6 shows a comparison of the N
CL for coastal wetlands with recent studies of ecological impacts of N (at N levels of
100-250 kg N/ha/yr).
Since the 2008 ISA, two N CLs for U.S. freshwater wetlands have been established. The
CL for wetland C cycling, quantified as altered peat accumulation and NPP, is between
2.7 and 13 kg N/ha/yr. The upper end of this CL range is based on measurements of wet
deposition only (10 to 13 kg N/ha/yr), and therefore, does not reflect total N loading.
There is also a CL to protect biodiversity based on morphology and population dynamics
of the purple pitcher plant (Sarracenia purpurea) between 6.8-14 kg N/ha/yr. A more
recent study across an N deposition gradient suggests that purple pitcher plant
populations experience negative effects ofN deposition at rates lower than this CL, but
the more recent research has not yet been incorporated into the CL framework. A
comparison of freshwater wetland CLs to observed ecological impacts of N from recent
studies (4.4-500 kg N/ha/yr) is given in Figure 11-7.
IS-92
-------
IS.9 Freshwater and Wetland Ecosystem Sulfur Enrichment
New evidence from wetland and freshwater aquatic ecosystems strengthens and extends
the causal findings of the 2008 ISA regarding nonacidifying sulfur effects and provides
the basis for a new causal determination (Table IS-1). New research together with the
information included in the 2008 ISA shows that the evidence is sufficient to infer a
causal relationship between S deposition and the alteration of Hg methylation in surface
water, sediment, and soils in wetland and freshwater ecosystems. New evidence is
sufficient to infer a new causal relationship between S deposition and changes in
biota due to sulfide phytotoxicity, including alteration of growth and productivity,
species physiology, species richness, community composition, and biodiversity in
wetland and freshwater ecosystems.
SOx deposition can have chemical and biological effects other than acidification,
particularly in flooded wetland soils and aquatic ecosystems. The 2008 ISA described
qualitative relationships between SO42 deposition and a number of ecological endpoints,
including altered S cycling, sulfide phytotoxicity, internal eutrophication of aquatic
systems, altered methane emissions, increased mercury (Hg) methylation, and increased
Hg loading in animals, particularly fish. Table 12-11 summarizes the chemical
concentrations that alter ecological endpoints and the quantitative relationships
describing the effects of SO42 deposition. Recent research supports these relationships
between S deposition and ecological endpoints and provides the basis for SOx deposition
levels, water column SO42 concentrations, and water column sulfide concentrations
protective of plants and animals.
IS.9.1 Biogeochemistry
SOx deposition alters biogeochemical processes via S enrichment. The processes include
S cycling (see Appendix 12.2.1). P cycling (see Appendix 12.2.4). C cycling (see
Appendix 12.2.5). and Hg cycling (see Appendix 12.3). The primary chemical indicator
for nonacidifying or enrichment effects of S in wetland and aquatic ecosystems is surface
water SO42 concentration, as it is for acidifying effects. The 2008 ISA reported that
chemical reduction of SO42 was an important indicator of SOx effects on water
chemistry because the process generates ANC. There are no new studies on ANC
generation through SO42 reduction, although microbial SO42 reduction remains an
active area of research. In aquatic ecosystems for which atmospheric and terrestrial S
inputs are similar in magnitude to rates of microbial SO42 reduction, the products of
microbial SO42 transformation may be more reliable indicators of S enrichment effects
IS-93
-------
than surface water SO42 concentrations. These chemical indicators include
methylmercury (MeHg), sulfide, and phosphate.
MeHg is the most persistent and toxic form of Hg in the natural environment. It is
measured in surface water or aquatic sediments (MeHg concentration or the percentage of
MeHg in total Hg) to predict its effects on biota. Several new studies demonstrate
significant positive relationships between surface water SO42 concentrations and water
or sediment MeHg concentrations (see Appendix 12.3.5). Another product of SO42
reduction, sulfide (measured as surface water or sediment pore water S2 concentrations),
is also a water quality indicator of deposition effects on biota. In freshwater ecosystems
with iron-rich sediments, sulfide may react with iron bound to phosphates in the sediment
to release phosphate into the water column, increasing primary productivity recent
literature refers to this process as internal eutrophication (Appendix 12.2.4).
In terms of S enrichment effects on carbon cycling, the 2008 ISA documented the
suppression of methane emissions in wetland soils by SO42 addition in several studies
and noted that 15 kg S/ha/yr suppressed methane emissions. Recent research has
confirmed that S enrichment increases the abundance or metabolic activity of
SO42 -reducing prokaryotes (SRPs), which under some conditions compete with
methanogens by suppressing their activity, and in turn, suppressing methane emissions
(Appendix 12.2.4). However, there are no new studies documenting S deposition effects
on methane emissions in U.S. ecosystems.
IS.9.2 Biological Effects of Sulfur Enrichment
Nonacidifying S effects upon biota include plant toxicity, changes in plant growth and
biodiversity, and increased Hg concentrations in biota. The toxicological effects of Hg
accumulation in animals were documented in the 2008 ISA and newer studies.
IS.9.2.1 Sulfur Nutrient and Toxicity to Plants
Plants and other organisms require S as an essential nutrient. The deposition of S can
affect plant protein synthesis by affecting S availability for S containing amino acids,
which in turn will affect N uptake. The 2008 ISA documented the effects of SO42
toxicity on plant development and reproduction at very high S loads. There is no new
evidence of S deposition effects upon plant S nutrition or SO42 toxicity. The product of
microbial SO42 reduction, sulfide, is an important plant toxin, and the 2008 ISA
documented sulfide phytotoxicity in European systems. Together with new research
showing sulfide phytotoxicity in North American wetlands, the body of evidence is
IS-94
-------
sufficient to infer a causal relationship between S deposition and changes in biota
due to sulfide phytotoxicity including alteration of growth and productivity, species
physiology, species richness, community composition, and biodiversity in wetland
and freshwater ecosystems.
The 2008 ISA showed that sulfide toxicity decreased the biomass of wetland plants and
aquatic macrophytes in mesocosms under aquatic S concentrations higher than current
U.S. concentrations. In Europe, research showed that a threshold value of <48 mg
S042 /L in surface water would protect the sensitive aquatic species Stratiotes aloides and
Potamogeton acutifolius (not native to contiguous US), as well as to protect P.
zosteriformis and Utricidciria vulgaris, which are both native and widely distributed in
contiguous US. New research has demonstrated sulfide phytotoxicity effects at current
ambient sulfide concentrations in multiple ecosystems within the U.S. (Appendix 12.2.3).
Sulfide decreased total plant cover and cover of dominant species in a New York fen and
decreased the growth rate of Cladium jamcticense (sawgrass), a keystone species in the
Florida Everglades. Zizaniapalustris (wild rice) is an economically and culturally
important species sensitive to sulfide, and the Minnesota Pollution Control Agency has
developed a model for this species that calculates protective levels of water SO42
concentrations, given (specific) iron and DOC concentrations in water bodies. A recent
review identifies sulfide thresholds between 0.3-29.5 mg S27L for altered growth,
productivity, physiology, or increased mortality of 16 freshwater wetland emergent plant
and aquatic submerged macrophyte species native to North America (see Table 12-2).
IS.9.2.2 Sulfur Effects on Mercury Methylation
In the 2008 ISA, evidence was sufficient to infer a causal relationship between S
deposition and increased methylation of Hg in aquatic environments where the value of
other factors is within an adequate range for methylation. In the 2008 ISA,
sulfur-reducing bacteria (SRB) were identified as the organisms responsible for Hg
methylation. New evidence shows the ability to methylate Hg is more broadly distributed
phylogenetically, including both bacteria and archaea, which is why this document refers
to S042 -reducing mercury methylators as sulfur-reducing prokaryotes (SRPs) rather than
SRB (Appendix 12.3.2). In the 2008 ISA, wetland and lake-bottom sediments were
identified as habitat for mercury methylating SRPs. Recent research documents microbial
mercury methylation in lakes, in wetland sediments and moss, within periphyton, in
marine ecosystems, and within disturbed terrestrial forest soils (Appendix 12.3.2 and
Appendix 12.3.3). Microbial mercury methylation responsive to SOx deposition occurs in
freshwater lakes, freshwater wetlands, freshwater reservoirs, and freshwater agricultural
areas (Appendix 12.3.4). Between the 2008 ISA and new research, the body of evidence
IS-95
-------
is sufficient to infer a causal relationship between S deposition and the alteration of
Hg methylation in surface water, sediment, and soils in wetland and freshwater
ecosystems.
Hg methylation is determined in part by surface water S042 . because many strains of
SRPs possess the recently identified gene pair hgcAB, and pair their metabolism of C
with both dissimilatory S042 reduction and mercury methylation (see Appendix 12.3.2
and Figure 12-5). Microbial methylation rates are determined by other environmental
requirements of SRPs, including seasonality and temperature, pH, salinity, amount of
organic matter in the water and sediments, and concentrations of iron and nitrate
(Appendix 12.3.3). New research demonstrates that Hg methylation occurs at current
ambient SO42 concentrations within U.S. water bodies. Multiple lines of evidence
support a relationship between SO42 surface water concentrations and MeHg
concentration or production in various freshwater systems. Linear relationships between
S042 concentrations and MeHg concentrations were observed in sediments of the South
River, VA, across peat bogs in Minnesota and Ontario, and across prairie pothole lakes in
Saskatchewan (Figure 12-17). In addition to the studies of lake and wetland sediments
reviewed in the 2008 ISA, studies employing lab incubations show that SO42 increases
Hg methylation in samples from Adirondack peat bogs, from South River, VA sediments,
from periphyton growing in North American lakes and wetlands, and from leaf packs in
Minnesota river water (Appendix 12.3.3.1). Experimental addition of S to field
mesocosms or whole ecosystems has shown that S enrichment as wet S deposition
increases MeHg in water, sediment, or biota, in Little Rock Lake, WI; Bog Lake Fen,
MN; the Experimental Lakes Area, Ontario; and the bog experiment at Degero Stormyr,
Sweden (Appendix 12.3.4.1). In observational studies of S and Hg deposition, fish Hg
concentrations decline with temporal declines in SOx deposition in Isle Royale (a Class I
area). Fish Hg concentrations correlate positively with Hg and S deposition across Texas
ecoregions, and a 12-year study found that fish Hg in Voyageurs National Park (a Class I
area) declined in lakes with decreasing S deposition only when lake DOC remained
constant (Appendix 12.3.5.1). New research is consistent and coherent with the research
presented in the 2008 ISA in demonstrating that sulfur enrichment from SOx deposition
stimulates mercury methylation in North American ecosystems. Current research
suggests that mercury methylation generally peaks between 10 and 100 mg SO42 /L in
surface water, and quantitative relationships between S and Hg, such as target values or
thresholds, are reported in Table 12-12.
IS-96
-------
IS.9.2.3 Sulfur, Mercury, and Animal Species
Mercury is a developmental, neurological, endocrine, and reproductive toxin across
animal species. The 2008 ISA documented Hg accumulation in fish, songbirds, four turtle
species, insectivorous passerine birds, and the common loon (Gcivici immer). Recent
research also documented Hg accumulation in insectivore songbirds, bats, and fish in
agricultural wetlands. The 2008 ISA reported that 23 states had issued fish advisories by
2007 in response to the U.S. EPA's fish tissue criterion of 0.3 (.ig MeHg/g fish (0.3 ppm),
set to protect human health. The 2008 ISA reported on the negative impacts of Hg on the
development, morphology, survival, or reproduction in the following fish species:
walleye (Stizostedion vitreum), grayling (Thymallus thymallus), mummichog (Fundulus
heteroclitus), rainbow trout (Oncorhvnchus mykiss), fathead minnows (Pimephcdes
promelets), and zebrafish (Danio rerio). However, a recent report on Hg in streams of the
U.S. by the USGS summarizes current research indicating that birds, fish, and fish-eating
wildlife experience negative effects of Hg at lower concentrations than the 0.3 ppm
criterion set to protect human health on the basis of fish consumption.
The 2008 ISA documented a link between decreased S deposition and decreased fish
MeHg concentrations. Recent research in Voyageurs National Park (a Class I Area)
supports this finding, and there is supporting evidence from fish surveys of Texas
reservoirs across regions with different S deposition loads. There is also supporting
evidence from an S addition experiment in a peat bog in the Marcell Experimental Forest
in northern Minnesota, where increased S loading increased Hg concentrations in larval
Culex spp. (mosquitoes), which are an important food source for both aquatic and
terrestrial species (Appendix 12.4 and Figure 12-18). In addition to the studies that
consider S deposition, there are recent studies that consider SO42 concentrations in water
in relation to fish Hg concentrations in six lakes in South Dakota, and in the marshes of
the Everglades (Appendix 12.4). In the freshwater marshes of the Everglades, recent
work indicates a concentration of 1 mg/L S042 to keep water MeHg low
(Appendix 12.3.4.3) and protect fish from elevated Hg burdens in that system
(Figure 12-14).
IS.9.3 National-Scale Sensitivity and Critical Loads
The 2008 ISA identified ecosystems in the Northeast as particularly sensitive to Hg
methylation in response to S deposition because many watersheds in this region have
abundant wetlands and freshwater water bodies with high DOC and low pH. The U.S.
EPA national stream surveys found that MeHg in predator fish exceeded the Hg criterion
in a quarter of stream miles and half the lakes surveyed. Fish MeHg levels were highest
IS-97
-------
in streams in watersheds with considerable wetland area, and surveys showed highest fish
MeHg concentrations in the southeastern U.S., suggesting that ecosystems sensitive to
SOx deposition effects on Hg methylation extend beyond the Northeast (Figure 12-15).
Recent studies confirm that Hg methylation is more widespread than was documented at
the time of the 2008 ISA. New research conducted in agricultural wetlands in California
suggests Hg methylation in these systems may provide a route to animal and human Hg
exposure through food, specifically MeHg concentrations in rice seeds.
There are no CLs for S to prevent sulfide phytotoxicity or Hg methylation, although there
are SO42 and sulfide water quality values that represent protective levels against toxic
effects of sulfide and Hg to biota (see Table 12-12). There are European CLs for Hg
concentrations in soil and fish tissue targeted to protect human health, drinking water
quality, and terrestrial soils, but these CLs are not framed in terms of SOx, Hg, or PM
deposition (see Appendix 12.6).
IS.10 Ecological Effects of Particulate Matter Other Than Nitrogen
(N) and Sulfur (S) Deposition
Since publication of the 2009 PM ISA, new literature builds upon the existing knowledge
of ecological effects associated with PM components other than those associated with N
and S deposition, especially metals and organics. In some instances, new techniques have
enabled further characterization of the mechanisms of PM on soil processes, vegetation,
and effects on fauna. New studies provide additional evidence for community-level
responses to PM deposition, especially in soil microbial communities. However,
uncertainties remain due to the difficulty in quantifying relationships between ambient
concentrations of PM and ecosystem response. Overall, the body of evidence is
sufficient to infer a likely causal relationship between deposition of PM and a
variety of effects on individual organisms and ecosystems, based on information from
the previous review and new findings in this review. However, the new findings are
limited in scope.
PM deposition comprises a heterogeneous mixture of particles differing in origin, size,
and chemical composition. Exposure to a given concentration of PM may, depending on
the mix of deposited particles, lead to a variety of toxic responses and ecosystem effects.
Effects of PM on ecological receptors can be both chemical and physical (U.S. EPA.
2009a. 2004). As described in the 2009 Integrated Science Assessment for Particulate
Matter (2009 PM ISA), particulates that elicit direct and indirect effects on ecological
receptors vary by size, origin, and chemical composition. Ecological outcomes are
attributed more to particle composition than to particle size (Grantz et al.. 2003).
IS-98
-------
PM-associated metals and organics are linked to responses in biota; however, the
heterogeneous nature of PM composition and distribution coupled with the variability
inherent in natural environments confound assessment of the ecological effects of
particulates. Although most effects are from chemical composition of PM, there are some
effects of particle size such as changes to flux of solar radiation and soiling of leaves by
large coarse particles in areas near industrial facilities and unpaved roads. Atmospheric
deposition of PM from crustal material may be a source of base cations (especially Ca2+,
Mg2+, and K+) that can partially ameliorate the effects of acidifying deposition. Base
cations are important plant nutrients that in some locations are in short supply (U.S. EPA.
2009a).
In general, new studies on PM deposition to vegetation support findings in previous PM
reviews on altered photosynthesis, transpiration, and reduced growth. Since the 2009 PM
ISA, additional characterization of PM effects at the leaf surface has led to a greater
understanding of PM foliar uptake. Alterations in leaf fatty acid composition are
associated with metals transferred to plant tissues from PM deposition on foliar surfaces
(Appendix 15.4.2).
An important characteristic of fine particles (0.1 to 1.0 |im) is their ability to affect the
flux of solar radiation increases in the diffuse component. A newly available research
method links changes in expression of proteins involved in photosynthesis to increases in
the diffuse component due to aerosols and PM. Although this method has not been
widely applied, it may represent an important way to study mechanistic changes to
photosynthesis in response to more diffuse radiation resulting from PM in the air column
(Appendix 15.2).
Several studies published since the 2009 PM ISA show PM chemical constituent effects
on soil physical properties and nutrient cycling. Previous findings in the PM ISA of
changes to microbial respiration and biomass are further supported by new studies.
Microbial communities respond to PM in various ways depending on their tolerance to
heavy metals and organics (Appendix 15.5.3).
In fauna, results from ecotoxicity assays with PM extracts using bacteria, rotifers,
nematodes, zebrafish, and earthworms support findings in the 2009 PM ISA that toxicity
is not related to the total mass of PM in the extract, but to the chemical components of the
PM. In nematodes exposed to PM from air filters, the insulin-signaling pathway was
identified as a possible molecular target. Use of wildlife as PM biomonitors has been
expanded to new taxa since the last PM review. Several studies in invertebrates and birds
report physiological responses to air pollutants, including PM (Appendix 15.6).
IS-99
-------
For ecosystem-level effects, a gradient of response with increasing distance from PM
source was reported in the 2009 PM ISA. Newly available studies from long-term
ecological monitoring sites provide limited evidence for recovery in areas such as those
around former smelters due to the continued presence of metals in soils after operations
ceased. A novel experimental microecosystem using microbial communities living in
terrestrial mosses indicates that PM deposition alters responses of primary producers,
decomposers, and predators (Appendix 15.3).
IS.11 Recovery of Ecosystems from Nitrogen (N) and Sulfur (S)
Deposition in the U.S.
Evidence from across the U.S. of ecosystem recovery from N nutrient enrichment and
acidification corresponding to long-term trends in N and S emissions varies. Most studies
of recovery focus on ecosystem acidification recovery due to decreases in S emissions
and deposition. Overall N emissions and deposition have been increasing or relatively
steady, although a few areas have seen some decrease (Appendix 2.7). Consequently, the
amount of new information available and reported here on N enrichment recovery is
small.
IS.11.1 Overarching Concepts of Ecological Recovery from Acidification
Both chemical and biological indicators are used to assess the degree of ecological
degradation associated with environmental stressors and document responses in
ecosystems where improved conditions allow for recovery. Recovery can be documented
by measurement of indicators and projected/modeled recovery trajectories.
Chemical recovery of aquatic and terrestrial ecosystems is characterized by trends in
water quality indicators (NCh , SO42 . pH, ANC, inorganic monomeric Al, MeHg)
towards inferred preindustrial values or, in the case of inorganic Al and MeHg, below
water quality threshold values protective of biota and human health. Preindustrial
conditions varied across the U.S. depending on climate, geology, and biological
communities, and preindustrial chemical indicator values are currently inferred from
models, paleolimnology samples, or historical samples. When evaluating ecosystem
recovery from acidification, it is important to note that different chemical pools within
the soil or water column may recover at different rates with the same decreases in
atmospheric deposition. For example, the soil solution CaAl ratio, SO42 . or NO,
respond more quickly than will total N. Indicators of slowly recovering pools (such as the
percentage of base saturation in the soil or soil C to N ratio) will have long response
IS-100
-------
times with regard to changes in atmospheric deposition. An indicator such as
acid-neutralizing capacity (ANC), which is influenced by both fast and slow pools, has an
intermediate response time. Chemical indicators such as ANC or pH may not necessarily
follow a recovery path that mirrors the reverse of the acidification path due to dynamic
relationships among ANC, pH, DOC, and inorganic Al; depletion of soil base cation
pools; and/or pH-dependent S adsorption on soils. In addition, the ANC level that reflects
recovery of pH or Al may differ between the acidification and recovery phases
(Hcsthagcn et al.. 2008).
Biological recovery may follow chemical recovery of such water and soil quality
constituents; however, there may be a lag of decades between the onset of chemical
recovery and biological recovery ITJ.S. EPA (2008a); Appendix 81. As observed in some
of the early studies on formerly acidified systems, the biological recovery trajectory may
exhibit hysteresis, where a system does not follow the same path from acidification to
recovery (Frost et al.. 2006). Complete biological recovery would entail a return to the
same species make-up, richness, and abundance as existed in the ecosystem in question
prior to the advent of human-caused acidic deposition (around the year 1860 in North
American ecosystems). In a practical sense, complete biological recovery is probably not
attainable at most acidified locations within a reasonable management time frame
(perhaps 100 years) because soil reserves of base cations at many locations have been
depleted in response to many decades of acidic deposition and because other stressors, in
addition to acidic deposition, have also altered ecosystem structure and/or function or
will do so in the coming decades. Such stressors include changes in climate, land use, and
other perturbations. More commonly, partial biological recovery may be possible.
Ecosystems deemed to be on a recovery trajectory are those found to be moving towards
a mix of species presence and abundance that approximates the undisturbed state. There
is substantial evidence that recovery rates from acidification differ between taxonomic
groups [e.g., rotifers vs. crustaceans; Frost et al. (2006); Mallev and Chang (1994)1. In
general, recovery in freshwater ecosystems is characterized by populations of plankton
and benthic invertebrates prior to the recovery of fish populations, although most
biological communities studied to date have not returned to preacidification conditions,
even after recovery of chemical parameters.
IS.11.2 Acidification Recovery in the U.S.
Long-term monitoring has been very important in tracking the ecological response to N
and acidifying deposition (Appendix 7 and Appendix 4.4). Experimental liming studies
have also provided some evidence for biological recovery, although these types of studies
are limited in the U.S. (Appendix 4.3.4 and Appendix 8.4.6). The historical focus on
IS-101
-------
aquatic acidification has resulted in more data to evaluate recovery in aquatic than
terrestrial ecosystems (Appendix 7.1.3). Fewer studies have tracked the potential
recovery of terrestrial ecosystems; however, since the early 1990s, increasing evidence
indicates that soils in some areas are beginning to recover, yet most sensitive regions
continue to acidify in response to deposition (Appendix 4.6.1). In areas where N and S
deposition has decreased, chemical recovery must first create physical and chemical
conditions favorable for growth, survival, and reproduction of the pre-1860 assemblage
for biological recovery to occur.
The northeastern U.S. and southern Appalachians are two regions of the U.S. where a
large body of research has evaluated recovery. In the Northeast, evidence for chemical
recovery is primarily from soils (Appendix 4.6.1) and freshwater lakes and streams
(Appendix 7.1.5.1). In regard to biological recovery (Appendix 8.4). newer studies have
documented some evidence for zooplankton recovery and the successful reintroduction of
brook trout in previously acidified Adirondack water bodies or recolonization of
previously acidic lakes from refugia (Appendix 8.6.6). In addition to decreased
acidification, a few studies report declines in methylmercury concentrations in biota or
water in response to decreasing S, which is suggestive of ecosystem recovery
(Appendix 12.5).
In contrast to the northeastern U.S., there is little evidence for recovery in the southern
Appalachian Mountain region (Appendix 4.6.1 and Appendix 16.3). This area is
characterized by an abundance of low-ANC streams situated on acidic, highly weathered
soils. Streams in this region are strongly affected by SO42 adsorption on soils, and
long-term monitoring studies suggest that soil base cation depletion has prevented
chemical recovery (Appendix 7.1.5.1.4). Biogeochemistry modeling scenarios suggest
that even with large decreases in SO42 deposition, it may take decades for soil base
cation levels to recover in this region.
New studies continue to support findings in the 2008 ISA that biological response to
water chemistry recovery varies among taxa and water bodies, and that most biological
communities studied have not returned to preacidification conditions, even after recovery
of chemical parameters (Appendix 8.4). Since the 2008 ISA, research has demonstrated
that the DOC of many lakes and streams has risen, with the source of the DOM and
associated DOC likely to be the soils in the terrestrial watershed (Table IS-2;
Appendix 4.3.9 and Appendix 7.1.2.9). The mechanism causing the observed increase in
DOC is unclear; it may be a combination of soil recovery from acidification, changes in
climate (e.g., temperature and precipitation), and N deposition, among other mechanisms.
DOC interacts like a weak acid; therefore, DOC concentration may affect pH and ANC
levels and constrain the extent of recovery from acidification. At the same time, the
IS-102
-------
acidic properties of DOC make it a host for binding trace metals such as toxic inorganic
A1 (for additional discussion on inorganic A1 and DOM see Appendix 4.3.5) and
decreases the toxicity of dissolved A1 to aquatic organisms. Overall, current research
indicates DOC increases are inconsistent across surface waters in the U.S., with large
increases in DOC with acidification recovery in some locations and no increases in other
recovering sites.
IS.11.3 Nitrogen (N) Driven Nutrient Enrichment Recovery in the U.S.
Most freshwater systems sensitive to nutrient effects of atmospheric deposition of N have
shown no evidence for biological recovery, although decreases in NO;, concentrations
consistent with declines in N deposition have been reported in some regions of the U.S.,
notably the Appalachian, Adirondack, and Rocky Mountains (Appendix 7.1.5). Some
estuaries have shown improvements in biological indicators, such as increases in the
extent of SAV, in response to decreases in N inputs from atmospheric deposition and in
wastewater and agricultural runoff. For an example, see the Tampa Bay case study
(Appendix 16). In other coastal areas of the U.S., biological indicators of nutrient
enrichment have remained relatively unchanged or declined. In the well-studied
Chesapeake Bay watershed where extensive restoration efforts have been implemented,
water quality and measures of ecological condition have shown little improvement during
a 23-year period (Williams et al.. 2010). The one exception to the pattern of no
improvement in water quality was an observed increase in the amount of SAV
(Appendix 10.2.5).
IS.12 Climate Modification of Ecosystem Response to Nitrogen (N)
and Sulfur (S) Deposition
Nitrogen and S deposition occur in many ecosystems concurrently experiencing multiple
stressors, including human-driven climate change. Climate change effects on U.S.
ecosystems were recently summarized in the U.S. National Climate Assessment
(Galloway et al.. 2014; Groffman et al.. 2014). Each appendix of the ISA evaluating N
enrichment or acidification includes a section on how climate modifies the ecosystem
response. In the context of this section of the ISA, climate refers to meteorological factors
over a 5-year horizon (because NAAQS are reviewed every 5 years) in contrast to
long-term climate change, or associated changes to CO2 concentrations. Additionally, to
serve as a foundation for the discussion, text in Appendix 13 is excerpted from Greaver et
al. (2016). a current review of how climate (e.g., temperature and precipitation) modifies
ecosystem response to N that focuses on empirical observations.
IS-103
-------
Anthropogenic emissions of greenhouse gases are likely to cause a global-average
temperature increase of 1.5 to 4.0°C and a significant shift in the amount and distribution
of precipitation by the end of the 21st century (Collins et al.. 2013). Recent work has
focused on the effects of anthropogenic N on the Earth's radiative forcing (Pindcr et al..
2012) and how temperature and precipitation alter ecological responses to N exposure
(Grcavcr et al.. 2016). Most work is conducted on the effects of climate interactions with
N or acidifying deposition (N + S); relatively little work is conducted on how climate
modifies ecosystem response to S nutrient-related effects.
Understanding climate effects on ecosystems is a rapidly expanding field with many new
empirical studies, meta-analyses, and modeling work published since the 2008 ISA.
General patterns of how climate affects some biogeochemical processes are known and
how climate alters growth rates and biodiversity of some species have been identified,
Figure 13-1 is an example of how processes relevant to N enrichment and acidification
may be altered with either wetter or drier conditions. In addition to the excerpt from
Greaver et al. (2016). additional studies are summarized for effects of climate on N
transport and transformation (Table 13-1). N and C cycling (Table 13-2). acidification
(Table 13-3). and biodiversity (Table 13-4). Our understanding of the effects of climate
on ecosystem response to N and S deposition varies; for many ecological endpoints, data
are insufficient to quantify either the direction or magnitude of how climate may alter
ecosystem response with certainty.
IS.13 Ecosystem Services
"Ecosystem services" refers to the concept that ecosystems provide benefits to people,
directly or indirectly (Costanza et al.. 2017). and that ecosystems produce socially
valuable goods and services deserving of protection, restoration, and enhancement (Bovd
and Banzhaf. 2007). The concept of ecosystem services recognizes that human
well-being and survival are not independent of the rest of nature, and that humans are an
integral and interdependent part of the biosphere (Costanza et al.. 2017). In some cases,
and in line with more conventional economic thinking, ecosystem services analysis can
result in attaching monetary values to ecosystem outcomes. However, because ecosystem
services are often public goods their benefits can be difficult to monetize. We emphasize
that this practical difficulty in no way implies that ecosystem service benefits are small or
without value. At a minimum, ecosystem services analysis involves discussion and,
ideally, quantification of ecological outcomes understood by households, communities,
and businesses. Explicitly linking ecosystem services to social and economic welfare
measures has proven difficult because of the broad definition of ecosystem services and
the numerous types of services that could be affected. An analysis of ecosystem services
IS-104
-------
specifically altered by NOx, SOx, and PM would translate the effects of ambient
concentrations and deposition into biological, physical, or monetary metrics that give
insight to public welfare effects.
For acidification, the ecosystem service literature since the 2008 ISA includes studies that
better characterize ecosystem service valuation by pairing biogeochemical modeling and
benefit transfer equations informed by willingness-to-pay surveys, especially for the
Adirondacks and Shenandoah regions (Appendix 14). Aside from valuation studies, there
is an improved understanding of the numerous causal pathways by which N and S
deposition may affect ecosystem services, supported by studies that relate deposition to
final ecosystem services under the FEG-CS (Bell et al.. 2017; Clark et al.. 2017; Irvine et
al.. 2017; O'Dea et al.. 2017; Rhodes et al.. 2017b). However, for many regions and
specific services, poorly characterized dose-response between deposition, ecological
effect, and services are the greatest challenge in developing specific data on the economic
benefits of emission reductions (NAPAP. 2011).
In the 2008 ISA there were no publications that specifically evaluated the effects of N
deposition on ecosystem services associated with N driven eutrophication. Since then
several comprehensive studies have been published on the ecosystem services related to
N pollution in the U.S. (Appendix 14). These include an evaluation of services affected
by multiple N inputs (including N deposition) to the Chesapeake, a synthesis of the
cost-benefits on N loading across the nation, and analysis of the amount of N that leaked
out of its intended application area causing effects on adjacent ecosystems and ecosystem
services, two calculations of the social cost of nitrogen (Minnesota and the Mississippi
Alluvial Valley), and an estimate of the cost to remove N from the White River Basin in
Indiana (this work specifically identified the costs of the atmospheric portion of total N
loading). The estimate of the total number of ecosystem services affected by N is better
quantified by the new studies that use FEG-CS (Bell et al.. 2017; Clark et al.. 2017;
Irvine et al.. 2017; O'Dea et al.. 2017; Rhodes et al.. 2017b). In these analyses, CL
exceedances for N related air pollution were used as a model stressor from which a total
of 1,104 unique chains linking stressor to beneficiary were identified.
The conclusions considering the full body of literature are that (1) there is evidence that
N and S emissions/deposition have a range of effects on U.S. ecosystem services and
their social value; (2) there are some economic studies that demonstrate such effects in
broad terms; however, it remains methodologically difficult to derive economic costs and
benefits associated with specific regulatory decisions/standards; and (3) there is an
improved understanding of the numerous causal pathways by which N and S deposition
ay affect ecosystem services, though most of these causal relationships remain to be
quantified.
IS-105
-------
IS.14 Key Scientific Uncertainties
Evaluation of uncertainty is an important part of ecosystem assessment. Uncertainty
refers to the absence of information and is a way to describe how certain we are in
scientific knowledge. As described by Curry and Webster (2011). the nature of
uncertainty can be expressed by the distinction between ontic uncertainty and epistemic
uncertainty. Ontic uncertainty is associated with inherent variability or randomness and is
an irreducible form of uncertainty. Epistemic uncertainty is associated with imperfections
of knowledge, which may be reduced by further research and empirical investigation.
Walker et al. (2003) [as summarized in Curry and Webster (2011)1 characterized
uncertainty as a progression from deterministic understanding to total ignorance:
"Statistical uncertainty is the aspect of uncertainty that is described in
statistical terms. An example of statistical uncertainty is measurement
uncertainty, which can be due to sampling error or inaccuracy or
imprecision in measurements.
"Scenario uncertainty implies that it is not possible to formulate the
probability of occurrence of one particular outcome. A scenario is a
plausible but unverifiable description of how the system and/or its
driving forces may develop overtime. Scenarios may be regarded as a
range of discrete possibilities with no a priori allocation of likelihood.
"Recognized ignorance refers to fundamental uncertainty in the
mechanisms being studied and a weak scientific basis for developing
scenarios. Reducible ignorance may be resolved by conducting further
research, whereas irreducible ignorance implies that research cannot
improve knowledge."
The understanding and reporting of uncertainty is not consistent across scientific
disciplines, and uncertainty may be quantified by various methods. Csavina et al. (2017)
provided an overview of terminology and definitions of 41 different terms used to
describe uncertainty. Here we provide a summary of some of the key methods that may
be used to evaluate the uncertainty of the relationships between NOx, SOx, and PM
pollutants and ecological effects. This summary presents uncertainties associated with
several specific concepts, including source emissions measurements, atmospheric
deposition estimates, empirical measurements of CLs, models used to estimate CLs, and
uncertainties in the aquatic acidification index. Quantified estimates of uncertainty vary
according to the number of decision points (Section IS. 14.2.3). including the method used
and the input parameters under consideration; therefore, the analyses and discussion of
IS-106
-------
quantified uncertainty values will occur in the Risk and Exposure Assessment as scoped
in the 2017 IRP (U.S. EPA. 2017c).
IS.14.1 Atmospheric Science
Estimating atmospheric deposition involves quantification of emissions, atmospheric
concentrations, and deposition fluxes of the various species that make up atmospheric
SOx, NOy, and NHx. This is accomplished with environmental measurements, model
predictions, or hybrid approaches that combine measurements and modeling methods.
There are a wide range of uncertainties across the environmental measurements and
model parameters used to estimate atmospheric deposition fluxes. The largest
uncertainties are those for dry deposition and ammonia emissions, whether measured or
modeled. The smallest uncertainties are associated with ambient concentration
measurements and continuously monitored stationary emissions like electric power
plants.
IS.14.1.1 Emissions Uncertainty
Quantitative uncertainty estimates are not documented in the National Emissions
Inventory (NEI), but uncertainties are often evaluated through separate efforts by
comparing inventory predictions with measured long-term trends, statistical source
apportionment methods, inverse chemical transport modeling, and comparison with
satellite data (Appendix 2.2.2). SO2 and NOx emission uncertainties for
electricity-generating units, the major source of SO2 and an important source of NOx, are
in the 10-15% range because emissions are usually continuously monitored
(Appendix 2.2.3). NOx emission uncertainties for mobile sources, the largest source of
NOx, arise from differences in engine type, size, age, and maintenance, as well as fuel
composition and emission control equipment. Overestimation of NOx emissions from
mobile sources was proposed as an explanation for modeled NOx concentration bias in
several studies. However, mixed results have been observed across several studies when
modeled concentrations were compared with measurements. Estimates of NOx emissions
uncertainties are in the 10-20% range for on-road gasoline and diesel vehicles, and up to
30% for off-road vehicles like ships, airplanes, and locomotives (Appendix 2.2.3). Spatial
and temporal variability in soil NOx emissions can lead to uncertainty in emissions
estimates. Soil emissions occur mainly during summer and across the U.S., but some
areas, such as the central Corn Belt of the U.S., release more NOx emissions than others
(Appendix 2.2.3).
IS-107
-------
In contrast, total NH3 emissions uncertainties appear to be greater, underestimated by as
much as a factor of two or more according to several recent studies (Appendix 2.2.3). The
predominant sources, livestock operations and fertilizer application, exhibit large
temporal and regional variability due to differences in climate conditions and farming
practices. As a result, detailed models are required for estimating NH3 emissions
(Appendix 2.2.2). but data on local environmental conditions and farming practices
necessary for good model performance are often not available. Large discrepancies
between modeled and measured N concentrations and deposition rates have been
attributed to uncertainties in NH3 emissions (Appendix 2.2.3). Activity rates, including
those for mobile source emissions, are also difficult to quantify, contributing to
uncertainty in NH3 emission estimates (Appendix 2.2.3).
IS.14.1.2 Atmospheric Measurement Uncertainty
Uncertainties in concentration and deposition measurements from network-based
measurements are generally under 20%, and surface concentration uncertainties from
satellite-based measurements typically somewhat higher. Concentration and deposition
data are derived from several specialized national monitoring networks, including the
national SO2 monitoring network, the NCore network for multipollutant concentration
monitoring including NOy, the Ammonia Monitoring Network, CASTNet for estimating
dry deposition, and the National Trends Network for wet deposition (Appendix 2.4.1).
Uncertainties are estimated from reports of precision in data quality reports where
available, and otherwise from network data quality objectives.
For air concentration measurements used to estimate dry deposition, CASTNet measured
precision was 2-5% for SO42 , 5-13% forN03~, and 2-6% for NH3 in 2016
(Appendix 2.4.5). Additional uncertainty is associated with estimating dry deposition
from NTN concentration data. Uncertainties of 30% for SO2 and 40% for HN03 have
been reported using a simple inferential approach (Clarke et al.. 1997). However, single
site determinations are of limited use because dry deposition fluxes are determined by
several factors and can vary considerably over small spatial scales. In most recent efforts,
dry and total deposition on a regional or national scale is usually modeled with CTMs
(Section IS. 14.1.3).
Precipitation concentration measurement precision and estimated wet deposition
precision in the National Trends Network were less than 7% for SO42 and N03 and less
than 20% for NH3. PRISM (Parameter-elevation Regression on Independent Slopes
Model) enhances spatial resolution using National Trends Network data to improve the
creation of wet deposition maps (Appendix 2.6). Uncertainty for PRISM data sets has
IS-108
-------
been evaluated in the literature using cross validation and a 70% prediction interval for
different data sets. NH3 air concentration measurement methods used in AMoN were
evaluated and found to have a precision of 10% (Appendix 2.4.3). Minimum performance
specifications for SO2 monitoring from the national SO2 monitoring network include a
precision of 2.0% (Appendix 2.4.4). Data quality objectives for NOy in the NCore
network include a precision of 15% (Appendix 2.4.2). Uncertainty in satellite-based
measurements depend on vertical profile, cloud fraction, cloud-top pressure, surface
reflectivity, and extent of aerosol scattering. Estimates of 20% forNC>2 (Appendix 2.4.2)
and 10-45% for SO2 (Appendix 2.4.4) have been reported for cloud-free conditions.
IS.14.1.3 Atmospheric Modeling Uncertainty
The Community Multiscale Air Quality modeling system is probably the most widely
used model in the U.S. for estimating atmospheric deposition. CMAQ accurately
modeled total SOx, but partitioning resulted in overpredicting SO2 and underpredicting
SO42 . In a recent CMAQ evaluation, SO2 concentrations were overestimated by 39 to
47%, and SO42 concentrations were underestimated by 9 to 17%, as annual averages
over a range of 4 years compared to surface-based measurements. In addition,
atmospheric NO;, concentrations were overestimated by 22 to 26%, as annual averages
over a range of 4 years compared to surface-based measurements (Appendix 2.5.3).
Mixed results have been observed in several recent comparisons of CMAQ wet
deposition estimates to network-based measurements, with average differences in
modeled results and measurements ranging from <15 to 99% for NO, . SO42 . and <15 to
60% for NH3 (Appendix 2.5.3). Modeling methods for estimating dry and total deposition
are still under development, and uncertainties have not been extensively evaluated or
quantified. Recent sensitivity analysis results found less than 5% differences in total
deposition estimates because of compensation of competing model processes, but
extensive comparison of model results and measurements are not available
(Appendix 2.5.3).
Horn et al. (2018) used deposition and forest inventory data (from 2000 to 2016) to assess
the relationship between deposition and growth and survival of 71 tree species across the
contiguous U.S. in a correlational analysis. Authors attempted to reduce uncertainty by
accounting for other variables, either directly in their model or by quantifying and
avoiding instances with high collinearity. The authors isolated the effects of N deposition
from S deposition by adding S deposition explicitly into their models. Using variance
inflation factors (VIFs), they also quantified the collinearity of N and S deposition against
a suite of environmental variables that might have an effect. The analysis focused on the
relationships of tree growth and survival to N and S deposition where the VIF was less
IS-109
-------
than or equal to 3. VIF values of 3-10 have been proposed as thresholds above which
there is a potential for high collinearity (Horn et al.. 2018). To what extent variables not
included could have varied with deposition and had an effect, including ozone and
drought, remained a key uncertainty.
Clark et al. (2018) analyzed exceedances of multiple types of CLs for the contiguous U.S.
since 1800 and projecting out to 2025. The study authors discussed the uncertainty
around CMAQ deposition estimates using CMAQ estimates starting in 1980. They noted
that CMAQ may underestimate hot spots of deposition in space (e.g., concentrated
deposition because of an orographic effect) or in time (e.g., from cloudbursts). CLs are an
ecosystem response to deposition, and so any errors associated with deposition estimates
would propagate through CLs. Fenn et al. (2010) found that CMAQ estimates and N in
throughfall were similar under low throughfall conditions, but CMAQ underestimated N
deposition when throughfall was high. Clark et al. (2018) noted that CMAQ is corrected
using NADP data, but NADP sites do not provide complete spatial coverage. Remote
sites are likely underrepresented.
In addition to measurable uncertainties associated with measurement precision or
comparisons between models and measurements, there are also structural uncertainties
due to incomplete understanding of the underlying science related to atmospheric
deposition that are not possible to quantify. The main structural uncertainties associated
with deposition estimates are canopy effects on NOx (including both bidirectional gas
exchange and canopy reactions), bidirectional exchange of NIL with biota and soils, and
processes determining transference ratios that relate average concentration to deposition
(Appendix 2.5).
IS. 14.2 Ecological Effects
Evaluation of ecological effects caused by acidification or eutrophication involves a suite
of parameters and dose-response functions, both empirical and modeled. The quantitative
uncertainty of empirically observed variables in ecology is determined by using statistics.
A suite of mathematical statistical models is available to describe the variability among
empirical observations and the strength of a cause and ecological effect relationship, the
appropriate method to apply depends on the experimental design. Statistics for empirical
data include calculation of probability, distributions, standard deviation, variance, /-tests.
ANOVA, linear regression, spatial statistics, Bayesian analysis, and multivariate analysis,
among others. In general, ecological endpoints determined by empirical studies to be
affected by deposition were reported in the ISA if they were statistically significant; this
means the magnitude of effect was larger than the estimated uncertainty.
IS-110
-------
Models of chemical and ecological processes, including biogeochemistry, provide
representations of biological and geochemical interactions through mathematical
expressions. The models used to characterize aquatic and terrestrial biogeochemistry
response to N and S deposition can be complex, including many interacting variables.
Model results are often compared to empirically collected data to confirm the model.
Each of the input variables used in a biogeochemical model entails uncertainty. Model
uncertainty is governed, in part, by how close the model predictions are to actual
observations. Uncertainty in modeled results may arise from limitations in input data or
from limitations in model assumptions. Statistical inference methodologies enable
uncertainty analysis and determine the strength of the relation between a given uncertain
input and the output (i.e., sensitivity analysis). For biogeochemistry models these
methods include first-order sensitivity index, Monte Carlo technique, extended Fourier
amplitude sensitivity test, Morris one-factor-at-a-time, and Bayesian analysis.
IS. 14.2.1 Empirical Critical Loads
Empirical N CLs for terrestrial and aquatic ecosystems reported in this ISA have been
estimated using empirical data sets. The exact effects threshold may be determined using
expert judgement. For example, if three levels of N addition are applied to a study site
(10, 20, and 30 kg N/ha/yr) and an effect is noticed at 20 kg N/ha/yr, then the CL is
estimated at <20 kg N/ha/yr. Another approach would be to fit a mathematical function to
the observations, and a scientific judgement made to identify the level of deposition
and/or N addition, or threshold, at which the ecological effect is considered to occur and
which is likely to be biologically adverse.
There are some challenges associated with developing CLs that can result in uncertainty.
First, because biological responses are often continuous, there can be a lack of an obvious
cutoff between adverse and nonadverse effects. As a result, individual author groups have
selected different response thresholds. For example, N CLs for lichens have been
calculated for (1) deposition values associated with thallus N concentrations above the
97% distribution quantile observed for clean sites (Fenn et al.. 2008). (2) community
composition shifts from oligotroph to eutroph dominance (Fenn et al.. 2008). (3) low
probability of detecting regionally distributed sensitive species (Root et al.. 2015; Geiser
et al.. 2010). or (4) extirpation of oligotrophs (Fenn et al.. 2008). Secondly, clean site data
can be lacking in some ecoregions. For instance, few empirical data are available for sites
in the eastern U.S. with deposition rates <4 kg N/ha/yr. This makes it difficult to quantify
physiological or community compositional conditions that may have occurred in this
region at deposition rates of 1-4 kg N/ha/yr.
IS-111
-------
The Pardo et al. (2011a) study provided a compilation of terrestrial and aquatic N CLs
reported since the 2008 ISA. Uncertainty in the derivation of empirical CLs for N input
as presented by Pardo et al. (2011a) arises in estimating the ambient (and perhaps
historical) deposition loads and in estimating the biological effects caused by those
deposition levels. According to Pardo et al. (2011a). sources of uncertainty in N
deposition estimates for N CLs at the Ecoregion Level 1 scale include ""(1) the difficulty
of quantifying dry deposition of nitrogenous gases and particles to complex surfaces;
(2) sparse data, particularly for arid, highly heterogeneous terrain (e.g., mountains); and
(3) sites with high snowfall or high cloud water/fog deposition, where N deposition tends
to be underestimated." Examples of high uncertainty include high-elevation sites in the
Rockies and Sierra Nevada mountains, due in part to highly uncertain estimates of dry
deposition (Appendix 2). For sensitive receptors such as phytoplankton, shifts in
high-altitude lakes, N deposition model bias may be close to, or exceed, predicted CL
values (Williams et al.. 2017a).
Physical, chemical, and ecological variability across lakes affect their response to N
deposition and contribute to uncertainty of CL estimates (Appendix 9.1.1.2). A review by
Bowman et al. (2014) noted that current N CLs for sensitive alpine systems may not be
protective under future climate scenarios of warmer summer temperatures and a shorter
duration of snow cover.
Between the publication of Pardo et al. (2011a) and the cutoff date for literature in this
ISA (May 2017), some additional aquatic and terrestrial N CLs have been published
(Appendix 4; Appendix 6.5). Simkin et al. (2016) was not based on field addition or N
gradient of deposition studies; instead, the methods were a spatial analysis of plant
diversity using a large data set of over 15,000 forest, shrubland, and herbaceous sites
across the U.S. Atmospheric N deposition varied nearly 20-fold across the site gradient.
The study authors found that N deposition was negatively correlated with plant species
richness at many locations, but positively correlated at others with most of the positive
correlations in areas with low N deposition averaging 3 kg N ha/yr or less. Simkin et al.
(2016) also estimated the uncertainty surrounding the mean CL estimates. For open
canopy ecosystems, for example, they estimated a mean of 8.7 kg N ha/yr and provided
95% confidence intervals, which can be used as estimates of uncertainty, of 6.4 to
11.3 kg N ha/yr. For closed canopy systems, the mean of 13.4 kg N ha/yr was surrounded
by a 95% confidence interval of 6.8 to 22.2 kg N ha/yr.
Clark et al. (2018) noted that many of the CLs used are empirically derived. Some of the
uncertainties with these CLs are that they are often from one or two studies at a given
location or area and extrapolated to a larger area, such as an entire Level 1 ecoregion.
Thus, there is uncertainty about how representative these are for larger areas. As noted in
IS-112
-------
Horn et al. (2018). there also can be covariates unaccounted for that could affect
estimates of CLs. CLs also do not generally account for historical effects that already
might have affected the ecosystem. There are also uncertainties regarding process-based
CLs, such as the terrestrial acidification CL. Clark et al. (2018) specifically pointed
towards the existence of poor estimates of soil weathering despite the importance of soil
weathering estimates for acidification CLs.
The majority of studies that evaluate terrestrial N CLs for N enrichment effects are based
on observed response of a biological receptor to N deposition (or N addition as a proxy
for deposition), without a known soil chemistry threshold that causes the biological
effect. In contrast, CLs for acidification are typically based on the deposition amount that
gives rise to a soil chemical indicator value which is known to cause an adverse
biological effect. The link between soil chemical indicator and biological effect is based
on empirical evidence (Appendix 5). The relationship between deposition and the
biogeochemistry that causes effects on soil chemistry is typically modeled (Appendix 4;
Section IS. 14.2).
IS.14.2.2 Modeled Critical Loads
IS.14.2.2.1 Terrestrial and Aquatic Acidification: Biogeochemistry
A variety of process models have been used to estimate past and future resource
conditions under scenarios of acidification/recovery responses and critical and target
loads, both aquatic and terrestrial. Models include simple approaches such as the simple
mass-balance equation (SMBE), and dynamic models, such as PnET-BGC and ForSAFE,
MAGIC, VSD, and VSD+ (Appendix 4.5). CLs for terrestrial and aquatic acidification
are calculated by the model to determine the amount of deposition that alters soil or water
chemistry to a threshold value known to have detrimental effects on a biological receptor.
Each of the several well-established models of terrestrial biogeochemistry used to
evaluate soil acidification (Appendix 4.5) rely heavily on input or simulated values for
base cation weathering (BCw) rate, one of the most influential yet difficult to estimate
parameters in the calculation of critical acid loads of N and S deposition for protection
against terrestrial acidification (Appendix 4.5.1.1). Obtaining accurate estimates of
weathering rates is difficult because weathering is a process that occurs over very long
periods of time, and the estimates on an ecosystem's ability to buffer acid deposition rely
on accurate estimates of weathering. Various approaches can be used to estimate BCw,
including the empirical soil clay approach, the PROFILE model [e.g., Phelan et al.
(2014)1. the F-factor approach (U.S. EPA. 2009c). and calibration of a dynamic model
IS-113
-------
such as MAGIC [e.g., Povak et al. (2014); McDonnell et al. (2014b)l. There are new
studies on estimating BCw, including evaluation of uncertainty (Whitfield et al.. 2018;
Futter et al.. 2012). When applying PROFILE to upland forests in the U.S., Whitfield et
al. (2018) found the greatest uncertainty in BCw estimate was due to the particle size
class-based method used to estimate the total specific surface area on which weathering
reactions can take place.
The uncertainty of forest soil CLs for acidification in U.S. calculated using simple
mass-balance equations (SMBE) was investigated by Li and McNultv (2007). The results
included a quantification of how 17 of the model's parameters contributed to the
uncertainty and indicated that uncertainty in the CLs came primarily from components of
base cation weathering and acid-neutralizing capacity, whereas the most critical
parameters were BCw base rate, soil depth, and soil temperature. The study authors
concluded that improvements in estimates of these factors are crucial to reducing
uncertainty and successfully scaling up SMBE for national assessments (see
Appendix 4.6).
Several dynamic models are commonly used to model terrestrial soil acidification
(Appendix 4.5). Tominaga et al. (2009) conducted a Monte Carlo multiple-model
evaluation of the dynamic models MAGIC, SAFE, and VSD and found that given the
same deposition scenario, the three models (without calibration) simulate changes in soil
and soil solution chemistry differently, but the basic patterns were similar. The study
authors also found the greatest differences in model outputs were attributed to the cation
exchange submodel. Bonten et al. (2015) compared how well the common types of
dynamic models used to evaluate terrestrial soils (VSD, MAGIC, ForSAFE, and
SMARTml) quantified several variables including soil S, soil pH, soil ANC, BC, base
saturation, and Al (Appendix 4.5.3).
Uncertainty analysis of a dynamic model (VSD) used for CL based on soil chemistry
chemical limits showed that the main drivers of uncertainty were largely dependent on
the chemical criterion selected rAppendix 5.5.3.3; Reinds and de Vries (2010)1. For
example, base cation weathering, deposition, and the parameters describing the H-Al
equilibrium in the soil solution were the main sources of uncertainty in the estimates of
maximum CLs for S (Clmax[S]) based on the Al:Bc criterion of 1.0, and uncertainty in
Clmax(S) based on ANC was completely determined by base cation inputs. The
denitrification fraction was the most important source of uncertainty for the maximum
CLs of N (Clmax[N]). Calibration of VSD reduced the levels of uncertainty for all CLs
and criteria.
Fakhraei et al. (2017b) reviewed sensitivity and uncertainty analysis techniques
(e.g., first-order sensitivity index, Monte Carlo technique, extended Fourier amplitude
IS-114
-------
sensitivity test, Morris one-factor-at-a-time, and Bayesian analysis) in the context of a
biogeochemistry model. The study authors applied these techniques to determine the
uncertainty and sensitivity of the PnET-BGC model calculation of TMDLs of acidifying
deposition that occur in high-elevation, acid-impaired streams in GSMNP (Fakhraci et
al.. 2017a). Sensitivity analyses showed that modeled estimates of maximum allowable
acidifying deposition loads were most sensitive to uncertainty in model input parameters
of air temperature, precipitation quantity, and rate of calcium weathering. Importantly, as
more uncertainty was incorporated into model input parameters (±5 to ±10 to ±20%
uncertainty), estimates of allowable deposition loads to protect aquatic ecosystem
recovery decreased in magnitude (Fakhraei et al.. 2017a).
15.14.2.2.2 Biogeochemistry and Plant Biodiversity Linked Modeling
Plant biodiversity models, such as VEG and PROPS, have been coupled to dynamic
biogeochemical models, such as ForSAFE and VSD± (Mcdonncll et al.. 2018b;
Mcdonnell et al.. 2018a; Phelan et al.. 2016). ForSAFE-VEG is an older and more
broadly applied model than VSD + PROPS. There are some key differences between
VEG and PROPS. Plant species in the VEG component of ForSAFE-VEG are defined by
mathematical equations based on expert opinion regarding such parameters as plant needs
for moisture, sunlight, and N supply to represent unobservable fundamental niches. In the
PROPS, statistical relationships based on empirical data are used to characterize plant
species, which are more likely to approximate real-world niches influenced by
competition among species. These model chains are subject to the same constraints and
uncertainties as the biogeochemical models on their own, plus those of the plant response
modules.
15.14.2.2.3 Aquatic Eutrophication Modeling
Many of the models that estimate N loads to the coastal zone from land-based inputs
(agricultural practices, sewage, atmospheric deposition, natural lands) and freshwater
inflow have been compared, and there is a good deal of knowledge about their limitations
and uncertainties (McCrackin et al.. 2013; Alexander et al.. 2008). A National Research
Council review determined that these models are hydrodynamically complex and tend to
be site specific. Thus, they are difficult to apply broadly (NRC. 2000).
The SPARROW model application used only wet N deposition. A large amount of N
from nonpoint source urban influences (most likely due primarily to the dry deposition of
exhaust N gases) often approximately doubles the importance of N deposition as an N
source to higher order river systems (Howarth. 2008a. b).
IS-115
-------
IS. 14.2.3 Additional Key Considerations for Critical Loads
The choice of model for CL estimation, or for scenario projection, depends largely on the
availability of time, data, and resources. Major decisions inherent in the modeling efforts
include:
• Empirical observation or application of a model
• Steady-state or dynamic model
• Statistical or process-based model
• Protection against acidification or nutrient N enrichment
• Site-specific, regional, or national spatial scale
• Resources to be protected (i.e., stream, lake, soil, vegetation, aquatic biota)
• Chemical indicator(s) of adverse effects (e.g., water ANC, water NO3 , soil BS)
• Critical level(s) for selected indicator(s)
• Time frame of evaluation (i.e., ambient, 2050, long-term steady state)
Each of these decision points introduces additional uncertainties, data needs, and
potential assessment errors. U.S. EPA (2008a) summarized CL research and monitoring
needs identified by U.S. EPA (2006b) at the time of the previous (2009) U.S. EPA Risk
and Exposure Assessment.
IS.14.3 Aquatic Acidification Index
Detailed analysis of uncertainty in the AAI equation can be found in Appendix F of the
2011 Policy Assessment for the Review of the Secondary National Ambient Air Quality
Standards for Oxides of Nitrogen and Oxides of Sulfur (U.S. EPA. 2011a). The AAI is
made up of components including ecosystem effects; dose-response relationships;
underlying ecosystem sensitivity to acid deposition, biogeochemical, atmospheric and
deposition processes; and characterization of ecosystem services. Some degree of
uncertainty exists in all of the components of the AAI. Overall, the 2011 Policy
Assessment found, on balance, low uncertainty in the information and processes
associated with linkages from ecological effects to atmospheric conditions through
deposition and ecosystem modeling. However, it acknowledged the need to improve
certainty of several components including nitrogen and sulfur deposition processes in
CMAQ, natural emissions of NOx from lightning processes, and improving the amount
of samples of CL estimates at several ecoregions (U.S. EPA. 2011a).
IS-116
-------
APPENDIX 1 QUALITY ASSURANCE AND
INTRODUCTION TO APPENDICES
1.1 Quality Assurance Summary
The use of quality assurance (QA) and peer review helps ensure that the U.S. EPA
conducts high-quality science assessments that can be used to help policymakers,
industry, and the public make informed decisions. Quality assurance activities performed
by the U.S. EPA ensure that environmental data are of sufficient quantity and quality to
support the Agency's intended use. The ISA for Oxides of Nitrogen, Oxides of Sulfur
and Particulate Matter-Ecological Criteria is classified as a Highly Influential Scientific
Assessment (HISA), which is defined by the Office of Management and Budget (OMB)
as a scientific assessment that is novel, controversial, or precedent-setting, or has
significant interagency interest (OMB, 2004). OMB requires a HISA to be peer reviewed
before dissemination. To meet this requirement, the U.S. EPA engages the Clean Air
Scientific Advisory Committee (CASAC) as an independent federal advisory committee
to conduct peer reviews. Both peer-review comments provided by the CASAC panel and
public comments submitted to the panel during its deliberations about the external review
draft were considered in the development of this ISA.
Agency-wide, the U.S. EPA Quality System provides the framework for planning,
implementing, documenting, and assessing work performed by the Agency, and for
carrying out required quality assurance and quality control (QA/QC) activities.
Additionally, the Quality System covers the implementation of the U.S. EPA Information
Quality Guidelines (U.S. EPA, 2002). This ISA follows all Agency guidelines to ensure a
high-quality document.
Within the U.S. EPA, Quality Assurance Project Plans (QAPPs) are developed to ensure
that all Agency materials meet a high standard for quality. U.S. EPA has developed a
Program-level QAPP (PQAPP) for the ISA Program to describe the technical approach
and associated QA/QC procedures associated with the ISA Program (PQAPP ID# L-
HEEAD-0030253-QP-1-0). All QA objectives and measurement criteria detailed in the
PQAPP have been employed in developing this ISA. Quality assurance checks were
conducted on numerical entries used in the appendices, and at a minimum, the numbers
obtained from every tenth reference cited in the appendices were verified against the
original source by an independent scientist for accuracy. Furthermore, publicly available
databases (e.g., National Emissions Inventory, Air Quality System database) from which
data was used in analyses were verified to have their own QA processes in place. U.S.
EPA QA staff are responsible for the review and approval of all quality-related
1-1
-------
documentation. Because this is a Highly Influential Scientific Assessment, U.S. EPA QA
staff performed a Technical System Audit on the ISA in June 2018 and August 2019.
These audits verified that the appropriate QA/QC procedures and reviews were
adequately performed and documented.
1.2 Introduction to the Appendices
The appendices and Integrated Synthesis serve different purposes in this ISA. The
Integrated Synthesis is meant to summarize the key messages derived from assessment of
the policy relevant science of NOy, SOx, and PM in this review of the secondary National
Ambient Air Quality Standards. It provides a general introduction to the purpose,
process, development, and organization of the ISA as well as highlights connections,
concepts, and changes based on new evidence and causality. In addition, the Integrated
Synthesis provides a discussion of uncertainty and a synthesis of information on the
recovery of ecosystems from N and S deposition.
While the purpose of the Integrated Synthesis is to synthesize, integrate, and provide key
messages, the purpose of the appendices is to provide a more detailed description of the
state of science for specific topic areas. Appendix 1 is an introduction to the purpose and
organization of Appendix 2-Appendix 16. Appendix 2 characterizes the sources,
atmospheric processes, and the trends in ambient concentrations and deposition of NOy,
SOx, and PM. Appendix 3 describes the direct effects of NOy and SOx gases on plants
and lichens. Appendix 4-Appendix 6 describe the effects of N and S deposition on
biogeochemistry and the biological effects of acidification and N enrichment in terrestrial
environments. Appendix 7 describes the effects of N and S deposition on aquatic
biogeochemistry. Appendix 8-Appendix 10 characterize the biological effects of
freshwater acidification, freshwater N enrichment, and N enrichment in estuaries and
near-coastal systems. Appendix 11 describes the effects of N deposition on wetlands, and
Appendix 12 characterizes the ecological effects of S as a nutrient. Appendix 13 presents
information on climate modification of ecosystem response to N and S, while
Appendix 14 discusses ecosystem services. Appendix 15 is a review of the ecological
effects of forms of PM, not related to N or S deposition. Finally, Appendix 16 presents
case studies for six locations in the U.S. (southern California, northeastern U.S., Rocky
Mountain National Park, southeastern Appalachia, Tampa Bay, and the Adirondacks)
where data are sufficient to well characterize the ecological effects of N and S deposition.
These sites would therefore make good candidates to assess risk and exposure by
exploring linkages across various effects and ecosystems-types in a specific location.
1-2
-------
APPENDIX 2 SOURCE TO DEPOSITION
2.1 Introduction
In this appendix, emphasis is placed on those species subject to atmospheric processes
relevant for review of the air quality criteria and associated welfare-based secondary
National Ambient Air Quality Standards (NAAQS) for oxides of nitrogen, oxides of
sulfur, and/or particulate matter (PM). As such, this appendix largely focuses on
examining the fundamental and applied science of atmospheric processes relevant to
assessing environmental exposures and effects associated with atmospheric deposition of
nitrogen (N) and sulfur (S) species, including those present in PM.1 Together this
information serves as a prologue to the detailed descriptions of the evidence of ecological
effects from oxidized and reduced N and oxidized S, including those present in PM, that
follow in later appendices.
Recent advances in research on N and S emissions sources, atmospheric transformation
and transport, measurement and modeling techniques, atmospheric loadings, and
deposition processes relevant to this review of the NAAQS are evaluated in this
appendix. N and S species of interest are generally classified into three groups: oxidized
nitrogen, reduced nitrogen, and oxidized sulfur. While NO2 and SO2 are the most well
known as air pollutants, research on the entire range of oxides of nitrogen and sulfur is
considered for review of the air quality criteria. Reduced nitrogen is also discussed
because it strongly influences the atmospheric deposition of NOy and SOx as well as the
chemistry of PM formation. Oxidized nitrogen, reduced nitrogen, and oxidized sulfur all
have particulate forms (NO, . NH44", SO42 ). which together account for a large fraction
of PM mass (U.S. EPA. 2019). as well as gas-phase components that act as major
precursors to PM. Thus, a consideration of the combined effects of oxides of nitrogen,
oxides of sulfur, and PM requires an understanding of the atmospheric processes
involving oxidized nitrogen, reduced nitrogen, and oxidized sulfur.
Oxidized nitrogen species considered here range from nitric oxide (NO) and nitrogen
dioxide (NO2), collectively referred to as NOx, to higher order organic and inorganic
oxidation products, collectively referred to as NOz (e.g., pNOs, HNO3, HONO, PAN,
other organic nitrates). NOz is especially relevant when considering nutrient addition to
ecosystems and the acidification of surface waters. NOx and NOz together are referred to
1 Since ecological effects of PM are governed mainly by PM composition, the most relevant PM species (nitrate and
sulfate) are also species that are derived from sulfur and nitrogen oxide precursors, and there is a high degree of
overlap in the discussion of the impacts of NOy and SOx, and the impacts of PM2.5. The PM ISA (U.S. EPA. 2009a)
provides more extensive information about the atmospheric processes for total PM mass.
2-1
-------
as NOy (i.e., NOx + NOz = NOy). Nitrous oxide (N2O) is an oxide of nitrogen, but it is
not included as a component of NOy. N2O contributes to stratospheric ozone depletion
and climate forcing [AR5; IPCC (2013)1. but it contributes little to N deposition, and is
not included in this appendix.
Reduced nitrogen species are NH3 and NH4 as well as reduced organic nitrogen
compounds. NH3 and NH44" together are referred to as NHX (i.e., NH3 + NH4+ = NHX).
Reduced nitrogen contributes to acidification and N enrichment, and it also has a key role
in neutralizing acidity in cloud, fog, and rainwater as well as aqueous aerosol particles
formed from atmospheric oxidation of SO2 and NOx. Additionally, NH3 is a precursor for
atmospheric particulate matter, reacting with gas-phase nitric acid (HNO3) to form
ammonium nitrate (NH4NO3), a major component of N deposition in many areas of the
U.S. For this assessment, NOy and NHx are grouped together as total reactive nitrogen,
Nr (i.e., NOy + NHx = Nr). Nr does not include nitrous oxide and reduced organic
nitrogen compounds. However, to the extent it is available, information on the sources,
abundances, and fate of reduced organic nitrogen is included in the sections that follow.
Gaseous oxides of sulfur (SOx) is defined to include sulfur monoxide (SO), sulfur
dioxide (SO2), sulfur trioxide (SO3), and disulfur monoxide (S2O). Of these only SO2 is
present in the lower troposphere at concentrations relevant for environmental
considerations. However, SO2 interacts with particles and cloud drops and is oxidized to
sulfate. SO2 and sulfate (SO42 ) account for much of the acidification of surface water in
the U.S. and together these make up total oxides of sulfur discussed in this appendix
(SOx).
Particulate matter (PM) impacts discussed in this document are also mainly focused on N
and S containing species, which together usually make up a large fraction of the fine PM
mass in many areas of the U.S. and have better understood and potentially greater
ecological impacts than other PM components. PM is usually classified into two size
fractions which differ in their physical and chemical characteristics, atmospheric
behavior, and health and environmental effects. These are PM2.5, particles smaller than
2.5 (.un in diameter, and PM10-2.5, particles with diameter between 2.5 and 10 |_im.
Ecological impacts of PM depend largely on its composition (U.S. EPA. 2009a. 2004).
Together PM2 5 and PM10-2.5 make up PMi0. PM contains numerous individual
components representing a wide range of chemical and physical properties. However, in
most areas of the U.S. PM2.5 mass is composed mainly of sulfate, nitrate, and organic
materials. In contrast, PM10-2.5 is composed of crustal material similar in composition to
soil in the area where the PM10-2.5 is found, as well as sea salt in coastal areas. There is
little discussion of PM10-2.5 effects in this document because in most rural and remote
areas PM10-2.5 is largely due to natural sources like soil and sea salt.
2-2
-------
There are several other reasons for focusing on the sulfate and nitrate fraction of PM2 5
mass in this document in addition to the observation that they often account for most
PM2.5 mass. Together they also usually account for an even greater fraction of PM2 5 in
rural and remote areas that make up most of the U.S. land mass over which effects in this
document are relevant. While organic matter can also account for a large fraction of
PM2.5, it is composed of a wide variety of individual compounds that cannot be identified
at a molecular level, making it difficult to assess ecological impacts. Also, there is little
information on organic PM impacts, except for individual compounds that make minor
contributions to mass. As a result, the main contributors to PM2 5 mass for which
ecological impacts can be readily assessed are limited to sulfate and nitrate, which are
also components of total oxides of sulfur and oxides of nitrogen, respectively. Since
ecological effects of PM are governed mainly by PM composition, the most relevant PM
species (nitrate and sulfate) are also species that are derived from sulfur and nitrogen
oxide precursors, and there is a high degree of overlap in the discussion of the impacts of
NOy and SOx, and the impacts of PM2.5.
In addition to nitrogen (N) and sulfur (S) and their transformation products, other PM
components such as trace metals and organics are deposited to ecosystems and may
subsequently impact biota. Evidence for effects of PM on ecological receptors include
direct effects of airborne PM on radiative flux and both direct and indirect effects of
deposited particles. Direct effects include alteration of leaf processes from deposition of
PM ("dust") to vegetative surfaces (U.S. EPA. 2009a). Indirect effects encompass
physiological responses associated with uptake of PM components and alterations to
ecosystem structure and function (see Appendix 15).
Much of the discussion in sections dealing with chemistry, measurement, and deposition
(both wet and dry) focus on sulfuric acid (H2SO4) and HNO3, which have been long
established as the major species contributing to acid rain. Other N and S species that
either hydrolyze to form acids are also included, along with organic acids, to the extent
that they contribute substantively to acidification of terrestrial (see Appendix 5) and
aquatic environments (see Appendix 8) and/or N enrichment (see Appendix 6.
Appendix 7. Appendix 10. and Appendix 11).
Major sources of the precursors (NOx, SO2) to the formation of HNO3 and H2SO4 include
on- and off-highway vehicles and electricity-generating units (EGUs). SO2 is oxidized to
H2SO4 either in the gas phase or cloud water by several well-known mechanisms. NO2 is
oxidized to HNO3, which can either deposit as HNO3 or interact with NH3 to form
particulate NH4NO3. NH4NO3 can exhibit semivolatile behavior that can substantially
alter the distance over which NH3 and HNO3 can travel. Reduced organic nitrogen
species, which could have large agricultural sources, can constitute a substantial fraction
2-3
-------
of N delivered to ecosystems by precipitation. In areas where the rainwater pH is greater
than about 4.5, pH is not exclusively controlled by N and S species, as organic acids such
as formic, acetic, and oxalic acids can be major contributors to the acidity of rainwater.
A wide variety of N containing compounds, consisting of oxidized and reduced organic
and inorganic species contribute to wet and dry deposition. In general, deposition of
reduced (organic + inorganic) N exceeds that of oxidized N across the contiguous U.S.
(CONUS). Nationwide, deposition of N species occurs mainly by dry deposition of
HNO3 and NH3. The pattern is more complex for S in that for large areas, mainly in the
central U.S., wet deposition tends to dominate over dry deposition. However, in some
regions mostly in the west, dry deposition of mainly SO2 makes a greater contribution
than wet deposition.
Precipitation chemistry has been monitored at a large number of sites across the U.S. for
several decades as part of the National Atmospheric Deposition Program (NADP).
Concentrations of inorganic gas and particulate phase N and S species have been
monitored across the U.S. since 1990 by the Clean Air Status and Trends Network
(CASTNET). These concentrations are used in atmospheric models to infer dry
deposition (i.e., the transfer of gaseous and particulate pollutants from the atmosphere to
the surface by impaction through turbulent motions and gravitational settling). Starting in
2007, monitoring of NH3 was initiated at a subset of CASTNET sites. Cloud deposition,
which can account for the bulk of deposition at high elevations in mountainous areas, is
monitored at two locations on a regular basis, but has been the subject of shorter term
field studies in various locations in the U.S.
Although the pH of rainwater has increased noticeably across the U.S., coincident with
notable decreases in the wet deposition of nitrate and sulfate since 1990, there are still
widespread areas, mainly in the eastern U.S., affected by acid precipitation. Deposition of
total nitrogen has not reflected the continuing decrease in NOx emissions, largely because
deposition ofNH3 has increased. Large areas, at least one-third of the contiguous U.S.
(CONUS), are estimated to receive at least 10 kg/ha/yr wet + dry deposition of reactive
nitrogen species. This estimate is likely too low because reduced organic nitrogen species
are not measured by the routine monitoring networks or considered in air quality models
such as U.S. EPA's Community Multiscale Air Quality (CMAQ) modeling system.
Three of the four major contributors to inorganic N deposition are included in the
definitions of either oxides of nitrogen or particulate matter: HNO3 is an oxide of
nitrogen, NH44" is a PM component, particulate NO3 is both a PM component and an
oxide of nitrogen. The fourth major contributor, NH3, is neither an oxide of nitrogen nor a
component of particulate matter. In a recent comparison, the contribution of NH3 to total
inorganic N deposition ranged from 19% in locations in the Northwest U.S. to 63% in
2-4
-------
locations in the Southwest U.S., and was generally higher in summer than in winter (Li et
al.. 2016d). In general, the majority of inorganic N deposition was accounted for by
oxides of nitrogen and particulate matter in the Northwest, Northeast, Southeast, and
Rocky Mountains, but the contribution of oxides of nitrogen and particulate matter was
roughly equivalent to contributions from NH3 in the Upper Midwest, Florida, and smaller
in the Southwest (Li et al.. 2016d). However, because NH3 is a PM precursor, it can also
be definitively stated that inorganic N deposition is entirely accounted for by oxides of
nitrogen, PM components, and PM precursors.
Since the 2008 Integrated Science Assessment for Oxides of Nitrogen and
Sulfur—Ecological Criteria [hereafter referred to as the 2008 ISA; U.S. EPA (2008a)l.
there have been a number of new developments. These apply to methods, such as
data-model fusion to integrate information for deposition across the U.S.; the use of
chemistry-transport models linking deposition to ambient air quality; the expansion of
CASTNET monitoring to include NH3 and NOy at selected sites and intercomparisons of
monitoring methods with research grade instruments; and advances being made in
satellite-based measurements in conjunction with chemistry-transport model simulations
of tropospheric NO2, SO2, and NH3 that will allow mapping of dry deposition over
remote areas with spatial resolution of -10 km / 10 km. These new developments are
described in the following sections of this appendix. Appendix 2.2 considers sources and
emissions of N, S, and PM to the atmosphere. Appendix 2.3 summarizes the atmospheric
chemical transformations of N and S compounds and formation of PM that occur during
transport from sources to deposition to the surface. Appendix 2.4 describes measurement
of relevant atmospheric species, including national monitoring networks and methods.
Appendix 2.5 discusses the use of chemical transport models to estimate deposition.
Appendix 2.6 shows the geographic distributions of atmospheric concentrations and
deposition of N, S, and PM.
2.2 Sources of Nitrogen and Sulfur Compounds and Particulate
Matter to the Atmosphere
This section describes advances in our understanding of NOx, NH3, and SO2 emissions.
Appendix 2.2.1 describes annual national emissions of each species based mainly on the
2014 National Emissions Inventory (NEI). Appendix 2.2.2 describes methods of
estimating emissions. Methods for major sources of SO2 and NOx are reliable and have
remained largely the same since the last review. The same is true to an extent for direct
PM emissions. However, there have been fundamental changes in methods for estimating
NH3 emissions, and these are described in some detail. Appendix 2.2.3 describes
emission uncertainties, including recent comparisons between NEI data and alternative
2-5
-------
methods of estimating emissions. Geographic distributions of emissions are presented in
Appendix 2.5. where they can be more directly compared with concentration and
deposition data.
2.2.1 National Emissions by Source
N and S containing compounds contributing to deposition can be either primary
(i.e., directly emitted from sources) or secondary (i.e., produced from atmospheric
reactions involving precursor species directly emitted from sources). Major primary
species include NOx, SO2, NH3, and reduced organic nitrogen (RON).
Table 2-1 shows nationwide emissions of NOx, SO2, and NH3 by source category
compiled from the 2014 National Emissions Inventory (NEI) and other sources.
Emissions estimates are not available for RON. For the most part, NOx, SO2, and NH3
are each emitted by different sources. NOx emissions come from several important
sources. Highway vehicles are the largest source category of NOx emissions nationwide,
but off-highway vehicles, EGUs, other stationary fuel combustion, industrial processes,
fires, and biogenic emissions from soil are all substantial contributors to total NOx
emissions. Lightning is not included in the NEI, but it can also contribute substantially to
total NOx emissions. Although lightning is shown as a relatively modest source of NOx,
most production by lightning occurs during the summer and is the highest in the
south-central and southwestern U.S. (Zhang et al.. 2012a).
NH3 originates mainly from agricultural sources, which account for -80% of its
emissions nationwide. Major sources are livestock waste, including confined animal
feeding operations, and soils, after addition of N containing fertilizers. Fertilizer
application occurs mainly during spring and summer. Emissions from livestock waste are
roughly three times those from fertilizer application. In addition to NH3, reduced organic
compounds such as urea and a wide range of proteins and other biological components
are also emitted as the result of agricultural activity. Xing et al. (2013) observed that in
contrast to SO2, NOx, and other pollutants, total national emissions of NH3 increased
from 1990 to 2010. The authors attributed this to agricultural emissions, including
livestock, which they also identified as the dominant source of NH3 emissions in the
continental U.S. However, regionally the relative importance of agricultural and vehicle
emissions is likely to be variable. The deposition of reduced nitrogen can be three times
higher near roads (Bettez et al.. 2013). and motor vehicles can be a substantial contributor
to total NH3 emissions in urban areas (Baum et al.. 2001).
2-6
-------
Table 2-1 Emissions of NOx (nitric oxide + nitrogen dioxide), sulfur dioxide,
and ammonia by source category for 2017 (Teragrams3 N, S/yr).
NOx
SO2
NHs
Highway vehicles
1.0
0
0.1
Off-highway
0.7
0.1
<0.1
Utilities (fuel combustion)
0.3
0.6
<0.1
Other stationary (fuel combustion)
0.3
0.1
<0.1
Industrial and other processes
0.4
0.4
0.2
Livestock waste
0
0
1.9
Fertilizer application
0
0
0.7
Fires: wild, prescribed and agricultural
0.4
0.1
0.3
Biogenic
1.4
b
<0.1c
Total
3.3
1.2
3.2
N = nitrogen; NH3 = ammonia; NOx = the sum of nitric oxide and nitrogen dioxide; S = sulfur; S02 = sulfur dioxide; yr = year.
a1 Teragram = 1 * 109kg.
bNot applicable.
°Bouwman et al. (1997V
Source: httDs://www.eDa.aov/air-emissions-inventories/air-Dollutant-emissions-trends-data update except as noted.
SO2 emissions are dominated by stationary sources burning fossil fuels, particularly
EGUs, which contribute about half of total nationwide SO2 emissions. SO2 emissions
densities in most counties east of the Mississippi River are larger than in most counties in
the West (see Appendix 2.6.5).
Emissions of NOx and SO2 have decreased appreciably in recent years. National
emissions of NOx have decreased by 61%, and national SO2 emissions by 89% from
1990 to 2017 (OAQPS-Emissions Inventory and Analysis Group. 2016). Further details
of declining emissions for these species can be found in the ISAs for health effects for
NOx and SO2 (U.S. EPA. 2017d. 2016f). In contrast, nationwide primary PM2.5 and PM10
emissions estimates have changed little between the 2002 NEI (U.S. EPA. 2009a) and the
2017 NEI (U.S. EPA. 2020a). with national PM2.5 emissions estimates decreasing from
5.8 to 5.7 MMT, and PM10 emissions estimates decreasing from 21.6 to 17.1 MMT.
National annual NH3 emissions have fluctuated as a result changes in both emissions and
2-7
-------
methods of estimating emissions. However, no clear trend is evident in national NH3
emissions, with estimates for 1990 and 2017 differing by less than 1%.
Not included in Table 2-1 are primary emissions of PM2.5 and PM10. Nationwide PM10
emissions according to the 2014 NEI totaled 17.0 MMT. This is somewhat higher than
the SO2, NOx, and NH3 emissions in Table 2-1. but 85% of these were from dust and
fires, dominated by soil and organic matter. By comparison, total nationwide PM2 5
primary emissions totaled 5.3 MMT based on the 2014 NEI. This is comparable to the
totals for NOx, SO2 in Table 2-1 when it is considered that emissions in Table 2-1 are
given in terms of mass of N and S and does not include oxygen, while PM2.5 emissions
are based on total mass. However, primary PM2.5 emissions are also dominated by dust
(i.e., agricultural dust and road dust) and fires (i.e., wildfires, prescribed fires, and
agricultural fires), which together account for two-thirds of total nationwide PM2.5
primary emissions. PM2.5 from these source categories are mainly crustal material (dust)
and organic matter (fires).
As described in Appendix 2.1. in rural and remote areas secondary PM2.5 formed from
NOx and SO2 account for a greater fraction of PM2.5 than primary PM2.5. The fraction of
PM2.5 accounted for by NO;, and SO42 formed from SO2 and NOx in various U.S.
locations is discussed in Appendix 2.3. The NOx, SO2, and NHx emissions listed in
Table 2-1 cannot be used to quantitatively estimate the amount of secondary PM formed
because the precursors are not completely converted to PM. However, they provide not
only an estimate of emissions that lead to total NOy and SOx, but also provide an
estimate of the emissions that can be used in conjunction with atmospheric models to
estimate PM2.5 concentrations (see Appendix 2.5). The emissions estimated in Table 2-1
are ultimately responsible for a large fraction of PM2.5 in many areas, and it is the fraction
of PM mass (i.e., SO42 , NO;, . NH4 ) for which ecosystem impacts are best understood.
2.2.2 Methods of Estimating Emissions
The source categories used in Table 2-1 represent groups of similar NEI source sectors.
Highway Vehicles comprise all on-road vehicles, including light-duty as well as
heavy-duty vehicles, both gasoline and diesel powered. Off-highway vehicles and
engines include aircraft, commercial marine vessels, locomotives, and nonroad
equipment. Utilities (Fuel Combustion), also identified as electric power-generating units
(EGUs), are mostly coal burning, but some facilities burn natural gas and other fuels.
Other Stationary (Fuel Combustion) includes commercial/institutional, industrial, and
residential combustion of biomass, coal, natural gas, oil, and other fuels. Industrial and
Other Processes include a variety of different industries, including chemical
2-8
-------
manufacturing, cement manufacturing, and oil and gas production. The other processes
included in this category include gasoline stations and terminals, commercial cooking,
road and construction dust, solvent use, and waste disposal. Agriculture includes both
fertilizer application and livestock waste emissions. Fires include wildfires, prescribed
fires, and agricultural field burning. The biogenic category includes emissions from
vegetation and soil. Both nitrifying and denitrifying organisms in the soil can produce
NOx, mainly in the form of NO.
Emissions data for each source category listed in Table 2-1 are from the 2014 National
Emissions Inventory (NEI), Version 1. The NEI is a national compilation of criteria air
pollutant and hazardous air pollutant emissions. The process of estimating emissions is
explained for each source in a detailed technical support document for the inventory
(U.S. EPA. 2016b). The NEI is maintained to support the NAAQS, and the Clean Air Act
requires states to submit emissions to the U.S. EPA as part of their State Implementation
Plans (SIPs). The Air Emissions Reporting Rule (AERR) requires agencies to report all
sources of emissions, except fires and biogenic sources. Reporting of open fire sources,
such as wildfires, is encouraged, but not required. Data in the NEI come from a variety of
sources. The emission values are predominantly from state, local, and tribal agencies and
are used wherever they are available, unless there are gaps or problems with submitted
data. U.S. EPA quality assures and augments the data provided by states to assist with
data completeness using separate augmentation procedures for each source as described
in detail in a technical support document to fill in gaps for sources and/or pollutants that
are often not reported by state, local, and tribal agencies. The intent is to create the most
complete inventory for use in air quality modeling, national rule assessments,
international reporting, and other reports. QA procedures and acceptance criteria are
detailed in the NEI technical support document (U.S. EPA. 2016b).
For nonpoint sources, U.S. EPA provides tools that state, local, and tribal agency staff
can use to generate emission estimates. For the 2014 NEI, the U.S. EPA developed
emission estimates for many nonpoint sectors in collaboration with a consortium of
inventory developers from various state agencies regional planning organizations called
the Nonpoint Method Advisory (NOMAD) Committee. More detailed NOMAD
subcommittees were established to collaborate on methods and emission factors for key
nonpoint source categories/sectors, including oil and gas exploration and production,
residential wood combustion, agricultural NH3 sources (including fertilizer and
livestock), and industrial and commercial/institutional fuel combustion, among other
sources. The U.S. EPA also generates emission estimates as stand-alone data sets
covering biogenics, agricultural livestock, fertilizer application, nonroad mobile sources,
rail emissions, and commercial marine vessel ports and in-transit (underway) sources.
U.S. EPA data sets for sources and pollutants are used only for sources not provided by
2-9
-------
state, local, and tribal data. Tools and methods for estimating emissions from a given
source, including EGUs, agricultural livestock, fertilizer application, mobile sources,
agricultural and wildland fires (wildfires + prescribed fires), and wood combustion, were
described in the 2014 NEI technical support document (U.S. EPA. 2016b).
The 2014 NEI technical support document provides considerable detail on emission
factors and emission estimation methods by source used to generate the data in Table 2-1
(U.S. EPA. 2016b). Methods for estimating PM2.5 and PM10 emissions from dust and fire,
the two largest national sources, are derived from experimental emission factors along
with source specific information (e.g., crop type and tilling frequency for agricultural
dust, vehicle weight and miles traveled for unpaved road dust) using source specific
equations available in the NEI technical support document. These methods are largely
well-established, although they have been updated to accommodate satellite data and
emissions modeling improvements, particularly in the case of fire emissions.
Methods for estimating emissions from electric power-generating units and mobile
sources, the largest sources of SO2 and NOx, are also well established, and emissions
from these sources are decreasing. In contrast, ammonia emissions in the U.S. are
increasing (Butler et al.. 2016). and significant uncertainties in the magnitude as well as
spatial and temporal variability of NH3 emissions estimates were reported in the 2008
ISA (U.S. EPA. 2008a). Two new methods for estimating ammonia emissions from
fertilizer applications and livestock waste are highly relevant to understanding NHx
sources and deposition are provided here as examples, but a similar level of detail is
given in the NEI technical support document for other sources of NH3, SO2, and NOx,
and a thorough reading of that document is necessary for a full description of emissions
estimation methods used to construct Table 2-1 (U.S. EPA. 2016b).
Soil and fertilizer emissions are treated differently in the NEI for NH3 and NOx. For NH3,
fertilizer application is recognized as a major source for which emissions are specifically
estimated, and emissions from fertilizer application are estimated only for NH3. The
approach to calculating emissions from fertilizer application in the 2014 NEI is a
completely new methodology to estimate ammonia (NH,) emissions from agricultural
soils. The approach to estimate 2014 fertilizer emissions consists of these steps: (1) run
the Fertilizer Emissions Scenario Tool for CMAQ FEST-C (vl.2;
https://www.cmascenter.org/fest-c/) and the bidirectional version of CMAQ (v5.0.2;
https://www.cmascenter.org/) to produce year 2011 nitrate (NO3), ammonium (NH/,
including urea), and organic (manure) nitrogen (N) fertilizer usage estimates, and gaseous
ammonia NH3 emission estimates respectively; (2) calculate county-level emission
factors for 2011 as the ratio of bidirectional CMAQ NH3 fertilizer emissions to FEST-C
total N fertilizer application; (3) run FEST-C to produce year 2014 NO3, NH4+ (including
2-10
-------
Urea), and organic (manure) nitrogen fertilizer usage estimates; and (4) multiply
county-level 2014 FEST-C total fertilizer estimates by the 2011 emission factors to
estimate 2014 NH3 emissions. FEST-C reads land use data from the Biogenic Emissions
Land Use Dataset (BELD) Version 4, meteorological variables from the Weather
Research and Forecasting (WRF v3.7.1) model (https://www.mmm.ucar.edu/weather-
research-and-forecasting-model) and nitrogen deposition data from a previous or
historical average CMAQ simulation. The Environmental Policy Integrated Climate
(EPIC) modeling system (http://epicapex.tamu.edu/) provides information regarding
fertilizer timing, composition, application method, and amount. Figure 2-1 provides a
comprehensive flowchart of the complete EPIC/FEST-C/WRF modeling system.
The Fertilizer Emission Scenario Tool for CMAQ
(FEST-C)
Source: U.S. EPA (2016bl
Figure 2-1 Modeling system used to compute 2014 Fertilizer Application
Emissions.
2-11
-------
Fertilizer application can also lead to NOx emissions, but it does not dominate soil
emissions of NO. Biogenic emissions of NO are computed based on 2014 meteorology
data from the Weather Research and Forecasting (WRF) model version 3.8 (WRF v3.8)
and using the Biogenic Emission Inventory System, Version 3.61 (BEIS 3.61) model,
based on land use and vegetation data (U.S. EPA. 2016b). The contribution of fertilizers
to soil NOx emissions is not estimated in the NEI, but it has been estimated as 10%
globally (Hudman et al.. 2012). Biogenic emissions of NH3 are not estimated in the NEI,
but aside from fertilizer application it is a minor contributor, as shown in Table 2-1.
Further details on estimating biogenic NOx emissions are given in the NEI Technical
Support Document (U.S. EPA. 2016b).
Livestock waste is another important source of ammonia in the U.S. In the 2014 NEI, the
U.S. EPA has updated the methodology for ammonia emissions from the
housing/grazing, storage and application of manure from beef cattle, dairy cattle, swine,
broiler chicken, and layer chicken production. Cows, swine, and chickens account for
95% of national NH3 emissions from livestock waste in 2014. The approach to estimate
2014 livestock NH3 emissions from these animals consists of these general steps:
(1) estimate 2014 county-level animal populations using 2012 and 2014 USDA
agricultural census data; (2) use a model developed by Carnegie-Mellon University
(McQuilling and Adams. 2015; Pinder et al.. 2004a; Pinder et al.. 2004b) to produce daily
resolved, climate-level emission factors for a particular distribution of management
practices for each county and animal type, as expressed as emissions/animal; and
(3) multiply the county animal populations by the daily emission factor for each county
and animal type to estimate emissions per day and sum daily emissions across the entire
year for each county and SCC to produce annual emissions for use in the NEI. The model
inputs and outputs are shown in Figure 2-2.
2-12
-------
Source: U.S. EPA (2016bl
Figure 2-2 Process to produce specific location and practice specific daily
emission factors for livestock waste.
Similar details are given for estimating emissions for major sources of NOx and SO2 in
the NEI technical support document (U.S. EPA. 2016b). including EGUs, on-road mobile
sources, marine vessels, locomotives, other nonroad sources, airports, rail yards, landfills,
agricultural and wildland fires, wood combustion, other fuel residential and industrial
fuel combustion, charcoal grilling, waste disposal, vegetation and soil, and other sources.
Air emissions data from the 2014 Toxic Release Inventory (TRI) were also used in the
2014 NEI to supplement point source NH3 emissions provided to the U.S. EPA by state,
local, and tribal agencies. The TRI is a U.S. EPA database containing data on disposal or
other releases including air emissions of over 650 toxic chemicals from approximately
21,000 facilities. Data are submitted annually by U.S. facilities that meet TRI reporting
criteria.
2.2.3 Evaluation and Uncertainty
As described in the 2008 ISA (U.S. EPA. 2008a). emissions from different sources in the
NEI are estimated with a wide range of methods that include direct measurements,
indirect measurements, model predictions, and assumptions. Because there are unknown,
incomplete, and variable emission rates, as well as unknown sources that are not
represented, the NEI reflects an on-going process of updating increasing or declining
emissions, improving estimation methods, and filling data gaps as measurements become
available or understanding of emissions changes. Often, steps are taken to reduce errors
in estimation as they are discovered, resulting in improved estimates as uncertainties are
found. For example, the estimate of Gilliland et al. (2003) that annual NEI NH3 was 37%
2-13
-------
higher than estimates based on modeled NH44" deposition led to the development of lower
emission factors for nondairy cows and swine (U.S. EPA. 2008a). Because of these
unknowns and limitations, quantitative uncertainty estimates based on probability density
functions or other statistical methods are not provided with NEI data. Instead, emission
factors receive a rating based on the reliability of methods used for determining the
emission factor. The rating is based on both the number of representative sources and the
characteristics of the data used to determine the emission factor, but it does not imply
statistical error bounds or confidence intervals for the emission factor (U.S. EPA. 1996).
As an alternative, uncertainties are often evaluated through separate efforts using a
variety of techniques. These techniques include comparing inventory predictions with
measured long-term trends, comparing emission estimates derived from principle
component analysis or other statistical methods, comparing emissions estimated by
inverse modeling of chemical transport models, and comparison with satellite data (U.S.
EPA. 2008a). The distinction between the inventory compilations like the NEI and
alternative satellite- and model-based methods of estimating emissions is generally
described in terms of bottom-up and top-down estimates. The entries in emissions
inventories are obtained using a bottom-up approach, in which entries are based on
emissions factors, activity rates, and control device efficiency for various source types.
This contrasts to a top-down approach in which measurements of pollutant concentrations
from satellites, aircraft, or surface monitors are used to constrain a priori estimates of
emissions using a chemistry-transport model (CTM).
Because of this variety of top-down approaches and the number of separate studies
resulting in a wide range of estimated uncertainties, there is no single estimate of
uncertainty that applies to either total emissions or emissions from individual sources in
Table 2-1. However, reports from numerous publications on emissions inventory
evaluation have resulted in a wide range of uncertainty estimates for application to NEI
data, and these were summarized in the 2008 ISA (U.S. EPA. 2008a). Across all sources,
total NOx emissions estimates from satellite data ranged from "highly consistent" with
NEI estimates (Martin et al.. 2006) to 68% higher than NEI estimates (Jaegle et al..
2005). Fewer estimates of individual source emissions were evaluated, with NOx
emissions both higher and lower than estimates using other methods.
Some recent work has shown summertime over-prediction of model NOx estimates using
recent U.S. EPA inventories (e.g., 2008 and 2011) when compared against monitored
ambient concentrations (Cantv et al.. 2015; Anderson et al.. 2014a). Appel et al. (2017)
found in their simulations that NOx model over-estimates in summer were greatly
diminished or reversed in other seasons. Anderson et al. (2014a) and Travis et al. (2016)
concluded that emissions of NOx from mobile sources are being overestimated and are
2-14
-------
the source of this bias. Studies over Texas (Souri et al.. 2016; Tang et al.. 2015)
suggested a smaller or no bias in on road NOx emissions compared to Anderson et al.
(2014a) and Travis et al. (2016). and suggested that both high and low biases in other
NOx source categories (e.g., area sources, point sources, soil NO) impact
model/measurement discrepancies. Marr et al. (2013) used near-road measurements to
also conclude that mobile source NOx emissions in U.S. EPA's 2008 NEI agreed well
with measurements (i.e., within 3%). The cause of discrepancies between measured and
modeled concentrations are difficult to diagnose because the emission modeling process
and associated photochemical modeling is complex. Researchers are continuing to
investigate this question.
There can be higher uncertainties for specific sources. For example, about 60% of the
total NOx emitted by soils nationwide is estimated to occur in the central Corn Belt of the
U.S. Spatial and temporal variability in soil NOx emissions can lead to uncertainty in
emissions estimates. Soil emissions occur mainly during summer and across the entire
country, including areas where anthropogenic emissions are low. Emission rates depend
primarily on fertilization amount, soil temperature, and moisture. Models of NOx
emissions from soils [e.g., Hudman et al. (2012)1 include these dependencies, but most
measurements on which they are based are made at temperatures <30°C. However, in
agricultural areas subjected to very high temperatures (>40°C) like the Imperial Valley,
CA, emissions factors for NO following fertilizer application ranged from 1.8 to 6.6%, as
compared to estimates of typically ~1 to 2% (Oikawa et al.. 2015). Oikawa et al. (2015)
also suggested that in many areas of the Southwest, the NEI overestimates anthropogenic
emissions at the expense of soil emissions and that these soil emissions have a noticeable
effect on ozone formation. Travis et al. (2016) estimated that combustion accounts for
68% of NOx emissions in the Southeast in summer, with the remainder from soils. These
results indicate that soil emissions need to be better understood. Estimating emissions
from highway vehicles can also be challenging because there is a wide variation in
emissions between different vehicles.
Activity rates and uncertainties for NH3 are difficult to quantify, and estimates have yet
to be made for reduced organic nitrogen. NH3 emission estimates are generally more
uncertain than NOx and SO2 emission estimates because of the variety of agricultural
practices used, re-emission after deposition, and the dispersed nature of agricultural
processes, as well as the complex influences of meteorology on processes controlling
transformation and removal of nitrogen species on spatial and temporal emission patterns.
As described in Appendix 2.5.2. NO2 and NH3 can be both emitted from and deposited to
soils, water, or vegetation depending on their atmospheric concentrations and
characteristics of the underlying surface.
2-15
-------
Mixed results were reported in evaluating the total NH3 emissions estimates in the 2008
ISA (U.S. EPA. 2008a). However, as described above, there have been major changes in
estimation methods for the most important ammonia sources since that report, and more
recently Paulot et al. (2014) showed that a range of published bottom-up and top-down
estimates of annual total U.S. ammonia emissions agreed within 10%. The NEI national
estimate of 2.3 Tg NH3 from Table 2-1 was within 20% of the average across all
estimates of 2.8 ± 0.2 Tg NH3 reported by Paulot et al. (2014). However, in the same
comparison there was a divergence in the timing of the seasonal maximum, and
agreement varied considerably temporally and spatially (Paulot et al.. 2014).
Most SO2 emissions originate from point sources having well-known locations and
identifiable fuel streams. Uncertainties in annual emissions were estimated to range from
4 to 9% for SO2 and slightly larger for NOx from the same point sources identified in the
1985 NAPAP inventories for the U.S. (Placet et al.. 1990).
2.3 Atmospheric Chemistry of Nitrogen and Sulfur Species and
Particulate Matter (PM)
The atmospheric chemistry ofN and S species relevant for the production of ecosystem
nutrients and acidic species was extensively reviewed in the 2008 ISA (U.S. EPA.
2008a). The main findings from that review and key findings from more recent studies
and reviews are included here. Appendix 2.3.1 describes atmospheric NOx chemistry and
the formation of HNO3. Appendix 2.3.2 describes atmospheric sulfur oxide chemistry and
the formation of H2SO4. Appendix 2.3.3 reviews the role of ammonia as the most
important atmospheric base for neutralizing atmospheric nitric and sulfuric acids and
forming PM. The chemistry of all of these species largely controls the extent of acid
deposition as well as the fraction of nitrogen in particulate matter, which in turn
determines deposition rate and transport distance. The remaining sections review
atmospheric organic sulfur and nitrogen compounds (Appendix 2.3.4). atmospheric
organic acids (Appendix 2.3.5). and formation of PM2.5 (Appendix 2.3.6).
2.3.1 Nitrogen Oxides
NOx (NO + NO2) is the precursor for oxidized nitrogen species that contribute to acidic
deposition. More specific details on the chemistry and transformation of NOx can be
found in the 2016 ISA for Oxides of Nitrogen—Health Criteria (U.S. EPA. 2016f) and
for SOx in the 2008 ISA for Sulfur Oxides (U.S. EPA. 2017d). Hence, those topics are
only briefly recounted here with special reference to the secondary NOx and SOx
2-16
-------
NAAQS. Oxidized nitrogen species (NOy) are introduced into the atmosphere as NOx,
mainly from fossil fuel combustion as described in Appendix 2.2. Figure 2-3 summarizes
the atmospheric reactions of NOx, showing rapid inter-conversion of NO and NO2 in
sunlight, with slower formation of more oxidized organic and inorganic products (NOz).
A large number of oxidized nitrogen species in the atmosphere are formed from the
oxidation of NO and NO2 (shown in the inner box). These include nitrate radicals (NO3),
nitrous acid (HONO), nitric acid (HNO3), dinitrogen pentoxide (N2O5), nitryl chloride
(CINO2), peroxynitric acid (HNO4), peroxyacetyl nitrate (PAN) and its homologues
(PANs), other organic nitrates, such as isoprene- and monoterpene-derived nitrates, and
particulate nitrate (pMV). These species (and NH3) are characterized by large
differences in their solubilities (Table 2-2). which determine their ability to be taken up
by cloud droplets, airborne particles, and moist surfaces.
2-17
-------
emissions
Ca2+ = calcium ion; HN03 = nitric acid; H02 = hydroperoxy radicals; hv= energy of solar photon with frequency v;
Mg2+ = magnesium; NH3 = ammonia; NH4+ = ammonium; NO = nitric oxide; N02 = nitrogen dioxide; N03" = nitrate ion; NOx = the
sum of NO and N02; NOz = oxidation products of NOx; NOY = NOx + NOz; 03 = ozone; PAN = peroxyacetyl nitrate; R02 = organic
peroxy radicals.
Note: The inner shaded box contains NOx (NO + N02). The outer box contains other species (NOz) formed from reactions of NOx.
All species shown in the outer and inner boxes are collectively referred to as NOY by the atmospheric sciences community.
Source: CPHEA.
Figure 2-3 Schematic diagram showing pathways for reactive nitrogen
species in ambient air.
2-18
-------
Table 2-2
Henry's law coefficients for selected reactive nitrogen species at
25°C in water.
Compound
Coefficient (mol/kg/bar)
HNOs
2.6 x 10®
HONO
49
NO
0.0019
NO2
0.012
PAN
4.1
NHs
61
1 bar = 105 Pa; C = Celsius; HN03 = nitric acid; HONO = nitrous acid; NH3 = ammonia; NO = nitric oxide; N02 = nitrogen dioxide;
PAN = peroxyacetyl nitrate.
Source: adapted from Sutton etal. (2011).
Reactions producing more oxidized forms of nitrogen (NOz) involve mainly O3, OH, and
organic radicals with NO and NO2. The reaction of NO2 with OH leads directly to HNO3:
NO2 + OH + M HNOs
Equation 2-1
The reaction ofN02 with O3 produces nitrate radical (NO3), which reacts further to form
dinitrogen pentoxide (N2O5), and ultimately also produces HNO3:
NO2 + 03"^ NO3 + O2
Equation 2-2
NO2 + NO3 N2O5 (equilibrium)
Equation 2-3
N2O5 + H2O 2HN0s
Equation 2-4
The relative importance of these two paths for producing HNO3 is strongly location and
seasonally dependent, with the first path dominating when OH radicals are abundant
(during the day) and the second during the night and under cold conditions. Warneck
(1999) estimated that most HNO3 is formed in the sunlit portions of clouds by the
reaction of NO2 with OH, with much smaller amounts from the pathway involving N2O5
hydrolysis. Because it is highly soluble, HNO3 is taken up by particles or cloud droplets
to form NO3 and is also deposited onto moist surfaces, such as on vegetation. HNO3 also
2-19
-------
recycles back to NO2 in the gas phase by photolysis and reaction with OH radicals, but on
timescales longer than that for uptake by cloud droplets, particles, and the surface.
Whereas photolysis of HNO3 is slow (1 ~ 106 s) in the gas phase, it can be two orders of
magnitude faster on moist surfaces (Ye et al.. 2016). releasing NO2 and/or HONO back to
the atmosphere.
NO2 reacts with organic peroxy radicals to form organic nitrates such as peroxyacetyl
nitrate (PAN) and its homologues as shown on the right side of Figure 2-3; other RO2NO2
compounds are much less stable than PANs. NO and NO3 radicals also react with organic
radicals produced by the oxidation of isoprene and monoterpenes to form a wide range of
organic nitrates. Considering the troposphere as a whole, most of the mass of NOz shown
in Figure 2-3 is in the form of PAN and HNO3. However, organic nitrates such as
isoprene- and monoterpene-derived nitrates increase in importance in the planetary
boundary layer (PBL), and are likely to be dominant in vegetated areas (Kim et al..
2015a: MinetaL 2014).
In forested areas, the initial step in the production of isoprene nitrates (INs) is most often
the reaction of isoprene with OH radicals to produce isoprene peroxy radicals. These can
react with HO2 radicals, other RO2 radicals, or isomerize to produce a variety of organic
compounds. They can also react with NO to produce multifunctional organic nitrates.
Lifetimes on the order of one to a few hours can be estimated for these first generation
INs based on their reactions with OH radicals and O3 (Lockwood et al.. 2010; Paulot et
al.. 2009). The reaction products can further react with NO (after internal rearrangement)
to form secondary organic nitrates such as ethanal nitrate, methacrolein nitrate,
propanone nitrate, and methyl vinyl ketone nitrate. The second-generation organic
nitrates are more stable than the first-generation INs because they lack a double carbon
(C = C) bond. Obviously, the relative importance of pathways forming nitrates or other
products depends on the ambient concentrations of NO and other oxides of nitrogen for
which many key experimental details are still lacking. During the SEAC4RS
measurement campaign, which took place in the summer of 2013 in the southeastern
U.S., Travis et al. (2016) found that these two pathways were of comparable importance.
In addition to oxidation initiated by OH radicals, isoprene is also oxidized by NO3
radicals. Rollins et al. (2009) determined a yield of first-generation carbonyl nitrates of
70% based on experiments in large reaction chambers. These first-generation nitrates can
further react with NO, leading to the production of second-generation organic (alkyl)
nitrates. Mao et al. (2013) estimated that the global mean lifetime is ~5 days for these
second-generation organic nitrates. Mao et al. (2013) also suggested that the export of
INs and other organic nitrates followed by their decomposition is potentially a larger
source of NOx to the boundary layer of the western North Atlantic Ocean than the export
2-20
-------
of PANs. Some INs are low enough in volatility that they can partition to the particle
phase [e.g., Rollins et al. (2009)1. Once in the particle phase, the INs hydrolyze to form
HNO3 and an alcohol, with a rate constant that correlates strongly with the acidity of the
particles (Rindclaub et al.. 2015; Jacobs et al.. 2014).
In addition to considering the chemistry of isoprene-derived nitrates during SEAC4RS,
Fisher et al. (2016) considered the formation of organic nitrates derived from the
oxidation of monoterpenes with either one or two double bonds. Their modeling results
suggest that isoprene- and monoterpene-derived nitrates account for 25 to 50% and -10%
of total organic nitrates and that production of isoprene- and monoterpene-derived
nitrates account for -20% of the net loss of NOx emitted in the Southeast during summer.
Fisher et al. (2016) also noted that production of organic nitrates involving biogenic
VOCs is the dominant NOx sink only in areas where elevated levels of biogenic VOCs
coincide with low NOx levels (otherwise the major sink would be formation of HNO3).
As a result, these processes will represent only a minor pathway for NOx loss. In any
event, as with isoprene-derived nitrates, monoterpene-derived nitrates are also mainly
taken up by particles with formation of HNO3. Uptake by particles was estimated by
Fisher et al. (2016) to account for -60% of the removal of gas-phase organic nitrates,
with -20% recycled back to NOx and another 15% deposited to the surface.
2.3.2 Sulfur Oxides
SO2 is the only gas phase form of SOx (SO2 + SO42 ) emitted in the tropospheric
boundary layer at concentrations of concern for environmental exposures (U.S. EPA.
2008c). It reacts in both the gas phase and in aqueous solution in clouds and particles to
form SO.f As described in the 2008 ISA (U.S. EPA. 2008a). the steps involved in
aqueous-phase oxidation of SO2 begins with dissolution of SO2 following Jacobson
(2002):
S02(g) S02(aq)
Equation 2-5
and is followed by formation and dissociation of H2SO3:
S02(aq) + H2O H2S03(aq) H+ + HSOs" 2H+ + SOs2"
Equation 2-6
Dissolved SO2 thus rapidly partitions into four forms with the same oxidation state, with
their relative concentrations dependent on pH:
2-21
-------
S(IV) = S02(aq) + H2S03(aq) + HS03"(aq) + S032"(aq)
Equation 2-7
S(IV) is then oxidized to SO42 in cloud water primarily by either H2O2, O3, and O2 in the
presence of dissolved Fe(III). Reaction with H2O2 is most important at pH less than about
5.3, and reaction with either dissolved O3 or with O2 catalyzed by Fe(III) becomes most
important at pH greater than about 5.3, as shown in Figure 2-4 (Seinfeld and Pandis.
1998).
pH
aq = aqueous; Fe(lll) = iron (oxidation number III); H202 = hydrogen peroxide; Mn(ll) = mangnanese (oxidation number II);
N02 = nitrogen dioxide; 03 = ozone; S(IV) = sulfur (oxidation number IV); S02 = sulfur dioxide.
Concentrations assumed are: [S02(g)] = 5 ppb; [N02(g)] = 1 ppb; [H202(g)] = 1 ppb; [03(g)] = 50 ppb; [Fe(lll)(aq)] = 0. 3 |jM;
[Mn(ll)(aq)] = 0.03 |jM.
Source: Seinfeld and Pandis (1998).
Figure 2-4 Rate of conversion of sulfur (IV) to sulfur (VI) by different
oxidation paths as a function of pH.
The remaining SO2 is oxidized to H2SO4 in the gas phase with a characteristic timescale
of-10 days [based on OH = 106/cm3 and rate coefficient =1.3 x 10 l2/cm7molcc/s:
Sander etal. (2011)1 following a multistep process:
2-22
-------
S02 + OH + M -> HS03 + M
Equation 2-8
HSOs + 02 ^ SO3 + HO2
Equation 2-9
and/or by
502 + sCI -> SO3 + products
Equation 2-10
where sCI is a stabilized Criegee intermediate (Berndt et al.. 2012; Mauldin et al.. 2012;
Welz et al.. 2012) and products refer to other organic radicals. Criegee radicals are
produced by the reaction of alkenes with O3 during both night and day. The relative
importance of the OH and sCI pathways depends in large measure on the local
concentration of alkenes, in particular biogenic alkenes. Welz et al. (2012) also raised the
possibility that Criegee radicals might be important for the oxidation ofN02 to form
nitrate radicals. SO3 produced by either path further reacts to form gas-phase H2SO4 via
503 + H2O ^ H2SO4
Equation 2-11
Because H2SO4 is extremely soluble, it is removed rapidly by transfer to the aqueous
phase of particles and cloud droplets.
2.3.3 Acid Neutralization by Ammonia
As the most common soluble base in the atmosphere, NH3 plays a key role in neutralizing
the acidity in ambient particles and in cloud, fog, and rainwater resulting from dissolution
of H2SO4 and HNO3, and the weak acidity due to organic acids. The atmospheric lifetime
of NH3 with respect to oxidation by OH radicals is ~2 months [based on
OH = 106 molec/cm3 and rate coefficient =1.6/ 10 ' Vcm Vmolec/s; Sander et al. (2011)1.
As a result, uptake by cloud droplets, particles, and the surface is favored over reaction
with OH radicals. Xu and Penner (2012) estimated a globally averaged lifetime for NH3
of ~11 hours as a result of these processes, implying strong spatial and temporal
variability of NH3 concentrations.
Sulfuric acid can be partly or totally neutralized by NH3. Seinfeld and Pandis (1998)
defined two regimes: (1) ammonia poor [TA] <2 [TS] and (2) ammonia rich [TA]
>2 [TS], where TA and TS refer to total ammonia, ammonium and sulfate concentrations
in gas, aqueous, and solid forms. In the first regime, there is partial neutralization; sulfate
is in the form of (NH4)HS04, the vapor pressure of NH3 is very low, equilibrium favors
2-23
-------
formation of ammonium sulfate over ammonium nitrate, any nitrate is driven to the gas
phase, and ammonium nitrate levels are low (or even zero). In the second regime, sulfate
is in the form of (NH^SC^, and any NH3 left over could react with HNO3 to form
NH4NO3. However, this conceptual model neglects interactions with organic compounds.
Analysis of data by Kim et al. (2015a) from the SEAC4RS field study and the Chemical
Speciation Network (CSN) during August-September of 2013 indicates that the extent of
neutralization of sulfuric acid and acidic sulfate by ammonium was incomplete in the
Southeast despite an excess of atmospheric NH3. Kim et al. (2015a) suggested that uptake
of NH3 is inhibited by organic compounds in particles. This suggestion is in accord with
laboratory studies of Liggio et al. (2011) who found that organic compounds, especially
terpenes and «-alkancs on particle surfaces are effective in inhibiting NH3 uptake by
particles.
NH4NO3 is in thermodynamic equilibrium with gas-phase NH3 and HNO3. The
equilibrium constant is extremely sensitive to variations in relative humidity and
temperature such that it varies over several orders of magnitude depending on
atmospheric conditions, but in general, lower temperature and higher relative humidity
(e.g., during winter) shifts the equilibrium towards condensed phase NH4NO3. The effects
on phase partitioning are pronounced because of the large variation in the equilibrium
constant, K, (—10—103 ppb2) between summer and winter conditions in many locations.
Also, as noted by Malm et al. (2016). NH4NO3 can volatilize and reform multiple times
during transport away from sources of NH3 and HNO3. Because the atmospheric lifetimes
of NH3, HNO3, and NH4NO3 differ substantially from each other, local conditions of
temperature and relative humidity, by implication, control how far these species can
travel.
Although the above considerations apply to particles in general, it should be remembered
that the mass of airborne particles is present in two distinct size fractions, each with its
own characteristic composition [see U.S. EPA (2019)1. These differences determine the
size fraction in which pNO, will be found. Because SO42 is found mainly in the fine
particle mode these considerations tend to apply more to the atmospheric fine mode.
Displacement of HC1 (and other hydrohalic acids) from marine aerosol (found typically
in the coarse mode) by gas-phase HNO3 has long been known to occur, resulting in
particulate nitrate (pNO, ) being associated with sodium in the coarse mode in many
coastal areas. Brimblecombe and Clegg (1988) provided a detailed evaluation of the
thermodynamic data and a discussion of this process. Wolff (1984) found that
coarse-mode pNCh" is formed by adsorption of HNO3 on basic soil particles (i.e., those
containing Ca2+ and Mg2+). These distinctions between the behavior of pNO, in the fine
and coarse modes are important as deposition rates for these two size modes can differ
2-24
-------
appreciably and there can be large differences in the ratio of fine to coarse pNO, as
shown in Appendix 2.5.2.1.
The composition of rainwater and of particles is strongly affected by pH. As described
above, pH determines the distribution of S(IV) species in cloud water, rainwater and the
aqueous phase of particles. This in turn impacts their oxidation processes
(Appendix 2.3.2). the solubility of trace metals, and the partitioning of weak acids among
other factors. As the most abundant base in the atmosphere, NH3 has a strong influence
on pH of cloud water. However, the role of NH3 as a base is limited to atmospheric
processes. Once deposited in soil, oxidation of NH3 and NH44" to NO;, (during
nitrification) produces an amount of H+ equivalent to HNO3 deposition rScheffe et al.
(2014) and references therein].
2.3.4 Organic Nitrogen and Sulfur
In addition to deposition of NOy and NHx, the deposition of other nitrogen compounds,
in particular dissolved organic nitrogen (DON) also occurs. Proteins, amino acids, urea,
amines, and other DON compounds can contribute to acidification in soils and be an
important source of nutrients to terrestrial and aquatic environments (Jickells et al.. 2013;
Cape et al.. 2011; Cornell. 2011b; Sutton et al.. 2011). The content of organic nitrogen in
particles and rainwater can be characterized in two ways. First, it can be calculated as the
difference between total N as measured by total elemental analysis [e.g., Bronk et al.
(2000)1 minus NO3 and NH44". In the second way, the content of organic nitrogen in
particles and rainwater can be characterized by measuring the concentrations of
individual species. However, the number of species constituting DON at a particular
location can be quite large. For example, Altieri et al. (2009a) detected several thousand
organic N containing species in precipitation samples collected in New Jersey and found
the overall composition was consistent with oligomerization of amino acids; in most
compounds, N was in reduced form.
Cornell (2011a) estimated based on measurements reported in 58 published studies that
organic N constitutes 35% of total N in rainwater in North America. Jickells et al. (2013)
estimated based on data from a number of measurement sites (n = 115 globally), that
average DON in rainwater contributes -25% of the flux of total nitrogen. They also
reasoned that because it is correlated with total nitrogen in rainwater (if = 0.57), which
has a large anthropogenic component, DON might also have a large anthropogenic
component. Further description of DON measured at sites in the CONUS are deferred to
Appendix 2.6.2.
2-25
-------
A large number of organosulfates (R-O-SO3H) have been detected in rainwater samples
(Alticri et al.. 2009b). However, their abundances could not be determined. Tolocka and
Turpin (2012) estimated that organic sulfates could contribute up to 5 to 10% of organic
mass on average to particle mass based on measurements taken at 12 sites across the U.S.
Liao et al. (2015) found that organic sulfates accounted a few percent of particulate
sulfate, mainly from the two most abundant forms, isoprene epoxydiol sulfate and
glycolic acid sulfate. Organic sulfates such as these have high acid dissociation constants
and are expected to act as singly charged species.
Phytoplankton emit copious amounts of dimethyl sulfide, which can be oxidized to sulfur
dioxide and to methane sulfonic acid. The SO2 that is formed can then be a source of
H2SO4 in coastal areas.
2.3.5 Organic Acids
The effects of deposition of acidic sulfur and nitrogen should be considered in the context
of a more complete description of the composition of rainwater, including organic acids.
However, organic acids are not routinely measured by monitoring networks because they
are unstable with respect to microbial degradation following collection. As a result, data
for organic acids in rainwater are sparse. Formic and acetic acids are typically the most
abundant organic acids found in rainwater in the U.S. (Willev et al.. 2011; Avery et al..
2006; Talbot et al.. 1990). They are largely secondary in origin (i.e., produced in the
atmosphere by the photochemical oxidation of biogenic and anthropogenic
hydrocarbons). Paulot et al. (2011) suggested that isoprene oxidation is the largest global
source of formic and acetic acids. These acids are also produced by the oxidation of
ethane and propylene emitted in automobile exhaust. Their abundances in rainwater and
their effects on pH are not negligible. For example, Willev et al. (2011) found that formic
acid (pKa = 3.75) and acetic acid (pKa = 4.76) were the major organic acids present and
contributed -22 and 5%, respectively, of free acidity (mean pH = 4.65) in rainwater
samples collected at Wilmington, NC in 2008. In addition, other organic acids
(e.g., oxalic acid, lactic acid) have been found to be present at much lower levels at this
site (Avery et al.. 2006). Vet et al. (2014) noted that organic acids should be monitored in
areas where the concentration of H+ is <5 (j,eq/L (or pH > 5.3). As will be seen in
Appendix 2.6. this condition is met in areas like the Northwest where concentrations of
NO;, and SO42 in rainwater are low. Even in areas where the effects of organic acid
neutralization by NH4 might be small, the vapor pressures of some organic acids
(e.g., oxalic acid) would be reduced by orders of magnitude, resulting in increased uptake
of the organic acid from the gas phase and growth of particles (Ortiz-Montalvo et al..
2014; Paciga et al.. 2014).
2-26
-------
2.3.6
Particulate Matter (PM)
The chemistry of NOx and SO2 described in Appendix 2.3.1 and Appendix 2.3.2. and the
neutralization reactions of NH3 described in Appendix 2.3.3 are relevant not only to
understanding the distribution of NOy and SOx species, but also in explaining a large
fraction of PM2.5 in most of the U.S. Figure 2-5 shows PM2.5 composition in numerous
U.S. locations. In all locations SO42 and NO;, account for a substantial fraction, and in
many cases the majority, of PM2.5. In general, SO42 accounts for an increasing fraction of
PM2.5 moving east or south, and NO;, for a greater fraction moving west or north.
Figure 2-5 also shows that PM2 5 concentrations are lower and S042 accounts for a much
greater fraction of PM2.5 mass in 2003-2005 than in 2013-2015. This reflects the steep
decline in SO2 emissions over this period (Appendix 2.2.1) and demonstrates that it has
greatly impacted PM2.5 composition and concentration in the U.S.
The decrease in SO42 contribution is so large that in many locations where SO42 was the
greatest contributor to PM2.5 mass in 2003-2005, organic carbon was more abundant in
2013-2015. However, as described in Appendix 2.1. organic matter does not contribute
as much to acidification or nutrient enrichment as SO42 and NO; . and SO42 and NO;
still account for the majority of PM2.5 mass in many locations. The remaining mass of
PM2.5 is composed of elemental carbon, sea salt (mostly Na and CI), and crustal material
(Si and A1 are most abundant elements). Monitoring methods are described and spatial
and temporal trends for PM2.5 species are further developed in the 2019 PM ISA (U.S.
EPA. 2019). There is much less information on PM10 or PM10-2.5 composition because of
the lack of routine monitoring on the scale that has been implemented for PM2.5 species.
2-27
-------
Figure 2-5 Contributions of organic carbon (OC), elemental carbon (EC),
sulfate, nitrate, sea salt, and crustal components to PM2.5 at
selected sites (A) 2003-2005 (B) 2013-2015.
2-28
-------
2.4
Concentration and Deposition Measurements
An extensive review of techniques for measuring NOx, NOz, NOy, NHX, and SOx
appeared in the 2008 ISA (U.S. EPA. 2008a). Updates to techniques for measuring NOx,
NOy, and SOx species can be found in the latest ISAs for Oxides of Nitrogen (U.S. EPA.
2016f), Sulfur Oxides (U.S. EPA. 2017d). and PM, including PM2.5, PM10, and PM10-2.5
(U.S. EPA. 2019) Health Criteria, to which the reader is referred for details. In the
following sections, measurements of NOx, SO2, and PM in national networks are only
briefly discussed and the measurement of species most relevant to acid and nutrient
deposition and measurement methods for wet and dry deposition are the main focus.
Appendix 2.4.1 explains the roles of various national and regional monitoring networks
in place to support the NAAQS or to collect data used for estimating acid and nutrient
deposition. Appendix 2.4.2. Appendix 2.4.3. and Appendix 2.4.4 describe methods used
to measure gas-phase oxides of nitrogen, reduced nitrogen, and sulfur oxides that are not
based on filter collection. Each of these sections is divided into separate discussions of
the methods used in monitoring networks, remote sensing methods, and recent advances
in research methods and other methods that are effective for intensive field studies but
impractical for routine monitoring. Satellite-based remote sensing methods are useful
because network coverage is often sparse and satellite-based measurements are becoming
a more widely used alternative to ground-based measurements. Appendix 2.4.5 describes
filter-based methods used in CASTNET and other networks for mainly particulate
species, but also for some gases, including HNO3 and SO2. Appendix 2.4.6 describes wet
and dry deposition measurement and recent advances.
2.4.1 Monitoring Networks
Federal Reference Methods (FRMs) have been established and national monitoring
networks put in place for NO2 as the indicator of oxides of nitrogen, SO2 as the indicator
of sulfur oxides, and PM2.5 and PM10 as indicators for PM. These methods and networks
are described in detail in recent ISAs for Oxides of Nitrogen (U.S. EPA. 2016f). Sulfur
Oxides (U.S. EPA. 2017d). and Particulate Matter (U.S. EPA. 2019) Health Criteria.
However, in general the large fractions of N and S deposition accounted for by species
other than NO2 and SO2 make measurements of these indicator species alone inadequate
for estimating deposition amounts of total oxides of nitrogen and total sulfur oxides.
Similarly, it has long been established that wet deposition of S is usually dominated by
S042 rather than SO2 (Dana. 1980). In this respect, PM2.5 monitoring is potentially useful
because it efficiently collects the range of PM species involved in acidification and N
deposition. However, variability of SO42 and NO;, as a fraction of total PM2 5 presents a
2-29
-------
challenge for relating PM2 5 mass to these effects. An additional challenge is the estimate
of deposition from PM2.5 concentrations, because of the strong dependence of deposition
flux on particle size (see Appendix 2.5.2) and general unavailability of size distribution
measurements.
In principle, a multipurpose, multipollutant monitoring network could efficiently meet the
needs of estimating N and S deposition and air monitoring for ecosystem protection,
while at the same time addressing other national air monitoring priorities. Such a network
could include measurements of other N and S species besides NO2 and SO2 as well as
other species that are otherwise not routinely monitored to better understand a variety of
air pollution processes. This is the overall concept behind the National Core Network
(NCore), a newly developed multipollutant monitoring network, and measurements of
NOy and NH3 were included as NCore monitoring in part because of their relevance to
atmospheric deposition (Scheffe et al.. 2009). NCore has been operating since January 1,
2011 and has 80 monitoring sites designed for measuring multiple pollutants (We in stock.
2012). The network provides a core of sites that measure SO2, NO2, NOy, and PM
components including ammonium, nitrate, and sulfate, but with sparser coverage than the
FRM networks for SO2 or NO2. Because NOy is measured rather than NOx, and because
of collocated SO2 and SO42 measurements, ambient concentrations of both NOy and SOx
can be determined from NCore data, so that these data can be used to estimate total
deposition of oxides of nitrogen and sulfur. However, because of the wide range of
deposition velocities for different species, NOy measurements alone are not sufficient for
estimating deposition and species concentrations are also necessary. A further
disadvantage is that most NCore sites are located in urban areas.
Instead of using NCore or the national NO2, SO2, and PM monitoring networks, national
scale N and S deposition have relied on monitoring networks specifically designed for
estimating deposition. Table 2-3 lists monitoring networks that have been used for recent
estimates of atmospheric N and S deposition for the NADP (Schwede and Lear. 2014a).
2-30
-------
Table 2-3 Summary of monitoring networks used by Schwede and Lear
(2014a).
Network
Chemical Species
Period of Record
Website
CASTNET
Concentration: HNO3,
SO2, pSCM2"3, pN03~a,
pNH4+a
2000-2012
httD://eDa.aov/castnet/iava web/index, html
AMoN
Concentration: NH3
2008-2012
http://nadD.slh. wisc.edu/amon/
SEARCH
Concentration: HNO3,
SO2, NH3
2005-2011
No longer in operation
NTN
Wet deposition:
S042", NO3-, NH4+
2000-2012
http://nadD.slh.wisc.edu/ntn/
apS042- is particulate sulfate concentration, pN03" is particulate nitrate concentration, pNH4+ is particulate ammonium concentration.
Note: Summary of data from monitoring networks used in the methodology.
Source: Schwede and Lear (2014a1.
Wet deposition is estimated as the product of pollutant concentration in precipitation and
precipitation depth (e.g., in rain or snow). Concentration in precipitation is currently
measured as a weekly average by the National Atmospheric Deposition Program/National
Trends Network (NADP/NTN) across a national network of 250 sites using a standard
precipitation collector described in the 2008 ISA (U.S. EPA. 2008a). The NADP
precipitation network was initiated in 1978 to collect data on amounts, trends, and
distributions of acids, nutrients, and cations in precipitation. It expanded to meet the
needs of the National Acid Precipitation Assessment Program established in 1981 to
understand causes and effects of acid precipitation. The NTN is the only network
providing a long-term record of precipitation chemistry across the U.S. Sites are mainly
located away from urban areas and pollution sources. An automated collector ensures that
the sample is exposed only during precipitation (wet-only sampling). Species measured
are free acidity (H+ as pH), conductance, calcium (Ca2+), magnesium (Mg2+), sodium
(Na+), potassium (K+), sulfate (SO42 ). nitrate (NO, ). chloride (CO, and ammonium
(NH4+). Relatively high confidence has been assigned to wet deposition estimates because
of established capabilities for measuring relevant chemical components in precipitation
samples (U.S. EPA. 2011a). The Atmospheric Integrated Research Monitoring Network
(AIRMoN) started in 1992 and measures the same species as the NTN, but on a daily
rather than weekly basis.
In contrast, direct measurements of dry deposition flux are rare and difficult, and dry
deposition fluxes of gases and particles are estimated from concentration measurements
by an inferential technique described in the 2008 ISA (U.S. EPA. 2008a). In the
2-31
-------
inferential model approach, the deposition of a pollutant is accomplished by introducing a
resistance component to account for the individual chemical and biological processes that
control pollutant adsorption and capture at natural surfaces (Hicks et al.. 1987).
Concentrations are measured in the Clean Air Status and Trends Network (CASTNET),
which was established under the 1991 Clean Air Act Amendments to assess trends in
acidic deposition. CASTNET is a long-term environmental monitoring network with
95 sites located throughout the U.S. and Canada, managed and operated by the U.S. EPA
in cooperation with other federal, state, and local partners (www.epa. gov/castnet)
including six Native American tribes. CASTNET is the only network in the U.S. that
provides a consistent, long-term data record of acidic dry deposition fluxes. It
complements the NTN, and nearly all CASTNET sites are collocated with or near an
NTN site. Together, these two monitoring programs are designed to provide data
necessary to estimate long-term temporal and spatial trends in total deposition (dry and
wet) as well as ecosystem health. Species measured in CASTNET include: O3, SO2,
HNO3 in the gas phase and SO42 . NO;, . NH/, Ca2+, Mg2+, K+, Na+, and Cl~ in particles.
While CASTNET data are more useful for estimating dry deposition than data from FRM
networks, monitors are generally sparse and deposition is only determined for discrete
locations. Also, not all of the species that contribute to total sulfur and nitrogen
deposition are measured in CASTNET (Schwede etal.. 2011). Despite these
disadvantages, CASTNET data still be very useful if used in combination with modeled
data (Schwede et al.. 2011). NH3 is not measured in CASTNET, but the National
Atmospheric Deposition Program (NADP) deployed a separate NH3 monitoring network
(AMoN) using Radiello® passive samplers starting in the fall of 2007 at 16 sites;
currently there are more than 60 active AMoN sites, two-thirds of which are located at
CASTNET sites.
A limitation of dry deposition derived from CASTNET and other dry deposition
networks is that results cannot be spatially interpolated because of the complexity of the
deposition field (Schwede and Lear. 2014a; Baumgardner et al.. 2002). Combined with
the sparse coverage of the network, this complexity restricts the capability of routine
monitoring networks to provide data on dry deposition. To some extent, this limitation
can be addressed by considering data from other networks.
The remaining network in Table 2-3 is the Southeastern Aerosol Research and
Characterization network (SEARCH), which was a highly instrumented network of four
urban and four rural stations in Alabama, Florida, Georgia, and Mississippi that was
terminated in 2016 (Hansen et al.. 2003). The four rural SEARCH sites have been used
for dry deposition estimates. SEARCH began as a public-private collaboration in early
1998 and has continued operation with several objectives, including understanding
2-32
-------
processes governing PM2 5 and copollutants emissions, transport, and deposition in the
southeastern U.S.
One additional network that has been identified as potentially suitable for use in future
deposition estimates (Schwede et al.. 2011) is the Interagency Monitoring of Protected
Visual Environments (IMPROVE) network. The IMPROVE network consists of more
than 100 monitoring sites in national parks and other remote locations and is primarily
focused on visibility impairment, but has also provided a reliable, long-term record of
particulate mass and species components. Several other monitoring networks are operated
either by the U.S. EPA or jointly with other federal agencies; species measured and other
details for networks making measurements relevant for deposition are shown in Naess
(2016). Even if concentration data from other networks are combined with CASTNET
data, large areas of the U.S. are still relatively far away from, or in a different
environment than, the nearest monitor.
Another deficiency of both the NTN and CASTNET is that not all species that contribute
to total sulfur and nitrogen deposition are measured. Reliable measurements of NOy and
NO2 concentrations, especially at the low concentrations observed in many areas far from
sources, are crucial for evaluating the performance of three-dimensional, chemical
transport models of oxidant and acid production in the atmosphere. To meet this need,
NOy monitors have been installed at six sites in CASTNET as part of the NCore
program. At most sites, however, NO2 is not currently monitored in CASTNET. The
same is true for HNO2 and peroxyacyl nitrates, which can also contribute significantly to
total gas-phase reactive nitrogen. These species can be important contributors to N
deposition locally, especially near populated areas. Neither the NTN nor CASTNET
monitor reduced organic nitrogen compounds, which can also contribute significantly to
N deposition (see Appendix 2.3.4). The sparse geographic coverage and lack of
measurements for key species in these networks along with the awareness of modeling
uncertainties led to the initiation of the Total Deposition Science Committee [TDEP;
NADP (2016)1 to develop hybrid approaches to improve estimates of atmospheric
deposition. The TDEP approach is described in Appendix 2.6.
2.4.2 NO2, NOx, and NOy
2.4.2.1 Network Monitoring
As described in Appendix 2.4.1. a nationwide monitoring network is in place for routine
monitoring of NO2, and NOy is measured in the nationwide NCore network. NO2 is
2-33
-------
routinely measured using the FRM chemiluminescence method based on the catalytic
reduction of NO2 to NO, followed by reaction NO with O3. However, the reduction of
NO2 to NO on the MoOx catalyst substrate also reduces other oxidized nitrogen
compounds (i.e., NOz compounds shown in the outer box of Figure 2-3) to NO. This
interference by NOz compounds has long been recognized following Winer et al. (1974)
who found NO was also produced by catalytic reduction of HNO3, PAN, and organic
nitrates using this method. As a result of their experiments, Winer et al. (1974) concluded
that, "the NOx mode of commercial chemiluminescent analyzers must be viewed to a
good approximation as measuring total gas-phase 'oxides of nitrogen," not simply the
sum of NO and NO:.' Numerous later studies, as noted in the ISA for Oxides of Nitrogen
(U.S. EPA. 2016f). have confirmed this conclusion. Further details were also described in
the 2008 ISA (U.S. EPA. 2008a).
Commercially available NOx monitors have been converted to NOy monitors by moving
the molybdenum oxide catalyst substrate to interface directly with the sample inlet to
improve the efficiency of reduction of NOz compounds susceptible to loss on inlet
surfaces. NOx concentrations cannot be considered as a universal surrogate for NOy.
However, near sources of fresh combustion emissions, such as highways, most of the
NOy is present as NOx. To the extent that all the major oxidized nitrogen species can be
reduced quantitatively to NO, measurements of NOy concentrations should be more
reliable than those for NOx concentrations, particularly at typical ambient levels of NO2.
Exceptions might apply in locations near NOx sources, where NOx measurements are
likely to be less biased and confidence in measurement accuracy increases.
2.4.2.2 Remote Sensing
Satellite-based methods have also been used to measure NO2. Remote sensing by
satellites is especially useful in areas where surface monitors are sparse. Retrieving NO2
column abundances from satellite data typically involves three steps: (1) determining the
total NO2 integrated line-of-sight (slant) abundance by spectral fitting of measurements
of backscattered solar radiation, (2) removing the stratospheric contribution by using data
from remote regions where the tropospheric column abundance is small, and (3) applying
an air mass factor to convert tropospheric slant columns into vertical columns. The
retrieval uncertainty is largely determined by steps 1 and 2 over remote regions where
there is little tropospheric NO2, and by step 3, over regions of elevated tropospheric NO2
(Boersma et al.. 2004; Martin et al.. 2002). Satellite retrievals are largely limited to cloud
fractions <20%. A hybrid approach using data for NO2 tropospheric column abundances
obtained by the Ozone Monitoring Instrument (OMI) on the Aura satellite coupled with
results from the GEOS-Chem, global-scale, three-dimensional, chemistry-transport model
2-34
-------
has been developed by Lamsal et al. (2008) with updates by Lamsal et al. (2010). In this
approach, the surface mixing ratio divided by the column abundances calculated by the
GEOS-Chem model are used as the scaling factors to derive surface mixing ratios from
satellite-measured column abundances. This method provides estimates of surface NO2
concentrations that are especially useful in data-sparse regions. The algorithm used to
derive the tropospheric columns of NO2 is given in Bucsela et al. (2013). Note that this
algorithm was recently shown to produce NO2 column abundances that are too high by
-20% (Marchenko et al.. 2015). There have also been advances in satellite remote
sensing of NH3 with the implementation of Cross-Track Infrared Sounder Satellite
measurements (Kharol et al.. 2018).
2.4.2.3 Research and Nonroutine Methods
Alternatively, multiple methods for observing components of NOy have been developed
and evaluated in some detail. As a result of these methods, as applied in the field and the
laboratory, knowledge of the chemistry of odd-N species has evolved rapidly. Recent
evaluations of methods can be found in Arnold et al. (2007) for HNO3, Wooldridge et al.
(2010) for speciated PANs, and Pinto et al. (2014) for HONO. However, it is worth
reiterating that the direct measurements of NO are still the most reliable of all.
2.4.3 Ammonia
2.4.3.1 Network Monitoring
The recently implemented AMoN for monitoring ammonia was described in
Appendix 2.4.1. The passive sampling method relies on diffusion across a membrane
onto an absorbing substrate, which for NH3 is H3PO3. The sampling period in AMoN is
2 weeks. Puchalski et al. (2011) compared the results from three passive samplers with
annular denuder systems (taken to be the reference method). The median relative
percentage difference between the Radiello passive samplers and the denuder systems
was -37% and the coefficient of variation among triplicate Radiello samplers was 10%.
Puchalski et al. (2015) further compared 2-week samples collected at five sites over the
course of a year by Radiello passive samplers with collocated annular denuder systems
(ADS) with different configurations. The mean relative percentage difference between
the ADS and AMoN samplers was -9% to be compared to a precision of 5% for both the
ADS and AMoN samplers. Ammonia was also been measured as a part of the SEARCH
2-35
-------
network (see Appendix 2.4.1) from 2004 to 2016 by 24-hour collection on citric acid
denuders and laboratory analysis with U.S. EPA method 350.1, based on
chemiluminescence from reaction with indophenol (Edgerton et al.. 2007).
2.4.3.2 Remote Sensing
In addition to these in situ techniques, remote sensing techniques have also been used to
measure NH3. The Tropospheric Emission Spectrometer (TES) on the Aura satellite
rShephard et al. (2011). and references therein; Beer et al. (2008)1 and the Infrared
Atmospheric Sounding Interferometer (IASI) on the MetOp-A satellite (Van Damme et
al.. 2014) measured spectral features in the v2 vibrational band centered at around
950/cm (the so-called atmospheric window in the infrared). Operating specifications for
TES (spectral resolution, 0.06/cm; footprint 5.3 x 8.3 km2; 0.15-0.20 K noise) are
generally better than for IASI (spectral resolution, 0.50/cm; footprint 12 x 12 km2;
0.15-0.20 K noise). Although TES has higher spectral resolution, it has less dense spatial
coverage. Unlike satellite detection of atmospheric molecules by backscattered solar
radiation (e.g., NO2 and SO2), NH3 is detected in the thermal infrared spectral range, so
data for both day and night can be obtained (satellite overpasses at the Equator at
approximately 1:30 a.m. and 1:30 p.m. for TES and 9:30 a.m. and 9:30 p.m. for IASI).
The sensitivity of the IR sounding technique for NH3 increases with the thermal contrast
between the surface and the temperature of the air in the lower troposphere, and thus the
daytime crossing allows for increased detectability of NH3 (Clarisse et al.. 2010). NH3 is
confined largely to the planetary boundary layer (PBL), with much lower concentrations
aloft in the free troposphere. TES retrievals are most sensitive to NH3 at atmospheric
pressures between 700 and 900 hPa. The TES level of detectability for NH3 is given by a
profile with a peak concentration of 1 ppbv, or equivalently a constant mixing ratio of
0.4 ppbv distributed over the pressure range of maximum sensitivity, provided there is
substantial thermal contrast.
Pinder et al. (2011) found that TES retrievals of NH3 in the PBL captured the spatial and
seasonal variability of NH3 over eastern North Carolina measured by surface
observations. Similarly, Sun et al. (2015a) found that column abundances measured by
TES over the San Joaquin Valley agreed with those measured by upward looking
instruments at the surface to within 2% and to within 6% for aircraft measurements. TES
columns were also shown to be reasonably well correlated (R2 = 0.67) with median NH3
measured at the surface by quantum cascade laser, thereby demonstrating the ability of
the satellite-based measurements to capture spatial variability in NH3 between individual
pixels.
2-36
-------
Alvarado et al. (2011) derived emissions factors for NH3 in biomass burning plumes over
Canada using data from TES. Using data from IASI R'Honi et al. (2013) found total
column abundances for NH3 in the plumes of Russian wildfires during the summer of
2010 that were two orders of magnitude larger than background values.
2.4.3.3 Research and Nonroutine Methods
Ambient instruments with much higher time resolution were compared by Schwab et al.
(2007) and von Bobrutzki et al. (2010). Schwab et al. (2007) conducted a
laboratory-based intercomparison of ambient NH3 instruments with seven instruments
using six methods sampling from a common manifold, including tunable diode laser
(TDLAS) absorption spectrometer, wet scrubbing long-path absorption photometer
(LOPAP), wet effusive diffusion denuder (WEDD), ion mobility spectrometer (IMS),
Nitrolux laser acousto-optical absorption analyzer, and a modified CL analyzer. Schwab
et al. (2007) reported that all instruments agreed to within -25% of the expected
calibration value, with the exception of the CL analyzer which suffered from problems
related to its MoOx conversion of NOz to NO.
von Bobrutzki et al. (2010) conducted a field intercomparison of ambient NH3
measurements with 11 instruments using 8 methods including 3 wet techniques (annular
rotating batch denuders, 1 with offline analysis and 2 with online analysis [AMANDA,
AiRRmonia]), 2 Quantum Cascade Laser Absorption Spectrometers (c-QCLAS,
DUAL-QCLAS), 2 photo-acoustic spectrometers, a cavity ring down spectrometer, a
chemical ionization mass spectrometer, an ion mobility spectrometer, and an open-path
Fourier transform infrared spectrometer. This study was unique in that the surrounding
field was fertilized with urea halfway through the campaign to increase average
concentrations of NH3 from 10 to 100 ppb. Overall, R2 was >0.84 with respect to the
ensemble mean for all instruments over the entire range of concentrations (<120 ppb),
with slopes ranging from 0.67 to 1.13. Higher variability was found at lower
concentrations (<10 ppb) with R2 > 0.52 and slopes ranging from 0.42 to 1.15. Perhaps
the most consistent agreement between two instruments was found for the c-QCLAS and
AiRRmonia (R2 = 0.91, slope = 0.86, intercept = 0.84 ppb forNH3 <10 ppb; and
R2 = 0.91, slope = 0.83, intercept = 0.34 ppb over the entire range ofNHa concentrations).
Nitrolux-100 denuders were used both in the intercomparison study of von Bobrutzki et
al. (2010) and in the one by Puchalski et al. (2011). Compared to the ensemble mean, the
slope was 0.97, the intercept was 1.86 ppb, and R2 = 0.98 over the entire concentration
range. As noted by von Bobrutzki et al. (2010). comparisons of this sort only show
relative performance of the instruments and not a functional relationship to a standard.
2-37
-------
2.4.4
Sulfur Dioxide
2.4.4.1 Network Monitoring
In the nationwide monitoring network for SO2, SO2 is routinely measured by pulsed UV
fluorescence. This technique is a Federal Equivalence Method (FEM). The method's
principles and potential inteferences have been described in detail in the 2008 ISA (U.S.
EPA. 2008a). However, measurements using this method are not as widely used as
measurements from the CASTNET. In the CASTNET, SO2 is collected by capturing it on
filters and measured as sulfate following procedures described in Appendix 2.4.5.
2.4.4.2 Remote Sensing
In addition to the above in situ methods, satellite-based measurements have also been
used to measure tropospheric SO2 and to infer surface SO2 concentrations with the aid of
the GEOS-Chem chemistry-transport model (Nowlan et al.. 2014; Lee et al.. 2011).
Tropospheric column abundances of SO2 are obtained by the Ozone Monitoring
Instrument (OMI) on the Aura satellite or the Scanning Imaging Absorption Spectrometer
for Atmospheric Cartography (SCIAMACHY) on Envisat and are combined with results
from the GEOS-Chem, global-scale, three-dimensional, chemistry-transport model to
derive surface concentrations of SO2 (as they are for NO2). Lee et al. (2011) associated
annual mean surface mixing ratios of SO2 derived from the hybrid satellite/model
technique with ambient measurements of SO2, (R2 = 0.66 and 0.74, slope = 0.70 and 0.93,
n = 121 and 115, for OMI and SCIAMACHY, respectively).
The algorithms used to derive vertically integrated SO2 abundances in the troposphere
undergo continuing refinement. For example, Thevs et al. (2015) applied an algorithm
based on differential optical absorption spectroscopy (Piatt and Stutz. 2008) combined
with a radiative transfer model. Li et al. (2013) developed an algorithm based on
principal components analysis, which has replaced the earlier standard algorithm
developed by Krotkov et al. (2008). The methods applied by Thevs et al. (2015) and Li et
al. (2013) are designed to be operational for retrieving SO2 in the PBL with an estimated
detection limit of -0.5 Dobson units (1 DU = 2.69 x 1016 molec/cm2 corresponding to a
concentration of ~3 ppb if SO2 is well mixed in a 2-km-deep mixed layer), or about half
that in the older standard method. Table 2-4 summarizes sources of uncertainty for
individual OMI measurements of NO2 and SO2.
2-38
-------
Table 2-4 Sources of uncertainty for individual Ozone Monitoring Instrument
measurements in the studv of Nowlan et al. (2014).
Source
NO2
SO2
OMI fitting error
~1015 molec/cm2
-1016 molec/cm2
Air mass factor
20%
15-45%
Stratospheric correction
2 x 1014 molec/cm2
N/A
SO2 offset correction
N/A 2
x 1014 molec/cm2
Profile shape
30%
10%
molec = molecules; N/A = not applicable; N02 = nitrogen dioxide; OMI = Ozone Monitoring Instrument; S02 = sulfur dioxide.
Source: Nowlan et al. (20141.
The errors in the column measurements result mainly from uncertainties in the vertical
profiles of NO2 and SO2, cloud fraction, cloud pressure, surface reflectivity, and particles
used in the calculation of air mass factor. A correction is required to account for NO2 in
the stratosphere (produced from N2O oxidation and cosmic ray interactions dissociating
with N2). The SO2 offset correction refers to a global background correction arising from
issues in spectral fitting, such as spectral correlations with O3 and stray light within the
instrument.
2.4.5 Filter-Based Concentration Measurements
As described in Appendix 2.4.1. most measurements used for estimating deposition are
from CASTNET, rather than from monitoring networks based on FRM and FEM
methods. The CASTNET filter pack is shown in Figure 2-6. Particulate matter is
collected on the open-face Teflon filter, extracted in deionized water, and analyzed by ion
chromatography (IC) for sulfate, nitrate, ammonium, and other species identified in
Figure 2-6. In the CASTNET filter pack, gases are collected downstream of the
particulate species, with nitric acid collected on nylon filters and analyzed as NO;, by ion
chromatography, and SO2 on carbonate impregnated cellulose filters and analyzed as
S042 by ion chromatography. Extensive intercomparisons of CASTNET methods with
other measurement methods were described in the 2008 ISA (U.S. EPA. 2008a).
2-39
-------
Figure I-4 Three-Stage Filter Pack
Cellulose (2)
Nylon
Teflon®
Gaseous
Gaseous
Particulate
•so2
• hno3 • so2
• so2 • no3 • nh; • K*
• Ca2* • Mg2* • Na* • CI"
Quick Disconnect
Two
Cellulose Nyton Teflon8
Filters Filter Filter
Teflon* Spacers
Ca2+ = calcium ion; CI" = chloride; HN03 = nitric acid; K+ = potassium ion; Mg2+ = magnesium ion; Na+ = sodium ion;
NH4+ = ammonium; N03" = nitrate; S02 = sulfur dioxide; S042" = sulfate.
Air flows from right to left.
Source: MACTEC (2010V
Figure 2-6 Clean Air Status and Trends Network filter pack.
As can be seen in Figure 2-6. SO2 is measured by the CASTNET filter pack by IC
analysis of extracts from the cellulose filters. Because the nylon filter adsorbs some of the
SO2 (Sickles et al.. 1999; Sickles and Hodson. 1999). SO42 is also measured on nylon
and added to the SO2 (expressed as SO42 ) collected on the backup cellulose-fiber filters.
Uncertainties in CASTNET data are reported quarterly in a quality assurance report (U.S.
EPA. 2016c). Precision is determined as the absolute value of quarterly or annually
aggregated relative percentage difference for duplicate sample pairs collected with
collocated samplers at two sites. Data quality objectives for ammonium, nitrate, and
sulfate are within 20%, but reported precision for 2016 was well under this target, 2-5%
for sulfate, 5-13% for nitrate, and 2-6% for ammonium, where the range reflects that
there are two sites. Analytical accuracy was reported within 2% based on spiked
calibration verification samples. Further detail on uncertainty and data quality can be
found in (U.S. EPA. 2016c). Additional unknown uncertainty is associated with
volatilization of NFU+ from collected PM. Substantial loss during sampling can occur
because collected NH4NO3 in PM is in a temperature dependent equilibrium with NH3
and HNO3 (see Appendix 2.3.3). leading to volatilization of both species after PM
2-40
-------
collection. Loss of NH44" during sampling was thoroughly reviewed in the 2008 ISA (U.S.
EPA. 2008a).
Results of an intercomparison of weekly average SO2 data (ppbv) collected by the
CASTNET filter pack and trace level SO2 monitors during all of 2014 at Bondville, IL
and Beltsville, MD are shown in Figure 2-7 (AMEC Environment & Infrastructure.
2015).
In addition to CASTNET, pSC>42~, pNO, are monitored in the Chemical Speciation
Network (CSN), and the Interagency Monitoring of Protected Visual Environments
(IMPROVE) network. Sampling and measurement methods for these networks were
described in detail by Solomon et al. (2014). In the CSN network, pNH/ is also
measured, but as for CASTNet, it may be subject to volatilization error (volatilization of
NO3 is corrected).
2-41
-------
Trace UV Fluorescence S02
0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5
Trace UV Fluorescence SCI;
Filter Pack vs Trace U V Fluorescence S02
Beltsville, MD
±0.05
Filter Pack vs Trace UV Fluorescence S02
Bondville, IL
±0.07
S02 = sulfur dioxide; UV = ultraviolet.
Results for Beltsville, MD are shown in the upper panel and results for Bondville, IL are shown in the lower panel.
Figure 2-7 Comparison between weekly average measurements of sulfur
dioxide using the Clean Air Status and Trends Network filter pack
and the trace ultraviolet pulsed fluorescence monitor in 2014.
2-42
-------
2.4.6 Deposition Measurements
Wet deposition is estimated as the product of pollutant concentration in precipitation and
precipitation depth (e.g., in rain or snow; see Appendix 2.4.1). Concentration in
precipitation is currently measured by the National Atmospheric Deposition
Program/National Trends Network (NADP/NTN) across a national network of sites using
a standard precipitation collector described in the 2008 ISA (U.S. EPA. 2008a).
Relatively high confidence has been assigned to wet deposition estimates because of
established capabilities for measuring relevant chemical components in precipitation
samples (U.S. EPA. 2011a). Measurement precision determined as the average absolute
percentage differences of replicate samples in the 2016 annual quality assurance report
was 1% or less for sulfate, nitrate, and ammonium, and absolute percentage difference
was no greater than 5% for sulfate and ammonium or 7% for nitrate for any single sample
pair (U.S. EPA. 2016c). Bias determined from internal blind samples was 2% or less for
sulfate and nitrate, and 6% or less for ammonium (U.S. EPA. 2016c).
In contrast, direct measurements of dry deposition flux are rare and difficult. Methods for
estimating dry deposition from field measurements fall into two major categories: surface
analysis methods, which include all types of estimates of contaminant accumulation on
surfaces of interest, and atmospheric deposition rate methods, which use measurements
of contaminant concentrations in the atmosphere and micrometeorological measurements
of atmospheric turbulence (U.S. EPA. 2008a). Emphasis here is placed on the latter class
of methods, which are more widely applicable because the accumulation methods are
subject to limitations such as the site specificity of the measurements and the restriction
to elements that are largely conserved within the vegetative system. Dry deposition
estimates using atmospheric deposition rate methods are based on field measurements of
a species or particle concentration gradient along with a measurement or estimate of its
turbulent diffusivity under the field conditions of the measurement (Mvles et al.. 2012;
Businger. 1986). Examples include eddy covariance and aerodynamic gradient
techniques (U.S. EPA. 2008a). In the eddy covariance method, flux is calculated from the
covariance between fluctuations in wind velocity and concentration. Historically,
empirical estimates of deposition for wind tunnel and field conditions often have not
agreed well with theoretical predictions, probably because transport phenomena and
turbulence structure near surfaces are not well characterized (U.S. EPA. 2004). However,
improvement of dry deposition measurements is an active research area, both to reduce
measurement uncertainties and to improve modeling capabilities by better understanding
deposition processes and their parameterization in chemical transport models.
Progress in improving measurement capabilities has been triggered by the development
of continuous air sampling measurement techniques with higher sensitivity and temporal
2-43
-------
resolution, with the objective of improving costs and measurement quality using
atmospheric deposition rate measurement methods. Concluding that the main reason for
the shortage of dry deposition measurements is the expense and complexity of
measurement methods, Almand-Hunter et al. (2015) tested a dynamic flux chamber using
automated, inexpensive multispecies gas flux monitoring system for measurement of a
variety of pollutants, including NOx, that would be needed to contend with the large
spatial and temporal variability in air-surface exchange rates of reactive compounds. A
dynamic chamber system was also developed to allow measurements of NO, NO2, and O3
measurements so that compensation points and deposition velocities (see Appendix 2.5.2)
could be estimated (Breuninger et al.. 2012).
As part of a series of measurements to be tested at several CASTNET sites, Rumsev and
Walker (2016) evaluated the ability of the MARGA 2S (Monitoring for AeRosols and
GAses) system to simultaneously measure fluxes of NH3-HNO3-NH4NO3 using the
aerodynamic gradient method to allow for an assessment of the errors due to the
instability of the particle phase, as well as SO2 and SO42 . to allow for the investigation of
ammonium sulfate neutralization and codeposition between SO2 and NH3. Over a range
of meteorological conditions, median flux uncertainty was found to range from ~3 1 % for
NH3 to ~ 120% for NH4+. The flux gradient technique was also applied to a forest clearing
as an example of a complex ecosystem with the objectives of improving deposition rate
accuracy and model parameterization for SO2 (Mvles et al.. 2012). Deposition velocities
fluctuated considerably with a mean of 1.00 ± 0.48 cm/s, and the large variation was not
fully captured by estimates from widely used models (Mvles et al.. 2012). Uncertainties
in canopy resistance, including stomatal and nonstomatal processes were identified as
probable sources of uncertainty (Mvles et al.. 2012).
This approach has been successfully applied to measurements of a wide variety of
species. However, it is less suitable for HNO3 and NH3 because they are subject to
interactions with inlet surfaces of measuring devices. For example, in one study a
correction factor of 1.62 was reported for inlet surface interactions (Breuninger et al..
2012). To address the problem of loss of HNO3, NH3, and other substances that interact
with sampling inlet surfaces during measurement, Roscioli et al. (2016) developed a
method to passivate inlet surfaces and thereby overcome this difficulty by allowing for
more rapid response measurements. Min et al. (2014) determined the net flux of NO and
NO2 (at a frequency of 5 Hz) over a forest with estimated total systematic uncertainties of
<8 and <6%, and random errors of <25 and <21%, respectively.
Measurement methods are well developed for ideal conditions of flat, homogeneous, and
extensive landscapes and for chemical species for which accurate and rapid sensors are
available (U.S. EPA. 2008a). However, the strong dependence of dry deposition on
2-44
-------
surface characteristics, which are highly spatially variable, causes in situ measurements
to be limited in applicability, especially in areas of high topographic relief [e.g., Weathers
et al. (2006)1. The lifetime in the boundary layer (~1 km depth) for rapidly depositing
species is ~1 day, implying that distances from source to deposition are relatively short
and that the deposition flux depends strongly on the nature and strength of nearby
sources. Deposition also shows strong temporal variability on timescales ranging from
diurnal to seasonal, as do wind regimes affecting a particular site. In addition, cost and
logistics make the eddy covariance and aerodynamic gradient techniques impractical for
monitoring networks.
Instead, dry deposition fluxes of gases and particles are estimated in CASTNET and by
chemistry-transport models, such as CMAQ, by an inferential technique described in the
2008 ISA (U.S. EPA. 2008a). In the inferential model approach, the deposition of a
pollutant is estimated by introducing a resistance component to account for the individual
chemical and biological processes that control pollutant adsorption and capture at natural
surfaces (Hicks et al.. 1987). Ambient pollutant concentrations of O3, SO42 . NO;, . NH4+,
SO2, and HNO3 are routinely collected at CASTNET dry deposition sites. Deposition
velocities based on local meteorological measurements were calculated using the
Multi-Layer Model (MLM) at U.S. EPA-sponsored CASTNET sites until 2010, when
meteorological measurements were discontinued at all but five U.S. EPA CASTNET
sites. Dry deposition fluxes are still reported at sites with discontinued meteorological
measurements using historical data. A disadvantage to this approach is that relevant
atmospheric species are not routinely measured. For example, NO2 and peroxyacetyl
nitrate, which together consistently contribute 15 to 25% of estimated oxidized nitrogen
dry deposition, are not measured at CASTNET sites. Even for those species that are
routinely measured, network spatial coverage is sparse (U.S. EPA. 2011a).
Satellite-based measurements offer a potential means of greater coverage, but only a
limited number of deposition related applications have been described. Nowlan et al.
(2014) combined satellite data with modeled NO2 and SO2 surfaces and vertically
integrated concentrations and deposition velocities to estimate deposition fluxes. SO2 dry
deposition fluxes compared well with surface network-based deposition fluxes.
Uncertainties in depositional flux estimates in this approach result from the combined
uncertainties in the satellite-derived surface concentrations and model-derived deposition
velocities and were estimated to be -30% on average for both NO2 and SO2 over land. In
the absence of routine measurements, dry deposition is often modeled with CMAQ or
other modeling tools using relevant emissions, meteorological, and land use data, rather
than estimated from measured concentration measurements (U.S. EPA. 2011a).
2-45
-------
Modeling dry deposition is particularly challenging over varying terrain and under
extremely stable conditions such as those occurring at night. Under optimal conditions
over a relatively small area where dry deposition measurements have been made,
uncertainties on the order of ±30% have been reported and larger uncertainties are likely
when the surface features in the built environment are not well known or when the
surface comprises a patchwork of different surface types, as is common in the eastern
U.S. (U.S. EPA. 2008a). For this reason, dry deposition is routinely estimated from
concentration measurements, usually from CASTNET data (Appendix 2.4.5) using a
hybrid approach based on both measured and modeled data (Appendix 2.6).
2.5 Modeling Chemistry, Transport, and Deposition
In this section, advances in modeling transport and deposition of species relevant to acid
and nutrient deposition are discussed, along with research progress on understanding
underlying transport and deposition processes. The use of chemical transport models
(CTMs) to model deposition was discussed extensively in the 2008 ISA (U.S. EPA.
2008a). Relevant new research and improvements in CTM modeling in general are
described in Appendix 2.5.1. Appendix 2.5.2 discusses environmental processes relevant
to understanding and modeling acid and nutrient deposition. Appendix 2.5.2.1 begins
with an overview of fundamental processes of atmospheric deposition of gases and
particles, along with deposition velocities for some key gas-phase species. It also contains
discussions of research advances in three key processes that serve as major structural
uncertainties (lack of knowledge of the underlying science) in modeling deposition: NOx
canopy processes, which involve both bidirectional gas exchange and NOx chemistry
(Appendix 2.5.2.2); bidirectional exchange of ammonia (Appendix 2.5.2.3); and
transference ratios relating average ambient concentration to deposition flux
(Appendix 2.5.2.4). Appendix 2.5.3 discusses model evaluation and uncertainty,
including comparisons between CTM and network-based wet deposition results.
2.5.1 Advances in Chemistry-Transport Model (CTM) Modeling
To understand the lifetime and fate of the varied forms of atmospheric sulfur and nitrogen
from emission to deposition, it is necessary to account for both atmospheric transport and
chemical transformations. CTMs simulate the relevant atmospheric transport processes
(e.g., horizontal and vertical advection and diffusion), as well as chemistry, aerosol
physics, deposition, and cloud processes. The 2008 ISA (U.S. EPA. 2008a) provided a
detailed description of the CTM models and their application to estimating deposition.
Continental-scale CTMs include CMAQ (Appel et al.. 2017) and CAMx (Koo et al..
2-46
-------
2015). while GEOS-Chem (Zhang et al.. 2012a) is an example of a global-scale model.
Most major regional-scale, air-related modeling efforts at U.S. EPA use the Community
Multiscale Air Quality (CMAQ) modeling system (Bvun and Schere. 2006; Bvun and
China. 1999). Recent updates to CTM model design, and in particular to CMAQ, are
described here.
A number of complex atmospheric processes influence pollutant behavior between
emission and deposition and must be taken into account to achieve good model
performance. A variety of mechanisms operating over a wide range of spatial and
temporal scales transport heat, water, and pollutants horizontally and vertically through
the atmosphere. These mechanisms range from local-scale circulations (e.g., urban heat
islands) to hemispheric-scale transport by the jet stream. Long-range transport of
pollutants in the lower free troposphere associated with large-scale synoptic systems is
possible because flows are largely uncoupled from surface friction. Flows in the upper
planetary boundary layer (PBL), especially during the day, might not be as effective for
transporting pollutants over long distances because air can be mixed down to the surface
by turbulence. If these pollutants react with surface material or are taken up by
vegetation, they can be removed within a relatively short distance from their sources.
In addition to wind velocity, the distance scale for transport of a pollutant that is
relatively stable in the troposphere with respect to gas-phase reactions (i.e., chemical
lifetime > a few days) depends strongly on the pollutant's interactions with solid and
liquid surfaces and subsequent chemical transformations. Because NO and NO2 are only
slightly soluble, they can be transported over longer distances in the gas phase than can
more soluble species like HNO3 (and its anhydride, N2O5) and NH3 that are depleted by
deposition to moist surfaces or taken up more readily on aqueous surfaces of particles or
on cloud drops. For example, measurements of the ratio of NH3 to CH4 in the San
Joaquin Valley indicate substantial loss of NH3 to the surface within a few km of sources
of these gases (Miller et al.. 2015). consistent with an NH3 lifetime of minutes to a few
hours in this environment. On the other hand, a combination of models, remote sensing,
and in situ measurements over the eastern U.S. indicate an atmospheric lifetime for SO2
of 19 ± 7 hours in summer increasing to 58 ± 20 hours in winter (Lee et al.. 2011). which
indicates the potential for much longer-range transport of SO2.
Numerous advances in atmospheric science have been codified in CTMs, including
gas-phase oxidant chemistry relevant for the formation of aerosol precursors and dry
deposition by gravitational settling (Nolte et al.. 2015). improved representation of
meteorological processes in CTMs and interactions with aerosols (Tuccella et al.. 2015).
and improved algorithms for understanding the influence of weather on emissions of
ammonia from agricultural lands (Flechard et al.. 2013). Over the U.S. and Europe,
2-47
-------
substantial reductions in sulfur dioxide and nitrogen oxides have created an opportunity
to compare the model results with the trends in ambient observations (Banzhaf et al..
2015; Xing et al.. 2015; Civerolo et al.. 2010). Studies have shown that CMAQ is skilled
at capturing the seasonal and long-term changes in sulfate PM2 5 and nitrate PM2 5 as well
as total PM2.5 mass; however, the model performs less well for seasonal variability in
nitrate PM2 5, owing to uncertainties in ammonia emission trends (Banzhaf et al.. 2015;
Xing et al.. 2015).
2.5.2 Modeling Deposition
2.5.2.1 Wet and Dry Deposition of Gases and Particles
Considerable advances in both our understanding of atmospheric deposition and
modeling approaches to characterize it have taken place recently. Deposition is a
complicated process influenced by numerous atmospheric and deposition surface
properties, as well as chemical reactions and other processes that take place within
canopies of vegetation.
In Figure 2-8. Moller (2014) illustrated the pathways that transfer gaseous and particulate
pollutants from the atmosphere to the surface by deposition. Wet deposition occurs when
particulate and gaseous species are removed by cloud drops or by falling precipitation
(washout). Dry deposition occurs when they are removed without precipitation by
processes like turbulence and gravitational settling. In mountainous areas, a third
important type of deposition occurs, referred to as occult deposition (Pollard et al..
1983). which is not shown in Figure 2-8. and results from the impaction of droplets in
fogs or clouds on vegetation. Wet deposition is determined as the simple product of
concentration in precipitation and precipitation rate. Receptor (i.e., vegetation) surface
properties have little effect on wet deposition. Dry deposition is more difficult to
determine. It can be described as a flux Fd, the mass of pollutant deposited per unit area
of the Earth's surfaces where it deposits, or a deposition velocity that relates the dry
deposition flux Fd to a pollutant's ambient concentration:
Fa = v&C
Equation 2-12
where C is the pollutant's concentration in mass in per unit volume, and Vd is the
deposition velocity, which relates a pollutant's deposition flux to its ambient
concentration.
2-48
-------
Dry deposition of gases depends on leaf area, surface resistance to gas uptake,
interactions with biota through both stomatal and plant surface pathways, and
atmospheric reactivity, which can vary among depositing gases. Measurements of
average dry deposition velocities for gases over land surfaces are shown in Table 2-5. and
an indication of the seasonal variability of deposition velocities over different land
surface types can be seen in Table 2-6. which shows the variability in Vd for SO2.
Deposition velocities of other species are also expected to be spatially and temporally
variable.
Source: Moller (20141.
Figure 2-8 Schematic diagram showing mechanisms for transferring
pollutants from the atmosphere to the surface.
2-49
-------
Table 2-5 Average dry deposition velocities (cm/s) for several gases over land
surfaces.
Substance
SO2
NO
NO2
HNOs
Os
H2O2
CO
NHs
Vd
0.8
<0.02
0.02
3
0.6
2
<0.02
1
CO = carbon monoxide; H202 = hydrogen peroxide; HN03 = nitric acid; NH3 = ammonia; NO = nitric oxide; N02 = nitrogen dioxide;
03 = ozone; s = second; S02 = sulfur dioxide; vd = deposition velocities.
Source: Moller (20141.
Table 2-6 Deposition velocity (cm/s) for sulfur dioxide averaged over different
land use types for summer and winter.
Farmland
Grassland
Dec. Forest
Con.
Forest
Urban
Water
Su
Wi
Su
Wi
Su
Wi
Su
Wi
Su
Wi
Su
Wi
Wet
1.0
1.0
1.0
1.0
3.0
1.5
2.0
2.0
1.0
1.0
0.5
0.5
Dry
0.7
0.5
0.6
0.4
1.5
0.5
0.7
0.5
0.1
0.1
0.5
0.5
Snow
-
0.
-
0.1
-
0.2
-
0.2
-
0.1
-
0.1
Con. = coniferous; Dec. = deciduous; Su = summer; Wi = winter.
Source: Moller (20141.
A wide range of deposition velocities is observed among different atmospheric gas-phase
species. HNO3 is an example of gas with straightforward deposition behavior. It sticks
easily to vegetative surfaces, i.e., there is a negligible surface resistance to HNO3 uptake
by vegetation, and its deposition rates are independent of leaf area or stomatal
conductance, implying that deposition occurs to branches, soil, and the leaf cuticle as
well as leaf surfaces. The HNO3 Vd typically exceeds 1 cm/s and exhibits a diel pattern
controlled by turbulence characteristics of midday maxima and lower values at night in
the more stable boundary layer (U.S. EPA. 2008a. 2004). In contrast, NO2 interaction
with vegetation is more complicated. The uptake rate by foliage is related to stomatal
conductance and is more variable. It may also be associated with concentrations of
reactive species such as ascorbate in the plant tissue that react with NO2. At very low
NO2 concentrations, emission from foliage is observed. Internal NO2 appears to derive
from plant N metabolism, and there is evidence for a compensation point, typically near
2-50
-------
~1 ppb, at which uptake and emission rates are equal and net flux is zero (U.S. EPA.
2008a).
Dry deposition of PM is influenced by a number of variables, including particle diameter,
atmospheric stability, deposition surface roughness, and the shape, stickiness, roughness,
and cross-sectional area of leaves. Greater roughness and leaf shape complexity increase
deposition. The diversity of particle sizes, atmospheric conditions, and surface
characteristics makes it difficult to estimate dry deposition (U.S. EPA. 2008a. 2004). The
appreciable effects of particle size, local micrometeorological conditions and surface
characteristics on deposition velocity can be seen in Figure 2-9. The key to measurements
of Vd over surfaces covered by low vegetation is given in the left column and over forest
in the right column. These measurements are compared to six model formulations which
are shown as lines in the center column.
For particles >10 (mi, Vd varies between 0.5 and 1.1 cm/s, and a minimum particle Vd of
0.03 cm/s exists for particles in the size range 0.1 to 1.0 |im. while deposition of particles
from the atmosphere to a forest canopy has been estimated as 2 to 16 times greater than
deposition in adjacent open terrain like grasslands or other low vegetation (U.S. EPA.
2004).
2-51
-------
A
~
O
o
grass (Chamberlain 67)
grass (WT) "
sticky grass (WT) "
grass (WT) (Clough 75)
moorland (Gallagher 88)
moorland (Nemitz02)
semi arid savannah (Lamaud 94)
Davidson 82 u»=45 cm.s"''
Legg 79-Aylor 82 "
Wiman 85 "
Slinn 82 "
Zhang 01 "
• Spruce (Beswick 91)
® Spruce (Bleyl 01)
O Pine (Lorenz 89)
~ Pine (Buzorius 00)
O Pine (Gaman 04)
v Pine (Lamaud 94)
o Pine (Gronholm 07)
A Fir (Gallagher 97)
01 0.1 1 10 100
dp (nm)
01 0.1 1 10
dp (nm)
0.001
100
0.001
0.
i—i—i— 1—
: Grass
dp = aerodynamic diameter of particle; vdc = deposition velocity; 1/14 = Stokes settling velocity.
Notes: Chamberlain (1967): Clough (1975): Gallagher et al. (1988): Nemitz et al. (2002): Lamaud et al. (1994b): Davidson et al.
(1982): Legg and Powell (1979): Avlor (1982): Wiman and Agren (1985): Zhang et al. (2001): Beswick et al. (1991): Lorenz and
Murphy (1989): Buzorius et al. (1998): Gaman et al. (2004): Lamaud et al. (1994a): Gronholm et al. (2007): Gallagher et al. (1997).
Closed symbols correspond to wet or sticky surfaces or liquid particles; open symbols to dry surfaces or solid particles.
Source: Petroff et al. (2008).
Figure 2-9 Modeled and measured deposition velocities over grass (left
figure) and coniferous forest canopies (right figure) for particles
of density 1 am/cm3 depositing under similar friction velocity (u*)
(35 < u* < 56 cm/s).
2.5.2.2 NOx Canopy Processes
There are a number of ways that landscape characteristics influence the deposition
process (U.S. EPA. 2008a. 2004). In terrain containing extensive vegetative canopies,
any material deposited via precipitation to the upper stratum of foliage is likely to be
intercepted by several foliar surfaces before reaching the soil. Deposition velocity (vd) is
usually greater for a forest than for a nonforested area and greater for a field than for a
water surface. The upwind leading edges of forests, hedgerows, and individual plants are
2-52
-------
primary sites of coarse particle deposition, and upper canopy foliage tends to receive
maximum exposure to coarse and fine particles, but foliage within the canopy tends to
receive primarily fine particles (U.S. EPA. 2008a. 2004). Several Nr species are deposited
to vegetation, among them HNO3, NO2, PAN (and other RONO2), and NH3.
Wet, dry, and occult deposition all contribute N and S species to the forest canopy in
varying proportions. Deposited species can react on surfaces within the canopy, be taken
up by vegetation through stomata, be resuspended during stormy weather, or simply pass
through the canopy to the forest floor. Surface characteristics can influence foliar uptake,
chemical transformation, and resuspension. Landscape characteristics can affect wet
deposition via orographic effects and by the closer aerodynamic coupling to the
atmosphere of tall forest canopies as compared to the shorter shrub and herbaceous
canopies. The rainwater that passes directly through a canopy or is initially intercepted by
aboveground vegetative surfaces and subsequently drips from the canopy is measured as
throughfall. The fraction of the precipitation that drains from outlying leaves and
branches and is channeled to the stem of plants is classified as stemflow. Throughfall and
stemflow inputs constitute the majority of incident precipitation in forests and can
account for 70 to 90% of incoming precipitation in most cases, with the remainder lost to
interception within the canopy (Levia and Frost. 2003). Compared to wet deposition
measurements in the open, the magnitude of deposition from throughfall and stemflow
can either be smaller (e.g., from evaporation from canopy surfaces) or larger (e.g., from
resuspension of previously deposited material). The type of vegetation is important for
characterizing throughfall. For example, Templer et al. (2015a) found that the cycling of
N, particularly the rate of throughfall for NFU+, is significantly different in conifer
compared to deciduous forest sites. Rainfall introduces new wet deposition and also
redistributes previously dry-deposited particles throughout the canopy. Intense rainfall
may contribute substantial total particulate inputs to the soil, but it also removes
bioavailable or injurious pollutants from foliar surfaces, while low-intensity events may
enhance foliar uptake through the hydrating of some previously dry-deposited particles
(U.S. EPA. 2004).
Chemistry within the canopy can also be important. Very fast measurements of NO2 flux
are confounded by the rapid interconversion of NO and NO2 with O3, and the biosphere
also interacts with NOx through hydrocarbon emissions and their subsequent reactions to
form multifunctional RONO2, including isoprene nitrates, which can account for a
substantial fraction of total N deposition. NO2 emissions also show UV dependence, and
both photo-induced and dark production of HNO2 from NO2 have been observed on leaf
surfaces, especially wet surfaces, although there is no consensus concerning chemical
mechanisms (U.S. EPA. 2008a). It was recently reported that NO2 deposition velocities
would have been overestimated by up to 80% if NO2 photolysis had not been considered
2-53
-------
(Breuninger et al.. 2012). Both biotic interactions and reaction chemistry further
complicate our understanding of N deposition and our ability to estimate deposition
velocity.
Some species, most prominently HNO3, can be characterized by unidirectional exchange,
whereas bidirectional exchange is more appropriate for most other species. Bidirectional
exchange is often described in terms of a compensation point, defined as the ambient
concentration above which a net uptake of the trace gas occurs and below which the trace
gas is released (Ganzeveld et al.. 2002). A two-pathway process description can be used
to describe bidirectional exchange in a forest canopy [e.g., Fowler et al. (2009); Loubet et
al. (2001)1: (1) a stomatal pathway, which is bidirectional and modeled using a stomatal
compensation point and (2) a plant surface pathway, which denotes exchange with water
or waxes on the plant surface.
Understanding the exchange of reactive nitrogen species in the forest canopy has always
been challenging. One of the most comprehensive studies focusing on this question has
been the Biosphere Effects on AeRosols and Photochemistry Experiment (BEARPEX)
conducted in 2009 (Min et al.. 2014) that examined fluxes and transformations of NOx
within a forest canopy on the western slope of the Sierra Madre Range in 2009. The
study's results, shown schematically in Figure 2-10. indicate the existence of active
chemical interactions within the forest canopy in which NO emitted from soil or
transported from elsewhere is oxidized to NO2 and then to peroxy and alkyl nitrates and
HNO3. These pathways represent alternative mechanisms to plant uptake that have the
net effect of reducing the soil NO that escapes the forest canopy as NOx is converted to
peroxy nitrates and alkyl nitrates that can be transported to the atmosphere above the
canopy on very short time scales (~100s of seconds). The fraction of NO emitted by soil
that can be lost to the atmosphere above the canopy depends on the relative time scales
for transport through the canopy versus chemical transformation and foliar uptake.
Likewise, NO or NO2 transported form elsewhere can also be oxidized to organic nitrates
within the canopy. The organic nitrates formed can either be taken up within the canopy
or transported upward through the forest canopy to act as reservoirs of NOx that can
reform downwind.
2-54
-------
HN03 = nitric acid; Uv= solar photon; OH = hydroxyl radical; NO = nitric oxide; N02 = nitrogen dioxide; NOx = sum of NO and N02;
03 = ozone; RO = alkoxy radicals; R02 = organic peroxy radicals; R0N02 = alkyl nitrates; R02N02 = peroxy nitrates.
Bold arrows in blue (downward) and red (upward) represent the direction of the flux of each species across the canopy surface. Red
thin arrows within canopy indicate the nitrogen oxides removal processes within the canopy in addition to plant uptake.
Source: Min et al. (20141.
Figure 2-10 Schematic of the interactions involved in the exchange of
nitrogen oxides between the atmosphere and the forest canopy
as identified by Min et al. (2014).
The organic nitrates consist in large measure of isoprene- and monoterpene-derived
nitrates. Nguyen et al. (2015) measured fluxes of organic nitrates on a tall tower in the
Talladega National Forest (AL) in June 2013 as part of the Southern Oxidant and Aerosol
Study (SOAS). They found that fluxes of organic nitrogen formed by reactions of
nitrogen oxides (NO, NO2, NO3) with isoprene and monoterpene oxidation products
constituted -15% of the flux of oxidized N to the forest canopy, with most of the rest
from HNO3.
2.5.2.3 Bidirectional Exchange of NH3
NH3 can also be both emitted and deposited from plants and soils in a bidirectional
exchange. Farquhar et al. (1980) observed the existence of a compensation point for
ammonia due to gas exchange through the stomata of leaves. NH3 in the stomata results
from dissociation equilibria of NFU+ produced physiologically in the leaves, followed by
equilibrium partitioning into air in the stomata (Sutton et al.. 1998). Further research
indicated that NH3 deposition rates to leaf surfaces were often faster than stomatal uptake
(Sutton et al.. 1993) and that NH3 can both react to form particulate NFU+ and evaporate
from deposited PM within the canopy (Nemitz et al.. 2004; Brost et al.. 1988). Moisture
and plant type are strong influences, because deposition is more efficient on wet surfaces
(Sutton et al.. 1995). evaporation occurs under drying conditions (Fowler et al.. 2009).
2-55
-------
and plant emissions are controlled by the physiological importance of NH4+ in
intercellular fluids of plants (Sutton et al.. 1998). Copollutants also play a role because
NH3 deposition is enhanced by the presence of atmospheric acids like SO2 (Sutton et al..
1995: McLeodetal.. 1990).
The complex pattern of NH3 sources and sinks with strong horizontal gradients of NH3
concentration (Fowler et al.. 2009) presents problems for simple bidirectional exchange
models (Sutton et al.. 1998). These problems have been addressed through the concept of
a canopy compensation point to include exchange with leaf surfaces (Flechard et al..
1999; Sutton et al.. 1998). as well as decomposing leaf litter and soil surfaces (Nemitz.
2000) in addition to stomatal exchange to describe bidirectional exchange (Burkhardt et
al.. 2008; Sutton et al.. 1998). Emission flux is particularly high from recently fertilized
soils (Fowler et al.. 2009; Sutton etal.. 1998) and after leaf-cutting events (Nemitz et al..
2009).
According to the 2008 ISA (U.S. EPA. 2008a). large areas of the U.S. are very near the
NH3 compensation point for most of the year, resulting in a highly dynamic air-surface
flux, which is prone to shifts in magnitude and direction. Bidirectional NH3 fluxes with
some periods of deposition and some periods of emission are typical for fertilized and
grazed agricultural ecosystems, while forests and other unfertilized ecosystems are
usually sinks for NH3 (Fowler et al.. 2009). Smaller emissions can also occur in semi
natural ecosystems (Fowler et al.. 2009).
Recently, regional scale modeling studies began to include canopy compensation points
and parameterize bidirectional exchange (Kruit et al.. 2012; Dennis et al.. 2010; Kruit et
al.. 2010). and a bidirectional exchange model for NH3 based on observations from North
Carolina field sites (Walker et al.. 2013) was developed for the CMAQ modeling system
and an agroecosystem model was included in CMAQ Version 5.0 to estimate NH3
emissions, transport, and deposition from agricultural practices (Bash et al.. 2013).
Including bidirectional exchange in deposition modeling substantially improves
agreement between modeling results and ambient observations. A large bias of-19% has
been observed in annual wet deposition of NH4+ when modeling results were compared
with ambient measurements without bidirectional exchange included in the model (Appel
et al.. 2011). NH4+ was underestimated throughout the year, but the largest
underestimations were for winter and spring in the Eastern U.S. (Appel et al.. 2011). The
NH4+ wet deposition bias was reduced by a factor 3, from -19 to -6%, by including
bidirectional exchange in CMAQ (Appel et al.. 2011).
Poor temporal and spatial representation ofNFb emissions in areas with fertilizer
application was also identified as a source of bias (Appel et al.. 2011). When CMAQ was
coupled with the U.S. Department of Agriculture's (USDA) Environmental Policy
2-56
-------
Integrated Climate (EPIC) agroecosystem model to improve characterization of fertilizer
emissions in annual simulations, NH3 dry deposition decreased by 45%, total NHx
deposition decreased by 15%, and total N deposition decreased by 5% compared to
modeling without bidirectional exchange (Bash et al.. 2013). In sensitivity tests of key
parameters in dry deposition modeling, the largest uncertainty was observed for the
change of unidirectional to bidirectional flux, but uncertainties of 5% or less in total
nitrogen deposition were reported (Dennis et al.. 2013). Although this is a small
difference nationwide, changes can locally be up to 50%, and only 66% of the 12 x 12
grid cells modeled showed changes of less than 10% (Dennis et al.. 2013).
Recent modeling studies have also improved insight into local areas and conditions under
which bidirectional flux most strongly affects deposition estimates. For example,
accounting for bidirectional flux resulted in a 17% increase in NH3 emissions from
agricultural operations (Massad et al.. 2010) compared with a 5% increase in
domain-wide NH3 emissions (Dennis et al.. 2013). Increases in NH3 emissions from
including bidirectional flux in semi natural ecosystems mostly occurred in areas of the
western U.S. with low emissions, where emissions were not included in existing
inventories. Seminatural ecosystems in the eastern U.S. isolated from agricultural
emissions exhibited changes of less than 1%. Seasonal differences were also observed,
with greater NH3 emissions observed in summer and winter, but emissions up to 45%
lower in fall and in spring when bidirectional exchange was included (Bash et al.. 2013).
2.5.2.4 Transference Ratios
Ratios of modeled or measured concentrations of SOx and NOy to their deposition, or
transference ratios (TRsox, TRnoy) were proposed by Scheffe et al. (2011) as a means to
link ambient air quality to deposition, and can be extended to NHx. Transference ratios
for NOy, SOx, and NHx are given by:
• TRnoy = (annual wet + dry deposition of NOy)/annual average ambient
concentration of NOy
• TRsox = (annual wet + dry deposition of SOx)/annual average ambient
concentration of SOx
• TRnhx = (annual wet + dry deposition of NHx)/annual average ambient
concentration of NHx
These ratios are expressed in units of distance/time (as a velocity). In the 2011 Policy
Analysis for the NOx/SOx NAAQS review (U.S. EPA. 2011a). the transference ratios
were multiplied by measured ambient concentrations of NOy and SOx to estimate a flux.
2-57
-------
The transference ratio is an aggregate of multiple forms of nitrogen or sulfur. For
example, the transference ratio for NOy includes NO2 and HNO3, which have very
different deposition rates, chemical reaction rates, and atmospheric lifetimes. Close to
emission sources where fresh NOx emissions have had little time to react, the ratio of
HNO3 to NO2 is smaller than farther from emission sources. Accordingly, there is
considerable spatial and temporal variability (Sickles et al.. 2013) in transference ratios,
in part governed by the relative abundance of compounds with short and relatively longer
atmospheric lifetimes.
In practice, a regional modeling approach using a modeling system such as CMAQ or
CAMx is used in the calculation of transference ratios by simulating the relevant
transport processes discussed in Appendix 2.5.1. Sickles and Shadwick (2013) estimated
that TRsox and TRnoy could be given to within 25-35% of observed values using
observations of atmospheric concentration and deposition at some monitoring sites in the
eastern U.S. The study noted that accounting for year-to-year variability in precipitation
could lower the uncertainty. The transport processes described in Appendix 2.5.1 imply
that wet deposition should not necessarily be well correlated with surface concentrations
due to differences in the direction or spatial extent of transport in the boundary layer
compared to cloud levels. Dry deposition fluxes are more directly related to surface
concentrations. In a follow-on study, Sickles et al. (2013) calculated transference ratios
using measurements from CASTNET and NADP monitoring networks with CMAQ
model results and found the relative difference ranged from -37 to 64%. The authors
caution that this range should not be considered a definitive assessment of uncertainty
because relative differences do not reflect the extent to which the monitor is
representative of the grid-cell average modeled by CMAQ.
Koo et al. (2012) and Koo et al. (2015) raised the issue of model dependence on the
calculation of depositing species and transference ratios. Koo et al. (2012) compared
model results for concentrations of SO2, SO42 . HNO3, and NO;, and corresponding dry
deposition fluxes from CMAQ and CAMx to those measured at CASTNET sites. They
also compared model results for wet deposition to NADP/NTN measurements. On an
annualized basis, mean normalized errors (MNEs) in gas-phase concentrations ranged
from -25 to -100%. MNEs in dry deposition were much larger and ranged from -50 to
>300% and MNE in wet deposition ranged from -40 to -100% with no clear preference
for one model over another. MNE for NH44" in dry and wet deposition ranged from -35 to
70%. Koo et al. (2012) also found evidence for spatial variability in TRsox and TRnoy
across the U.S. and within selected ecosystems (roughly a few hundred km across).
Koo et al. (2015) compared simulations of transference ratios computed using CMAQ
and CAMx for two model years, 2005 and 2014, (see Figure 2-11) and found that TRsox
2-58
-------
was much higher in CMAQ than in CAMx; however, differences were much smaller for
TRnoy- R2 values for TRnoy between the two models was 0.37 for 2005 and 0.33 for
2014. For TRsox, R2 was 0.073 for 2005 and 0.072 for 2014. Note that each point shown
in Figure 2-11 represents the average over 365 x 24 model entries and are dissimilar to
measurement artifacts. The outlying point (in the lower right) is from the Weminuche
Wilderness IMPROVE site (WEMI) in the Rockies and likely is the result of the
difficulty mesoscale models have in simulating precipitation and flow patterns in areas of
high relief and indicates further work is needed in this regard. Note that the disagreement
in TRsox by CMAQ is mostly due to an error in CMAQ emissions of SO2 that has since
been corrected. There is no consistent geographic pattern of agreement or disagreement
between the two simulations.
A more complete understanding of the causes of differences between model simulations
requires understanding the differences in how major chemical and physical processes
have been parameterized, thus underscoring the importance of accurately representing
emissions, transport, chemistry, and deposition. Both models used different modules to
represent these processes. In addition, these results might imply that the metrics used
need further scrutiny. Note again that the results shown in Koo et al. (2012) and Koo et
al. (2015) were obtained using older versions of CMAQ and CAMx and that CTMs are
continually undergoing improvement. In this regard, detailed comparisons with
observations and intercomparisons between the most current versions of these models
might help explain these findings.
Koo et al. (2015) also found very small differences between simulations for model years
2005 and 2014 from either CMAQ or CAMx for both NOy and SOx, indicating the ratios
are relatively invariant at least over an annual time scale. This result is not surprising,
because with the long averaging time, concentrations and deposition rates can better track
emissions changes.
2-59
-------
o 2005
+ 2014
o 2005
+ 2014
o CAMx
+ CMAQ
O CAMx
+ CMAQ
(a) TRj1[d 50x [CMAQ vs. CAMx)
{b)TRJtl0 SOx (2014vs. 2005)
(c) TR„io NOv (CMAQ vs. CAMx)
1.4
1.5 2
CAMx
2005
CAMx
(dJTR^NOy [2014 vs. 2005)
i i i i r—
0.4 0.6 0.8 1 1.2
2005
CAMx = Comprehensive Air Quality Model with Extensions; CMAQ = Community Multiscale Air Quality; NOY = oxides of nitrogen;
SOx = sulfur oxides; TRatio = transference ratio.
Source: Adapted from Koo et al. (2015).
Figure 2-11 Scatterplots showing transference ratios for oxidized nitrogen
and sulfur oxides comparing the Community Multiscale Air
Quality model to the Comprehensive Air Quality Model with
Extensions in (a) and (c) and comparing 2005 to 2014 in (b) and
(d).
To summarize, recent studies have found that using transference ratios for estimating the
deposition flux from atmospheric measurements has lower uncertainty when applied at an
annual timescale. The transference ratio can vary spatially, and an estimate of uncertainty
and variability depends on the spatial scale of interest. Finally, previous studies have
highlighted some of the uncertainties when using models to calculate the transference
2-60
-------
ratio, but a comprehensive assessment of uncertainty is not available. When the
transference ratio is calculated using observations of atmospheric concentrations from
monitoring networks, the deposition flux for oxidized sulfur and oxidized nitrogen can be
estimated to within 25 to 35%.
2.5.3 Model Evaluation and Uncertainty
CMAQ model estimates were recently compared to monitoring network observations and
results were reported as normalized mean bias statistics (U.S. EPA. 2011a. 2009c). Total
nitrate concentrations were overestimated by CMAQ, with predictions averaged over
each of 4 years ranging from 22 to 26% higher than observed concentrations (U.S. EPA.
2011a). Model performance was described as good for total SOx, but the ability to
partition SOx into sulfate and SO2 was identified as an area that needed improvement
(U.S. EPA. 2011a). SO2 concentrations were overestimated, with CMAQ predictions
ranging from 39 to 47% higher than observed concentrations. Sulfate concentrations were
underestimated, with CMAQ predictions ranging from 9 to 17% lower than observed
concentrations (U.S. EPA. 2011a). A disadvantage of this type of comparison is that
modeled concentrations are outputted for a 12 * 12 km grid, while measured
concentrations are from a single point within that grid.
Wet deposition estimates from CTMs were extensively evaluated using wet deposition
data collected over the U.S. as part of the National Atmospheric Deposition Program's
National Trends Network. Simon et al. (2012) summarized model evaluation studies
published before 2010, and Table 2-7 summarizes the studies since 2010.
While measurements comparing model outputs to observations provide one perspective
on the uncertainty in the fate and transport of atmospheric N and S, another approach to
quantifying uncertainty is to estimate the sensitivity of the model results with respect to
the uncertain range of parameters relevant to deposition calculations. A study by Dennis
et al. (2013) examined a range of uncertainties relevant to dry deposition using the
CMAQ model. This study found little change (<5%) in total deposition, despite changes
in dry deposition parameters because competing processes in the model tended to
rebalance and compensate. Changes in a single grid cell were as large as 20%.
There are also structural uncertainties that are difficult to assess in applying CTM models
to estimate deposition. The main structural uncertainties are associated with canopy
effects of NOx, bidirectional exchange of NH3, and transference ratios that relate average
concentration to deposition. These factors are discussed in Appendix 2.5.2.
2-61
-------
Table 2-7 Reported comparisons of chemical transport models and
observations of nitrogen and sulfur wet deposition
Reference
Bias/Error
Model
Metric
ADDel etal. (2011)
7.9% normalized mean bias
CMAQ
Annual, wet deposition, sulfate,
continental U.S.
Accel etal. (2011)
-12.8% normalized mean bias
CMAQ
Annual, wet deposition,
ammonium, continental U.S.
ADDel etal. (2011)
-12.8% normalized mean bias
CMAQ
Annual, wet deposition,
ammonium, continental U.S.
Aooel etal. (2011)
-15% normalized mean bias
CMAQ
Annual, wet deposition, nitrate,
continental U.S.
ADDel etal. (2011)
0.5 kg/ha median bias
CMAQ
Annual, wet deposition, nitrate,
continental U.S.
Zhanq et al.
(2012a)
6.5% mean normalized bias
GEOS-Chem
Annual, wet deposition, sulfate,
continental U.S.
Zhana et al.
(2012a)
10% mean normalized bias
GEOS-Chem
Annual, wet deposition, nitrate,
continental U.S.
Zhana et al.
(2012a)
7.4% mean normalized bias
GEOS-Chem
Annual, wet deposition,
ammonium, continental U.S.
Koo et al. (2012)
45 to 99% normalized mean error
CMAQ
Annual, wet deposition, sulfate,
continental U.S.
Koo et al. (2012)
38 to 99% normalized mean error
CMAQ
Annual, wet deposition, nitrate,
continental U.S.
Koo et al. (2012)
45 to 66% normalized mean error
CMAQ
Annual, wet deposition,
ammonium, continental U.S.
Williams et al.
(2017a)
0.34 kg ha-1 mean error
CMAQ
Annual, wet deposition, inorganic
nitrogen, Pacific Northwest
2.6 Geographic Distribution of Concentration and Deposition
Maps of national distributions of emissions, atmospheric concentrations, and deposition
fluxes of relevant species are presented in this section. The first two sections are limited
to deposition maps and are intended to provide a broad overview of the extent and recent
trends for acid deposition (Appendix 2.6.1) and total nitrogen deposition, including
relative contributions of reduced and oxidized nitrogen (Appendix 2.6.2). The subsequent
three sections contain data on geographic distributions of emissions, ambient
2-62
-------
concentrations and deposition of oxidized nitrogen (Appendix 2.6.3). reduced nitrogen
(Appendix 2.6.4). and sulfur oxides (Appendix 2.6.5). including key species in each of
these classes.
Emission maps in Appendix 2.6.3. Appendix 2.6.4. and Appendix 2.6.5 are from U.S.
EPA's National Emissions Inventory described in Appendix 2.2. Ambient concentration
maps in these sections are from a variety of sources, depending on the availability of
data. Deposition maps in Appendix 2.6.1. Appendix 2.6.2. Appendix 2.6.3.
Appendix 2.6.4. and Appendix 2.6.5 are based on the approach of Schwede and Lear
(2014a) which combines measured and modeled values to produce spatially aggregated
maps of wet, dry, and total (wet plus dry) deposition of nitrogen and sulfur species across
the U.S. Wet deposition is based on concentrations measured in rainwater collected at
NADP/NTN monitoring sites combined with precipitation estimates interpolated by
PRISM (Parameter-elevation Regression Slopes Model) using inverse distance weighting
(IDW). In their approach to dry deposition, Schwede and Lear (2014a) measured values
of species concentrations in air at monitoring site locations and used bias-corrected
modeling results from CMAQ (currently at 12-km horizontal resolution) to fill in gaps
between sites and provide composition and deposition information for species not
measured (PANs, NO2, and HONO) in the routine monitoring networks. Distributions of
species that undergo dry deposition are derived mainly by fusion of data from the Clean
Air Status and Trends Network (CASTNET), the National Atmospheric Deposition
Program (NADP), Ammonia Monitoring Network (AMoN), and the Southeastern
Aerosol Research and Characterization (SEARCH) network. Parameterizations in CMAQ
are used to calculate deposition velocities for gases and particles (Pleim and Ran. 2011).
Note that CMAQ includes bidirectional exchange for NH3 (Bash et al.. 2012). but not for
other species such as NO2. Dry deposition fluxes are then combined with wet fluxes to
estimate total deposition. Details on interpolation, special treatment for particulate
species calculations, and procedures to correct for bias are described in detail by Schwede
and Lear (2014a).
Efforts have also been made to achieve greater consistency by relying more heavily on
WRF/CMAQ simulations for estimating wet deposition. Effects of biases in CMAQ wet
deposition are corrected by adjusting the modeled wet deposition by the ratio of observed
precipitation interpolated by PRISM to WRF precipitation. In this approach, it is assumed
that the ratio of observed to modeled precipitation is well correlated with the ratio of
observed to modeled wet deposition, but not (necessarily) that wet deposition scales
linearly with precipitation (Appel et al.. 2011). Likewise, estimates of dry deposition
could be obtained using CMAQ evaluated by comparison with monitoring results.
2-63
-------
This approach resulted from efforts to improve estimates of atmospheric deposition by
advancing the science of measuring and modeling atmospheric wet, dry, and total
deposition of atmospheric species. Recognizing that the thin geographic coverage and the
lack of species measurements described in Appendix 2.4.6 are not ideal for estimating
dry deposition on a national scale, the NADP established the Total Deposition Science
Committee (TDEP) with this mission in 2011, with an initial goal of providing estimates
of total S and N deposition across the U.S. for use in estimating critical loads and other
assessments, where loading results in the acidification and eutrophication of ecosystems
(NADP. 2016). Following this hybrid approach to mapping total deposition that
combines measured and modeled values, measured values are given more weight at the
monitoring locations, and modeled data are used to fill in spatial gaps and provide
information on chemical species that are not measured by routine monitoring networks.
This effort provides continuous spatial and temporal coverage of total deposition
estimates in the U.S., something previously unavailable (NADP. 2016).
Limitations to the TDEP approach are: (1) interpolation leads to a minimization of
extreme values and a lower than actual variability, (2) data are limited to sites that meet
network completion criteria, (3) discontinuities in trends can occur for intermittent
monitoring data, (4) characterization of wet and dry organic nitrogen components is
uncertain and likely incomplete, (5) deposition in urban areas is not well represented
because the monitoring sites used are primarily in rural areas, and (6) occult deposition is
not well understood and might not be characterized accurately. An additional potential
drawback is that a mass balance is not maintained, although the mass balance error was
small in a similar effort combining measured wet deposition and bias corrected modeled
deposition (Schwede and Lear. 2014a).
Differences in wet deposition of NH4+, NO;, . and SO42 and N + S expressed as H+
equivalents between the two, 3-year periods 1989-1991 and 2012-2014 across the U.S.
are shown in Figure 2-13. Figure 2-21. Figure 2-34. and Figure 2-43. These figures are
based on data obtained by the NADP/NTN. The maps were constructed by summing
gridded values and then taking the difference between the two, 3-year averages using data
from the NADP website. These maps are meant to provide a general indication of
large-scale features in the patterns and long-term changes in deposition, with potential for
error from extensive interpolation between monitoring sites, which are often distant from
each other. The TDEP values shown on the maps are derived from NADP/NTN wet
deposition measurements at 4-km resolution coupled with CMAQ results for dry
deposition at 12 km. Weathers et al. (2006) found evidence for substantial spatial
variability in deposition with altitude, generally at smaller scales than used for TDEP. In
particular, they found evidence for factors of 4-6 variability in N and S deposition
throughout Acadia National Park (121 km2) and Great Smoky Mountains National Park
2-64
-------
(2,074 km2); deposition rates averaged over the two parks were -70% higher than
inferred using only in situ point measurements. As might be expected, deposition
increased with elevation because of the increased importance of cloud deposition and dry
deposition (due to higher winds and hence increased turbulence). As noted by the NADP,
uncertainty within maps of wet deposition varies geographically and has not been
quantified.
Appendix 2.6.6 and Appendix 2.6.7 summarize concentration and deposition data from
other approaches. Appendix 2.6.6 describes distributions of dry deposition of NO2 and
SO2 from satellite-based measurements of tropospheric vertical column abundance and
model input derived by Nowlan et al. (2014). Appendix 2.6.7 describes recent estimates
of background concentrations, deposition fluxes, and sources and methods used to obtain
them.
Additional maps on the portion of the NADP website dedicated to the Total Deposition
(TDEP) program (http://nadp.slh.wisc.edu/committees/tdep/tdepmaps/) are shown in
Appendix 2.7. They present a comprehensive overview of changes in various parameters
related to deposition from 2000-2018. Each map shows two 3-year periods, 2000-2002
and 2016-2018. Although uncertainty has not been fully characterized using the TDEP
approach, they are instructive because they give an indication of how various estimates
carried out with the same approach have changed over the past decade.
2.6.1 pH and H+ Equivalents
Long-term trends in rainwater pH over the CONUS between the two periods 1989 to
1991 and 2016 to 2018 are shown in Figure 2-12. Substantial improvement in the quality
of rainwater in terms of pH has occurred from the earlier to the later period. Figure 2-12
is a remarkable demonstration of the effectiveness of the Clean Air Act Amendments,
showing that the steep reductions of NOx and SO2 emissions described in Appendix 2.2
coincide with a sharp decline in pH.
2-65
-------
Source: NADP/U.S. EPA/CAMD. We acknowledge the Total Deposition (TDep) Science Committee of the National Atmospheric
Deposition Program (NADP) for their role in making the TDep data and maps available.
Figure 2-12 (Left) pH of rainwater, 1989-1991; (Right) pH of rainwater,
2016-2018.
However, there are areas, especially in the eastern U.S. near the upper Ohio River Valley,
where the pH of rainwater is still much lower than the reference pH of -5.65 (for
equilibrium with CO2) that has often been used as a benchmark when characterizing the
excess acidity of rainwater. (Note that the reference CO2 concentration used to determine
this value is 316 ppm, atmospheric CO2 concentrations are now over 400 ppm, resulting
in a lowering of pH by -0.05 units.) As noted by Galloway et al. (1976). the major
contributors to free acidity in rainwater for pH <5.6 are the strong mineral acids HNO3
and H2SO4. However, weak acids (e.g., organic acids) can contribute substantively to free
acidity at pH levels seen throughout much of the U.S. (see Appendix 2.3.5). For example,
concentrations of formic acid and acetic acid (pKa = 3.75, 4.76) measured in rainwater at
pH -5 by Avery et al. (2006) are on the order of 10 |iM. which is comparable to
concentrations of NO, and SO42 measured in rainwater. Additionally, in areas like the
Northwest where the pH of stream water can be around or even larger than 5.6,
acidification of streams by CO2 might also need to be considered (Ou et al.. 2015).
The change in acid loading (H+ equivalents) due to wet deposition of NO, . NFU+, and
S042 ions in precipitation expressed as H+ equivalents between the two, 3-year periods
1989-1991 and 2016-2018 across the U.S. based on data obtained by the NADP/NTN is
shown in Figure 2-13. Substantial decreases in acid loading are seen in the eastern U.S.,
with most of the central and western U.S. showing smaller negative changes or
essentially no change.
2-66
-------
Change in Nitrogen and Sulfur Wet Deposition from Nitrate, Ammonium,
and Sulfate in the U.S. between (1989-1991) and (2016-2018)
Source: NADPNTN Annual Gradients
eq = H+ equivalent; H+ = hydrogen ion.
Source: CPHEA analysis of CASTNET/CMAQ/NTN/AMON/SEARCH data.
Figure 2-13 Difference in wet deposition of nitrate, ammonium, and sulfate
expressed as hydrogen ion equivalents (eq/ha/yr) over the
contiguous U.S. between 1989 to 1991 and 2016 to 2018.
This map was constructed by summing gridded values and then taking the difference
between the two, 3-year averages using data from the NADP website. Although
instructive, these results should be viewed with some caution, as errors are incurred
because development of a map spanning the CONUS requires extensive interpolation
between monitoring sites, which are often distant from each other. As noted by the
NADP, uncertainty within maps of wet deposition varies geographically and has not been
quantified. Therefore, these maps are meant to provide a general indication of large-scale
features in the patterns and long-term changes in deposition.
2-67
-------
2.6.2
Total Nitrogen
Figure 2-14 shows total deposition of N averaged over 2016-2018. This is determined
from the sum of dry and wet deposition for all NOy and NHX species considered by
CMAQ. Fluxes are based on the method developed by Schwede and Lear (2014a) as
outlined above and are given in terms of kg N/ha/yr.
As can be seen from Figure 2-14. the highest deposition of nitrogen occurs in a broad
swath across the Midwest, and in more localized patches across the U.S. Total N
deposition can be put in context by comparing the deposition amounts in Figure 2-14 to
estimates of critical load, which is the N deposition amount below which no significant
harmful effects on sensitive elements of the environment occur. These are typically
below 10 kg N/ha/yr and can be as low as 2-3 kg N/ha/yr in both eastern and western
locations (Lee et al.. 2016; Ellis et al.. 2013). amounts which are firmly below the
estimated deposition amounts in Figure 2-14 over wide areas of the U.S. Ellis et al.
(2013) observed that critical loads were exceeded in 24 of 45 parks, and Lee et al. (2016)
observed that critical loads were exceeded at more locations in the western U.S., but by
larger amounts in the eastern U.S.
Inspection of Figure 2-15a and Figure 2-15b shows that many of these areas are
dominated either by deposition of NHx emitted mainly by agriculture (e.g., California's
Central Valley, Upper Midwest), or NOy resulting from oxidation of nitrogen oxides
emitted mainly by combustion sources (e.g., the Northeast, Southwest). As might be
expected when considering all forms of NOy, total deposition tends to be much higher
near urban and suburban areas. Note also that deposition of Nr, based on the hybrid
measurements/modeling approach, is dominated by reduced forms across the CONUS as
a whole. This is consistent with Cross-Track Infrared Sounder Satellite measurements
(Appendix 2.4.2.2) combined with dry deposition modeling to show that NH3 deposition
fluxes were greater than NOy in most regions of North America (Kharol et al.. 2018).
Figure 2-16 illustrates how this has changed over time by comparing trends in oxidized
and reduced nitrogen deposition and changes in contributions of major species and type
of deposition from 2000-2017 (NADP. 2019).
2-68
-------
Source: CASTNET/CMAQ/NADP
Total deposition of nitrogen 1618
USEPA 10/21/19
Total N
(kg-N/ha)
N = nitrogen; Nr = reactive nitrogen.
Source: CASTNET/CMAQ/NTN/AMON/SEARCH. We acknowledge the Total Deposition (TDep) Science Committee of the National
Atmospheric Deposition Program (NADP) for their role in making the TDep data and maps available.
Figure 2-14 Total (wet + dry) deposition of nitrogen (kg N/ha/yr) over the
contiguous U.S. 2016-2018.
Dry deposition of gas-phase N (as HNO3 and NH3) exceeds dry deposition of particulate
forms (pNOs - NH41") over most of the CONUS according to the hybrid method. Overall,
deposition of N is mainly as reduced forms, with a maximum over the north-central U.S.
The Central Valley of California, northern Utah, and eastern North Carolina are among
other areas of high deposition of reduced N. In general, dry deposition of N, in either
oxidized or reduced form, exceeds wet deposition across the CONUS. However, as
discussed in earlier sections, uncertainties for dry deposition are likely much larger than
for wet deposition.
Also as mentioned earlier, several species potentially important for deposition are not
measured in CASTNET. Figure 2-17 shows dry deposition for oxidized nitrogen species
(e.g., PAN, other organic nitrates, HONO) calculated by CMAQ.
2-69
-------
Souk*. CASTNtXt WwQ/S A UP
Pet of idal N av reduced N IfilS
I'SEFA KMZ1AV
B *
Swarf CASTS! TvMAQ^E*
Pel of KV-i[ \ .n in: iiii/L'd \ 16! S
l si r.\ gnaw
Note: We acknowledge the Total Deposition (TDep) Science Committee of the National Atmospheric Deposition Program (NADP)
for their role in making the TDep data and maps available.
Figure 2-15 (A) Percentage of total nitrogen deposition as reduced inorganic
nitrogen over the contiguous U.S. 2016-2018. (B) Percentage of
total nitrogen deposition as oxidized nitrogen over the
contiguous U.S. 2016-2018.
2-70
-------
30%
Total N
Oxidized N
Reduced N
—HN03 Dry
-*-NH4 Dry
-^Unmeasured Dry
N03 Dry
NH4 Wet
¦N03 Wet
-NH3 Dry
0%
2000 2002 2004 2006 2008 2010 2012 2014 2016
Source: CNADP. 2019V
Figure 2-16
Trends in U.S. total deposition flux of total nitrogen, oxidized
nitrogen, reduced nitrogen, and major nitrogen species
2000-2017.
2-71
-------
Pet of total N as unmeasured species 1618
USEPA 10/21/19
Source: CASTNET/CMAQ/NADP
(Pet of Total)
Other N
N = nitrogen.
Source: CASTNET/CMAQ/NTN/AMON/SEARCH. We acknowledge the Total Deposition (TDep) Science Committee of the National
Atmospheric Deposition Program (NADP) for their role in making the TDep data and maps available.
Figure 2-17 Three-year average percentage of total nitrogen deposition by
species (i.e., those species that are not measured in the
networks) simulated by the Community Multiscale Air Quality
modeling system for 2016-2018.
As seen in Figure 2-17. deposition of these species can contribute substantially to N
deposition, especially near strong sources, in particular large urban areas. Turnipseed et
al. (2006) also indicated that not accounting for these species can result in significant
underestimates of N deposition.
Figure 2-18 illustrates the long-term trend in total wet deposition of N (NFL + NO3 ) by
showing maps for two 3-year periods (2016-2018 and 1989-1991) binned in increments
of 2 kg N/ha/yr for comparison to critical loads estimates. Although it is apparent that N
wet deposition has decreased overall across the U.S., there are areas showing increases.
Also shown are NTN sites active during either period.
2-72
-------
Sum N03' and NH/ Wet Deposition by 3-Year Averages
N = nitrogen; NH4+= ammonium; N03 = nitrate; NTN = National Trends Network.
Source: CPHEA analysis of CASTNET/CMAQ/NTN/AMON/SEARCH data.
Figure 2-18 Wet deposition of ammonium + nitrate (kg N/ha/yr) over the
contiguous U.S. in two, 3-year periods, 2016 to 2018 and 1989 to
1991. Also shown are active National Trends Network sites in
either period.
2-73
-------
Studies at individual sites (e.g., on the coast of North Carolina) have shown that about
30% of wet deposition of N consisted of organic N, 20-30% of which was then available
to primary producers on timescales of hours to days (Peierls and Paerl. 1997). In addition,
Benedict et al. (2013) found that wet deposition of organic nitrogen contributed 18% of
total quantified reactive nitrogen deposition and 25% of wet nitrogen deposition at Rocky
Mountain National Park between November 2008 and November 2009. Trends in total
(wet + dry) deposition of total (NOy + NHX) nitrogen between 2000 and 2018 are
described with maps in Appendix 2.7.
2.6.3 Oxidized Nitrogen
The geographic distribution of annual U.S. NOx emissions for 2017 is shown in
Figure 2-19. Areas of higher emissions are apparent in urban areas and along the major
travel routes between them, especially in the eastern U.S. and states along the west coast.
The distribution of ambient NOy concentrations is shown in Figure 2-20. However,
because NOy is only measured at a small number of sites, this map is based solely on
CMAQ model output. Although these model results for NOy are not as up to date as the
network concentration measurement results or the modeled deposition of NOy
component species that follow in this section, the same basic patterns observed for
emissions also apply to ambient concentrations, with higher concentrations in the eastern
U.S. than in the western U.S., as well as along the west coast, in urban areas and along
major travel routes. The distribution of NO2 (shown in Figure 2-21) is derived from
satellite data (OMI) and output from the GEOS-Chem model using the method outlined
in Appendix 2.6.3. Distributions of HNO3, pNO, . pNFU+, SO2, and pSO-f are based on
data from the CASTNET. Note, however, that because of artifacts relating to
measurement of HNO3 and pNO, (Appendix 2.4.5). the measurement of total nitrate
(TN = HNO3 + pNO;, ) is judged to be more reliable than measurements of its
components.
2-74
-------
> 2210
1964
1719
1473
>>
1228 Ł
o
982
'736
491
-< 245
Source: OAQPS analysis of U.S. EPA (2020a1 data.
Figure 2-19 Geographic distribution of annual U.S. NOx emissions in 2017.
2-75
-------
12-18
19-32
33-86
NOy = oxidized nitrogen species.
Source: U.S. EPA/OAQPS.
Figure 2-20 Distribution of annual average total oxidized nitrogen species
concentrations for 2011 simulated by Community Multiscale Air
Quality modeling system.
2-76
-------
OMI-derived surface N02 (ppb)
I
7.00
6.22
5.44
4.67
3.89
3.11
2.34
1.56
0.78
0.01
!%¦¦ fit"
JJA
4>
DJF = December, January, February; JJA = June, July, August; N02 = nitrogen dioxide; OMI = Ozone Monitoring Instrument.
Note: Images shown were constructed by Dr. Lok Lamsal of Universities Space Research Association from data obtained by the
OMI on the Aura satellite (http://aura.gsfc.nasa.gov/scinst/omi.htmn using the algorithm described in Bucsela et al. (2013). Output
from the GEOS-Chem, global-scale, three-dimensional, chemistry-transport model to derive surface concentration fields from the
satellite data as described in Lamsal et al. (2008) and Lamsal et al. (2010).
Top panel (winter: December, January, February). Lower panel (summer: June, July, August).
Figure 2-21 Seasonal average surface nitrogen dioxide mixing ratios in parts
per billion for winter (upper panel) and summer (lower panel)
derived by the Ozone Monitoring Instrument/GEOS-Chem model
for 2009-2011. The Ozone Monitoring Instrument has an overpass
at approximately 1:30 p.m. local standard time.
2-77
-------
As seen in the figure, the highest values are found in and around urban areas. In general,
much broader areas of high concentrations (>~5 ppb) are found in the eastern U.S., with
many areas in the western U.S. subjected to concentrations <1 ppb, which implies that
concentrations of components that might pose a hazard are also lower than 1 ppb because
NOy refers to the sum of oxidized N species.
Figure 2-21 through Figure 2-23 describe geographic concentration patterns for NOy,
HNO3, and pNO;, . Figure 2-21 shows seasonal average NO2 concentrations derived using
the hybrid (OMI-satellite/GEOS-Chem-model) approach described in Appendix 2.4.2.2.
Large variability in NO2 concentrations is apparent in Figure 2-21. As expected, the
highest NO2 concentrations are seen in large urban regions, such as in the Northeast
Corridor, and lowest values are found in sparsely populated regions located mainly in the
West. Minimum hourly values can be less than -10 ppt, leading to a large range between
maximum and minimum concentrations. Although overall patterns of spatial variability
are consistent with the current understanding of the behavior of NO2, there are limitations
in the satellite retrievals (see Appendix 2.4.2.2). Surface NO2 concentrations tend to be
higher in January than in July, largely reflecting lower planetary boundary layer heights
in winter. Such seasonal variability is also evident on a local scale, as measured by
surface monitors. For example, in Atlanta, GA, NOx measurements also exhibited higher
concentrations in winter and lower concentrations in summer, when NOx is more rapidly
removed by photochemical reactions. For example, see U.S. EPA (2008b). 98th
percentile NO2 concentrations from the national NO2 monitoring network, which has
monitors mainly in urban areas, decreased by 53% from 1990 to 2017.
Figure 2-22 shows ambient concentrations of HNO3. Elevated concentrations of HNO3
are notable in southern California, the Midwest, the south-central U.S., and the
Mid-Atlantic states. Conversion of NO2 to HNO3 takes place over a timescale of 1 to
several hours, during which time appreciable transport can occur.
Figure 2-23 shows 3-year average concentrations of particulate nitrate (pNO;, ) across the
CONUS. Average pNO;, concentrations were highest in the Upper Midwest with a
notable maximum at the junction of Iowa, Wisconsin, Missouri, and Illinois. The high
values in the Upper Midwest are expected to be found during winter for reasons noted in
Appendix 2.3.3. Elevated levels were also observed in central California in the San
Joaquin Valley, central Pennsylvania, central Florida, and through much of the Midwest.
Based on IMPROVE and CSN monitoring network data, ammonium nitrate
concentrations are highest in California and the Midwest (Hand et al.. 2012c).
Figure 2-24 shows decreasing trends in particulate NO3 based on CASTNET monitoring
data for 34 eastern U.S. and 16 western U.S. monitoring sites, showing on average a 51%
decrease in the eastern U.S. and a 37% decrease in the western U.S. from 1989 to 2017
2-78
-------
(U.S. EPA. 2017b). It is worth noting that at several monitoring sites in the IMPROVE
network in the central and northern Great Plains, NO;, and SO42 are increasing at a rate
of over 5% per year (Hand et al.. 2012a). Hand et al. (2012a) suggested that this increase
might be related to oil and gas exploration and production in the region, transport from
oil and gas fields in Alberta, and also to expansion of EGUs to meet the demands of
population growth.
Source: CASTNET USEPA/CAMD 09/10/19
^Bta^TC/mtnet^m^l618iluio3_c461S
HN03 = nitric acid.
Concentrations of nitric acid (|jg/m3) can be converted to mixing ratios (parts per billion) to rough approximation at normal
temperature and pressure by multiplying by 0.38.
Source: CASTNET/U.S. EPA-CAMD.
Figure 2-22 Three-year average (2016-2018) surface concentrations of nitric
acid based on monitoring data obtained at Clean Air Status and
Trends Network sites (black dots).
2-79
-------
N03 = nitrate.
Source: CASTNET/U.S. EPA-CAMD.
Figure 2-23 Three-year average (2016-2018) surface concentrations of
particulate nitrate based on monitoring data obtained at Clean Air
Status and Trends Network sites (black dots).
Source: CASTNET
USEPA/CAMD 09/10/19
2-80
-------
Ł
*5)
O
6H
5-
4-
3-
2-
1-
o-
90th Percentile
75th Percentile
Mean
Median
25th Percentile
10th Percentile
O^CMCO^ir)(DNCOC!)0-<-(SJ(OtinCDSCOO)OT-CMCO^lOCDS
-------
a field study in Tampa Bay, FL,, measured and modeled (using CMAQ) concentrations
of HNOs and pNO, in PM10-2.5 were much higher than pNO;, in PMaa. Wolff (1984)
found that most pNO; (as NH4NO3) is found in the fine mode in Denver but in the
coarse mode (associated with Ca2 and Mg2 ) in measurements made in Detroit and rural
South Dakota, Louisiana, and Virginia. Blanchard et al. (2013', found a range of 32 to
63% for the fraction of pNCh in PM10-2.5 versus PM-? particles in the Southeast. Lee et
al. (2008) found that most pNO: was in the coarse mode at Grand Canyon and Great
Smoky Mountains, corroborating earlier findings at Yosemite and Big Bend national
parks. They also found that both coarse and fine mode pNO3 were important at
Brigantine National Wildlife Refuge, NJ and San Gorgonio Wilderness Area, CA. Lefer
and Talbot (2001) also found that NO;, sampled at Harvard Forest, MA between March
and October was mainly found in the coarse mode with a mass median diameter of
4.8 ± 1.5 pin. These results indicate considerable regional variability in the ratio of
pNOj in the fine and coarse modes and consequently additional uncertainty in estimates
of pNO ; deposition.
XJL
V
' 1
*
VC\T
-A\ J
VV\ V
' . - - otai oxN
/ (kg-N/ha)
_-o
• " 7' 'J--*.-
' 4. a? mk~
v" Ir
f
^ 1 yv
nY
t
1
2
3
4
5
-6
7
Sourprt: CASTNITT/CMAQ/NA DP
Total deposition of oxidized N L618
IJSEPA 10/21/19
oxN = oxidized nitrogen.
Source: CASTNET/CMAQ/NTN/AMON/SEARCH. We acknowledge the Total Deposition (TDep) Science Committee of the National
Atmospheric Deposition Program (NADP) for their role in making the TDep data and maps available.
Figure 2-25 Total oxidized nitrogen deposition over the contiguous U.S.
2016-2018.
2-82
-------
Figure 2-26 and Figure 2-27 show decreases in wet deposition of NOa" at most locations
in the U.S. These decreases are associated with NOx emissions control measures since
the passage of the 1990 Clean Air Act amendments. However, some areas, located
mainly in the West show increases. As noted earlier, Hand et al. (2012a) suggested that
this increase might be related to oil and gas exploration and production in the region,
transport from oil and gas fields in Alberta, and also to expansion of motor vehicles and
EGUs to meet the demands of population growth. The large area with the strongest
increases in the north-central U.S. corresponds to oil and gas operations in the Bakken
Shale region. Trends in total (wet + dry) deposition of oxidized nitrogen between 2000
and 2013 are described with maps in Appendix 2.7.
Overall, total U.S. oxidized nitrogen deposition has decreased since 2000, reflecting the
substantial decline in nitrogen oxides emissions, as indicated in Figure 2-28 (NADP.
2019).
Source: NADP/U.S. EPA/CAMD.
Figure 2-26 (Left) nitrate wet deposition, 1989-1991; (Right) nitrate wet
deposition, 2016-2018.
2-83
-------
Change in NQ3' Wet Deposition in the U.S.
Nrtutfyr for 3-Year Averages Between 1BB&-1091 and 2016-2018
-12.0 -Ł.0 -4.0 -1.0 -0.5 -0.1 0.1 0.5 1.0 1.5 2.0
a 12S 250 50C A.
Source: NADP NTN Annual Gradients
N = nitrogen; N03 = nitrate.
Source: CPHEA analysis of CASTNET/CMAQ/NTN/AMON/SEARCH data.
Figure 2-27 Difference in wet deposition of nitrate (kg N/ha/yr) over the
contiguous U.S. between 1989 to 1991 and 2016 to 2018. The
range of positive values is smaller than that for negative values.
2-84
-------
30000
to
to c
O
E r
LU O
— to
= 1
.2 ro
to
5 3
2 o
TO ¦*-"
25000
Ł 20000
15000
10000
5000
0
2000 2002 2004 2006 2008 2010 2012 2014 2016 2018
CD
X
c
o
t/l
o
a
a>
Q
2000 2002 2004 2006 2008 2010 2012 2014 2016
Source: (A) U.S. EPA (2020a V (B) NADP (2019V
Figure 2-28 Trends in oxidized nitrogen emissions and deposition 2000-2017:
(A) total national emissions; (B) national average total deposition
flux.
2.6.4 Reduced Nitrogen
Figure 2-29 shows the geographic distribution of annual total NH3 emissions. The
observed pattern is somewhat different than for NOx emissions (Appendix 2.6.3).
reflecting the widely distributed and rural nature of NH3 emissions, compared to NOx
emissions, which are largely urban or from large point sources. Widespread areas of high
emissions are in the Midwest, California, and several other hot spots.
Figure 2-30 and Figure 2-31 show maps for the concentrations of the reduced inorganic
nitrogen species, NH3 and pNH4+. The distribution of NH3 was obtained from the
Ammonia Monitoring Network (AMoN). The highest concentrations of NH3 were
2-85
-------
measured in Salt Lake City, UT at 16.9 ^g/m\ All other annual average concentrations
for 2017 were lower than 5 ug/nr except eastern North Carolina. The Salt Lake City site
is located near feed lots, perhaps explaining in large part why levels were much higher
there than at other sites. In general, areas with the highest NH3 concentrations in
Figure 2-30 correspond well with areas of the highest NIL emissions, as shown in
Figure 2-29. Note that confidence in the magnitude and inter-monitor precision of NFL
measurements has increased since the 2008 ISA (U.S. EPA. 2008a). also see
Appendix 2.4.3.1. However, sparseness of the monitoring network still presents
uncertainty 111 describing the nationwide distribution of NH3 concentrations. National
ML monitoring is too recent for evaluating long-term concentration trends. In the
Southeastern U.S., SEARCH network observations indicated slight upward trends in
ammonia concentrations from 2004-2012, but the trend was statistically significant at
only monitoring site (Savior et al.. 2015).
Particulate NLLf concentrations were obtained from CASTNET and were highest in
Illinois-Indiana-western Ohio, along with high values in central Pennsylvania and central
California. These locations correspond generally to the highest concentrations of pNO <
and moderate-to-high concentration locations for NFL.
Figure 2-32 shows the depositional flux of NHx over the CONUS 2016-2018.
Source: OAQPS analysis of U.S. EPA (2020a) data.
Figure 2-29 Geographic distribution of annual U.S. ammonia (NH3) emissions
in 2017.
2-86
-------
0.7
o
• 0.4
0.4
0.3
o
1%
Site not pictured:
Puerto Rico (PR20)
AMoN = Ambient Ammonia Monitoring Network; NH3 = ammonia.
Source: (U.S. EPA. 2017b).
f/% '/ool
^ \ 2.4 0?„O oo *0S-
J :i-oQ)o.7,1:0c ®rf0-6
toocm%3
nW„ • v\ Concentration
{ ss^osii^^ o 0.2-1.0
'•6 O 0.8 \ jQG o 1.0-2.0
O 2.0-3.0
O 3.0-4.0
4.2-5.0
>5
Figure 2-30 Average (2017) surface concentration of ammonia obtained by the
Ambient Ammonia Monitoring Network at select Clean Air Status
and Trends Network sites. Concentrations of ammonia (pg/m3)
can be converted to mixing ratios (parts per billion) to rough
approximation at normal temperature and pressure by multiplying
by 1.4.
2-87
-------
Source: CASTNET
USEPA/CAMD 09/10/19
NH4+ = ammonium.
Source: CASTNET/U.S. EPA-CAMD.
Figure 2-31 Three-year average (2016-2018) surface concentrations of
particulate ammonium (jjg/m3) based on monitoring data obtained
at Clean Air Status and Trends Network sites (black dots).
2-88
-------
Total deposition of reduced N 1618
USBPA10/21/19
Source: CASTNET/CMAQ/NADP
Total reN
(kg-N/ha)
reN = reduced nitrogen.
Source: CASTNET/CMAQ/NTN/AMON/SEARCH. We acknowledge the Total Deposition (TDep) Science Committee of the National
Atmospheric Deposition Program (NADP) for their role in making the TDep data and maps available.
Figure 2-32 Total reduced inorganic nitrogen deposition over the contiguous
U.S. 2016-2018.
Figure 2-33 and Figure 2-34 shows large increases in wet deposition of NH4 throughout
the north-central U.S. with hot spots along the eastern Great Lakes, eastern Pennsylvania
and North Carolina, the Texas gulf coast, and parts of Utah and California. The situation
is more nuanced than shown in that some sites show small increases and others small
decreases. In general, large-scale increases in wet deposition of Ni I.;'. rather than
decreases, are seen across the U.S. in agreement with the analysis of Li et al. (2016d).
Trends in total (wet + dry) deposition of reduced nitrogen between 2000 and 2013 are
described with maps in Appendix 2.7.
Overall, total U.S. reduced nitrogen deposition has slightly increased since 2000, even
though annual national NH3 emissions data have fluctuated with no obvious trend dunng
this period. However, during the last few years both national NH? emissions and total
reduced nitrogen deposition estimates have increased, as indicated in Figure 2-35
(NADP. 2019).
2-89
-------
Source: NADP/U.S. EPA/CAMD.
Figure 2-33 (Left) ammonium wet deposition, 1989-1991; (Right) ammonium
wet deposition, 2016-2018.
2-90
-------
»*
Change in NH/ Wet Deposition in the U.S.
between (1989-1991) and (2016-2018)
Change in Kg Whafyr for 3- Year Averages Between 1080-1961 and 2016-2018
-0.2 0.2
-2.0 -1.0 -0.5
0.5 1.0 2.0 3.0
Q 125 250 5CC ,0^
Source: NADP NTN Annua! Gradients
N = nitrogen; NH4+ = ammonium.
Source: CPHEA analysis of CASTNET/CMAQ/NTN/AMON/SEARCH data
Figure 2-34 Difference in wet deposition of ammonium (kg N/ha/yr) over the
contiguous U.S. between 1989 to 1991 and 2016 to 2018.
2-91
-------
6000
C
5000
O
'vT
'vi
c
0
4—1
4000
E
LU
*-»—
O
V)
"ro
c
-O
C
3000
0
ro
(D
ro
S
Wl
3
O
2000
"ro
-C
4—1
|2
1000
to
JC
en
x
3
c
o
'«
o
CL
O
Q
0
2000 2002 2004 2006 2008 2010 2012 2014 2016 2018
9
8 i
7
6
5 '
4
3
2
1
0
2000 2002 2004 2006 2008 2010 2012 2014 2016
Source: (A) U.S. EPA f2020aV (B) NADP (20191
Figure 2-35
Trends in reduced nitrogen emissions and deposition 2000-2017:
(A) total national NH3 emissions; (B) national average reduced
nitrogen deposition flux.
2.6.5 Sulfur Oxides
Figure 2-36 shows the west-to-east increasing gradient in SO2 emissions, with greater
emissions in most areas east of the Mississippi than in the West. Widespread areas of
high emissions are in the Northeast and Mississippi Valley. Figure 2-37 shows that sulfur
dioxide emissions have declined by 89% from 1990 to 2017 and continue to decline
steeply (U.S. EPA. 2020a).
2-92
-------
Source: OAQPS analysis of U.S. EPA (2020a).
Figure 2-36 Geographic distribution of annual U.S. sulfur dioxide (SO2)
emissions by county from the 2017 National Emissions Inventory.
total national emissions
(thousands of tons)
M M N) IV)
Ln 0 <-n O <-n
OOOOO
OOOOO
M O O O O O O
KD
89 1994 1999 2004 2009 2014
Year
Source: (U.S. EPA. 2020al
Figure 2-37 Trends in total national sulfur dioxide emissions.
2-93
-------
Figure 2-38 and Figure 2-39 show the distribution of atmospheric concentrations of
gas-phase SO2 and particulate phase SO42 . Concentrations have decreased substantially
over the last decade throughout the eastern U.S. Comparison between the national SO2
distributions (Figure 2-38) for 2016-2018 and the ones for 1989-1991 and 2003-2005
presented in the 2008 ISA (U.S. EPA. 2008a) demonstrated continual decreases in SO2
concentrations across the nation.
Both concentrations and seasonal variability of sulfate are substantially higher in the
eastern U.S. than in the West (Hand et al.. 2012c). The higher concentrations in the
eastern U.S. are consistent with greater emissions indicated in Figure 2-36. Based on air
pollution monitoring network data (IMPROVE and CSN), sulfate concentrations on a
national scale are steadily decreasing across the U.S. Between 1992 and 2010, annual
mean sulfate concentrations at rural sites decreased fairly consistently at a rate of
-2.7% per year. This decline has become even steeper more recently, with annual mean
concentrations decreasing by an average of-4.6% per year from 2002 to 2010. The
decrease appears to be due to decreasing SO2 emissions from power plants (Hand et al..
2012b). While the nationwide trend is for a reduction in sulfate concentrations, there are
seasonal and regional increasing trends, specifically in the central and northern Great
Plains in winter, and in the western U.S. in spring (Hand et al.. 2012a). Both sulfate and
nitrate are increasing at a rate of over 5% per year at several monitoring sites in the
central and northern Great Plains (Hand et al.. 2012c).
Figure 2-40 shows trends based on CASTNET data for SO2 and SO42 concentrations for
34 eastern and 16 western U.S. monitoring sites. An average SO2 concentration decline of
89% was observed in the eastern U.S. and a 45% decrease was observed on average for
the western U.S. from 1989-2017. For the same monitoring sites, SO42 concentrations
decreased by 75% in the eastern U.S. and 35% in the western U.S. on average (U.S. EPA.
2017b).
Figure 2-41 shows wet plus dry deposition of SOx (SO2 + S042 ) over the CONUS.
Greater deposition occurs over the Ohio River Valley (southeastern Ohio, West Virginia,
and western Pennsylvania), Gulf coast (Texas and Louisiana), and Northern Great Plains
(North Dakota), than in other areas of the U.S. Figure 2-42 shows a good deal of spatial
variability in the percentage of dry deposition across the CONUS. In the Mid-Atlantic
states, dry deposition of SO2 is dominant and dry deposition of pS042 is very minor
component of dry deposition of SOx. Wet deposition dominates in the Pacific Northwest,
northern New England and in general in the central U.S. Note that deposition of organic
sulfur species (e.g., methane sulfonic acid) or SO2 and/or SO42 produced by the
oxidation of organic S species is not included.
2-94
-------
Source: CASTNET
USEPA/CAMD 09/10/19
S02 =sulfur dioxide.
Source: CASTNET/U.S. EPA-CAMD.
Figure 2-38 Three-year average (2016-2018) surface concentrations of sulfur
dioxide obtained by fusion of monitoring data obtained at Clean
Air Status and Trends Network sites (black dots) and Community
Multiscale Air Quality model system results. Concentrations
(jjg/m3) can be converted to mixing ratios (parts per billion) at
normal temperature and pressure) to rough approximation by
multiplying by 0.37.
2-95
-------
Source: CASTNET USEPA/CAMD 09/10/19
/daWiirc/castnct/pivg/1613/io4_c-161S
S042" = sulfate.
Source: CASTNET/U.S. EPA-CAMD.
Figure 2-39 Three-year average (2016-2018) surface concentrations of
particulate sulfate based on monitoring data obtained at Clean Air
Status and Trends Network sites (black dots).
2-96
-------
Source: (U.S. EPA. 2017bl
Figure 2-40 Trends in oxides of sulfur oxides concentrations 1990-2017:
(A) average eastern U.S. SO2 concentration based on 34 sites;
(B) average western U.S. SO2 concentration based on 16 sites;
(C) average eastern U.S. sulfate concentration based on 34 sites;
(D) average western U.S. sulfate concentration based on 16 sites.
2-97
-------
Source: CASTNET/CMAQ/NADP
Total deposition of sulfur 1618
USEPA 10/21/19
Total S
(kg-S/ha)
-8
-10
-12
|
¦->20
S = sulfur.
Source: CASTNET/CMAQ/NTN/AMON/SEARCH. We acknowledge the Total Deposition (TDep) Science Committee of the National
Atmospheric Deposition Program (NADP) for their role in making the TDep data and maps available.
Figure 2-41 Total deposition of sulfur (kg S/ha/yr) over the contiguous U.S.
2016-2018.
2-98
-------
Source: CASTNET/CMAQ/NADP
Pet of total S as dry deposition 1618
USEPA 10/21/19
Dry S
(Pet of Total)
¦
-0
-10
-20
-30
-40
-50
-60
-70
-80
-90
->100
S = sulfur.
Source: CASTNET/CMAQ/NTN/AMON/SEARCH. We acknowledge the Total Deposition (TDep) Science Committee of the National
Atmospheric Deposition Program (NADP) for their role in making the TDep data and maps available.
Figure 2-42 Percentage of deposition of total sulfur as dry deposition over the
contiguous U.S. 2016-2018.
Figure 2-43 and Figure 2-44 show that the pattern for changes in wet deposition of SO f
is similar to that for NO( with strongest decreases in the East, but with many areas in the
western U.S. showing some increase. Reasons for this increase are similar to those for
NO.r as noted by Hand et al. (2012a).
2-99
-------
Scarce: NAD3WTN & WUSM USHWCAMD 02/iOI 18
. — Sowcc:NADKNTK & PRISM USBWCAMD IQfl&'l?
Figure 2-43 (Left) sulfate wet deposition, 1989-1991; (Right) sulfate wet
deposition, 2016-2018.
S = sulfur; S042 = sulfate.
Source: CPHEA analysis of CASTNET/CMAQ/NTN/AMON/SEARCH data.
Figure 2-44 Difference in wet deposition of sulfate (kg S/ha/yr) over the
contiguous U.S. between 1989 to 1991 and 2016 to 2018. The
range of positive values is much smaller than for negative values.
Change in SO42' Wet Deposition in the U.S.
2-100
-------
The increases in wet deposition of S042 in the north-central U.S. correspond to those for
NO;, and are in the immediate vicinity of the Bakken Shale. There is a high degree of
interannual variability in deposition in some areas, especially those showing increases
(e.g., Logan, UT/Idaho), making source attribution difficult. Trends in total (wet + dry)
deposition of sulfur between 2000 and 2018 are described with maps in Appendix 2.7.
Overall, U.S. sulfur deposition has decreased since 2000, reflecting the steep decline in
SO2 emissions. This is especially evident in the eastern U.S., where SO2 emissions are
higher (see Figure 2-36) and concentrations have decreased more rapidly than in the
western U.S. (see Figure 2-40). Figure 2-45 shows the decreasing trend of sulfur
deposition for 34 CASTNET monitoring sites in the western U.S.
25
1988 1990 1992 1994 1996 1998 2000 2002 2004 2006 2008 2010 2012 2014 2016
Year
Coverage: 34 monitoring sites in the eastern U.S.
Source: (U.S. EPA. 2020b).
Figure 2-45 Trends in average sulfur deposition flux for 34 monitoring sites in
the eastern U.S. 1989-2017.
2.6.6 Particulate Matter (PM)
Figure 2-46 shows the 3-year mean of the 24-hour PM2 5 concentrations for PM2 5
network monitoring sites across the U.S. from 2013-2015. Emissions are not shown
2-101
-------
because the majority of PM2 5 mass is often produced by atmospheric reactions (see
Appendix 2.3). Some of the highest PM2 5 concentrations are in the San Joaquin Valley
and the Los Angeles-South Coast Air Basin of California. However, in general 3-year
average 24-hour PM2 5 concentrations are higher in the Eastern U.S. than in the western
U.S. An area of the highest concentrations in the Eastern U.S. can be seen in the Ohio
Valley. From Appendix 2.3.6 is the same area where SO42 and NO;, account for the
greatest fraction of PM25, indicating that at least in the eastern U.S., the highest PM25
concentrations also correspond to PM with greatest the fraction of mass accounted for by
S042 and NO3 .
Figure 2-46 Three-year average concentrations of particulate matter smaller
than 2.5 |jm diameter (PM2.5) 2013-2015.
A further indication that the fraction of mass contributed by SO42 and NO;, may increase
with increasing PM2 5 concentration is demonstrated by Figure 2-47. Figure 2-47 shows
that a decrease in national average PM2 5 between 2000 and 2016 is paralleled by a
similar decrease in SO42 . Since SO42 has been the most abundant component of PM2 5,
the steep decline in SO2 emissions (Appendix 2.2) has led to a sharper decrease in SO42
2-102
-------
concentrations than in concentrations of other PM2 5 components. As indicated by
Figure 2-46. in those areas where SO42 concentrations remain high, the SO42 fraction of
PM2.5 mass is also still high in the eastern U.S.
2001 2003 2005 2007 2009 2011 2013 2015
2001 2003 2005 2007 2009 2011 2013 2015
Notes: Black = mean, gray = 90th percentile.
Source: Chan et al. (20181.
Figure 2-47 National monthly average concentrations of particulate matter
smaller than 2.5 |jm diameter (PM2.5; top) and sulfate in PM2.5
(bottom) from 2000-2016 (concentrations in |jg/m3).
For completeness, Figure 2-48 shows PM10-2.5 concentrations. The highest concentrations
are observed in Southwest and Great Plains. These are the same areas where crustal
material accounts for the greatest fraction of PM2.5 as described in Appendix 2.3.6. and
PMi 0-2.5 is also largely composed of crustal material, which has little impact on acid or
nutrient deposition.
2-103
-------
Source: (U.S. EPA. 2019V
Figure 2-48 98th percentile concentrations for PM10-2.5 between 2013-2015.
2.6.7 Distributions of Dry Deposition of Nitrogen Dioxide and Sulfur Dioxide
Derived Using Satellite-Based Measurements and Chemistry Transport
Models
Figure 2-49 shows the annual average dry deposition velocities and fluxes of NO2 and
SO2 for 2005 to 2007 and their estimated uncertainties derived by Nowlan et al. (2014)
using data derived from the Ozone Monitoring Instrument (OMI) on board the Aura
satellite and model parameters from the GEOS-Chem three-dimensional,
chemistry-transport model.
As shown in Figure 2-49. higher fluxes for both NO2 and SO2 occur in the East than in
the West. In particular, there is a band of high dry deposition for NO2 and SO2 along the
Ohio River. High depositional fluxes for NO2 are also seen along the Northeast Corridor
and in scattered locations throughout the East. In addition, there is a noticeable plume of
SO2 over the western Atlantic Ocean. Average, relative uncertainty in the flux estimates
for both NO2 and SO2 are -30% over land and are not much higher over the Atlantic
Ocean south of Massachusetts and Nova Scotia. Increased SO2 deposition, especially
near shore, is expected based on the likelihood of off-shore transport of SO2 and NO2
along with other pollutants by synoptic weather systems. Note that bidirectional exchange
for NO2 (and a number of other gases) has not been implemented yet in GEOS-Chem or
2-104
-------
in CMAQ. Note also that in this study, the algorithms used to derive NO2 and SO2
columns are older than more recent ones with lower detection limits. These results,
however, do illustrate the potential of the hybrid, satellite/model approach for mapping
deposition at the continental scale.
2-105
-------
NOa Dry Deposition Flux [kg N ha"' yr'1]
S02 Dry Deposition Flux [kg S ha'1 yr'1]
0 0.1 0,2 >0,3 0 1 2 3 4 >5
N02 Dry Deposition Uncertainty [kg N ha"1 yr'1] S02 Dry Deposition Flux Uncertainty [kg S ha'1 yr'1]
N = nitrogen; N02 = nitrogen dioxide; S = sulfur; S02 = sulfur dioxide.
Source: Nowian et al. (20141.
Figure 2-49 Top panel: modeled deposition velocities for nitrogen dioxide and
sulfur dioxide for 2005 to 2007; middle panel satellite-model
estimates of annual mean dry deposition fluxes of nitrogen
dioxide and sulfur dioxide; bottom panel: uncertainties in
estimates.
2-106
-------
2.6.8
Background Concentrations and Deposition
Background refers here to those concentrations or fluxes that do not result from U.S.
anthropogenic emissions. Background sources of N include natural sources like lightning,
wildfires, and emission from soils. Background sources of SOx include natural sources,
such as volcanos or oxidation of reduced sulfur species (H2S, [CH3]2S) emitted in
anaerobic environments, and anthropogenic sources from outside the U.S. Background
levels so defined facilitate separation of pollution levels that can be controlled by U.S.
regulations (or through international agreements with neighboring countries) from levels
that are generally uncontrollable by the U.S.
Kim et al. (2014a) found increases in nitrate concentrations in the mixed layer of the
North Pacific Ocean, extending to near-shore areas off the west coast of the U.S., with
attendant changes in the status of N limitation. Because NH3 and NH44" are so highly
soluble, they are likely to be removed in rain during ascent before trans-Pacific transport.
Because SO2 is much less soluble than NH3, it can be transported to the free troposphere
by the warm conveyor belt system before it is oxidized to SO42 in cloud droplets or on
the surfaces of mineral dust particles. The survivability of nitrate is intermediate,
depending on the form the nitrate takes.
Zhang et al. (2012a) computed N deposition rates from background sources and from
domestic anthropogenic sources using GEOS-Chem. According to their estimates, most
of the eastern U.S. and parts of states along the Pacific Coast received >10 kg/ha/yr N
deposition. The version of GEOS-Chem used (8.2.3) is the same as described in Zhang et
al. (2011a) and used in the 2013 ISA for Ozone and Other Photochemical Oxidants (U.S.
EPA. 2013c). Figure 2-50 shows contributions from domestic anthropogenic, foreign
anthropogenic, and natural emissions to total (wet + dry) annual nitrogen deposition over
the CONUS for 2006 calculated by Zhang et al. (2012a) using the GEOS-Chem global
scale CTM with a horizontal resolution of 1/2° by 2/3°.
2-107
-------
Nitrogen deposition enhancement Percentage contribution
US
anthropogenic
Foreign
anthropogenic
Natural
0.1 0.3 1 2 3 6 8 12 16 0 5 10 20 30 50 70 90 [%]
[kg N ha 1 a ]
Source; Zhang et al. i'2012a).
Figure 2-50 Contributions to oxidized and reduced nitrogen deposition from
U.S.: anthropogenic (top), foreign anthropogenic (middle), and
natural sources (bottom).
The upper panel of Figure 2-50 shows that the highest values from U.S. anthropogenic
sources are found in the eastern U.S., in and downwind of the Ohio River Valley, and in
and around urban areas. The middle panel of Figure 2-50 shows the highest contributions
from foreign anthropogenic sources in regions of the CON US bordering Canada and
Mexico. Note the band of the highest contributions in upper New York State as a result of
emissions in southern Canada. There is also some indication in the Pacific Northwest of
smaller contributions due to transport from Eurasia. The pattern of N deposition in the
simulation for natural sources in Figure 2-50 (bottom panel), however, shows maximum
2-108
-------
deposition throughout the central U.S., with the highest values over the Midwest,
reflecting a combination of NOy emissions from lightning in the south-central U.S., from
biomass burning throughout the Southeast, and from soils, mainly in the Midwest. The
background contribution to N deposition is typically <30% over the eastern U.S. and
typically 30 to 50% in the western U.S. where N deposition is already lower. Overall,
according to these simulations, U.S. anthropogenic emissions account for 78% of Nr
deposition over the CONUS. Foreign anthropogenic emissions and natural emissions
account for 6 and 16% respectively of total N deposition in this model simulation.
Background concentrations of SO2 were calculated using the MOZART-2 global model
of tropospheric chemistry (Horowitz et al.. 2003) and were presented in the 2008 ISA for
Sulfur Oxides (U.S. EPA. 2008c). Background SO2 concentrations are estimated to be
below 10 parts per trillion (ppt) over much of the U.S. Maximum background
concentrations of SO2 of -30 ppt are found in the western U.S. In the Northwest, where
there are large geothermal sources of SO2, the contribution of background sources to total
SO2 is 70 to 80%; however, absolute SO2 concentrations are still on the order of ~2 ppb
or less. With the exception of the West Coast, where volcanic SO2 emissions cause high
background concentrations, background sources contribute <1% to present-day SO2
concentrations in surface air in the CONUS. Over the eastern U.S., the predicted
background contribution to SOx deposition was <10% and even smaller (<1%) where
present-day SOx deposition is the highest. The predicted contribution of background
sources to S deposition was the highest in the western U.S. at >20% because of the
geothermal sources of SO2 and oxidation of DMS in surface water of the eastern Pacific.
In comparison, values observed at several relatively remote sites cited in the 2008 ISA
for Sulfur Oxides (U.S. EPA. 2008c) ranged from 20 to 40 ppt.
As noted earlier, volcanic sources of SO2 in the U.S. are found in the Pacific Northwest,
Alaska, and Hawaii. The greatest potential domestic effects from volcanic SO2 occur on
the island of Hawaii. Nearly continuous venting of SO2 from Mauna Loa and Kilauea
produces SO2 in high concentrations of ~5 ppm lasting for periods of up to 1 hour [see
Figure 2-34 and Figure 2-35 in the 2008 ISA for Sulfur Oxides; U.S. EPA (2008c)l at
two national park sites near the Kilauea caldera and the nearby east rift zone. The latter
emits several times as much SO2 as the Kilauea caldera. The two measurement sites
within the national park are <3 km from the summit emission source and ~10 km from
the east rift source and are affected by the two sources during southerly and easterly
winds. A number of communities and population centers are within the same distance
from the east rift gas source that affects these two monitoring sites. When the normal
trade wind flows are disrupted, emissions from the sources can be brought directly to
these various communities. Because these communities are located at a similar distance
from the large east rift emission source as the national nark monitoring stations, it is
2-109
-------
probable that these communities are subjected to SO2 concentrations as high as those
measured within Hawaii Volcanoes National Park.
When considering emission sources further afield, intercontinental transport of O3 and
PM has been the focus of efforts by the Task Force on Hemispheric Transport of Air
Pollution (HTAP). To the extent that N and S species are transported along with O3 and
PM, they also contribute to deposition following subsidence to the surface. Modeling
studies estimate that only a small fraction of nitrogen and sulfur emissions are transported
to and deposited within a continent different than the source of the emissions (Stock et
al.. 2013; Roy et al.. 2012; Sanderson et al.. 2008). Global-scale modeling results
reported in the 2008 ISA (U.S. EPA. 2008a) and in the latest ISAs for Oxides of Nitrogen
Health Criteria (U.S. EPA. 2016f) and Sulfur Oxides Health Criteria (U.S. EPA. 2017d)
also indicate that intercontinental transport of oxidized and reduced nitrogen, SO2, and
S042 are likely minor background sources of these species. Of greater importance are
localized emissions from natural sources. These include emissions of NO from soils and
lightning and emissions of SO2 from geothermal and biogenic sources.
Background PM that would occur in the U.S. in the absence of anthropogenic emissions
originates from natural and international sources. Natural sources include windblown
dust, wildfires, and sea salt. International contributions include intercontinental transport
of dust, wildfire smoke, and pollution as well as transboundary transport of these
contributors from Canada and Mexico. Background PM includes both primary and
secondary natural and anthropogenic contributions and usually makes a relatively small
contribution to urban annual average PM2 5 concentrations. However, it is an important
contributor to PM2 5 concentrations in the southwestern U.S. and affects PM2 5
concentrations elsewhere on an episodic basis. Background contributions to PM10-2.5 can
be substantial, as it is generally dominated by dust and sea salt (U.S. EPA. 2019).
Estimated PM2.5 background concentrations were estimated to be less than 1 (ig/m3 on an
annual basis, with maximum daily average values in a range from 3 to 20 (ig/m3 and a
peak of 63 (ig/m3 at the nine national park sites across the U.S. (U.S. EPA. 2009a).
According to the 2019 PM ISA (U.S. EPA. 2019). there has not been a similar
national-scale effort to update background PM2.5 concentration estimates since the 2009
PM ISA, but there has been considerable research focused on better understanding the
sources and processes that influence background contribution to PM2.5 in the U.S.
Background PM can be viewed as a combination of two conceptually separate
components: 1) a baseline component characterized by reasonably consistent distribution
of daily values each year, with some variability by region and season, and 2) an episodic
component consisting of infrequent contributions from high-concentration events,
including volcanic eruptions, wildland fires, and dust storms. On average, natural sources
2-110
-------
including soil dust and sea salt have been estimated to account for approximately 10% of
U.S. urban PM2.5 (U.S. EPA. 2019). However, PM concentrations of several hundred
(ig/m3 in areas impacted by wildland fires and >5,000 (ig/m3 during intense dust storms
were reported in the 2019 PM ISA, and longer fire seasons resulting from invasive
species, historical fire management practices, more frequent droughts, and extreme heat
have led to more large fires (U.S. EPA. 2019). Intercontinental transport contributes 0.05
to 0.15 (ig/m3 to annual average PM2.5 concentrations in the U.S. Although concentrations
are typically less than 1 (.ig/rn3 at U.S. sites, episodic contributions as high as 20 (.ig/rn3
have been estimated (U.S. EPA. 2019).
Just as pollutants can be transported into the U.S., they can also be transported outward.
For example, wet deposition of pollutants emitted in eastern North America occurs over
the North Atlantic Ocean. Deposition of N species is an important source of nutrients to
the western North Atlantic (Zamora et al.. 2011). Dennis et al. (2013). based on CMAQ
modeling results, estimated that -1/3 of oxidized N emissions and slightly less than 1/3
of NH3 emissions in the U.S. are transported out over the North Atlantic Ocean. Although
the average pH of rainwater at Bermuda is ~5, reflecting deposition of acidic species
emitted in North America, this additional source of acidity is at most only -2% of that
due to anthropogenic CO2 (Bates and Peters. 2007).
With well-validated models, it is possible to compare the relative role of different
emission and removal processes. Adjoint models (Henze et al.. 2009) are particularly
useful for understanding the relative contribution of emission sources to dry and wet
deposition of different nitrogen and sulfur containing compounds. For example, Lee et al.
(2016) found that half of nitrogen deposition at Federal Class I areas, such as national
parks, can be attributed to emission sources within 500 km and 90% of nitrogen
deposition is due to emission sources within 1,500 km. Malm et al. (2013) simulated
conservative tracer transport from ammonia source regions with the GEOS-Chem model
to estimate that roughly equal amounts of ammonia deposition in Rocky Mountain
National Park (RMNP), CO was from within or outside Colorado, with most of the
transport into Colorado coming from the West. Thompson et al. (2015) reported that 40%
reduced nitrogen deposition in RNMP was from outside Colorado.
Background rainwater pH and background deposition in remote areas worldwide has
considerably lower H+ and N deposition levels than in more populated areas, as described
in Appendix 2.6. Galloway et al. (1982) measured the pH of rainwater at five remote sites
worldwide and measured pH values ranging from 4.8 to 5.0. At some sites, acidity was
attributed to long-range transport of acid sulfate, while at others a mixture of strong and
weak acids attributed to both anthropogenic and natural sources was observed. They
concluded that a pH of 5 was a good lower limit estimate for natural contributions. Curtis
2-111
-------
et al. (2018) estimated nitrogen and sulfur deposition in remote inland areas of Greenland
and reported 0.13 to 0.19 kg N/ha/yr for total nitrogen, 0.8 to 0.11 kg N/ha/yr for NO;, .
0.05 to 0.09 kg N/ha/yr for NH/, and 0.08 to 0.13 kg S/ha yr for SO42 .
2.7 Supplemental Material on Changes in Deposition since 2000
Maps on the portion of the NADP website dedicated to the Total Deposition (TDEP)
program (http://nadp.slh.wise.edu/committees/tdep/tdepmaps/) and shown in this section
present a comprehensive overview of changes in various parameters related to deposition
over 2000-2018. Changes between two, 3-year periods, 2000 through 2002 and 2016
through 2018 are summarized in this section. The most notable changes in the
geographical distribution of total (wet + dry) deposition of NOy + NHX between these
two time periods are the large increase in total deposition of N in the upper Midwest
centered in Iowa and Minnesota, with an accompanying decrease in a large area further
east, extending from Illinois and Indiana to Pennsylvania and further south (see
Figure 2-51). There are also shifts in the distribution of wet and dry deposition in these
areas (see Figure 2-52. Figure 2-53. and Figure 2-54). Deposition of oxidized nitrogen
has declined markedly throughout the eastern U.S. and southern California between the
periods 2000-2002 and 2016-2018 (see Figure 2-55. Figure 2-56. Figure 2-57. and
Figure 2-58) due mainly to large decreases in dry deposition of total nitrate
(TNO3 = HNO3 + pNO;, : see Figure 2-59). The decreases in total nitrate deposition
across the CONUS are generally due to decreases in HNO3 deposition (see Figure 2-60).
Decreases in dry deposition of pNOs" have generally been smaller, reflecting the smaller
contribution of pNO, to TNO3. Exceptions include areas such as Florida, Texas, and
southern California where dry deposition of pNO, has been much greater than HNO3 in
the earlier period and has decreased substantially (see Figure 2-61). Other N species,
mainly NO2 also show large decreases (see Figure 2-62 and Figure 2-63). especially near
urban source areas. Using OMI data, Krotkov et al. (2016) found decreases in column
(vertically integrated) abundances ofN02 of -40% from 2005 to 2014 over the U.S.
In contrast to deposition of oxidized N, deposition of reduced inorganic N has seen large
increases (see Figure 2-64 and Figure 2-65). Several areas, including the upper Midwest
and San Joaquin Valley, have increased markedly in size or intensity (see Figure 2-64)
between the two periods. Dry deposition ofNFb has been the major contributor to the
increase (see Figure 2-66). and although dry deposition of pNH/ has largely decreased
between the two periods (see Figure 2-67). the greater contribution of NH3 has resulted in
an overall increase of total reduced nitrogen deposition in affected areas. Between the
two periods, emissions of NOx have decreased resulting in lower formation rates of
HNO3 that could react with NH3 to form PNH4NO3.
2-112
-------
Deposition of oxidized and reduced N have undergone geographic shifts with
corresponding shifts in the contributions of each to total N deposition on regional and
smaller scales (compare Figure 2-55 and Figure 2-56 to Figure 2-64 and Figure 2-65).
For example, deposition of oxidized N in the Northeast has decreased substantially, but
deposition of reduced N has increased in the central U.S. Sizable shifts are also seen in
the fractional contributions of total N deposition as dry deposition of both oxidized and
reduced forms (compare Figure 2-58 and Figure 2-69).
Declines in the deposition of S since 2000 have also occurred, particularly in the Ohio
River Valley (see Figure 2-70). with generally much smaller declines in wet deposition
(see Figure 2-71) than for dry deposition (see Figure 2-72). The proportion of S dry
deposited has also decreased, especially in the Ohio Valley and the rest of the eastern
U.S. (see Figure 2-73). Dry deposition of both SO2 and pS042 have both decreased in
areas with the greatest deposition (Figure 2-74 and Figure 2-75). These decreases are
consistent with those derived by Krotkov et al. (2016) who detected decreases in the
column (vertically integrated) abundance of SO2 of -75 % for the period 2005 to 2014
over the Ohio River Valley and southwestern Pennsylvania. These decreases reflect
reductions in emissions mandated by the Clean Air Act Amendments and other
regulatory requirements.
As noted earlier, these estimates of change in deposition were derived from CMAQ
output and data from measurement networks. Each of these components has its own set
of uncertainties, and the estimates of deposition and the changes over time should be
viewed in this light.
2-113
-------
Source: CASTNET/CMAQ/NADP
Total deposition of nitrogen 0002
USEPA 02/19/19
Total N
(kg-N/ha)
I
-8
-10
-12
I
¦->20
Source: CASTNET/CMAQ/NADP
Total deposition of nitrogen 1618
USEPA 10/21/19
Total N
(kg-N/ha)
N = nitrogen.
Source: CASTNET/CMAQ/NTN/AMON/SEARCH. We acknowledge the Total Deposition (TDep) Science Committee of the National
Atmospheric Deposition Program (NADP) for their role in making the TDep data and maps available.
Figure 2-51 Wet plus dry deposition of total nitrogen over 3-year periods.
Top: 2000-2002; Bottom: 2016-2018.
2-114
-------
Wet N deposition 0002
USEPA 09/12/18
Source: CASTNET/CMAQ/NADP
(kg-N/ha)
Wet N
Source: NADP/NTN & PRISM
USEPA/CAMD 10/18/19
/datas'ac/piism/pii^lfilS/inoas^n-ieiS
N = nitrogen.
Source: CASTNET/CMAQ/NTN/AMON/SEARCH.
Figure 2-52 Wet deposition of total nitrogen over 3-year periods.
Top: 2000-2002; Bottom: 2016-2018.
2-115
-------
Source: CASTNET/CMAQ/NADP USEPA 09/12/18
Dry N deposition 1618
USEPA 10/21/19
Source: CASTNET/CMAQ/NADP
Dry N
kg-N/ha)
-0
-1
-2
-3
-4
N = nitrogen..
Source: CASTNET/CMAQ/NTN/AMON/SEARCH. We acknowledge the Total Deposition (TDep) Science Committee of the National
Atmospheric Deposition Program (NADP) for their role in making the TDep data and maps available
Figure 2-53 Dry deposition of total nitrogen over 3-year periods.
Top: 2000-2002; Bottom: 2016-2018.
2-116
-------
Pet of total N as dry deposition 0002
USEPA 09/12/18
Dry N
(Pet of Total)
¦
-0
-10
-20
-30
-40
-50
-60
-70
-80
-90
«s
->100
Source: CASTNET/CMAQ/NADP
Dry N
(Pet of Total)
¦
-0
¦
-10
-20
-30
-40
-50
-60
-70
-80
-90
¦
->100
Source: CASTNET/CMAQ/NADP
Pet of total N as dry deposition 1618
USEPA 10/21/19
N = nitrogen.
Source: CASTNET/CMAQ/NTN/AMON/SEARCH. We acknowledge the Total Deposition (TDep) Science Committee of the National
Atmospheric Deposition Program (NADP) for their role in making the TDep data and maps available.
Figure 2-54 Percent of total nitrogen as dry deposition over 3-year periods.
Top: 2000-2002; Bottom: 2016-2018.
2-117
-------
Total deposition of oxidized N 0002
USEPA 09/12/18
Source: CASTNET/CMAQ/NADP
Total oxN
(kg-N/ha)
-0
-1
-2
-3
A
— *+
-5
-6
-7
-
-8
-9
->10
Total deposition of oxidized N 1618
USEPA 10/21/19
Source: CASTNET/CMAQ/NADP
Total oxN
(kg-N/ha)
oxN = oxidized nitrogen.
Source: CASTNET/CMAQ/NTN/AMON/SEARCH. We acknowledge the Total Deposition (TDep) Science Committee of the National
Atmospheric Deposition Program (NADP) for their role in making the TDep data and maps available.
Figure 2-55 Wet plus dry deposition of oxidized nitrogen over 3-year periods.
Top: 2000-2002; Bottom: 2016-2018.
2-118
-------
of total N
as
Pet
oxidized N 0002
Total oxN
(Pet of Total)
I
ff-zo
-30
-40
70
80
90
>100
Source: CASTNET/CMAQ/NADP USEPA 09/12/18
Source: CASTNET/CMAQ/NADP
-50
-60
E70
80
90
>100
Pet of total N as oxidized N 1618
USEPA 10/21/19
Total oxN
(Pet of Total)
oxN = oxidized nitrogen.
Source: CASTNET/CMAQ/NTN/AMON/SEARCH. We acknowledge the Total Deposition (TDep) Science Committee of the National
Atmospheric Deposition Program (NADP) for their role in making the TDep data and maps available
Figure 2-56 Percent of total nitrogen as oxidized nitrogen over 3-year periods.
Top: 2000-2002; Bottom: 2016-2018.
2-119
-------
Source: CASTNET/CMAQ/NADP
Dry oxN
(kg-N/ha)
|
-4
Dry deposition of oxidized N 0002
USEPA 09/12/18
Dry deposition of oxidized N 1618
USEPA 10/21/19
So luce: CASTNET/CMAQ/NADP
Dry oxN
(kg-N/ha)
-0
oxN = oxidized nitrogen.
Source: CASTNET/CMAQ/NTN/AMON/SEARCH. We acknowledge the Total Deposition (TDep) Science Committee of the National
Atmospheric Deposition Program (NADP) for their role in making the TDep data and maps available
Figure 2-57 Dry deposition of oxidized nitrogen over 3-year periods.
Top: 2000-2002; Bottom: 2016-2018.
2-120
-------
Source: CASTNET/CMAQ/NADP
Pet of total N as dry oxidized N 0002
USEPA 09/12/18
Dry oxN
(Pet of Total)
I
ff-20
-30
-40
Source: CASTNET/CMAQ/NADP
Pet of total N as dry oxidized N 1618
USEPA 10/21/19
DryoxN
(Pet of Total)
I
¦L20
-30
-40
oxN = oxidized nitrogen.
Source: CASTNET/CMAQ/NTN/AMON/SEARCH. We acknowledge the Total Deposition (TDep) Science Committee of the National
Atmospheric Deposition Program (NADP) for their role in making the TDep data and maps available
Figure 2-58 Percent of total nitrogen dry deposited as oxidized nitrogen over
3-year periods. Top: 2000-2002; Bottom: 2016-2018.
2-121
-------
-4.0
L,
®->6.0
Dry deposition of HN03 + pN03 0002
TTQT7PA no/n/lff
TN03
(kg-N/ha)
-0.0
-1.0
-2.0
-3.0
-3.0
Source: CASTNET/CMAQ/NADP
Dry deposition of HN03 + pN03 1618
USEPA 10/21/19
TN03
(kg-N/ha)
-0.0
TNO3 = nitric acid and particulate nitrate.
Source: CASTNET/CMAQ/NTN/AMON/SEARCH. We acknowledge the Total Deposition (TDep) Science Committee of the National
Atmospheric Deposition Program (NADP) for their role in making the TDep data and maps available
Figure 2-59 Combined dry deposition of nitric acid and particulate nitrate over
3-year periods. Top: 2000-2002; Bottom: 2016-2018.
2-122
-------
HN03
(kg-N/ha)
-0.0
-0.5
-1.0
-1.5
-2.0
ti-a,
3.0
3.5
>4.0
Dry deposition of nitric acid 0002
USEPA 09/12/18
Source: CASTNET/CMAQ/NADP
Dry deposition of nitric acid 1618
USEPA 10/21/19
Source: CASTNET/CMAQ/NADP
HNO3
(kg-N/ha)
-0.0
-0.5
-1.0
-1.5
-2.0
HN03 = nitric acid.
Source: CASTNET/CMAQ/NTN/AMON/SEARCH. We acknowledge the Total Deposition (TDep) Science Committee of the National
Atmospheric Deposition Program (NADP) for their role in making the TDep data and maps available.
Figure 2-60 Dry deposition of nitric acid over 3-year periods. Top: 2000-2002;
Bottom: 2016-2018.
2-123
-------
Source: CASTNET/CMAQ/NADP
Dry deposition of particle nitrate 0002
USEPA 09/12/18
pN03
(kg-N/ha)
-0.0
-0.2
-0.4
-0.6
-0.8
-1.0
-1.2
-1.4
1
-1.6
-1.8
a
->2.0
Source: CASTNET/CMAQ/NADP
Dry deposition of particle nitrate 1618
USEPA 10/21/19
pN03
(kg-N/ha)
¦
-0.0
-0.2
-0.4
-0.6
-0.8
-1.0
-1.2
-1.4
-1.6
-1.8
I
->2.0
pN03 = particulate nitrate.
Source: CASTNET/CMAQ/NTN/AMON/SEARCH. We acknowledge the Total Deposition (TDep) Science Committee of the National
Atmospheric Deposition Program (NADP) for their role in making the TDep data and maps available.
Figure 2-61 Dry deposition of particulate nitrate over 3-year periods.
Top: 2000-2002; Bottom: 2016-2018.
2-124
-------
Source: CASTNET/CMAQ/NADP
Dry deposition of unmeasured N species 0002
USEPA 09/12/18
Other N
(kg-N/ha)
-0.0
-0.5
-1.0
-1.5
-2.0
-2.5
-3.0
-3.5
-4.0
-4.5
I
->5.0
Dry deposition of unmeasured N species 1618
USEPA 10/21/19
Source: CASTNET/CMAQ/NADP
Other N
(kg-N/ha)
-0.0
-0.5
-1.0
-1.5
-2.0
-2.5
-3.0
-3.5
-4.0
-4.5
¦
->5.0
N = nitrogen.
Source: CASTNET/CMAQ/NTN/AMON/SEARCH. We acknowledge the Total Deposition (TDep) Science Committee of the National
Atmospheric Deposition Program (NADP) for their role in making the TDep data and maps available.
Figure 2-62 Dry deposition of modeled (unmeasured) nitrogen species over
3-year periods. Top: 2000-2002; Bottom: 2016-2018.
2-125
-------
Source: CASTNET/CMAQ/NADP
Pet of total N as unmeasured species 0002
USEPA 09/12/18
Other N
(Pet of TotaJ)
¦
-0
1
-5
-10
-15
-20
-25
-30
-35
-40
-45
-50
*
-55
¦
->60
Source: CASTNET/CMAQ/NADP
Pet of total N as unmeasured species 1618
USEPA 10/21/19
Other N
(Pet of Total)
I
-20
N = nitrogen..
Source: CASTNET/CMAQ/NTN/AMON/SEARCH. We acknowledge the Total Deposition (TDep) Science Committee of the National
Atmospheric Deposition Program (NADP) for their role in making the TDep data and maps available
Figure 2-63 Percent of total nitrogen as modeled (unmeasured) species over
3-year periods. Top: 2000-2002; Bottom: 2016-2018.
2-126
-------
Source: CASTNET/CMAQ/NADP
Total reN
(kg-N/ha)
-0
-1
-2
-3
-4
-5
-6
-7
1
- 8
-9
. J
->10
Total deposition of reduced N 0002
USEPA 09/12/18
Source: CASTNET/CMAQ/NADP
Total reN
(kg-N/ha)
¦
-0
-1
-2
-3
-4
Total deposition of reduced N 1618
USEPA 10/21/19
reN = reduced nitrogen.
Source: CASTNET/CMAQ/NTN/AMON/SEARCH. We acknowledge the Total Deposition (TDep) Science Committee of the National
Atmospheric Deposition Program (NADP) for their role in making the TDep data and maps available.
Figure 2-64 Wet plus dry deposition of reduced (inorganic) nitrogen over
3-year periods. Top: 2000-2002; Bottom: 2016-2018.
2-127
-------
Pet of total N as reduced N 0002
Total reN
(Pet ol Total)
170
80
90
>100
Source: CASTNET/CMAQ/NADP USEPA 09/12/18
Source: CASTNET/CMAQ/NADP
-50
-60
170
80
90
>100
Pet of total N as reduced N 1618
USEPA 10/21/19
Total reN
(Pet ol Total)
reN = reduced nitrogen.
Source: CASTNET/CMAQ/NTN/AMON/SEARCH. We acknowledge the Total Deposition (TDep) Science Committee of the National
Atmospheric Deposition Program (NADP) for their role in making the TDep data and maps available
Figure 2-65 Percent of total nitrogen deposition by reduced (inorganic)
nitrogen over 3-year periods. Top: 2000-2002;
Bottom: 2016-2018.
2-128
-------
Source: CASTNET/CMAQ/NADP
Dry deposition of ammonia 0002
USEPA 09/12/18
NH3
(kg-N/ha)
¦
-0.0
-1.0
-2.0
-3.0
-4.0
-5.0
-6.0
-7.0
1
->8.0
Source: CASTNET/CMAQ/NADP
Dry deposition of ammonia 1618
USEPA 10/21/19
(kg-N/ha)
-0.0
-1.0
-2.0
-3.0
-4.0
-5.0
-6.0
-7.0
¦
->8.0
NH3 = ammonia.
Source: CASTNET/CMAQ/NTN/AMON/SEARCH. We acknowledge the Total Deposition (TDep) Science Committee of the National
Atmospheric Deposition Program (NADP) for their role in making the TDep data and maps available.
Figure 2-66 Dry deposition of ammonia over 3-year periods. Top: 2000-2002;
Bottom: 2016-2018.
2-129
-------
Source: CASTNET/CMAQ/NADP
pNH4
(kg-N/ha)
[0.0
0,
0.2
-0.3
-0.4
0.5
0.6
0.7
0.8
0.9
>1.0
Dry deposition of particle ammonium 0002
USEPA 09/12/18
pNH4
(kg-N/ha)
-0.0
-0.1
-0.2
-0.3
-0.4
-0.5
-0.6
-0.7
-0.8
-0.9
¦
->1.0
Dry deposition of particle ammonium 1618
USEPA 10/21/19
Source: CASTNET/CMAQ/NADP
pNH4 = particulate ammonium.
Source: CASTNET/CMAQ/NTN/AMON/SEARCH. We acknowledge the Total Deposition (TDep) Science Committee of the National
Atmospheric Deposition Program (NADP) for their role in making the TDep data and maps available.
Figure 2-67 Dry deposition of particulate ammonium over 3-year periods.
Top: 2000-2002; Bottom: 2016-2018.
2-130
-------
Source: CASTNET/CMAQ/NADP
Dry deposition of reduced N 0002
USEPA 09/12/18
Dry reN
(kg-N/ha)
-0
Source: CASTNET/CMAQ/NADP
Dry deposition of reduced N 1618
USEPA 10/21/19
r reN
(kg-N/ha)
reN = reduced nitrogen.
Source: CASTNET/CMAQ/NTN/AMON/SEARCH. We acknowledge the Total Deposition (TDep) Science Committee of the National
Atmospheric Deposition Program (NADP) for their role in making the TDep data and maps available.
Figure 2-68 Dry deposition of reduced (inorganic) nitrogen over 3-year
periods. Top: 2000-2002; Bottom: 2016-2018.
2-131
-------
Pet of total N as dry reduced N 0002
USEPA 09/12/18
Dry reN
(Pet of Total)
¦
-0
- 10
-20
-30
-40
-50
-60
-70
-80
-90
1
->100
Source: CASTNET/CMAQ/NADP
Pet of total N as dry reduced N 1618
USEPA 10/21/19
Source: CASTNET/CMAQ/NADP
reN = reduced nitrogen.
Source: CASTNET/CMAQ/NTN/AMON/SEARCH. We acknowledge the Total Deposition (TDep) Science Committee of the National
Atmospheric Deposition Program (NADP) for their role in making the TDep data and maps available
Figure 2-69 Percent of total nitrogen deposition by dry reduced (inorganic)
nitrogen over 3-year periods. Top: 2000-2002;
Bottom: 2016-2018.
2-132
-------
Souice: CASTNET/CMAQ/NADP
Total deposition of sulfur 0002
USEPA 09/12/18
Total S
(kg-S/ha)
-0
-2
-4
-6
-8
-10
-12
-14
-16
18
I
L >20
Souice: CASTNET/CMAQ/NADP
Total deposition of sulfur 1618
USEPA 10/21/19
otal S
g-S/ha)
S = sulfur.
Source: CASTNET/CMAQ/NTN/AMON/SEARCH. We acknowledge the Total Deposition (TDep) Science Committee of the National
Atmospheric Deposition Program (NADP) for their role in making the TDep data and maps available.
Figure 2-70 Wet plus dry deposition of total sulfur over 3-year periods.
Top: 2000-2002; Bottom: 2016-2018.
2-133
-------
Source: CASTNET/CMAQ/NADP
USEPA 09/12/18
Source: CASTNET/CMAQ/NADP USEPA 10/21/19
S = sulfur.
Source: CASTNET/CMAQ/NTN/AMON/SEARCH.
Figure 2-71 Wet deposition of total sulfur over 3-year periods.
Top: 2000-2002; Bottom: 2016-2018.
2-134
-------
Source: CASTNET/CMAQ/NADP
Dry S deposition 0002
USEPA 09/12/18
Dry S
(kg-S/ha)
¦
-0
-2
-4
-6
-8
Source: CASTNET/CMAQ/NADP
USEPA 10/21/19
S = sulfur.
Source: CASTNET/CMAQ/NTN/AMON/SEARCH. We acknowledge the Total Deposition (TDep) Science Committee of the National
Atmospheric Deposition Program (NADP) for their role in making the TDep data and maps available.
Figure 2-72 Dry deposition of total sulfur over 3-year periods.
Top: 2000-2002; Bottom: 2016-2018.
2-135
-------
Pet of total S as dry deposition 0002
USEPA 09/12/18
Source: CASTNET/CMAQ/NADP
Dry S
(Pet of Total)
IP
-0
-10
-20
-30
-40
-50
-60
-70
-80
-90
I
->100
Source: CASTNET/CMAQ/NADP
Pet of tota] S as dry deposition 1618
USEPA 10/21/19
Dry S
(Pet of Total)
¦
-0
-10
-20
¦
-30
-40
-50
-60
-70
-80
-90
1
->100
S = sulfur.
Source: CASTNET/CMAQ/NTN/AMON/SEARCH. We acknowledge the Total Deposition (TDep) Science Committee of the National
Atmospheric Deposition Program (NADP) for their role in making the TDep data and maps available.
Figure 2-73 Percent of total sulfur deposition by dry deposition over 3-year
periods. Top: 2000-2002; Bottom: 2016-2018.
2-136
-------
Source: CASTNET/CMAQ/NADP
Dry deposition of sulfur dioxide 0002
USEPA 09/12/18
S02
(kg-S/ha)
¦
-0
-2
-4
-6
-8
-10
-12
-14
-16
-18
I
->20
Source: CASTNET/CMAQ/NADP
Dry deposition of sulfur dioxide 1618
USEPA 10/21/19
S02
(kg-S/ha)
S02 = sulfur dioxide.
Source: CASTNET/CMAQ/NTN/AMON/SEARCH. We acknowledge the Total Deposition (TDep) Science Committee of the National
Atmospheric Deposition Program (NADP) for their role in making the TDep data and maps available.
Figure 2-74 Dry deposition of sulfur dioxide over 3-year periods.
Top: 2000-2002; Bottom: 2016-2018.
2-137
-------
Dry deposition of particle sulfate 0002
USEPA 09/12/18
Source: CASTNET/CMAQ/NADP
pS04
(kg-S/ha)
-0.0
pS04
(kg-S/ha)
1
-0.0
-0.1
-0.2
-0.3
-0.4
-0.5
-0.6
-0.7
I
-0.8
-0.9
1
->1.0
Dry deposition of particle sulfate 1618
USEPA 10/21/19
Source: CASTNET/CMAQ/NADP
pS04 = particulate sulfate.
Source: CASTNET/CMAQ/NTN/AMON/SEARCH. We acknowledge the Total Deposition (TDep) Science Committee of the National
Atmospheric Deposition Program (NADP) for their role in making the TDep data and maps available
Figure 2-75 Dry deposition of particulate sulfate over 3-year periods.
Top: 2000-2002; Bottom: 2016-2018.
2-138
-------
APPENDIX 3 DIRECT PHYTOTOXIC EFFECTS OF
GASEOUS OXIDIZED NITROGEN
AND SULFUR ON VEGETATION
This appendix provides a brief overview of the exposure and phytotoxic effects of
gaseous forms of oxidized nitrogen (N) and sulfur (S) compounds on vegetation. The
main focus of this ISA is understanding the ecological impact of oxidized N and S and
that the major effects of these compounds on ecosystems are through acidifying
deposition and N enrichment deposition. However, direct effects of gaseous oxidized N
and S could augment the effects of deposition on vegetation, and direct effects of gaseous
N and S may be apparent in some areas. The effect of sulfur dioxide (SO2) gas on
vegetation is discussed in Appendix 3.2. Appendix 3.3 discusses the effects of nitric
oxide (NO), nitrogen dioxide (NO2), and peroxyacetyl nitrate (PAN) on vegetation.
Appendix 3.4. presents information on the direct effects of nitric acid (HNO3) vapor on
vegetation, including lichens. A summary section with causal determinations based on a
synthesis of the body of information on the biological effects of exposure to these gases
is presented in Appendix 3.5.
3.1 Introduction
The effects of gaseous pollutants such as SO2, NO2, NO, HNO3, and ozone (O3) on
vegetation have been studied since at least the early 19th century (Holmes et al.. 1915;
Havwood. 1905). Methodologies have been developed to study the effects of gaseous
exposures to these pollutants in the laboratory, greenhouse, and in the field. The
methodologies to study the effects of gaseous pollutants on vegetation have been recently
reviewed in the 2013 Ozone ISA and 2006 Ozone AQCD [Air Quality Criteria
Document; U.S. EPA (2013c); U.S. EPA (2006a)l. A thorough description of the
methodologies used to expose vegetation to gaseous pollutants can be found in
Section 9.2 of the 2013 Ozone ISA (U.S. EPA. 2013c). AX9.1 of the 2006 Ozone AQCD
(U.S. EPA. 2006a). and Section 9.2 in the 1993 Oxides of Nitrogen AQCD (U.S. EPA.
1993).
Uptake of gaseous pollutants in a vascular plant canopy is a complex process involving
adsorption to surfaces (leaves, stems, and soil) and absorption into leaves. These
pollutants penetrate into leaves primarily in gaseous forms through the stomata. The
surface cuticle provides a protective barrier to gaseous pollutant exposure, although there
is evidence for limited uptake across the cuticle (Zhang et al.. 2003; Kerstiens et al..
1992). Pollutants must be transported from the bulk air to the leaf boundary layer to reach
3-1
-------
the stomata. The transport of pollutants through a boundary layer into the stomatal region
is by diffusion. Studies of transport through the boundary layer are based on aerodynamic
concepts and usually relate to smooth surfaces that are not typical of leaf-surface
morphology (Gates. 1968). Once through the boundary layer, the gas enters the leaf
through the stomata. The entry of gases into a leaf is dependent upon gas-phase chemical
processes and physical characteristics of surfaces, including stomatal aperture. The
aperture of the stomata is controlled largely by the prevailing environmental conditions,
such as humidity, temperature, light intensity, and water availability. When the stomata
are closed, as occurs under dark or drought conditions, resistance to gas uptake is very
high and the plant has a very low degree of susceptibility to injury (Figure 3-1). The
stomatal control of uptake of gaseous pollutants is described in more detail in AX9.2 of
the 2006 Ozone AQCD (U.S. EPA. 2006a) and Section 9.3.1.5 of the 1993 Oxides of
Nitrogen AQCD (U.S. EPA. 1993). Note that unlike vascular plants, mosses and lichens
do not have a protective cuticle barrier to gaseous pollutants, which is a major reason for
their sensitivity to gaseous S and N.
3-2
-------
Light
^3' NOx, SOx
Cuticle
Epidermis
Pallisade
Mesophyll
Spongy
Mesophyll
Epidermis
Cuticle
h2o
C0=[C02]
_ jj
Guard Cell
03, NOX) SOx
Vascular
-System
C, = internal C02 in leaf; C0 = C02 of the atmospheric air; C02 = carbon dioxide; H20 = water; SOx = sulfur oxides; NOx = oxides of
nitrogen; 03 = ozone.
Source: U.S. Environmental Protection Agency CU.S. EPA. 2008a1.
Figure 3-1 The microarchitecture of a dicot leaf. While details among species
vary, the general overview remains the same. Light that drives
photosynthesis generally falls upon the upper (adaxial) leaf
surface. Carbon dioxide, oxides of sulfur, oxides of nitrogen, and
ozone gases generally enter by diffusion through the guard cells
(or stomata) on the lower (abaxial) leaf surface, while water vapor
exits through the stomata (transpiration).
3.2 Direct Phytotoxic Effects of Sulfur Dioxide on Vegetation
It has been known since the early 1900s that exposure to SO2 can cause plant damage and
death (Wislicenus. 1914). The large sources of historic SO2 emissions were ore smelters.
Sulfides in the ore were oxidized during smelting and resulted in large releases of SO2.
Emissions from large ore smelters in the U.S. and Canada resulted in large areas denuded
of vegetation surrounding these facilities (Thomas. 1951; Swain. 1949). Much of the
3-3
-------
damage to the vegetation was due to acute effects of high concentrations of SO2.
However, as early as 1923, researchers recognized that SO2 might decrease plant growth
without producing acute symptoms of foliar injury (Stoklasa. 1923). In the 1950s through
the early 1980s, there was much research on the effects of SO2 on vegetation, as well as
the interaction with pollutants such as O3 and NO2. Since then, there has been much less
research on the effects of SO2 on vegetation, especially in the U.S., due to the decreasing
ambient concentrations of SO2 ITJ.S. EPA (2012a); see Figure 3-2 for max 3-hour SO2
concentrations for 2016], The effects of SO2 on vegetation are summarized below.
Currently, SO2 is the only criteria pollutant with a secondary NAAQS distinct from the
primary standard. Other criteria air pollutants produce adverse welfare effects, but the
secondary and primary NAAQS have been set to be identical. The SO2 NAAQS is to
protect acute foliar injury resulting from SO2 exposure. The standard is a 3-hour average
of 0.50 ppm and was promulgated in 1971 to protect against the adverse effects of acute
foliar injury in vegetation. The 1982 AQCD for Particulate Matter and Sulfur Oxides
concluded that controlled experiments and field observations of vegetation supported
retaining this secondary standard (U.S. EPA. 1982b. d, 1971).
Acute foliar injury usually occurs within hours of exposure, involves a rapid absorption
of atoxic dose, and involves a collapse or necrosis of plant tissues. Another type of
visible injury is termed chronic injury and is usually a result of variable SO2 exposures
over the growing season. After entering the leaf, SO2 is converted to sulfite (SOr ) and
bisulfite (HSO;, ) ions, which may be oxidized to sulfate (SO42 ). Sulfate is about
30 times less toxic than sulfite and bisulfite. The conversion of sulfite and bisulfite to
sulfate results in net H+ production in the cells. Kropff (1991) proposed that the
appearance of S02-induced leaf injury was likely due to a disturbance of intracellular pH
regulation. Kropff (1991) listed several studies in which the pH of homogenates of leaf
cells only shifted towards greater acidity when plants were lethally damaged from
long-term SO2 exposures (Jager and Klein. 1977; Grill. 1971; Thomas et al.. 1944). The
appearance of foliar injury can vary significantly among species and growth conditions
(which affect stomatal conductance). Currently, there is no regular monitoring for SO2
foliar injury effects in the U.S.
3-4
-------
Source: U.S. Environmental Protection Agency 2016 analysis of data from state and local air monitoring stations.
Figure 3-2 Map of maximum 3-hour daily max average sulfur dioxide
concentration reported at Air Quality System monitoring sites for
2016.
Besides foliar injury, long-term lower SO2 concentrations can result in decreased
photosynthesis, growth, and yield of plants. These effects are cumulative over the
growing season and are often not associated with visible foliar injury. As with foliar
injury, the effects of these injuries vary among species and growing environment. The
1982 Particulate Matter and Oxides of Sulfur (PM-SOx) AQCD summarized the
concentration-response information available at the time (U.S. EPA. 1982b). Effects on
growth and yield of vegetation were associated with increased SO2 exposure
concentration and time of exposure. However, that document concluded that more
definitive concentration-response studies were needed before useable exposure metrics
could be identified. Because ambient SO2 concentrations declined and focus on O3
vegetation effects research increased, relatively few studies have emerged to better
3-5
-------
inform a metric and levels of concern for effects of SO2 on growth and productivity of
vegetation.
SO2 is considered to be the primary factor causing the death of lichens in many urban and
industrial areas, with fruticose lichens being more susceptible to SO2 than many foliose
and crustose species (Hutchinson et al.. 1996). Damage to lichens in response to SO2
exposure includes reduced photosynthesis and respiration, damage to the algal
component of the lichen, leakage of electrolytes, inhibition of N fixation, reduced K+
absorption, and structural changes (Hutchinson et al.. 1996; Belnap et al.. 1993; Farmer et
al.. 1992). Significant reductions in lichen photosynthesis have been measured at
concentrations as low as 91 ppb over 2-4 hours (Sanz et al.. 1992; Huebert et al.. 1985).
Damage to the algal component of the thallus is evidenced by its discoloration. The entire
thallus dies soon after algal cells are damaged (Hutchinson et al.. 1996). At higher
concentrations, SO2 deactivates enzymes by chemical modification, leading to reduced
metabolic activity and loss of membrane integrity (Nieboer et al.. 1976; Ziegler. 1973). In
addition, SO2 binds to the central metal atoms of enzymes, adversely affecting membrane
function and cell osmolality. SO2 also competitively inhibits bicarbonate (HCO, ) and
dihydrogen phosphate (H2PO4 ) interactions with enzymes (Hutchinson et al.. 1996).
Low pH increases the toxicity of SO2 action (Farmer et al.. 1992). The toxic effects of
atmospheric deposition of SO2 are lessened when lichens are attached to a substrate,
typically bark or rock, that has high pH or superior buffering capacity (Richardson and
Cameron. 2004). van Herk (2001) evaluated relationships between bark pH and air
pollution levels as two significant variables affecting epiphytic lichen composition and
concluded that bark pH was the primary factor regulating the distribution of acidophilic
species in the Netherlands. In studies of unpolluted areas, differences in bark chemistry
also affect the presence and distribution of epiphytes (Farmer et al.. 1992). Indirect
changes to bark pH, caused by acidification and high SO2 concentrations, also affect
lichen distribution (Farmer et al.. 1992). More recently, Geiser and Neitlich (2007)
reported that direct SO2 damage to lichens in the Pacific Northwest may have been
confined to major urban areas such as Seattle, Portland, and Bellingham. However, lichen
monitoring plots were not colocated with SO2 monitors and the authors were not able to
quantify SO2 exposure. More information on the N effects reported in this study is found
in Appendix 6.3.7 and Appendix 6.5.2.
More recent research has been performed in areas of Europe where ambient SO2
concentrations are generally higher than in the U.S. Since the 2008 ISA, several studies
in Germany and some eastern European countries have indicated that direct effects of
SO2 caused growth reductions in trees during the last century (Cavlovic et al.. 2015;
Hauck et al.. 2012; Rvdval and Wilson. 2012; Elling et al.. 2009). Elling et al. (2009)
evaluated a large database providing long-term growth records of 1,010 silver firs (Abies
3-6
-------
alba), long-term climate records, and long-term air pollution data from 51 sites in
southern Germany. In this analysis, silver fir growth was influenced by SO2 pollution
more than any other factor in the second half of the last century. The study authors also
reported an almost immediate increase in growth in response to falling SO2 emissions in
the 1980s. This rapid response indicates a direct effect of gaseous SO2 rather than an
indirect effect of soil acidification, which would have a longer response time. Annual
average concentrations above 10 j_ig SCVm3 (approximately 4 ppb) appeared to reduce
growth of silver fir. (See Figure 2-15 for recent concentrations in the continental U.S.). In
a later publication, Bosela et al. (2014) argued a combination of SO2 and NOx gas
emissions has historically reduced growth in silver fir in the western Carpathian
Mountains of the Czech Republic.
A similar gaseous SO2 effect on tree growth may have been occurring in the eastern U.S.
Using tree ring analysis, Thomas et al. (2013) reported significant growth increases in
old-growth eastern red cedar (Juniperus virginiana) in West Virginia following decreases
in SO2 emissions since 1980. Growth continued to increase as SO2 emissions further
declined in the 1990s and 2000s. Thomas et al. (2013) also found evidence of
physiological changes in response to SO2 emissions. The authors attributed the growth
response to an indirect effect of decreasing acidifying deposition, and thus, recovery from
soil acidification. However, a historical record of acidifying deposition was not available.
As in Europe, the trees studied in West Virginia also had a relatively rapid recovery in
response to declining SO2 emissions that could indicate the effects were from direct
exposure to gases in the atmosphere rather than soil acidification. Further, a response to
this study from other researchers suggested that the eastern red cedars in the West
Virginia study were found on a limestone outcrop that could be well buffered from soil
acidification (Schaberg et al.. 2014). This study may indicate that gaseous SO2 alone or in
combination with other gases may have inhibited red cedar growth. See Appendix 5.2.1.3
for further discussion of this study.
3.3 Direct Phytotoxic Effects of Nitric Oxide, Nitrogen Dioxide,
and Peroxyacetyl Nitrate
In sufficient concentrations, nitric oxide (NO) and nitrogen dioxide (NO2) can have
phytotoxic effects on plants by decreasing photosynthesis and inducing visible foliar
injury (U.S. EPA. 1993). The current secondary (welfare) and primary (human health)
standard for oxides of nitrogen is aNCh annual mean of 0.053 ppm. See Figure 3-3 for
recent concentrations of NO2. The 1993 Oxides of Nitrogen AQCD concluded that
concentrations of NO2 or NO in the atmosphere are rarely high enough to have
phytotoxic effects on vegetation (U.S. EPA. 1993). Since the 1993 Oxides of Nitrogen
3-7
-------
AQCD, very little new research has been done on these phytotoxic effects to alter this
conclusion (Bender and Weigel. 2011). However, it is known that these gases alter the N
cycle in some ecosystems, especially in the western U.S., and contribute to N saturation
(Sparks. 2009; Fenn et al.. 2003a; Bvtnerowicz and Fenn. 1996). See Appendix 6.1 for a
discussion of the nutrient effects of N.
N02 = nitrogen dioxide.
Note: Concentrations indicated are the highest concentration in the county and do not represent countywide concentrations.
Source: U.S. Environmental Protection Agency 2014 analysis of data from state and local air monitoring stations CU.S. EPA. 2016f).
Figure 3-3 Map of U.S. annual average nitrogen dioxide concentrations for
2013.
In general, NO and NO2 enters leaves through the stomata (Saxe. 1986). However, the
leaf cuticle could be an important receptor for NO2, and there is evidence of transport of
NO and NO2 across isolated cuticles (Lendzian and Kerstiens. 1988). Several studies
3-8
-------
have demonstrated that plant canopies can directly assimilate N in the form of NO2, but
canopy uptake of NO2 is generally small relative to total plant uptake (Vallano and
Sparks. 2008; Ammann et al.. 1995; Nussbaum et al.. 1993; Segschneider et al.. 1993;
von Ballmoos et al.. 1993; Hanson et al.. 1989). After entering the leaves, NO2 dissolves
in the extracellular water of the substomatal cavity to form HNO2 and HNO3, which then
dissociate to form NO2 , NO;, . and H+ (Bvtnerowicz et al.. 1998). Both cell and tonoplast
membranes contain ATP-dependent H+ pumps, and the tonoplast pumps are strongly
inhibited by NO;, (Bvtnerowicz et al.. 1998). If extra protons are deposited in vacuoles of
the plant cells during normal cellular regulation, then additional acidity will occur in
combination with additional NO3 . This combination can cause disruptions in cellular
control (Taylor and MacLean. 1970). NO; and nitrite (NO2 ) are metabolized to amino
acids and proteins through a series of enzymatic reactions mainly involving NO; and
nitrite reductases (Amundson and MacLean. 1982). The ability of plants to reduce NO;
and N02 to amino acids and proteins determines the potential of the plant to detoxify
NO and NO2 (Wellburn. 1990). Reduction of NO; takes place outside of the chloroplast
while the reduction of NO2 is coupled with the light reactions of photosynthesis.
Therefore, when leaves are exposed to NO and NO2 in the dark, highly phytotoxic levels
of NO2 accumulate and may lead to greater toxicity to NO and NO2 at night (Amundson
and MacLean. 1982). Exposure to NO produces both NO3 and NO2 in the leaves, but
the rate of NO3 accumulation is much slower than NO2 . Thus, plants exposed to high
NO could be at risk to elevated concentrations of NO2 (Wellburn. 1990). More detailed
information on the cellular effects of NO and NO2 can be found in the 1993 Oxides of
Nitrogen AQCD.
The functional relationship between ambient concentrations of NO or NO2 and a specific
plant response, such as foliar injury or growth, is complex. Factors such as inherent rates
of stomatal conductance and detoxification mechanisms and external factors, including
plant water status, light, temperature, humidity, and the particular pollutant exposure
regime, all affect the amount of a pollutant needed to cause symptoms of foliar injury.
Plant age and growing conditions and experimental exposure techniques also vary widely
among studies quantifying the response of plants to NO2. An analysis conducted in the
1993 Oxides of Nitrogen AQCD of over 50 peer-reviewed reports on the effects of NO2
on foliar injury indicated that plants are relatively resistant to NO2, especially compared
to foliar injury caused by exposure to O3 (U.S. EPA. 1993). With few exceptions, visible
injury has not been reported at concentrations below 0.20 ppm, and these exceptions
occurred when the cumulative duration of exposures extended to 100 hours or longer. At
0.25 ppm, increased leaf abscission was reported on navel orange trees (Citrus sinensis),
but only after exposures in excess of 1,000 hours (Thompson et al.. 1970). Green bean
(Phaseolus vulgaris) plants used as bioindicators of NO2 injury in Israel developed foliar
injury symptoms when ambient concentrations exceeded 0.5 ppm (Donagi and Goren.
3-9
-------
1979). In most plants, injury occurred in less than 1 day only when concentrations
exceeded 1 ppm (U.S. EPA. 1993). In recent years (2011-2013), ambient hourly NO2
concentrations in the U.S. have been well below the exposures in the above studies with
maximum highest 1-hour daily concentrations less than 0.075 ppm [see Section 2.5.2 of
U.S. EPA (2016f)l.
Decreased rates of photosynthesis have been recorded in experimental exposures of
plants to both NO and NO2, but usually at concentrations significantly higher than would
normally be encountered in ambient air. For example, Sabaratnam et al. (1988) reported
that soybeans (Glycine max) exposed 7 hours/day for 5 days showed an increase in
photosynthesis at a concentration of 0.2 ppm but a decrease in net photosynthesis at a
concentration of 0.5 ppm. Short-term exposures of soybean to 0.6 ppm NO2 for 2 to
3 hours also had no effect on net photosynthesis (Carlson. 1983). Most plants appear to
be more susceptible to NO than to NO2, as shown by Saxe (1986). who exposed a variety
of horticultural plants raised in greenhouses (species of Hec/era. Ficus, Hibiscus,
Nephrolepis, and Dieffenbctchict) to both NO and NO2. Saxe (1986) reported that
decreases in net photosynthesis occurred at doses of NO that were 22 times less than that
for NO2. However, these decreases in net photosynthesis required concentrations as high
as 1 ppm NO for 12 hours to elicit a response in these plants.
In the 1970s and 1980s, hundreds of studies were conducted on the effects of NO2 on
growth and yield of plants. These studies varied widely in plant species, growing
conditions, exposure equipment, concentrations, durations, exposure regimes, and
environmental conditions during exposures. No clear dose-response relationships for
exposure to NO2 and reductions in growth and/or yield of plants emerged from these
experiments. Readers are referred to the analysis of over 100 studies conducted in the
1993 Oxides of Nitrogen AQCD. A few key studies are highlighted in this section. The
growth of several plant species appears to be susceptible to concentrations of NO2 less
than 0.2 ppm, particularly when exposure occurs during low light conditions. For
example, nearly continuous exposure to 0.1 ppm NO2 for 8 weeks significantly reduced
growth of Kentucky blue grass [Poa praiensis: Ashenden (1979); Whitmore and
Mansfield (1983)1. Eight species of tree seedlings were exposed to 0.1 ppm NO2 for
6 hours/day for 28 days, resulting in reduced shoot or root growth in two species, white
ash (Fraxinus americana) and sweetgum (Liquidambar styracifliia), reduced height
growth in two clones of loblolly pine (Finns taeda), and no effects on the other species
(Kress and Skellv. 1982). No effects of NO2 at 0.1 ppm or lower were observed on
numerous other species, including potato (Solamim tuberosum), black poplar (Populus
nigra), radish (Raphanus sativus), soybean, or peas [Pisum sativum; U.S. EPA (1993)1.
No effects of NO2 were observed on soybeans grown in field plots subjected to a series of
10 episodic exposures averaging 0.4 ppm for 2.5 or 3 hours (Irving et al.. 1982).
3-10
-------
Numerous studies have reported negative effects on growth of a variety of plants exposed
to 0.5 ppm NO2 and above (U.S. EPA. 1993). but these concentrations are very high
relative to current ambient levels of NO2 (see Appendix 2.6.1).
The 1993 Oxides of Nitrogen AQCD reviewed the extensive literature on the effects of
NO2 in combination with other gaseous air pollutants, particularly SO2 and O3, and
concluded that combinations of pollutants can cause foliar injury or decreases in
photosynthesis at concentrations lower than those associated with NO2 acting alone.
However, the plant responses occur at concentrations much higher than typically found in
ambient air in the U.S. (U.S. EPA. 1993). In addition, the presence of NO2 in studies
combining other gases did not produce symptoms different from those caused by the
dominant pollutant, either SO2 or O3, such that a plant response produced by
combinations of NO2 with other air pollutants in the field would be difficult, if not
impossible, to distinguish from those of the other single pollutants (U.S. EPA. 1993).
Since the 1993 Oxides of Nitrogen AQCD was completed, most new research on NO2
exposure to vegetation has taken place in Europe and other areas outside the U.S. For
example, foliar NO3 reductase activity was increased in Norway spruce (Picea abies)
growing near a highway in Switzerland with average exposures of about 0.027 ppm
compared to trees growing 1,300 m away from the highway with NO2 exposures less than
0.005 ppm (Ammann et al.. 1995). This result was consistent with other studies on
Norway spruce in the field and laboratory (von Ballmoos et al.. 1993; Thoene et al..
1991). Muller et al. (1996) found that the uptake rate of NO;, by roots of Norway spruce
seedlings was decreased by exposure to 0.1 ppm ofN02 for 48 hours. Similarly, soybean
plants grown in Australia had decreased NO3 uptake by roots and reduced growth of
plants exposed to 1.1 ppm ofN02 for 7 days (Qiao and Murray. 1998). In a Swiss study,
poplar (Populus x euramericana) cuttings exposed to 0.1 ppm ofN02 for approximately
12 weeks resulted in decreased stomatal density and increased specific leaf weight, but
did not result in other effects such as leaf injury or a change in growth (Giinthardtgoerg et
al.. 1996). However, NO2 enhanced negative effects of ozone on poplars, including leaf
injury, when the pollutants were applied in combination (Giinthardtgoerg et al.. 1996).
Since the 2008 ISA for Oxides of Nitrogen and Oxides of Sulfur-Ecological Criteria
[hereafter referred to as the 2008 ISA (U.S. EPA. 2008a)l. very few studies have been
published on the direct effects of NO and NO2 on vegetation. Hu et al. (2015b) exposed
clonal hybrid poplar (Populus alba x Populus berolinensis) saplings to 4 ppm of NO2.
The authors reported significant declines in photosynthesis and dark respiration with
exposures of 48 hours. They also reported stomatal dysfunction at this level of exposure
resulting in partial stomatal closure and a decline in stomatal conductance. However,
4 ppm of NO2 is very high relative to current ambient levels of NO2 in the U.S. (see
3-11
-------
Appendix 2.6.1). These results are consistent with past studies of plants with relatively
high NO2 exposure.
As part of a broad study of N deposition of a native serpentine grassland in California,
Vallano et al. (2012) exposed five native grasses and forbs (Plantcigo erectci, Layia
gaillardioides, Lasthenict californica, Viilpia microstcichvs, and Crvptantha flaccida) and
the most common invasive grass Loliiim miiltifloriim to NO2 and soil N addition. The
plants were exposed in a growth chamber to very low levels of NO2 (0.03 ppm) to
simulate recent ambient air concentrations for this ecosystem. At this relatively low NO2
exposure, no significant effects were found on shoot biomass, root biomass,
photosynthesis, or stomatal conductance. The authors reported that despite not finding
species responses to NO2 exposure, the additive effects of NO2 combined with soil N on
plant performance indicate that uptake of NO2 may play a role in species responses to
increasing N deposition. The authors found that the combined NO2 and N addition
resulted in a strong positive growth and competitive response in the invasive Loliiim. This
result is consistent with previous findings that low levels of NO2 that are not phytotoxic
to plants can add to the N load to an ecosystem from uptake through leaves (Sparks.
2009).
In a study in the Grand Canyon National Park, AZ, Kenkel et al. (2016) found that NOx
concentrations measured by Ogawa passive samplers were about 52% higher along
roadsides than 30 m away from the road. The pattern of the amount of 15N in pinyon pine
(Pinus edulis) mirrored the concentration gradient of NOx concentrations from the road,
indicating that the vegetation is taking up N from the vehicle traffic in the park. The
authors reported that sustained chronic N deposition on this arid environment could result
in deleterious effects for these ecosystems.
Peroxyacetyl nitrate (PAN) is a well-known photochemical oxidant that often co-occurs
with O3 during high photochemical episodes and that has been shown to cause injury to
vegetation [see reviews by Cape (2003). Kleindienst (1994). and Temple and Taylor
(1983)1. Acute foliar injury symptoms resulting from exposure to PAN are generally
characterized as a glazing, bronzing, or silvering of the underside of the leaf surface;
some sensitive plant species include spinach, Swiss chard, lettuces, and tomatoes
(Temple and Taylor. 1983). Petunias (Petunia hybrida) have also been characterized as
sensitive to PAN exposures and have been used as bioindicators in areas of Japan
(Nouchi et al.. 1984). Controlled experiments have also shown significant negative
effects on the net photosynthesis and growth of petunias and kidney beans (Phaseolus
vulgaris) after exposure of 30 ppb of PAN for 4 hours on each of 3 alternate days (Cape.
2003; Izuta et al.. 1993). As mentioned previously, it is known that oxides of N, including
PAN, could be altering the N cycle in some ecosystems, especially in the western U.S.,
3-12
-------
and contributing to N saturation (Fenn et al.. 2003a; Bvtnerowicz and Fenn. 1996).
Although PAN continues to persist as an important component of photochemical
pollutant episodes, there is little evidence in recent years to suggest that PAN poses a
significant risk to vegetation in the U.S.
3.4 Direct Phytotoxic Effects of Nitric Acid
Relatively little is known about the direct effects of HNO3 vapor on vegetation. Recent
information on HNO3 concentrations are given in Appendix 2.6.3. The deposition
velocity of HNO3 is very high compared to other pollutants (see Table 2-2) and HNO3
may be an important source of N for plants (Hanson and Garten. 1992; Hanson and
Lindberg. 1991; Yose and Swank. 1990). This deposition could contribute to N saturation
of some ecosystems close to sources of photochemical smog (Fenn et al.. 1998). For
example, in mixed conifer forests of the Los Angeles basin mountain ranges, HNO3 has
been estimated to provide 60% of all dry deposited N (Bvtnerowicz et al.. 1998). Since
the 2008 ISA, a controlled exposure study (Padgett et al.. 2009a) reported that 10 to 60%
of the HNO3 retained by foliage was incorporated into the biologically active N pool. The
remainder of the HNO3 was bound to foliar surfaces. This new study provides further
evidence for HNO3 as a contributor of biologically available N to southern California
forests.
Norbv et al. (1989) reported that exposure of 75 ppb of HNO3 for 1 day increased nitrate
reductase activity in red spruce (Picea mbens) foliage. In another study, foliar nitrate
reductase activity was also increased in California black oak (Onerous kelloggi), canyon
live oak (Quercus chrvsolepis), and ponderosa pine (Pinas ponderosct) seedlings with
exposure to 65 to 80 ppb of HNO3 for 24 hours (Krvwult and Bvtnerowicz. 1997).
Because the induction of nitrate reductase activity is a step in a process leading to the
formation of organic N compounds (amino acids), the nitrate from HNO3 could function
as an alternative source ofN for vegetation (Calanni et al.. 1999). However, in plants
under stress, the reduction of nitrate to amino acids consumes energy needed for other
metabolic processes.
At high ambient concentrations, HNO3 can cause vegetation damage. Seedlings of
ponderosa pine and California black oak subjected to short-term exposures from
50-250 ppb of HNO3 vapor for 12 hours showed deterioration of pine needle cuticle at
50 ppb in light (Bvtnerowicz et al.. 1998). Oak leaves, however, appeared to be more
resistant to HNO3 vapor, with 12-hour exposures in the dark at 200 ppb producing
damage to the epicuticular wax structure (Bvtnerowicz et al.. 1998). The observed
changes in wax chemistry caused by HNO3 and accompanying injury to the leaf cuticle
3-13
-------
(Bvtnerowicz et al.. 1998) may predispose plants to various environmental stresses such
as drought, pathogens, and other air pollutants. Because elevated concentrations of HNO3
and ozone co-occur in photochemical smog (Solomon et al.. 1988). synergistic
interactions between the two pollutants are possible (Bvtnerowicz et al.. 1998). Note,
however, that the experiments described above were observed at relatively short-term
exposures at above ambient concentrations of HNO3. Studies of long-term effects of
lower air concentrations that more closely approximate ambient HNO3 are needed.
Since the 2008 ISA, Padgett et al. (2009b) investigated dry deposition of HNO3 on the
foliage of ponderosa pine, white fir (Abies concolor), California black oak, and canyon
live oak in southern California. Using a chamber system within a greenhouse, leaves and
needles were exposed to control (0 (ig/m3 HNO3), moderate (30 to 60 (ig/m3 peak HNO3),
and high (95 to 160 (ig/m3 peak HNO3) concentrations. The high concentrations
represented a high ambient concentration that occurs periodically in California. The
experimental exposures resulted in a suite of damage symptoms that intensified with
increasing exposure. The exposures caused substantial perturbations to the epicuticular
surfaces of foliage of all four tree species studied. The damage caused by dry deposition
may leave foliage more vulnerable to other copollutants such as ozone.
It has been suspected that HNO3 may have caused a dramatic decline in lichen species in
the Los Angeles basin (Nash and Sigal. 1999). The suggestion was strengthened by
transplant of Ramalina lichen species from clean air habitats (Mount Palomar and San
Nicolas Island) to analogous polluted habitats in the Los Angeles basin and observing
death of the lichens over a few weeks in the summer (Boonpragob and Nash. 1991).
Associated with this death was a massive accumulation of H+ and NO3 in the lichen
thalli (Boonpragob etal.. 1989). Riddell et al. (2008) exposed healthy R. menziesii thalli
to moderate (8-10 ppb) and high (10-14 ppb) HNO3 in month-long fumigations and
reported a significant decline in chlorophyll content and carbon exchange capacity
compared to thalli in control chambers. Thalli treated with HNO3 showed visual signs of
bleaching and by Day 28 were clearly damaged and dead. The damage may have
occurred through several mechanisms, including acidification of pigments and cell
membrane damage (Riddell et al.. 2008). The authors concluded that R. menziesii has an
unequivocally negative response to HNO3 concentrations common to ambient summer
conditions in the Los Angeles air basin. They believed it was very likely that HNO3
contributed to the disappearance of this sensitive lichen species from the Los Angeles air
basin, as well as other locations with arid conditions with high HNO3 deposition loads
(Riddell et al.. 2008).
Since the 2008 ISA, there have been more studies published on HNO3 effects on lichen in
the Los Angeles basin. Riddell et al. (2012) studied six lichen species with differing
3-14
-------
morphology and physiology that were collected near the Los Angeles basin. All were
found to be sensitive to HNO3 exposures (daily peak levels near 50 ppb) in controlled
chambers. Measurements of effects included decreased chlorophyll content and
chlorophyll fluorescence, decreased photosynthesis and respiration, and increased
electrolyte leakage. The species showed differential sensitivity to the HNO3 exposures
(Riddell et al.. 2012). In the same study, lichens were not reported to be sensitive to
ozone exposure. This study adds to evidence that the main agent of decline of lichen in
the Los Angeles basin is HNO3 exposure.
In another study, Riddell et al. (2011) resampled 18 plots from a 1976-1977 study in the
Los Angeles basin. The 1976-1977 study (Sigal and Nash. 1983) had documented an air
pollution-related 50% decline of lichens described and collected in the same region in the
early 1900s (Hasse. 1913). In the 2008 resampling, Riddell et al. (2011) found
community shifts, declines in the most pollutant-sensitive lichen species, and increases in
abundance of nitrogen-tolerant lichen species compared to 1976-1977. The authors also
reported that these lichen communities have not recovered from the damaged state of the
late 1970s, and the 2008 survey data suggest that lichen communities are further
degrading. This recent observational field study further supports the evidence air
pollutants such as HNO3 may be causing declines in lichens in the Los Angeles basin.
3.5 Direct Phytotoxic Effects of Reduced Nitrogen Gases
Ammonia gas (NH3) can have direct phytotoxic effects. Section 2.6.4 discusses recent
concentrations of NH3 in the U.S. The literature on effects of NH3 exposure to vegetation
is limited in the U.S., where fumigation studies using NH3 are particularly lacking.
Reduced N gases such as ammonia are not criteria air pollutants or oxides of N and
therefore are not the focus of this review of the gas-phase effects. However, there are
similar phytotoxic effects to oxides of N, and NH3 can be a source of N nutrient effects.
This section briefly covers these ammonia effects.
As with other gases, NH3 is taken up through the leaf stomata in higher plants. The
uptake increases as ambient air concentrations rise. In higher plants, NH3 dissolves into
the mesophyll and becomes toxic if the rate of uptake exceeds the ability of a plant to
detoxify and assimilate NH3. This can depend on the N content of the plant, species of
plant, and age of leaf (Krupa. 2003). Direct visible damage to foliage has been reported at
relatively high concentrations of NH3 [annual averages of about 110 ppb; 77 (ig/m3;
Bvtnerowicz et al. (1998); Van der Eerden (1982)1. Historically, field studies on
phytotoxic effects of NH3 have generally been performed in Europe in close proximity to
livestock farms and manure storage facilities (Van der Eerden. 1982). Because of the
3-15
-------
rapid deposition and transformation of NH3, effects have been generally found close to
sources.
Ammonia exposure can also directly affect bryophytes and lichens. These nonstomatal
plants have acidic surfaces that facilitate the deposition ofNH3. Several European studies
have reported effects on lichen and bryophytes and have found them to be more sensitive
to NH3 than higher plants (Cape et al.. 2009). Often less tolerant lichens disappear and
are replaced by more tolerant species.
In a recent study, Sheppard et al. (2011) reported the effects of ammonia on vegetation in
a Whim peat bog in Scotland that included stomatal and nonstomatal plants. A free air
release system created a gradient of ammonia from 70 kg NHVha/yr down to 3-4 kg
NhMia/yr. Three years exposure to 20-56 kg NHVha/yr led to large declines in species
cover of Calluna vulgaris (a flowering shrub), Sphagnum capillifolium (a bryophyte), and
Cladoniaportentosa (a lichen). The authors reported that the effects were due to direct
uptake by the foliage and interactions with other stresses. Whim peat bogs do not occur in
the U.S. (see Appendix 11 for more discussion of peat bogs), but this study may provide
information about potential effects of NH3 in other areas relevant to U.S. ecosystems.
Studies at this site (Sheppard et al.. 2009; Sheppard et al.. 2008) and other European
studies (Wolselev et al.. 2006) have been referenced as evidence for changing the
European critical level to protect bryophytes, lichens, and herbaceous plants (Cape et al..
2009).
Besides being potentially phytotoxic to vegetation, NH3 exposure can lead to more N
inputs into plants and ecosystems through foliage uptake. In general, N content of leaves
increases with NH3 exposure (Krupa. 2003). Ammonia deposition that leads to N
enrichment is an important consideration when evaluating total N deposition. These N
nutrient effects to vegetation are discussed in Appendix 6.
3.6 Summary
3.6.1 Sulfur Dioxide
The current secondary standard for SO2 is a 3-hour average of 0.50 ppm, which is
designed to protect against acute foliar injury in vegetation. There has been limited
research on acute foliar injury since the 1982 PM-SOx AQCD, and there is no clear
evidence of acute foliar injury below the level of the current standard. The limited
3-16
-------
research since 2008 adds more evidence on acute effects of SO2 on vegetation but does
not change conclusions from the 2008 ISA on the levels producing the effects.
Effects on growth and yield of vegetation are associated with increased SO2 exposure
concentration and time of exposure. The 1982 PM-SOx AQCD concluded that more
definitive concentration-response studies were needed before useable exposure metrics
could be identified. Very few studies have been reported on the effects of SO2 on growth
of U.S. vegetation since the 1982 PM-SOx AQCD. Recent studies from eastern Europe
indicate recovery of tree growth correlated to falling SO2 concentrations since the 1980s.
Elling et al. (2009) reported that annual SO2 concentrations of 4 ppb appeared to reduce
silver fir (Abies alba) growth. There may be similar effects of SO2 emissions on trees in
West Virginia (Thomas et al.. 2013). but more research is needed to further investigate
the mechanisms of the apparent recovery of tree growth to link this phenomenon with
declines in ambient SO2 that have occurred since the 1980s.
Limited new evidence from 2008 to the present continue to support the causal findings of
the 2008 ISA. As a whole, the body of evidence is sufficient to infer a causal
relationship between gas-phase SO2 and injury to vegetation.
3.6.2 Nitrogen Oxide, Nitrogen Dioxide, and Peroxyacetyl Nitrate
It is well known that in sufficient concentrations, NO, NO2, and PAN can have
phytotoxic effects on plants through decreasing photosynthesis and induction of visible
foliar injury (U.S. EPA. 1993). However, the 1993 Oxides of Nitrogen AQCD concluded
that concentrations of NO, NO2, and PAN in the atmosphere are rarely high enough to
have phytotoxic effects on vegetation (U.S. EPA. 1993). Since the 1993 Oxides of
Nitrogen AQCD, very little new research has been performed on these phytotoxic effects
at concentrations currently observed in the U.S.
Since the 2008 ISA (U.S. EPA. 2008a). very few studies have been published on the
direct effects of NO, NO2, and PAN on vegetation; thus, the body of evidence is
sufficient to infer a causal relationship between gas-phase NO, NO2, and PAN and
injury to vegetation.
3.6.3 Nitric Acid
The 2008 ISA reported experimental exposure of HNO3 resulted in damage to the leaf
cuticle of pine and oak seedlings, which could predispose those plants to other stressors
such as drought, pathogens, and other air pollutants (U.S. EPA. 2008a). Since the 2008
3-17
-------
ISA, Padgett et al. (2009b) investigated dry deposition of HNO3 on the foliage, with
findings that supported the earlier research. The 2008 ISA also reported several lines of
evidence in lichen studies, including transplant and controlled exposure studies,
indicating that past and current HNO3 concentrations contributed to the decline in lichen
species in the Los Angeles basin (U.S. EPA. 2008a). Since the 2008 ISA, there have been
more exposure and field survey studies published on the effects of HNO3 on lichen in the
Los Angeles basin (Riddell et al.. 2012; Riddell et al.. 2011). These new studies continue
to support the causal findings of the 2008 ISA. As a whole, the body of evidence is
sufficient to infer a causal relationship between gas-phase HNO3 and changes to
vegetation.
3-18
-------
APPENDIX 4 SOIL BIOGEOCHEMISTRY
This appendix characterizes how nitrogen (N) and sulfur (S) deposition contribute to total
loading of N and S in the soils of nonagricultural terrestrial ecosystems (Appendix 4.2)
and how this loading causes soil acidification and eutrophication by altering soil chemical
pools and processes (Appendix 4.3). Additional topics are discussed, including soil
monitoring and databases (Appendix 4.4). soil biogeochemistry models (Appendix 4.5).
national-scale soil sensitivity to N and S deposition (Appendix 4.6). climate modification
of soil response to N (Appendix 4.7). and a summary (Appendix 4.8). The effects of
altered soil biogeochemistry on terrestrial biota due to soil acidification and
eutrophication are discussed in Appendix 5 and Appendix 6. respectively. Soil
biogeochemistry that has been altered by N may also be linked to aquatic
biogeochemistry, which is discussed in Appendix 7.
4.1 Introduction
The 2008 ISA for Oxides of Nitrogen and Oxides of Sulfur—Ecological Criteria
(hereafter referred to as the 2008 ISA) documented that the main effects of N and S
deposition on terrestrial soils were N enrichment and acidification. Since the 2008 ISA,
there is new evidence on how deposition contributes to total N and S loading in terrestrial
ecosystems, as well as the effects of deposition on soil chemical pools and processes.
This evidence is from addition, gradient, and time-series studies. Many of the new studies
focus on the effects of N deposition, with relatively little work focusing on S deposition.
There are improved models to evaluate ecosystem responses to deposition, most of which
are applicable at watershed scales. Some may be applied regionally. Soil N enrichment
and soil acidification occur in sensitive ecosystems across the U.S. at recent levels of
deposition. Decreasing emissions of S have led to early signs of recovery from soil
acidification in some northeastern watersheds; however, areas in the Southeast do not
show any appreciable recovery of soils. There are no signs of recovery of N enrichment
effects in soils. Critical load (CL) determinations for soils have been made at the
ecoregion scale for NO, leaching and some soil acidification indicators. The body of
evidence is sufficient to infer a causal relationship between N and S deposition and
alteration of soil biogeochemistry in terrestrial ecosystems, which is consistent with
the conclusions of the 2008 ISA.
4-1
-------
4.2
Nitrogen and Sulfur Sources to Soil
The 2008 ISA documented that atmospheric deposition is the main source of
anthropogenic N to nonagricultural and nonurban terrestrial ecosystems and headwater
streams. In 2008 it was well known that the global pool of reactive N (Nr) had increased
over the previous century, largely due to three main causes: (1) widespread cultivation of
legumes, rice, and other crops that support bacteria capable of converting diatomic
nitrogen gas (N2) to organic N through biological N fixation; (2) fossil fuel combustion
converting atmospheric N2 and fossil N to total oxidized N (NOy); and (3) the
Haber-Bosch process, which converts nonreactive N2 to Nr for N fertilizer production and
some industrial activities (Galloway et al.. 2003; Galloway and Cowling. 2002). Food
production was known to account for much of the conversion of N2 to Nr. N was shown
to be geographically redistributed through food shipment to meet human needs and often
returned to the environment via wastewater. Nr was known to accumulate in the
environment on local, regional, and global scales (Galloway et al.. 2003; Galloway and
Cowling. 2002; Galloway. 1998) in the atmosphere, soil, and water (Galloway and
Cowling. 2002). with a multitude of effects on humans and ecosystems (Townsend et al..
2003; Rabalais. 2002; van Egmond et al.. 2002; Galloway. 1998). The term the "N
cascade" was coined to refer to the sequence of transfers, transformations, and
environmental effects (Galloway et al.. 2003; Galloway and Cowling. 2002).
Since 2008, a number of estimates have been made of the relative contribution of sulfate,
oxidized N, and reduced N from atmospheric deposition. The most recent estimates are
summarized in Appendix 2. Maps showing the geographic distribution of deposition are
presented for total acidifying (N + S) deposition (Figure 2-12). total N deposition
(Figure 2-13). and total S deposition (Figure 2-31). Maps depicting how the relative
contribution of oxidized and reduced N species varies across the U.S. are presented in
Figure 2-14.
Several new studies have been published since the 2008 ISA on the source of N inputs to
ecosystems; however, no new studies on S sources to ecosystems have been identified.
Sobota et al. (2013) quantified sources and fluxes of reactive N inputs to U.S. lands and
waterways and found human-mediated N inputs are spatially heterogeneous across the
country, ranging up to 34.6-fold the background N input across all of the 8-digit
hydrologic unit codes (HUC-8s). Across the contiguous U.S. (CONUS), synthetic N
fertilizer and atmospheric N deposition are the largest and second-largest overall
human-mediated N inputs to ecosystems, and the single largest sources in 41 and 33% of
HUC-8s, respectively (Figure 4-1 and Figure 4-2).
4-2
-------
Largest human-mediated
N source
Synthetic fertilizer
Atmospheric N deposition
| Agricultural BNF
| Confined feedlot manure
| Centralized sewage
Most: Synthetic fertilizer (886 HUC-8s)
Least: Centralized sewage (32 HUC-8s)
8-digit USGS Hydrolog
BNF = biological nitrogen fixation; HUC-8 = 8-digit hydrologic unit code; N = nitrogen; USGS = U.S. Geological Survey.
Source: Map presented in Sobota et al. (2013).
Figure 4-1 Dominant sources of nitrogen across the U.S. at 8-digit
hydrologic unit codes.
Atmospheric N deposition, circa 2000
Proportion of total N input
N = nitrogen.
Source: Modified from data presented in Sobota et al. (2013).
Figure 4-2 Percentage of nitrogen input from nitrogen deposition at 8-digit
hydrologic unit codes.
4-3
-------
At the global scale, Fowler et al. (2013) estimated the relative contributions of five
sources of N that contribute to emissions and deposition, all of which are predicted to
change through the end of this century. Two new studies, both in the western U.S., have
confirmed that N from the weathering of sedimentary and metasedimentary rock can be a
source of N to some ecosystems (Montross et al.. 2013; Morford et al.. 2011). although
Sobota et al. (2013) suggested that anthropogenic sources of N are more significant at the
landscape scale.
4.3 Soil Pools and Processes
Eutrophication and acidification are two biogeochemical processes that can occur in
response to the inputs of N and S deposition. These processes can alter the
biogeochemistry in terrestrial ecosystems, and they may occur either in sequence or
simultaneously in a given geographic area. N driven eutrophication is typically indicated
by N accumulation (e.g., increased soil N concentrations or decreased C:N ratios). These
indicators of N accumulation are directly linked to biological effects in the soil, including
changes in microbial-mediated decomposition and nitrification. N added to terrestrial
ecosystems can also be lost through leaching from the soil, typically as nitrate (NO, ). or
through emissions to the atmosphere, primarily via denitrification (Galloway et al.. 2003;
Galloway and Cowling. 2002). Denitrification is a microbial process that reduces NO;, to
either unreactive N2 gas, nitrogen oxide (NO), or the potent greenhouse gas nitrous oxide
(N2O). S addition to an ecosystem typically causes S accumulation and variable amounts
of leaching from the soil, which in turn leads to acidification because demand for S as a
nutrient is low compared to human-caused soil stores of organic and inorganic S.
Soil acidification results from the accumulation of hydrogen ions (H+). This occurs
naturally through the production of carbonic acid and organic acids, as well as through
plant cation uptake (Charles and Christie. 1991; Turner et al.. 1991). The rate of soil
acidification can increase as a result of soil acidification caused by the deposition of the
inorganic acids HNO3 and H2SO4. In addition, NHx deposition contributes to soil
acidification by stimulating nitrifying bacteria that derive energy by oxidizing the NH44"
to NO;, . A byproduct of nitrification is the production of a H+ ion, but whether there is a
net effect on soil acidity depends upon the fate of the NO; . If the NO; is leached with a
base cation, then the H+ is left behind and the soil become more acidic. If, however, the
NO; is denitrified, then the H+ from nitrification is neutralized by OH generated by
denitrification. Likewise, if NOs" is taken up by a plant root, the root will exude an OH"
in exchange.
4-4
-------
Decreases in soil pH attributable to acid deposition have been documented throughout the
U.S. (Sullivan et al.. 2006b; Bailey et al.. 2005; Johnson et al.. 1994b; Johnson et al..
1994a). Inorganic and organic acids can be neutralized by soil weathering or base cation
exchange, in addition to denitrification. However, because soils vary in their capacity to
neutralize incoming acidity, the effects of acidification have been heterogeneous across
the U.S. In addition, pollutant loadings vary across the U.S. The primary chemical effects
of acidification in soils that have biological effects include the loss of important base
cation nutrients such as Ca and Mg, as well as the mobilization of aluminum (Al) cations
of varied speciation, several of which are toxic to many organisms. The quantities of
precipitation and runoff are important determinants of base cation leaching and
acidification (Van Breemen et al.. 1984). The accelerated loss of base cations through
leaching, decrease in base saturation, and decreased in soil solution Ca:Al ratio all serve
as indicators of soil acidification.
Studies published since the 2008 ISA augment our knowledge of previously identified
effects of N addition on soils. The following sections document the empirical evidence of
N effects on multiple pools, processes, and indicators associated with the general effects
of N enrichment and eutrophication (Table 4-1). These sections summarize the empirical
effects of N and S addition on soil biogeochemistry, often based on results from addition
or gradient studies. The publications summarized here present information on multiple
processes and indicators; therefore, individual papers are often discussed in more than
one section.
4-5
-------
Table 4-1 Summary of key soil geochemical processes and indicators
associated with eutrophication and acidification.
Endpoint
N-Driven Nutrient
Enrichment
Acidification
Appendix of ISA that Discusses
Each Endpoint
PROCESS
N saturation
X
X
4.3.2
Soil N accumulation
X
X
4.3.2
NO3" leaching
X
X
4.3.2
S accumulation and adsorption
X
4.3.3
SC>42"leaching
X
4.3.3
Base cation release/depletion
X
4.3.4
Al mobilization
X
4.3.5
Nitrification
X
X
4.3.6
Denitrification
X
4.3.6
Decomposition/mineralization
X
X
4.3.7 and 4.3.8
DOC leaching
X
X
4.3.9
INDICATOR
Soil [N]
X
X
4.3.2
Soil C:N ratio
X
X
4.3.6 and 4.3.7
Soil base saturation
X
4.3.4
Soil Bc:AI ratio
X
4.3.5
Fungi:bacteria ratio
X
4.3.11
Al = aluminum; Be = base cation; C = carbon; DOC = dissolved organic carbon; ISA =
N = nitrogen; N03" = nitrate; S = sulfur; S042" = sulfate.
Integrated Science Assessment;
4.3.1 Nitrogen Pathways and Pools
The 2008 ISA documented that N is stored primarily in the soil in forest ecosystems, and
soil N often exceeds 85% of the total ecosystem N (Cole and Rapp. 1981; Bormann et al..
4-6
-------
1977). Experimental 15N addition studies showed that trees typically take up only a small
fraction of added 15N; most is retained in the soil N pool (Providoli et al.. 2005; Templer
et al.. 2005; Nadelhoffer et al.. 1999a; Tietema et al.. 1998). However, these experiments
were criticized for applying 15N directly to the soil surface, thereby precluding direct
canopy uptake of N from wet or dry deposition and limiting inference for N deposition
effects (Sievering et al.. 2000; Sievering. 1999). Forest canopies take up an average of
16% of total atmospheric N input (Lovett. 1992). but this uptake could be considerably
higher (up to 90%) in some N limited forests with large epiphyte loads (Klopatck et al..
2006). It is unclear how much of the N from deposition retained by vegetation was used
in photosynthetic enzymes and would thus contribute to increased productivity (Bauer et
al.. 2004).
Most soil N is contained in organic matter, typically bound in organo-mineral complexes
or large-molecular-weight organic compounds (Schmidt et al.. 2011; Schimel and
Bennett. 2004). These pools of N are not directly available for uptake by plants or
microbes, and the relative immobility of these compounds means they contribute little to
the leaching loss of N into ground or surface waters. To enter the actively cycling portion
of the ecosystem N pool, recalcitrant organic N must be converted into inorganic forms
(e.g., NC>3~, NH4+) or small-molecular-weight organic compounds (e.g., amino acids,
amino sugars), typically through the activity of extracellular enzymes produced by soil
microorganisms. The size of this N pool often controls net primary productivity (NPP;
see Appendix 6). but plants compete with microorganisms for available N.
Since 2008 there is new research using 15N to trace pathways and pools in ecosystems
(Table 4-2). including a new paper confirming the previous conclusion that ecosystem N
is primarily stored in forest mineral soil (Perakis et al.. 2011). There is also new evidence
that litter is the largest sink for added N in grasslands, shrublands, and wetlands (Templer
et al.. 2012).
New studies have also been conducted tracing pathways of N in ecosystems through time.
In Swiss grasslands, N addition increased ecosystem N storage in plant biomass, both in
living plant biomass as well as litter, but did not affect soil N (Bassin et al.. 2015). In a
temperate forest ecosystem in Switzerland, application of 15N tracer in a solution to
simulate wet deposition showed increased dissolved organic N and enrichment of soil
microbes and plant roots within the first day. When plots were resampled a year later,
litter and soil pools, particularly soil organic matter, were enriched with the N tracer
(Morier et al.. 2008). These results confirm that deposited N is incorporated rapidly into
ecosystem pools and is stored primarily in litter and recalcitrant organic matter in the soil.
4-7
-------
Table 4-2
Pathways and pools.
Type of
Process/Indicator Ecosystem
Region
Deposition
(kg/ha/yr)
Addition
(kg/ha/yr)
Effect of Deposition
Reference
Mineralization
Semiarid
Southern
California
Up to 35-45
forN
Two
applications
per yr of
30 kg N
N fertilizer additions resulted in faster gross N cycling
rates. Greater net release or greater net
immobilization of N was observed at different
sampling times and likely related to seasonality of C
availability.
Sirulnik et al.
(2007b)
N pools
Temperate
hardwood forest
Swiss Prealps
16
The litter layer retained approximately 19-28% of the
15N tracer after 1 day. The authors concluded that the
processes relevant for the fate of atmospherically
deposited N take place rapidly and that N recycling
within the microbe-plant-soil organic matter system
prevents further losses in the long term.
Morier et al.
(2008)
N turnover times
Semiarid
grassland
Not specified
Not specified
5 as
NH4NO3
Model for N cycling was developed using
observations of 15N. The temporal dynamics of 15N
fractions (labeled-N fractions) in plant and microbial
biomass are closely tied to the turnover time of these
N pools
Diikstra (2009)
After two to three decades of high loads of N Kreutzer et al.
deposition, a new equilibrium was reached, (2009)
characterized by substantial losses of N to the
groundwater (approximately 20 kg NO3") and to the
atmosphere (16 kg N in form of N2O, NO, and N2).
Ecosystem N retention is dominated by microbial
immobilization, which was about a factor of three
higher than plant N uptake.
Tracing 15NC>3~ revealed that N accumulated in soil Zak et al.
organic matter by first flowing through soil (2008)
microorganisms and plants, and that the shedding of
15N-labeled leaf litter enriched soil organic matter.
N pathway Hoglwald, 45
Bavaria,
Germany
N pathway Simulated Michigan, Great 7 to 12 30
northern Lakes region
hardwood forest
4-8
-------
Table 4-2 (Continued): Pathways and pools.
Process/Indicator
Type of
Ecosystem
Region
Deposition
(kg/ha/yr)
Addition
(kg/ha/yr)
Effect of Deposition
Reference
N pools
Temperate forest
Oregon coast
range
2.0
None
Mineral soil accounted for 96-98% of total
ecosystem N. Soil water 15NC>3~ patterns suggested a
shift in relative N losses from denitrification to NO3"
leaching as N accumulated, and simulations
identified NO3" leaching as the primary N loss
pathway that constrains maximum N accumulation.
Perakis et al.
(2011)
N pools
Soil 15N
Global
Varied
across
48 studies
Varied
across
48 studies
A meta-analysis of studies at 48 sites across four
continents shows the largest sinks for 15N tracers
among ecosystem types were organic soil in forests
(35.5%, n = 31) and foliage in tundra (12.1%, n = 3).
Litter was the largest sink in grasslands (25.5%,
n = 9), shrublands (33.8%, n = 6), and wetlands
(34.1%, n =2).
TemDler et al.
(2012)
N pools
Common garden
experiment, five
broadleaved tree
species
Denmark
13 to 19 for
broadleaf
forest
18 to 26 for
Norway
Spruce
Tree species influenced N cycling and 15N patterns
through multiple species-specific traits. The type of
mycorrhiza association, light regime, and ground
vegetation differed between ash and sycamore and
beech, lime, oak, and Norway spruce.
Callesen et al.
(2013)
Pools and pathway
N uptake
Subalpine
(seminatural)
pasture
Alp Flix, a high
plateau near
Sur, Grosons,
Switzerland
4
0, 5, 10, 25,
and 50 as
NH4NO3
Plant N pools increased by 30-40% after N addition,
while soil pools remained unaffected.
Bassin et al.
(2015)
N cycling
All
Global
Not specified
None
Synthesis of recent literature that did not include new
quantitative analysis.
Niu et al.
(2016)
4-9
-------
Table 4-2 (Continued): Pathways and pools.
Process/Indicator
Type of
Ecosystem
Region
Deposition
(kg/ha/yr)
Addition
(kg/ha/yr)
Effect of Deposition
Reference
Throughfall
NH4+,
NOs", and S042"
Base cations
Dry N deposition
Quercus ilex
forests ranging
from
Mediterranean
climate to a drier,
more seasonal
climate
Not specified
None
Rainfall and net NhV throughfall were negatively
correlated at all sites. Rainfall and net NCV-N
throughfall correlation varied between negative and
positive. Rainfall and net SO42" throughfall were
positively correlated at wet sites and negatively
correlated at the drier site.
Aauillaume et
al. (2017)
Canopy throughfall
Holm oak forests
(Quercus ilex L.)
in the north,
center, and north-
east of the Iberian
Peninsula
Europe-Iberian
Peninsula
Wet
deposition
ranged
between 1.2
to 5.8, and
dry
deposition
ranged from
1.5 to 14
Canopies retained ranged between 0.5 and
12 ka/ha/yr. Both NHV and NO3" showed higher
retention at the agricultural and rural sites (50-60%)
than at sites located close to big cities (20-35%).
Avila et al.
(2017)
15N = Nitrogen-15; N = nitrogen; NO = nitric oxide; N20 = nitrous oxide; N02 = nitrogen dioxide; N03 = nitrate; 15N03 = Nitrogen-15-labeled nitrate; NH4N03 = ammonium nitrate;
yr = year.
4-10
-------
4.3.2 Nitrogen Accumulation, Saturation, and Leaching
In the 2008 ISA, a documented indicator of terrestrial N driven eutrophication was the
accumulation of N in soil. Atmospherically deposited N can accumulate in soil as
inorganic N or by incorporating N into organic matter. The ability of atmospheric N
pollution to cause an accumulation of N in soil is indicated by a positive correlation
between the atmospheric deposition rate and total N concentration in the Oa horizon
observed at sites in New York, Vermont, New Hampshire, and Maine (Driscoll ct al..
2001b). Accumulation of soil N as a result of N deposition has been either documented or
suggested to occur across large areas of the U.S. (Aber et al.. 2003) including semiarid
ecosystems (Padgett et al.. 1999). There is also evidence for soil N accumulation from
mass-balance studies of experimental N additions (Campbell et al.. 2004b). Further
evidence that atmospheric deposition increased the availability of N in soil is provided by
the strong negative correlation between atmospheric N deposition and the C :N ratio of
the Oa soil horizon across the northeastern U.S. (Aber et al.. 2003). Soil N accumulation
is linked to increased N leaching.
New studies (Table 4-3) confirm that, across terrestrial ecosystem types, N addition
increases soil N concentrations. In semiarid shrublands in southern California, Vourlitis
and Fernandez (2012) observed that N additions increased soil N. Lu et al. (2011a)
conducted a meta-analysis of N cycle responses to N additions using data from
206 peer-reviewed studies. They observed mean increases in N leaching, soil inorganic
N, soil total N pool, as well as increases in the litter N, organic horizon N, and mineral
soil N; the only pool that decreased was microbial N (Figure 4-3).
Thresholds of N deposition associated with the onset of elevated NO;, leaching have
been previously identified. Atmospheric deposition of 8 to 10 kg N/ha/yr resulted in the
onset ofMV leaching to surface waters throughout the eastern U.S. Slightly lower N
deposition levels (5-10 kg N/ha/yr) led to NO3 leaching in the Rocky Mountains, and
this was attributable to colder temperatures, shorter growing season, slow soil
development, extensive exposed bedrock, and rapid melting of large snowpacks
(Williams and Tonnessen. 2000; Williams et al.. 1996c; Baron etal.. 1994). Lastly,
deposition loads of 17 kg N/ha/yr led to the onset of NO3 leaching in the Sierra Nevada
and San Bernardino mountains (Fenn et al.. 2008).
4-11
-------
Den = denitrification; DON = dissolved organic N; N = nitrogen; N-min = net N mineralization; Nit = nitrification; Ps = photosynthesis;
SIN = soil inorganic N; SNP = soil N pool.
Source: Lu et al. (2011a).
Figure 4-3 A conceptual framework for the responses of the ecosystem
nitrogen (N) cycle to nitrogen (N) addition.
4-12
-------
The 2008 ISA documented that N saturation occurred when the input of N to the
terrestrial ecosystem chronically exceeded the uptake capacity of the soils and biota,
causing a large fraction of the incoming N to leach from soils to surface waters
(Stoddard. 1994; Aber et al.. 1989). N saturation had been observed or suggested to occur
across large areas of the U.S. (Adams et al.. 2000; Aber et al.. 1998; Adams et al.. 1997;
Peteriohn et al.. 1996; Cook et al.. 1994; Edwards and Helvev. 1991; Aber et al.. 1989).
New evidence (Table 4-3) from an N tracer study confirms that N retention varies across
ecosystem types and is highest in shrublands (89.5%) and wetlands (84.8%), followed by
forests (74.9%), and grasslands [51.8%; Templer et al. (2012)1. Other significant factors
affecting long-term 15N recovery (a proxy for N retention in N tracer studies) were
mycorrhizal association (ericoid > ecto > arbuscular), plant growth form, and site history
(less retention on former agricultural sites). The influence of biotic processes on N
retention is evident at smaller scales. Among nine forested sites along an urban-to-rural
landscape gradient in the Boston, MA area, throughfall inorganic N deposition increased
with proximity to the urban core, with inorganic N deposition rates positively correlated
with rates of soil inorganic N leaching across the sites (Rao et al.. 2014). Measurements
of S15N and SlsO in soil NO;, leachate indicated no clear relationship between microbial
nitrification and proximity to the urban core—evidence that factors other than N
availability influenced N processing. Overall, it is clear from recent research that N
retention is strongly influenced by biotic factors (e.g., mycorrhizae, plant growth form)
and environmental conditions (e.g., precipitation).
Leaching of N tends to increase with increasing N addition. Where 15N tracer studies
were conducted as part of N addition experiments, the N additions decreased 15N
retention. In 15N studies with multiple N addition treatments, a negative correlation
between retention and the rate of N additions was observed (Templer et al.. 2012). In
Europe, Pise et al. (2009) documented approximately 95% of forests receiving less than
8 kg N/ha/yr still had leaching, typically less than 1 kg N/ha/yr. Additional work on
monitoring data in Sweden by Khalili et al. (2010) showed a clear sudden increase in
NO;, leaching in regions where N deposition exceeded 7.5 kg/ha/yr (Appendix 4.4). In
the U.S. there are new studies modeling N leaching in eastern U.S. forests [Phelanetal
(2016); Fakhraei et al. (2016); Table 4-31.
N leaching is an indicator of ecosystem N saturation. New studies (Table 4-3) suggest the
N saturation concept (Aber et al.. 1998) may need revision in response to observations of
N cycling in temperate forests (Lovett and Goodale. 2011) and chaparral (Homvak et al..
2014). Lovett and Goodale (2011) proposed a new model of N saturation that
distinguished capacity N saturation, in which the vegetation and soil sinks for N have
been filled, from kinetic N saturation, in which the plant and soil sinks are accumulating
4-13
-------
N, but the rate of N accumulation is slower than the N input rate. One implication of this
new model is that NO;, leaching can occur even if the ecosystem N retention capacity
has not yet been saturated, as is observed at many sites [e.g., Lovett and Goodale (2011);
Talhelm et al. (2012); (Homvak et al.. 2014)1. New information confirms the applicability
of the capacity N saturation concept in northwest forests (Perakis and Sinkhorn. 2011).
although the new idea of kinetic N saturation is particularly useful in describing N
cycling processes in California chaparral (Homvak et al.. 2014). At chaparral sites, N
added during the dry season is lost from the ecosystem, while N added during the
growing season is retained in plant biomass (Vourlitis and Pasquini. 2009; Grulke et al..
2005). In these systems, water scarcity limits plant productivity, microbial C availability,
and denitrification, all of which cause NO;, to be flushed from surface soils during large
precipitation events (Menon et al.. 2010). Large leaching losses of N in arid ecosystems
are evidence for kinetic saturation.
New research highlights the importance of deposition of N from mobile sources along
roadsides (Fenn et al.. 2018; Bettez et al.. 2013). The higher deposition along roadsides in
Cape Cod, MA was associated with a two- to fourfold greater rate of nitrate leaching
from the soils. Approximately 15% less N from deposition was retained 10 m from the
road than sites more distant removed from the road. N deposition contribution to N in the
watershed may be underestimated by 13-25% when roadside deposition and the
associated leaching are not included. In a U.S. nationwide evaluation of on road
emissions of ammonia, Fenn et al. (2018) found NH4 :NO; ratios in urban deposition
reflected elevated NH3 from emissions, and that on-road NH3 emissions exceeded
agricultural emissions in locations where 40% of the population resides.
New research highlights the role of the microbial community in N saturation. Kopacek et
al. (2013) developed a conceptual model in which N saturation is associated with shifts in
the microbial community, namely a decrease in the fungi-to-bacteria ratio, and a
transition from N to C limitation. In N enriched systems, three mechanisms could lead to
lower amounts of bioavailable dissolved organic C (DOC) for the microbial community
and to C limitation: (1) lower plant allocation of nutrients to roots in response to
increased nutrient availability, leading to a decrease in plant exudates; (2) chemical
suppression of DOC solubility by soil acidification; and (3) enhanced bacterial
mineralization of DOC due to increased abundance of electron acceptors in the form of
NO3 in anoxic soil. In support of this model, recent studies indicate that N retention in
semiarid shrublands is driven more by spatial and temporal variations in labile C
availability than exceedance of N storage capacity (Vourlitis and Fernandez. 2015).
Furthermore, Hogbcrg et al. (2013) found that forest soils with low concentrations of
NO3 and Al had a higher fungi :bacteria ratio compared with stands having higher
concentrations of NO3 and Al (negative correlation, r = -0.857). Fungi:bacteria ratio,
4-14
-------
and a second indicator, stem growth, explained 70% of the variation in N and A1 leaching
(see broader discussion of N effects on microbial community composition in
Appendix 6).
Hogbcrg et al. (2013) proposed a hypothetical model to account for the effects of N
supply on plant N uptake and belowground C allocation, microbial production of
inorganic N, and N leaching (Figure 4-4). Shifts in N mineralization, nitrification, and
leaching from forests might occur in response to N loading as a result of decreasing tree
allocation of C to belowground roots, and ultimately, to ectomycorrhizal fungi and other
C limited soil microbes. When N supply increases, ectomycorrhizal fungi and other
rhizosphere microbes become progressively more C limited, and their abundance and
activity decline. Microbial assimilation of N diminishes, and N mineralization increases,
whereas the fungi:bacteria ratio decreases. Increasing NH/ levels and decreasing organic
C supply stimulate leaching, and denitrification. Increasing N loading can alter plant C
allocation, causing shifts in microbial activity and community composition that in turn
increase NO;, leaching from the ecosystem (Figure 4-4).
N supply to plants and microbes
Decreasing Increasing
I*
N leaching
C = carbon; N = nitrogen.
Thick arrows indicate an increasing rate whereas thin arrows mean decreasing rate. Nitrogen supply increases are indicated by
black arrows.
Source: Hoaberg et al. (20131.
Figure 4-4 A hypothetical model to account for the effects of nitrogen supply
on plant nitrogen uptake and belowground carbon allocation,
microbial production of inorganic nitrogen, and nitrogen leaching.
4-15
-------
Table 4-3 Nitrogen accumulation, saturation, and leaching.
Type of Deposition Addition
Process/Indicator Ecosystem Region (kg/ha/yr) (kg/ha/yr) Effect of Deposition Reference
N leaching Forest 248 sites (plot and Varied Not reported Gradient: The most consistent indicators of N Pise et al.
catchment scale) leaching were throughfall N deposition, organic (2009)
from 15 countries in horizon C:N ratio, and mean annual
Europe temperature. Sites receiving low levels of N
deposition (8 kg N/ha/yr) showed very low output
fluxes of N. In general, the models successfully
predicted N leaching (mean of ± 5 kg N/ha/yr
between observed and predicted) from forests at
early to intermediate stages of N saturation but
not from N saturated sites.
N accumulation Mesic desert Spanish Spring Not specified None Field experiment: Concluded that vadose soil Menon et al
N leaching Valley, NV resources (water and organic C) are rare. (2010)
Unused NO3" from low biological demand is
transported and accumulated in the deeper
vadose zone with occasional deep leaching
events.
N saturation
Oak forest
Southeastern New 9
York State
100 NH4NO3
(1996-1999)
50 NH4NO3
(2000-2006)
Conceptual model: New N saturation model
based on an N addition study of an oak forest in
southeastern New York State.
Lovett and
Goodale (2011)
N accumulation
Meta-analysis
Varied
Varied
Meta-analysis: N cycle responses to N
additions using data from 206 peer-reviewed
studies and observed a mean increase in N soil
pools except microbial biomass N (Fiaure 4-12).
Lu et al.
(2011a)
4-16
-------
Table 4-3 (Continued): Nitrogen accumulation, saturation, and leaching.
Process/I nd icator
Type of
Ecosystem
Region
Deposition
(kg/ha/yr)
Addition
(kg/ha/yr)
Effect of Deposition
Reference
N saturation
N accumulation
N mineralization
nitrification
Douglas fir
forest
Oregon coast range 2.0
None Gradient: This is a test of the N saturation
theory along a natural N gradient. Higher N
content of surface (0-10 cm) soil was linearly
related to higher net N mineralization. Lower pH
was related to lower nitrification. The ratio of net
to gross N mineralization to higher nitrification
increased along the gradient, indicating
progressive saturation of microbial N demand.
Perakis and
Sinkhorn (20111
N leaching
Douglas fir
forest
Oregon coast range 2.0
None Gradient: Hydrologic N losses were dominated
by dissolved organic N at low N sites, with
increased nitrate loss causing a shift to
dominance by nitrate at high N sites, particularly
where net nitrification exceeded plant N
demands.
Perakis and
Sinkhorn (2011)
N leaching
48 sites across
four continents
Grassland,
forest,
wetland,
shrubland
Global
Varied across Varied across Meta-analysis: The greatest recoveries of Templer et al.
48 studies 48 studies ecosystem 15N tracer occurred in shrublands (2012)
(mean, 89.5%) and wetlands (84.8%) followed
by forests (74.9%) and grasslands (51.8%).
Soil N, C:N
Chaparral and
coastal sage
scrub (CSS)
Southern California 6-8.1
56 to 58 Addition: In this 6-yr field experiment, chaparral
and CSS vegetation communities were found to
have the capacity to immobilize 6.2 and 11.9 g
N/m2/yr, respectively. Soil extractable N
increased significantly after 7-10 g/m2 of
cumulative N exposure, resulting in a
simultaneous increase in the N concentration
and a decline in the C:N ratio of shrub tissue.
Similar results were observed for the surface
litter pool and litter production but at a higher
cumulative N exposure.
Vourlitis and
Fernandez
(2012)
4-17
-------
Table 4-3 (Continued): Nitrogen accumulation, saturation, and leaching.
Type of
Deposition
Addition
Process/Indicator Ecosystem Region
(kg/ha/yr)
(kg/ha/yr)
Effect of Deposition
Reference
NO3" leaching Forest Cape Cod, MA
Annual
None
Roadside gradient: 25 to 30% higher N
Bettez et al.
throughfall
deposition to forests 10 m away from roads on
(2013)
deposition was
Cape Cod, compared to hundreds of meters
8.7 (±0.4) and
away from the road; the higher deposition was
6.8 (±0.5) TDN
associated with a two to four-fold greater rate of
at sites 10 and
nitrate leaching from the soils. 73% of the
150 m away
deposition was retained in the forest away from
from the road
roads, compared to 58% retention at 10 from the
road. Results scaled to the entire watershed
indicate an underestimate of the amount of N
deposition to the watershed by 13-25% by not
including roadside deposition and leaching.
Soil [NO3"] 19 Picea abies South Sweden
Throughfall N
Ammonium
Addition: Microbial community composition in
Hoabera et al.
Soil [Al] (L.) Karst.
Stands
Fungi:bacteria ratio
includes wet
NO3- at 20
the organic horizon and soil solution chemistry
(2013)
and dry inputs
and ranged
from 2.7 to 19
below the rooting zone were highly correlated.
Stands with low concentrations of NO3" and Al
had higher fungi:bacteria ratio compared with
stands with higher concentrations of NO3" and
Al. Microbial community composition in the soil
was more closely related to the soil solution than
to the soil chemistry. The study found a
significant negative correlation between the
fungi:bacteria ratio in the soil and NO3" and Al in
soil solution (r= -0.533 and -0.857,
respectively).
N saturation None None (theoretical) Not specified None Conceptual model: N addition alleviates N Kopacek et al.
(theoretical) limitation, and together with SO42" deposition, (2013)
causes soil acidification and increases
availability of electron acceptors for soil
microbial processes. Increasing N and SO42"
decreases fungal biomass, increases bacterial
DOC mineralization, and decreases DOC
leaching.
4-18
-------
Table 4-3 (Continued): Nitrogen accumulation, saturation, and leaching.
Process/I nd icator
Type of
Ecosystem
Region
Deposition
(kg/ha/yr)
Addition
(kg/ha/yr)
Effect of Deposition
Reference
Decomposition
Deciduous
forests
Catskills Mountains
of southeastern
New York
9.0
50 (NH4NO3)
Addition: NO3" leaching increased markedly in
response to the N addition in all species,
indicating that the addition rate of N exceeded
the N retention capacity of vegetation and soils
in these plots.
Lovett et al.
(2013)
NO3" leaching
Alpine
Rocky Mountain
National Park
(Niwot Ridge)
Not specified
Variable
Review: Nitrate leaching increases above 10 kg
N/ha/yr deposition.
Bowman et al.
(2014)
Rate of soil N cycling
Alpine
Rocky Mountain
National Park
(Niwot Ridge)
Not specified
Variable
Review: The rate of soil N cycling increases
above 15 kg/ha/yr deposition.
Bowman et al.
(2014)
Soil acidification and
soluble Al
Alpine
Rocky Mountain
National Park
(Niwot Ridge)
Not specified
Variable
Review: Soil acidification and soluble aluminum
increases above 25 kg/ha/yr deposition.
Bowman et al.
(2014)
N export Forest Ontario, Canada Not reported Not Isotope: Rain on snow, as it passes through the Casson et al.
applicable ecosystem, has a higher concentration of NO3" (2014b)
(throughfall/snowmelt average = 498 pg/L)
compared with baseflow (average = 7.3 |jg/L;
average = 41 pg/L) and as a result, throughfall
and snowmelt contribute the majority of NO3"
export (average = 62%) during rain on snow
events.
NO3 leaching
Urban gradient
Urban sites
Gradient: The source of N leaching from five of
Rao et al.
12.3 and
nine sites was almost entirely from nitrification,
(2014)
nonurban 5.7
indicating that the NO3" in leachate came from
biological processes rather than directly passing
through the soil. A significant proportion
(17-100%) of NO3" leached from the other four
sites came directly from the atmosphere.
4-19
-------
Table 4-3 (Continued): Nitrogen accumulation, saturation, and leaching.
Type of
Deposition
Addition
Process/I nd icator
Ecosystem
Region
(kg/ha/yr)
(kg/ha/yr)
Effect of Deposition
Reference
N saturation/NH4+,
Forest
White Mountains
6.6 to 6.7 N
Not
Isotope: There was no significant biological
TemDler et al.
NO3", total inorganic
National Forest,
8.1 to 8.5 SO42"
applicable
production of NO3" via nitrification in the canopy.
(2015b)
N, and S042"-S
Crawford Notch,
NO3" concentrations in streams were low and
NH; Lye Brook, VT
had natural 1sO abundances consistent with
in southern Green
microbial production, demonstrating that
Mountain National
atmospheric N is being biologically transformed
Forest
while moving through these watersheds and that
these forested watersheds are unlikely to be N
saturated.
N accumulation
Semiarid
Santa Margarita
6 to 8
50
Addition: N enrichment significantly increased
Vourlitis and
chaparral and
Ecological Reserve
N accumulation but not microbial respiration.
Fernandez
CSS
and the Sky Oaks
(2015)
Field Station
N retention
Beech and
Soiling, Germany
Direct total
None
Field Time Series: Two study plots observed
Meesenburq et
Norway spruce
deposition not
that N retention decreased from 40 to
al. (2016)
forests
specified
0-20 kg/ha/yr (from 1970s to current), indicating
increasing N saturation.
NO3" leaching and N
Eastern U.S.
Hubbard Brook
Four modeling
28.8 kg/hr/yr
Model: Modeled scenarios at Hubbard Brook,
Phelan et al.
availability
forests
Experimental
scenarios
S and
NH received twice the N and S at Bear Brook,
(2016)
Forest, NH, and
ranging from
25.2 kg/ha/yr
ME and resulted in the largest changes in soil
Bear Brook
7.4 to
(NH4)2S04
base saturation, ANC, NO3" leaching, and N
Watershed, ME
103.9 keq/ha N
availability. Modeled recovery of soil chemistry
and 7.8 to
and understory plant communities at both forests
149.1 keq/ha S
only occurred when N and S deposition were
from
modeled at preindustrial levels.
1850-2100 with
0-95%
decrease from
2010 to 2100
NO3" leaching
Mixed
Neuglobsow
Total N wet
None
Field observation: Seepage water (120 cm) is
Schulte-BisDina
beech-pine
Integrated
deposition:
estimated to contain at 2.38 kg N/ha/yr (96% as
and Beese
forest
Monitoring (IM) site,
13.26 ± 2.01 kg
NOs").
(2016)
Germany
N/ha/yr
(1998-2013)
4-20
-------
Table 4-3 (Continued): Nitrogen accumulation, saturation, and leaching.
Type of
Deposition
Addition
Process/Indicator Ecosystem
Region (kg/ha/yr)
(kg/ha/yr)
Effect of Deposition Reference
N immobilization Restored
Prairie invasion and -1/3 of N
5 g N/m2/yr
N Addition: Nitroaen addition durina Dlant Schuster (2016)
tallgrass
Climate Experiment added
urea.
growth accelerated subsequent mass loss of
prairie
(PRICLE) Loveland,
"Seasonal
Schizachyrium scoparium litter and litter
CO
maximum
produced with N addition had a 65% greater N
5-day
loss than ambient N litter in Solidago canadensis
cumulative
plots, indicating N addition accelerates N
rainfall"
cycling.
increased by
33% (2012)
and 9%
(2013)
N soil retention Forest Great Smoky 5.1 kg N/ha/yr None Model: PnET-BGC was used to model Fakhraei et al.
Mountains National (36.5 mmolc 30 stream watersheds characterized by (2016)
Park /m2/yr), decreased SO42" and NO3" deposition during
respectively. 1981-2014 (81 and 53%) and predict stream
recovery. Spruce-fir forests at higher elevations
have limited N retention and exhibit N saturation
due to elevated N deposition.
Al = aluminum; C = carbon; cm = centimeter; CSS = coastal sage scrub; DOC = dissolved organic carbon; g = gram; ha = hectare; kg = kilogram; L = liter; m = meter; N = nitrogen;
NH4+ = ammonium; N03" = nitrate; P = phosphorus; NH4NO3 = ammonium nitrate; r= correlation coefficient;; S042" = sulfate; yr = year.
4-21
-------
4.3.3
Sulfate Accumulation, Adsorption, and Leaching
The 2008 ISA showed that SOx deposition may be assimilated by vegetation or microbes,
accumulate in the soil, or act as a mobile ion and leach out of the soil to aquatic
ecosystems. Plant demand for S is typically low, particularly in comparison to the large
pools of S stored in the soil in recalcitrant organic compounds; consequently, almost all S
deposited in terrestrial ecosystems enters soil rather than plant pools. At many locations
in the U.S. that receive high levels of S deposition, notably the glaciated Northeast and
upper Midwest, much of the deposited S leaches through soils into streams and lakes. The
physical process of charge balance pairs sulfate (SO42 ) leaching with leaching of
countercharged cations from the soil matrix, and this process contributes to acidification
of soil, soil water, and surface water. As the base cations become depleted in the soil
matrix, charge balance in the soil is maintained by an increase in acidic cations (H+ and
inorganic Al), sometimes resulting in toxic conditions for plant roots and aquatic
organisms (Charles and Christie. 1991; Turner et al.. 1991).
In the 2008 ISA, regional trends of S042 soil accumulation and leaching were identified
in the U.S. In the Southeast, accumulation of atmospherically deposited S in soil resulted
from S042 adsorption to soil particles as well as incorporation of S into soil organic
matter through biological assimilation. Accumulated S can be slowly released from soil
pools into drainage water, and this process can temporarily delay ecosystem recovery in
response to decreases in S deposition (Sullivan et al.. 2004; Elwood et al.. 1991; Turner
et al.. 1991). In the Northeast, there was a demonstrated accumulation of S in soil
(Driscoll et al.. 2001a). Two new studies on SO42 accumulation, leaching, and
adsorption are described below (Table 4-4).
The net loss of S from soils is occurring in a number of northeastern watersheds in
response to decreased levels of atmospheric S deposition. In a new study evaluating
watersheds in the Northeast, Mitchell and Likens (2011) calculated that annual
discrepancies in the watershed S budgets (SO42 flux in drainage waters minus total
atmospheric S deposition) have become more negative, indicating the increasing
importance of the release of S from internal ecosystem sources. The release of S from
forest soils is controlled (57%) by water flux and soil moisture (Mitchell and Likens.
2011).
In the southeastern U.S., Rice et al. (2014) calculated SO42 mass balances for
27 forested, unglaciated watersheds from Pennsylvania to Georgia by using total
atmospheric deposition (wet plus dry) as input. Unlike their counterparts in the
northeastern U.S. and southern Canada, many of these watersheds still retain SO42 . Rice
4-22
-------
et al. (2014) predicted that many of the watersheds in the study will begin releasing SO42
over the next two decades. The specific years when the watersheds cross over from
retaining to releasing SO42 correspond to a general geographical pattern from north to
south of later net watershed release. For instance, the three watersheds in West Virginia
have crossover years that ranged from 2006 to 2011, and crossover years for five of the
watersheds in Virginia range from 2012 to 2021. The runoff ratio, computed as the ratio
of annual mean discharge to annual mean precipitation, was the single best
watershed-scale predictor of the crossover year (r2 = 0.72). Watersheds with higher
runoff ratios tend to convert sooner from net retention to net release of S042 (Rice et al..
2014). More recently, Fakhraei et al. (2016) used PnET-BGC to model 30 stream
watersheds during 1981-2014 in the Great Smoky Mountains National Park. Hindcast
modeling (beginning ca. 1850) to 2014 showed that the soil SO42 pools increased from
20.2 g/m2 (preindustrial) to 145.6 g/m2 (ambient median). Soils with high SO42
adsorption capacity also had a faster rate of base cation depletion. In general, soils with
high S042 adsorption capacity will have more SO42 release for longer periods of time
during the recovery period from acidification, thereby having a faster rate of base cation
depletion compared to soils with low SO42 adsorption capacity during the recovery
period.
4-23
-------
Table 4-4 Sulfate adsorption, accumulation, and leaching.
Process/Indicator Type of Ecosystem
Region
Deposition
kg/ha/yr
Addition
kg/ha/yr
Effect of Deposition
Reference
Sulfate adsorption Upland forests,
and cation low-lying areas and
leaching wetlands
Athabasca oil sands Not specified
region (AOSR) in
Alberta, Canada
None
Field comparison: SO42"
adsorption capacity was relatively
low (50 to 500 mg SC>42"/kg) in both
watersheds as compared to other
acid-sensitive soils in eastern North
America.
Jung et al. (2011)
S budget
Varied
15 sites
southeastern
Canada and
northeastern U.S.
-0.1 to 14
Long-term deposition: The net
annual fluxes of SO42" showed a
strong relationship with hydrology;
the sensitivity of S budgets is likely
greatest in watersheds with the
greatest wetland area, which are
particularly sensitive to drying and
wetting cycles.
Mitchell et al. (2011)
Sulfate
accumulation and
leaching
Forest
HBEF in the White
Mountains of New
Hampshire
-7 to 20
None
Long-term deposition: S released
from internal sources is increasing
overtime. Watershed wetness, as a
function of log-io annual water flux
explained 57% (n = 157) of the
annual variation for four
watersheds. The biogeochemical
control of annual SO42" export in
stream water draining from forested
watersheds has shifted from control
by atmospheric S deposition to soil
moisture driven by climate.
Mitchell and Likens
(2011)
4-24
-------
Table 4-4 (Continued): Sulfate adsorption, accumulation, and leaching.
Process/Indicator Type of Ecosystem
Region
Deposition
kg/ha/yr
Addition
kg/ha/yr
Effect of Deposition
Reference
SO42" leaching Watersheds ranging 27 forested,
from Pennsylvania to unglaciated
Georgia watersheds
Not specified
None
Long-term deposition: Calculated
SO42" mass balances for
27 watersheds showed that over
the next two decades, many of the
study's watersheds will begin
releasing SO42". The single most
important variable that explained
the crossover year was the runoff
ratio, defined as the ratio of annual
mean stream discharge to
precipitation.
Rice et al. (2014)
Soil S042"
Forest
Great Smoky
Mountains National
Park
3.1 kg S/ha/yr
(19.3 mmolc/m2/yr)
None Model: PnET-BGC used to model
30 stream watersheds during
1981-2014 when SO42" and NOs"
deposition decreased (81 and 53%,
resp.). Hindcast modeling
(beginning ca. 1850) increased soil
pools of SO42" from 20.2 g/m2
(preindustrial) to 145.6 g/m2
(current median).
Fakhraei et al. (2016)
Base saturation
Forest
Great Smoky 5.1 kg N/ha/yr
Mountains National (36.5 mmolc/
Park, U.S. m2/yr)
None Model: PnET-BGC used to model
30 stream watersheds during
1981-2014 when S042" and NOs"
deposition decreased (81 and 53%,
resp.). High capacity SO42"
adsorbing soils depleted base
saturation faster.
Fakhraei et al. (2016)
Soil S042"
Boreal forest
Sweden
Not specified None Time series: In an analysis of 10 yr
of data, riparian zone soil SO42"
was observed to decrease from
2003-2012.
Ledesma et al. (2016)
4-25
-------
Table 4-4 (Continued): Sulfate adsorption, accumulation, and leaching.
Deposition Addition
Process/Indicator Type of Ecosystem Region kg/ha/yr kg/ha/yr Effect of Deposition Reference
Soil SO42" Maritime pine forest Northwest Spain Not specified None Field experiment: The soil parent Eimil-Fraaa et al.
plantation material (slate, biotitic schist, mica (2016)
schist, and granite) did not
significantly affect soil solution
SCM2" (p = 0.39).
AOSR = Athabasca oil sands region; HBEF = Hubbard Brook Experimental Forest; kg = kilogram; N03 = nitrate; S = sulfur; S042 = sulfate; yr = year.
4-26
-------
4.3.4
Base Cation Leaching and Exchange
In the 2008 ISA, it was known that acidifying deposition changes concentrations of
exchangeable base cations in soil by accelerating natural rates of base cation leaching
until stores become depleted (Lawrence et al.. 1999; Cronan et al.. 1978). Base cations
include the essential plant nutrients (e.g., Ca, Mg, and K), and the loss of exchangeable
base cations from the soil may have adverse effects on flora. When S042 and NO;,
leaching occur in equal magnitude to base cation leaching, the drainage water is not
acidified. However, in the process of neutralizing the acidity of drainage water, base
cation release from soil eventually can cause decrease of the base saturation of the soil.
Soil base saturation expresses the concentration of exchangeable bases (Ca, Mg,
potassium [K], sodium [Na]) as a percentage of the total cation exchange capacity (which
includes exchangeable H+ and inorganic Al).
Under conditions of low soil base saturation (approximately <20%) and elevated
concentrations of strong acid anions, Al is mobilized from soil to drainage water (Cronan
and Schofield. 1990; Reuss. 1983). with potentially harmful consequences for sensitive
terrestrial plants and aquatic organisms (Appendix 7) throughout the food web
(Appendix 4.3.6).
Leaching of base cations by acidifying deposition has been documented in sensitive
regions in the U.S., including the Adirondack Mountains, New England, the Catskill
Mountains, and northwestern Pennsylvania (U.S. EPA. 2008a). Base cation loss increases
the sensitivity of the watershed to further acidifying deposition. Watersheds that were
capable of fully neutralizing a particular level of acidifying deposition in the past may no
longer be capable of fully neutralizing that level today or in the future due to the
cumulative effect of acidifying deposition on soil base saturation.
Base saturation values less than 10% predominate in the soil B-horizon in the areas in the
U.S. where soil and surface water acidification from acidifying deposition have been
most pronounced, including conifer and hardwood forests in the Adirondack Mountains
(Sullivan et al.. 2006b). red spruce forests throughout the Northeast (David and
Lawrence. 1996). hardwood forests in the Allegheny Plateau (Bailey et al.. 2004). and
conifer and hardwood forests in the southern Appalachian Mountains (Sullivan et al..
2003). In a study of sugar maple decline throughout the Northeast, Bailey et al. (2004)
found threshold relationships between base cation availability in the upper B soil horizon
and sugar maple mortality at Ca saturation less than 2%, and Mg saturation less than
0.5% (Bailey et al.. 2004). The authors concluded that base saturation varied as a function
of topography, geologic parent material, and acidifying deposition.
4-27
-------
New publications further support findings from the 2008 ISA that N and S deposition
cause base cation depletion from soils (Table 4-5). In the Rocky Mountains, Lieb et al.
(2011) observed that soil acid buffering capacity decreased as N inputs increased (40%
decrease at highest N input). An acidification threshold was calculated for significant loss
of soil acid buffering capacity of around 28 kg N/ha/yr. In another U.S. study, long-term
trends in base cation depletion at Bear Brook Watershed, ME, showed N and S addition
over a 17-year time period (while ambient S deposition was simultaneously decreasing)
resulted in little evidence of continued soil exchangeable base cation depletion or
recovery [expected because of decreasing S deposition; SanClements et al. (2010)1. A
study of the forest understory herb community in West Virginia found that N addition
lowered plant-available Ca and, to a lesser degree, Mg, but not K, illustrating how
biogeochemical cycling of forest ecosystems is altered (Gilliam et al.. 2016a).
A meta-analysis of 107 studies found N addition alters the availability of base cations in
terrestrial and aquatic ecosystems (Lucas et al.. 2011): although short-term N and S
deposition cause base cation depletion, long-term trends across all studies are unclear and
may be affected by confounding disturbances. Evaluating the strength of these results is
difficult because they are based on averages from various biome types and there are few
long-term studies.
Field studies in forests in Europe confirm that N deposition and N addition lower soil pH
and decreases base cations (Chen et al.. 2015; Ferretti et al.. 2014). As acidifying
deposition decreased, base cation concentration in the soil increased (Berger et al.. 2016).
Two studies from grasslands in Asia report mixed results, with N addition causing Be
levels to decrease in one study (Chen et al.. 2015) and increase in another (Tian et al..
2016b). A study of European grasslands found that base cation depletion increased with
N addition over a 10-year period, leading to a loss of 23 to 35% of total available bases
(Ca2+, Mg2+, K+, and Na+) from the soil and acidifying it by 0.2 to 0.4 pH units (Horswill
et al.. 2008).
Base cation weathering rates are uncertain, but substantial advancements have been made
in this field since the 2008 ISA (Appendix 4.5.1.1). New model estimates have been
published for two forested areas in Canada (Williston et al.. 2016; Watmough et al..
2014).
Major sources of Be to ecosystems are either from atmospheric deposition or weathering
from soils. Two additional studies from Spain considered the sources of Be in ecosystems
from deposition (Aguillaumc et al.. 2017) or as compared with sources from different
types of soil parent material (Eimil-Fraga et al.. 2016).
4-28
-------
Additional literature that evaluates how base cation depletion from acidification may
recover in response to the addition of base cations to the soil is also noted here; however,
this literature does not describe the effects of N and S deposition, but rather a method for
recovering ecosystems to a more natural state. This literature includes several
publications from a 15-year Ca addition study at Hubbard Brook Experimental Forest
(HBEF), NH (Shao et al.. 2016; Johnson et al.. 2014; Green et al.. 2013; Nezat et al..
2010).
4-29
-------
Table 4-5
Base cation leaching and exchange.
Process/
Indicator
Type of Deposition Addition
Ecosystem Region kg/ha/yr kg/ha/yr
Effect of Deposition
Reference
Addition: Treatments caused grassland soils to lose 23 to 35% of Horswill et
their total available bases (Ca2+, Mg2+, K+, and Na+), and they al. (2008)
became acidified by 0.2 to 0.4 pH units. Al, Fe, and Mn were
mobilized and taken up by limestone grassland forbs and were
translocated down the acid grassland soil. Mineral N availability
increased in both grasslands and many species showed foliar N
enrichment. N deposition depletes base cations from grassland
soils.
Base cation Grasslands Peak District Not specified Plots treated for
depletion National Park, 8 to 10 yr with 0,
30ll p|_| England 35, or 140 N as
Soil [Al], [Fe],
[Mg]
NH4NO3
Base cation
depletion
Hardwood
forest
Eastern U.S.,
Bear Book
watershed, ME
28.8 S and 25.2 Addition: Compared treated and untreated watersheds after N
N as (NH4)2SC>4 and S manipulation over a 17-yrtime period. Found little evidence
of continued soil exchangeable base cation concentration
depletion or recovery, possibly because a 1998 ice storm
increased litterfall and accelerated mineralization, obscuring
temporal trends in soil chemistry.
SanCleme
nts et al.
(2010)
Base cation
depletion
Critical load
Forest stands
with mature
white ash
Niwot Ridge in
southern
Rocky
Mountains
20, 40, 60 N Addition: Soil acid buffering capacity decreased with increasing
N inputs (40% decrease at highest input), and was associated
with a decrease in pH, loss of extractable Mg2+ and increases in
Mn and Al3+. The threshold at which acidification occurred was
around 28 kg N//ha/yr.
Lieb et al.
(2011)
Base cation Boreal forest, 107 sites
depletion temperate globally
forest, tropical
forest, and
grassland
Not specified
Median N
addition across
the studies was
38, and 71% of
the studies
added 70 or
less
Meta-analysis: Evaluation of 107 independent studies to Lucas et al.
determine whether N fertilization alters the availability of base (2011)
cations (Be) in terrestrial and stream ecosystems. Results suggest
N fertilization may accelerate Be loss from terrestrial ecosystems
over time periods less than 5 yr.
4-30
-------
Table 4-5 (Continued): Base cation leaching and exchange.
Process/
Indicator
Type of
Ecosystem
Region
Deposition
kg/ha/yr
Addition
kg/ha/yr
Effect of Deposition
Reference
Base cation Jack pine Athabasca oil Not specified 30 N, 30 S Addition: No evidence of N saturation in the studied forest
leaching (Pinus sands region forAOSR (2006 through ecosystem after 4 yr of N and S additions. No long-term increase
banksiana) and (AOSR), 2009) of inorganic N concentrations in the soil; leaching of N beyond the
aspen Alberta, main rooting zone in the soil profile was minimal, and tree growth
(Populus Canada was increased by N addition, all indications of N limitation in the
tremuloides) in studied forest stand. However, exchangeable Ca2+ and Mg2+
upland forests concentrations in the surface mineral soil layer were reduced by N
and black and S additions because of increased cation leaching associated
spruce (Picea with increased SO42" leaching caused by S addition and
mariana) in increased nutrient uptake associated with increased tree growth
low-lying areas resulting from N addition,
and wetlands.
Jung and
Chang
(2012)
Hydrologic
flowpath
Forest
HBEF, NH
Not specified 41 metric tons Addition: The flow path of the added Ca was followed through (Nezatet
of the mineral time. The deepest flowpaths to the streams were penetrated by
wollastonite 3-9 yr after application. It was estimated that only -360 kg out of
(CaSiC>3) was 19 metric tons of Ca applied as wollastonite had been exported
applied to an from the watershed in stream flow 9 yr after its application and it
11.8-hectare would take 1,000 yr for all of the added Ca to be transported from
watershed the watershed.
al. (2010)
Evapotran- Forest HBEF, NH Not specified 1,028 kg Ca/ha Field observation: 25, 18, and 19%, evapotranspiration,
spiration (as wollastonite) increased in years 1-3 after addition and then returned to
in 1999 pretreatment levels. Watershed soil retained Ca from the
wollastonite, indicating a watershed-scale fertilization effect on
transpiration. That response is unique in being a measured
manipulation of watershed runoff attributable to fertilization.
Green et
al. (2013)
4-31
-------
Table 4-5 (Continued): Base cation leaching and exchange.
Process/
Indicator
Type of
Ecosystem
Region
Deposition
kg/ha/yr
Addition
kg/ha/yr
Effect of Deposition
Reference
Total Ca,
exchangeable
Ca, cation
exchange
capacity, base
saturation
Forest Hubbard Brook Not specified 1,028 kg Ca/ha Field observation: Compared to conditions before the 1999 Ca Johnson et
Experimental (as wollastonite) addition, total Ca increased 596% in the Oie-horizon in 2000, 81% al. (2014)
Forest, NH in 1999 in the Oa-horizon in 2002, and 146% in the 0-10 cm mineral soil
in 2006.
Oie-horizon exchangeable Ca tripled in 2000 and remained
significantly higher through 2010. Significant increases occurred in
2002 in Oa; modest increases measured in upper mineral soil in
2006 and 2010. Total Ca pool in O-horizon and mineral soils was
250 kg/ha greater than pretreatment. In total, 92.4% of Ca added
was estimated to dissolve and enter the ecosystem as labile Ca.
Cation exchange capacity CEC increased by 106% in the
Oie-horizon in 2010 but decreased 52 and 32% in the Oa and
mineral soil. Oie, Oa, and mineral soil base saturation rose 69, 84,
and 58%, resp. by 2010.
PH
Base cations
Forest soils
Italy
4.5 to 28.8 Not applicable
throughfall N
(NOs" +
NH4+)
Gradient: Exchangeable base cations and pH decreased with Ferretti et
increasing N deposition, and foliar nutrient N ratios (especially N:P al. (2014)
and N:K) increased. Comparison between bulk open-field and
throughfall data suggested possible canopy uptake of N, levelling
out for bulk deposition >4-6 kg/ha/yr.
Be weathering Boreal plains NE Alberta, S = 118, Not applicable Modeling: base cation weathering rate estimated for 63 sites
rate Canada NH4+ = 93, using PROFILE was 17 mmolc m2/yr, however acidification was
NO3" = 49 in not expected because base cations from fugitive dust sources
units of were a comparatively high 250 mmolc m2/yr, offsetting much of
mmolc m2/yr the acidifying input of N and S deposition.
Watmouqh
et al.
(2014)
PH
Soil N
Base cations
Fungi:bacteria
ratio
Belowg round
biomass
Microbial
community
Structure
Semiarid Mongolia Not specified 0,17.5,52.5, Addition: Soil pH decreased across the N addition gradient by
grassland 105.0, 175, and 0.3-1.8 units in 2010 and by 0.1-1.7 units in 2011. Decreased
280 NH4NO3 concentrations of mineral cations Ca2+, Mg2+, and Na+ were
fertilizer observed. The observed increases in above- and belowground
biomass and changes in plant community structure were mainly
(57-69%) attributed to the increase in soil N availability and
changes in soil base cations. N addition increased the
fungi:bacteria ratio by 5-18% in 2010 and by 2-10% in 2011.
Chen et al.
(2015)
4-32
-------
Table 4-5 (Continued): Base cation leaching and exchange.
Process/ Type of Deposition Addition
Indicator Ecosystem Region kg/ha/yr kg/ha/yr Effect of Deposition Reference
Model: PnET-BGC was used to evaluate biophysical factors that Zhou et al.
affect CLs of acidity. Model simulations included a range of future (2015c)
scenarios of decreases in atmospheric nitrate, ammonium, and
sulfate deposition from the present to 2,200; historical forest
harvesting; supply of naturally occurring organic acids; and
variations in lake hydraulic residence time.
Assuming the current soil base saturation of 6.4%, a decrease in
SO42" deposition from 0 to 100% resulted in a soil percentage BS
range of 6.2 to 15.3%, respectively, after 200 yr. (Preindustrial soil
percentage BS -22%).
Time Series: In 1984 and 2012, soil samples were taken from Beraer et
20 cm downhill and 3 m away from the base of a beech tree stem. al. (2016)
Exchangeable Ca2+, Mg2+, and pH increased in 0-5 cm soil from
1984 to 2012. Recovery appeared delayed in deeper soils. Foliar
base cations Ca, Mg, and K decreased. (Foliar K declined the
most at 48%).
Be
Maritime pine
Northwest
Not specified
None
Field observation: The soil parent material (slate, biotitic schist,
Eimil-Fraaa
forest
Spain
mica schist, and granite) significantly affected base cation
et al.
concentrations.
(2016)
Foliar base
Eastern U.S.
Fernow
-10 kg
35 kg N/ha/yr
N Addition: Foliar measurements were used as a proxy for soil
Gilliam et
cation
temperate,
Experimental
N/ha/yr (wet)
(NH4)2S04
micronutrient availability under N addition. Excess N lowered
al. (2016a)
hardwood
Forest, WV
plant-available Ca and, to a lesser degree, Mg, but not K. N
forest
addition significantly affected Ca:AI ratios in Viola sp. and Rubus
herbaceous
sp.
layer (Viola
rotundifolia,
Rubus
allegheniensis)
Base Adirondack SO42" = None
saturation Mtns. region, 290.3-365.7
NY eq/ha/yr;
NO3- = 172.
5-233.5
eq/ha/yr
Be Predominantly Vienna Woods, Not specified None
beech Austria locally
4-33
-------
Table 4-5 (Continued): Base cation leaching and exchange.
Process/
Indicator
Type of
Ecosystem
Region
Deposition
kg/ha/yr
Addition
kg/ha/yr
Effect of Deposition
Reference
Soil solution Be Forest
(2016)
Hubbard Brook Not specified 1,028 kg Ca/ha Field observation: Significant increase in soil Ca concentrations. Shao et al.
Experimental (as wollastonite) Increases in pH and ANC and decreases in inorganic Al in the Oa,
Forest, NH in 1999 Bh, and Bs soil horizons. Ca:AI ratios ranging from 1 to 5.3 in the
Oa-, Bh-, and Bs-horizons before treatment increased significantly
after Ca addition, with the greatest increase occurring in the
Oa-horizon. Average 2000-2011 Ca fluxes in soil solution across
the study area's three subwatersheds ranged from
26.4 + 3.2 mmol/m2/yrto 29.9 + 4.2 mmol/m2/yr. The weighted
average percentage increase in Ca fluxes for the three
subwatersheds from 1999-2011 were 139, 91, and 97% for Oa-,
Bh-, and Bs-horizons, respectively.
Soil Be
Temperate
grassland
Mongolia,
China
1.6 mg
N/m2/yr (not
specified if
wet or total)
1, 2, 4, 8, 16,
32, 64 g N/m2/yr
(urea)
N Addition: Soil exchangeable Mn2+, Fe3+, and Al3+
concentrations increased linearly with N addition.
Tian et al.
(2016b)
Be weathering
Forest
NW British
Columbia,
CAN (Kitmat
and Prince
Rupert)
0-80 meq
S042"/m2/yr
and
0-30 meq
N/m2/yr
None
Model: Weathering rates were modeled by PROFILE and A2M
solver using empirical soil data to parameterize the models. Rates
ranged between 19 and 393 meq/m2/yr (average: 76) in the top
50 cm of soil in Prince Rupert and between 24 and 118 meq/m2/yr
(average: 57 meq/m2/yr) in Kitimat.
Williston et
al. (2016)
Be deposition
Mediterranean
holm-oak
(Quercus ilex)
forests ranging
from the typical
Mediterranean
climate to a
drier, more
seasonal
climate
Spain
Not specified None
Field observation: In wetter forest sites, 55-65% of total base
cation deposition was wet deposition. Rainfall and net throughfall
were positively correlated for leaching for K+ and uptake for NHV
at all sites. Variable response between sites was found for Na+,
Ca2+, SO42" and CI". The authors suggest that the interplay of dry
deposition, leaching, and uptake at the canopy was different
depending on site climate and air quality characteristics.
Aquillaume
et al.
(2017)
Al = aluminum; AOSR = Athabasca oil sands region; Be = base cations; BCE = exchangeable base cations; Ca:
kg = kilogram; m = meter; Mg2+ = magnesium ion; N = nitrogen; Na+ = sodium ion; NH4+ = ammonium; NH4NO3:
S = sulfur; S042" = sulfate; yr = year.
h = calcium ion; Fe = iron; g = gram; ha = hectare; K+ = potassium ion;
: ammonium nitrate; (NH4)2S04 = ammonium sulfate; N03" = nitrate;
4-34
-------
4.3.5
Aluminum Mobilization
The 2008 ISA documented that when soil base saturation is 15 to 20% or lower,
acidifying deposition can mobilize inorganic Al, which can lead to its leaching into soil
solution and into surface waters (Cronan and Schofield. 1990; Reuss and Johnson. 1985;
Reuss. 1983). Leaching of inorganic Al is an extremely important effect of acidifying
deposition because some forms of inorganic monomeric Al, including Al3+ and various
hydroxide species, are toxic to tree roots, fish, algae, and aquatic invertebrates
(Appendix 5 and Appendix 8). Increased concentrations of exchangeable inorganic Al in
the soil have been identified through repeated sampling in the U.S. and Europe over
periods ranging from 17 to 41 years in studies by Billett et al. (1990). Falkcngrcn-Grcrup
and Eriksson (1990). Bailey et al. (2005). and Lawrence et al. (1995).
The negative biological effects of Al mobilization are discussed in Appendix 5; in
general, Al disrupts Ca uptake by tree roots (Shortle and Smith. 1988). Substantial
evidence of this relationship has been provided through field studies (Kobe et al.. 2002;
Minocha et al.. 1997; Shortle et al.. 1997; McLaughlin and Tjoelker. 1992; Schlegel et
al.. 1992) and laboratory studies (Cronan and Grigal. 1995; Svcrdrup and Warfvinge.
1993). These studies make clear that high inorganic Al concentration in soil water can be
toxic to plant roots. The toxic response is often related to the concentration of inorganic
Al relative to the concentration of Ca, expressed as the molar ratio of Ca to inorganic Al
in soil solution. From an exhaustive literature review, Cronan and Grigal (1995)
estimated a 50% risk of adverse effects on tree growth if the molar ratio of Ca to Al in
soil solution was as low as 1.0. They estimated a 100% risk for adverse effects on growth
at a molar ratio value Ca:Al <0.2 in soil solution and minimal to no risk is thought to
occur at CaAl >10.
New studies on Al in soils are summarized in Table 4-6 and include an investigation of
long-term soil solution chemistry trends in Hubbard Brook Experimental Forest (Fuss et
al.. 2015). an investigation of the influence of increasing dissolved organic matter (DOM)
on toxic inorganic Al (Al,) concentration in 52 Adirondack and Hubbard Brook
watersheds (Fakhraei and Driscoll. 2015). an investigation of the effects of soil parent
material on Al (Eimil-Fraga et al.. 2016). a model comparison of three models"
predictions of soil solution Al concentrations in three monitored Swiss and German
forests (Bonten et al.. 2015). and a study showing that N addition increased Al
mobilization on grasslands in England (Horswill et al.. 2008).
Fakhraei and Driscoll (2015) evaluated the influence of increasing DOM on Al
concentration in the Adirondacks and Hubbard Brook. The authors linked a PnET-BGC
4-35
-------
model chemical equilibrium subroutine with an optimization algorithm. They determined
that accounting for the increasing concentration of DOM (as acid deposition decreases
and ANC increases) and its binding capacity for A1 is necessary in the model to avoid
substantial overestimation of available toxic A1 in waters. In an empirical study, Fuss et
al. (2015) found that A1 in soil solution in the mineral horizon decreased over the period
of 1984-2011 (also discussed in the recovery Appendix 4.6.1).
4-36
-------
Table 4-6
Aluminum mobilization.
Process/
Indicator
Type of
Ecosystem
Region
Deposition
kg/ha/yr
Addition
kg/ha/yr
Effect of Deposition
Reference
Base cation depletion Grasslands
Soil pH
Soil [Al], [Fe], [Mg]
Peak District
National Park,
England,
Not specified
Plots treated for 8
to 10 yr with 0, 35,
or 140 N as
NH4NO3
Addition: Treatments caused the
grassland soils to lose 23 to 35% of their
total available bases and acidify by 0.2
to 0.4 pH units. Al, Fe, and Mn were
mobilized and taken up by limestone
grassland forbs and were translocated
down the acid grassland soil.
Horswill et al.
(2008)
Ali, DOM, DOC
Lakes
Adirondack
Long-Term
Monitoring
(ATLM) Program
lakes
Variable
None
Model: A PnET-BGC chemical
equilibrium subroutine was linked with
an optimization algorithm. Accounting for
the increasing concentration of dissolved
organic matter (DOM) as acid deposition
decreases and ANC increases, DOM's
binding capacity for Al is necessary to
avoid substantial overestimation of toxic
Ali in waters.
Fakhraei and
Driscoll (2015)
Ali
Forest and Hubbard Brook Variable
streams Experimental
Forest, NH
None
Field observation/elevation gradient: Fuss et al.
Ali in soil solution in the mineral horizon (2015)
decreased over the period of
1984-2011.
Al
Forest
Three Swiss
forest monitoring
sites
Not specified None
Model: VSD modeled soil solution Al
concentrations are substantially smaller
than measured ones. However, VSD
only calculates free Al3+, whereas
measurements also include other Al
species as Al hydroxides, Al fluorides,
and Al complexed by DOM.
Bonten et al.
(2015)
Al
Spruce forest
Bechtel,
Switzerland
Not specified
None
Model: ForSAFE-modeled Al3+ was in Bonten et al.
the range of the measured soil solution (2015)
values at 20 cm depth (1990-2005)
4-37
-------
Table 4-6 (Continued): Aluminum mobilization.
Process/
Indicator
Type of
Ecosystem
Region
Deposition
kg/ha/yr
Addition
kg/ha/yr
Effect of Deposition
Reference
Al, Ca and protons
Norway spruce
forest
Long-term
monitoring site in
Germany
Not specified
None
Model: SMARTml results showed that
modeled Al3+, Ca2+, and H+ agree with
measured values in soil solution
(1990-2005).
Bonten et al.
(2015)
Total Al, reactive Al
Maritime pine
forest
Northwest Spain Not specified
None
Field Observation: The soil parent
material (slate, biotitic schist, mica
schist, and granite) significantly affected
total and reactive Al in soil solution.
Total Al ranged from 17.2 to 64.2 pmol/L
(p = 0.0006). The concentration was
significantly higher for soil developed
from mica schist developed soil than
granite and biotitic schist.
Eimil-Fraaa et
al. (2016)
Al = aluminum; Al: = toxic inorganic Al; AOSR = Athabasca oil sands region; Be = base cations; BCE = exchangeable base cations; Ca2+ = calcium ion; Fe = iron; g = gram;
H+ = hydrogen ion; ha = hectare; K+ = potassium ion; kg = kilogram; m = meter; Mg2+ = magnesium; N = nitrogen; Na+ = sodium ion; NH4+ = ammonium; NH4N03 = ammonium
nitrate; (NH4)2S04 = ammonium sulfate; N03" = nitrate; S = sulfur; S042" = sulfate; yr = year.
4-38
-------
4.3.6
Nitrification and Denitrification
Nitrification is the microbial oxidation of ammonia or ammonium to form nitrite
followed by the oxidation of the nitrite to form NO3 . The 2008 ISA documented that
oxidation of ammonia into nitrite is performed by two groups of organisms,
ammonia-oxidizing bacteria and ammonia-oxidizing archaea. The rate of nitrification is
controlled by numerous factors, including substrate availability (presence of NH/),
aeration (availability of O2, often as well-drained soils with <60% soil moisture), and
acidity (pH). N addition may increase nitrification, which is often stimulated in soils with
a C:N ratio below about 20 to 25 (Ross et al.. 2004; Aber et al.. 2003). The microbial
process of autotrophic nitrification is an acidifying process, releasing 2 moles of
hydrogen ion (H+) per mole of NH44" converted to NO;, (Reuss and Johnson. 1986). As
the N cycle becomes enriched through cumulative N addition, N becomes more abundant,
competition among organisms for N decreases, net nitrification rates often increase, and
NO;, can leach from the soil (Aber et al.. 2003; Aber et al.. 1989).
Soils with a C:N ratio below about 20 to 25 are associated with stimulated mineralization,
nitrification, and cation leaching (Ross et al.. 2004; Aber et al.. 2003). This observation
makes the C:N ratio an especially useful field measurement that provides a relative index
rather than a quantitative rate of N leaching (Ross et al.. 2004). C:N ratios in the forest
floor are generally inversely related to acidifying deposition levels, although the
relationship is stronger for hardwood stands than conifer stands (Aber et al.. 2003). The
C :N ratio is a reliable and relatively straightforward measure for identifying forest
ecosystems that may be experiencing soil acidification and base leaching as a result of N
input and increased nitrification.
Denitrification is the microbial process that transforms NO; by anaerobically reducing it
to nitrite (NO; ). nitric oxide (NO), the greenhouse gas nitrous oxide (N2O) and N2. The
2008 ISA documented that in terrestrial ecosystems, denitrification mainly occurs in
oxygen-depleted soils (e.g., during periods of water saturation), groundwater, and
riparian zones. Soil pH has a marked effect on denitrification, with lower rates in acidic
as compared with alkaline conditions (Yamulki et al.. 1997). Soil denitrification and N2O
production and consumption are extremely variable in time and space (McClain et al..
2003).
New information published since 2008 is summarized in Table 4-7. New empirical work
along an N deposition gradient in Oregon found that nitrification increased with
increasing N deposition (Perakis and Sinkhorn. 2011); likewise, denitrification increased
with N deposition in northeastern forests (Morse et al.. 2015a). In a deposition exclusion
4-39
-------
study in Europe, after two decades of deposition exclusion, net nitrification and NO;,
concentration in soils were not detectible, and in fact, the soil switched from a net source
of NOy to a next sink (Eickenscheidt and Brumme. 2012).
A growing body of information indicates that increased N deposition also alters the soil
microbial community (Frccdman et al.. 2013). Discussion of this is provided in
Appendix 6. Marusenko et al. (2013) explored the role of fungi in NO3 and N2O
production in soils from regions across the southwestern U.S. and found that fungi are
significant sources of N2O production in soils in semiarid grasslands and deserts. Russow
et al. (2008) found that soils with high soil organic matter adsorbed added NFL+, making
it difficult to determine microbial activity.
Several new syntheses evaluated N addition effects on denitrification and nitrification in
terrestrial ecosystems (Yang et al.. 2017; Bouwman et al.. 2013; Lu et al.. 2011a; Liu and
Greaver. 2009). Globally, the amount of N removed from ecosystems by denitrification
may be higher in terrestrial ecosystems than from groundwater or riparian zones
(Bouwman et al.. 2013); however, other older estimates have indicated more
denitrification may occur in riparian wetlands and in first-order streams than in terrestrial
ecosystems (Van Breemen et al.. 2002). although the authors acknowledged this estimate
likely underestimated terrestrial N. Liu and Greaver (2009) showed that N addition
significantly increased denitrification from all ecosystems tested (coniferous forest,
deciduous forest, tropical forest, wetland, grassland) except heathland. Among the five
chemical forms of N fertilizer added to ecosystems in studies, NO;, showed the strongest
stimulation of N2O emission (Figure 4-5). Lu et al. (201 la) further confirmed that N
addition stimulates nitrification and denitrification (Figure 4-6). Using data extracted
from 206 peer-reviewed papers, the meta-analysis showed that the largest changes caused
by N addition in the ecosystem N cycle were increased nitrification (154%), nitrous oxide
emission (134%), and denitrification (84%). In addition, Yang et al. (2017) evaluated N
cycling in five biomes in California and found a strong (r2 = 0.34) significant linear
correlation between NO; and nitrification rates. They also found a strong negative
relationship between gross nitrification and soil C:N in forests that had soil C:N ratios
greater than 20. However, the authors noted that not all forests have such high soil C:N
ratios as those in their study that were dominated by coniferous trees. Coniferous forests
generally have higher litter C:N ratios than deciduous forests, and deciduous forests with
lower C:N ratios may also have negative relationship with C:N and nitrification.
4-40
-------
Table 4-7
Nitrification and denitrification.
Ambient N/S
Deposition
N/S Addition
Process Endpoint Type of Ecosystem
Region
kg/ha/yr
kg/ha/yr
Effect of Deposition
Reference
Nitrification and
Agricultural black
Central
Not specified
81.3 (KNOs,
Isotopic tracer: Addition of 15N revealed
Russow et al.
denitrification
earth soils (haplic
Germany
Ca[N03]2);
denitrification of NO3" represents the main
(2008)
chernozem); two
80 kg (NH4)2S04
pathway of soil N2O release. On average,
sites: high and
76 and 54% of N2O was emitted during
normal SOM
denitrification from soils with high and
normal SOM content, respectively.
Denitrification contributed, on average, only
17 and 12% of released NO from soil with
high and normal SOM content,
respectively.
Nitrification
Spruce plantation
Hoglwald,
30 (two
None
Time series: Dynamic internal N cycle
Kreutzer et al.
Bavaria,
decades)
within the soil, driven by growth and death
(2009)
Germany
of the microbial biomass, which turns over
approximately seven-fold each year.
Denitrification
Agricultural crop,
Global
Not specified
10 to 562
Meta-analysis: Analysis of
Liu and
forest, grassland,
313 observations across all ecosystems
G re aver
wetland, tundra,
show N addition increased N2O emission
(2009)
heathland, and
by 216%.
desert
Nitrification
Agriculture and
Not specified
Mean = 105 Tg
Oto >100
Meta-analysis: included 206 papers on
Lu et al.
Denitrification
nonagriculture
N/yr
responses of ecosystem N cycle in
(2011a)
response to N addition. Increases in
nitrification (154%), N2O emission (134%),
and denitrification (84%) were found.
Increased N2O emissions and N leaching
under N addition tended to export the N out
of the systems rather than benefit plant
uptake over the long term, suggesting a
leaky terrestrial N system.
4-41
-------
Table 4-7 (Continued): Nitrification and denitrification.
Process Endpoint Type of Ecosystem
Region
Ambient N/S
Deposition
kg/ha/yr
N/S Addition
kg/ha/yr
Effect of Deposition
Reference
Nitrification and
denitrification
Microbial N demand
Temperate hardwood
and conifer forests
(unfertilized)
Nine sites in the
north-central
Oregon coast
range, U.S.
Not specified None
Perakis and
(2011)
Soil and foliar N gradient: As future
reductions in N deposition to polluted sites Sinkhorn
occurs, symptoms of N saturation are most
likely to persist where soil N content
remains elevated. Temperate and
hardwood forests of the north-central
Oregon coast range showed the ratio of net
to gross N mineralization and nitrification
increased along the gradient, indicating
progressive saturation of microbial N
demands at high soil N.
Nitrification
Common garden
Denmark
13 to 19 for
Isotopic tracer: Litter 815N was positively
(Callesen et
experiment, five
broadleaf
correlated with N status based on
al.. 2013)
broadleaved tree
forest; 18 to
nitrification, as well as other factors. A
species
26 for Norway
spruce
linear relationship was found between
fungal mycelia production and net
nitrification rate in lab incubations of soils
collected in the field.
Denitrification
Forested watershed
Pond Branch in
Maryland, U.S.
10 ± 4
None
Isotopic tracer: Spatial and temporal Duncan et al.
extrapolations of measured rates suggest (2013)
that a minimum of 16-27% of atmospheric
N deposition is lost to denitrification.
Fungal nitrification
and denitrification
Semiarid grasslands Arizona and Not specified
New Mexico
10 g soil to a
100 mL solution of
50 mmol/L
(NH4)2S04,
0.2 mol/L K2HPO4,
and 0.2 mol/L
KH2PO4, with a
pH of 7.2.
Moisture X biocide: Fungi are significant
sources of N2O production in soils from
semiarid grasslands and deserts,
expanding evidence that fungi play a vital
role in the N cycle of arid lands.
Marusenko et
al. (2013)
4-42
-------
Table 4-7 (Continued): Nitrification and denitrification.
Process Endpoint
Type of Ecosystem
Region
Ambient N/S
Deposition
kg/ha/yr
N/S Addition
kg/ha/yr
Effect of Deposition
Reference
Denitrification
Agricultural land and
natural ecosystems
Global
24 to 46 Tg
N/yr from
1900-2050
Not applicable
Model: N2 production from denitrification
increased from 52 to 96 Tg/yr between
1900 and 2000, and N2O emissions
increased from 10 to 12 Tg N/yr. The
scenarios suggest a further increase to
142 Tg N2 and 16 Tg N20-N/yr by 2050.
Riparian buffer zones are an important
source of N2O, contributing an estimated
0.9 Tg N20-N/yr in 2000.
Bouwman et
al. (2013)
Nitrification and
denitrification
Microbial community
Sugar maple
dominated northern
hardwood forest
Upper Michigan
15 to 20
30 kg N in the
form of NaNC>3
pellets delivered
to the forest floor
over the growing
season
Addition: NO3" addition to forest stands
across a 500-km climatic gradient
decreased the abundance and richness of
key protein-coding genes in archaea and
bacteria responsible for N fixation,
ammonification, denitrification, and
assimilatory NO3" reduction; the same was
true for bacterial genes mediating
nitrification and dissimilatory NO3"
reduction.
Freedman et
al. (2013)
Denitrification Northern hardwood HBEF, White 6 to 8 None Method comparison: Both the isotopic Kulkarni et al.
forest Mountain tracer and gas-flow soil core method (2014)
National Forest, indicate that denitrification is higher and
NH N20:N2 ratios are lower (<0.02) than
previously thought in the northern
hardwood forest and that short-term abiotic
and biotic transformations of atmospheric
N deposition to gas are significant in this
ecosystem.
4-43
-------
Table 4-7 (Continued): Nitrification and denitrification.
Process Endpoint Type of Ecosystem
Region
Ambient N/S
Deposition
kg/ha/yr
N/S Addition
kg/ha/yr
Effect of Deposition
Reference
Nitrification
Mixed hardwood
forested headwater
catchments
South-central
Ontario, Canada
Average NO3
N deposition
2.78 ± 1.22;
average NhV
N deposition
3.90 ± 1.39
Not applicable Time series: Seasonal differences in
nitrification were largely driven by
temperature, soil moisture, and inorganic N
concentration in soil.
Annual nitrification fluxes were almost two
orders of magnitude greater than N
deposition or NO3" leaching. Nitrification
rates scaled up to annual catchment-scale
production of NO3"; the resulting fluxes are
64.9 ± 8.7 and 59.7 ±3.1 kg N/ha/yr, which
greatly exceed seasonal inputs in
deposition.
Rates of nitrification and mineralization
were similar, indicating that almost all
mineralized N is converted to NO3"
(ranging from 71 to 99%).
Casson et al.
(2014a)
Nitrification
denitrification
Deciduous and
coniferous forests
Ontario,
Canada; New
Hampshire; and
Maine
Three sites
along a
gradient: 4.5,
7, and 11
Not applicable Gradient: N gas flux increased
systematically with natural N enrichment
from soils with high nitrification rates.
N gas fluxes were linked to patterns of N
availability in forests; results do not
suggest that these fluxes respond to
increases in atmospheric N deposition at
the study sites.
Morse et al.
(2015a)
4-44
-------
Table 4-7 (Continued): Nitrification and denitrification.
Process Endpoint Type of Ecosystem
Region
Ambient N/S
Deposition
kg/ha/yr
N/S Addition
kg/ha/yr
Effect of Deposition
Reference
Nitrification
C:N
Five biome classes in California 0.6-18.4 kg
California: desert, N/ha/yr
grassland, shrubland,
forest, wetland
Grassland soil:
76.4 kg N/ha/yr as
NPK (29:3:4);
desert soil: 60 kg
N/ha/yr as
NH4NO3;
shrubland soil:
60 kg N/ha/yr as
NH4NO3
N Addition: Across biomes, a positive
correlation of gross nitrification to soil NO3"
(r2 = 0.34) and a negative correlation to soil
C:N (r2 = 0.31) was observed.
No correlation was found between gross N
mineralization and nitrification. Deserts had
the lowest gross N mineralization rates and
exhibited similar nitrification rates to the
shrublands and grasslands. Only 15% of
NH4+ produced was nitrified in the forests
compared to 47 to 86% in the other
biomes. This suggests that NHV
production rates did not limit nitrifiers in the
forests, but rather, that another factor
limited nitrifier activity.
Yang et al.
(2017)
Ca(N03)2 = calcium nitrate; g = gram; HBEF = Hubbard Brook Experimental Forest; K2HP04 = dipotassium phosphate; KH2P04 = monopotassium phosphate; KN03 = potassium
nitrate; L = liter; mL = milliliter; mmol = millimole; mol = mole; N = nitrogen; N2 = molecular nitrogen; N20 = nitrous oxide; N20-N = nitrogen from nitrous oxide; NaN03 = sodium
nitrate; NH4+ = ammonium; (NH4)2S04 = ammonium sulfate; NO = nitric oxide; N03" = nitrate; SOM = soil organic matter; Tg = teragram yr = year.
4-45
-------
(a) aii
Mean (156)
i—*-
(b)
Agriculture (aerobic) (3S)
Agriculture (artaercttic:) (1B) I—•—I
Ooniterous (33)
Deciduous (15) f-
Tropicai forest (.11)
Wellarvd (19) j I-
Grassland (16) | t-
Heathland (3) (-
(c) N form
NH,NO= (44)
NH/ t2&)
NO," (24}
Urea (32)
UAN (23)
(d) Experimental length
Short term (61)
Long term (90)
(e) N addition level
<65 (36)
55-150(29>
>150 (51)
I—•-
3 4 5 6 7 6 9 1S 16 20
Response ratio
N = nitrogen; NH4N03 = ammonium nitrate; NH4+ = ammonium; N03 = nitrate; UAN = urea and ammonium nitrate fertilizer.
The data are expressed as the mean response ratio with 95% confident intervals. The numbers of studies included are indicated in
parentheses.
Source: Liu and Greaver (2009).
Figure 4-5 Effects of nitrogen addition on biogenic nitrous oxide emission.
4-46
-------
3
Leaf N
Aboveg round plant N
Belowground plant N
Litter N
Microbial N
Organic horizon N
Soil N
DON
Soil inorganic N
N mineralization
Immobilization
Nitrification
Denitrification
Leaching ¦
0.0 0.4 0.8 1.2 1.6 2.0
Weighted response ratio (RRJ
N = nitrogen; N20 = nitrous oxide; RR = response ratio.
Bars represent RR++ ± SE. The vertical line is drawn at logeRR = 0. The sample size for each variable is shown next to the bar.
Source: Lu etal. (2011a1.
Figure 4-6 The weighted response ratio for the responses to nitrogen
addition for fluxes and pools related to the ecosystem nitrogen
cycle in agricultural (open bars) and nonagricultural (closed bars)
ecosystems.
4.3.7 Decomposition
Decomposition is a general term that refers to the breakdown of organic matter
(Schlesinger. 1997). Decomposition is an important part of N and C cycling that can be
altered by N deposition. Decomposition rates correlate with ratios of C:N, lignin:N, and
4-47
-------
lignin:cellulose in litter, all of which may be altered by N deposition. The addition of N
can stimulate the decomposition of labile compounds that degrade during the initial
stages of decomposition, but added N can suppress the decomposition of more
recalcitrant material. Evidence for this is widespread in forests but has not yet been well
documented in grasslands and other ecosystems. Since 2008, there are new addition
studies and meta-analyses to better understand the mechanisms and response trends.
The 2008 ISA documented that the soil microbial community (bacteria and fungi) are the
main decomposers of organic matter. Both the microbial community composition and
microbial enzyme activity can dynamically respond to shifts in inorganic nutrient and
substrate availability (Compton et al.. 2004; Carreiro et al.. 2000); the shift reflects the
nutrient and energy limitation of the microbial community. Litter decay rates are also
well established to correlate with ratios of C:N, lignin:N, or lignin:cellulose in litter
(Hobbic. 2008: Aerts. 1997; Melillo et al.. 1982). These chemical traits are strong
predictors of litter decay, accounting for over 73% of the variation in litter decomposition
rates worldwide (Zhang et al.. 2008).
Traditionally, carbon dioxide (CO2) is measured as an indication of soil respiration and a
proxy for decomposition. Since the 2008 ISA, analysis of microbial enzymes and genes
in the soil have been used to identify microbial activity and determine how it relates to
decomposition. Numerous studies have been published since 2008 describing how N
addition affects the decomposition of organic C and N (Table 4-8).
4-48
-------
Table 4-8 Decomposition.
Process/ Type of Deposition Addition
Indicator Ecosystem Region kg/ha/yr kg/ha/yr Effect of Deposition Reference
Decomposition
Sugar maple
Bear Brook
1,800 eq/ha/yr
N + S Addition: Caused increased N concentration in
Hunt et al.
hardwood forest
Watershed in
of (NH4)2S04
leaves and faster short-term decomposition.
(2008)
eastern Maine
Enzyme activity
Eight annual
Controlled
2.0 and 44.0
Growth chamber: N deposition increased soil enzyme
Mannina et al.
Decomposition
herb species
ecosystem (litter
activity known to breakdown cellulose (cellobiosidase,
(2008)
bags placed in
p-glucosidase and (3-xylosidase).
annual
herb-based
microcosm
ecosystems)
Enzyme activity
Northern
Michigan
6.8 to 11.8 30 N03~
N Addition: N addition altered rates of organic matter
Preaitzer et
Decomposition
hardwood forest
decomposition by suppressing the soil enzymes
al. (2008)
responsible for litter degradation when litter has a higher N
concentration. This causes an increase in surface soil C
storage.
Enzyme activity
Forests
Central
Not 100(NH4N03)
N Addition: Generally stimulated activities of cellulose
Keeler et al.
Decomposition
grasslands
Minnesota
specified
degrading enzymes in litter and soil, but had no effect on
(2009)
lignin degrading enzyme activity. N addition had a negative
or neutral effect on litter and SOM decomposition in the
same sites, with no correspondence between effects of N
on enzyme activity and decomposition across sites.
Enzyme activity
Northern
Catskill
50 (NH4NO3)
N Addition: Identified that patterns in microbial community
Weand et al.
hardwood forest
Mountains of
structure and function were more strongly influenced by
(2010)
New York, U.S.
tree species than by fertilization.
Soil respiration
Forest
Global
Varied
Meta-analysis: N additions decreased root respiration,
Janssens et
Decomposition
heterotrophic respiration, and soil CO2 but had no effect on
al. (2010)
litter decomposition.
4-49
-------
Table 4-8 (Continued): Decomposition.
Process/
Indicator
Type of
Ecosystem
Region
Deposition
kg/ha/yr
Addition
kg/ha/yr
Effect of Deposition
Reference
Soil respiration
Temperate
broadleaf and
mixed forest,
temperate
conifer forest,
boreal forest,
tropical forest,
grassland,
wetland, tundra,
desert, and arctic
Global
Varied
Meta-analysis: N additions decreased heterotrophic
respiration but had no effect on total soil respiration.
Liu and
Greaver
(2010)
Enzyme activity
Decomposition
Forest and
grasslands
U.S.
None
None
Synthesis: Data from 28 ecosystems show resources to
produce the enzymes phenol oxidase and
(3-1,4-glucosidase are uncoupled. This indicates that the
increasing recalcitrance of organic matter decreases C and
nutrient availability and slows microbial growth.
Sinsabauqh
and Shah
(2011)
Enzyme activity
Mature black
Central Alaska
Not
Years
N Addition: N fertilization may alter decomposer
Talbot and
Decomposition
spruce forest in
specified
2009 = 200
community structure by favoring a shift toward cellulose-
Treseder
upland boreal
(NH4NO3)
and mineral-N users. Cellulose degrading microbes
(2012)
ecosystem
Years
(decomposers) were competitively dominant under N
2010 =100
fertilization.
(NH4NO3)
Enzyme activity
Decomposition
Sugar maple
forests
Michigan
Not
specified
30
Meta-analysis: N addition increased cellulose
decomposition by 9% and decreases lignin decomposition
rates by 30%. Overall, N increases the amount of litter
mass entering the humus pool and leads to increases in
soil C storage under experimental N deposition.
Whittinahill et
al. (2012)
Enzyme activity
Decomposition
Northern
hardwood forest
Minnesota
N Addition: Accelerated the initial decomposition rate.
Faster initial decomposition rates corresponded to higher
activity of polysaccharide-degrading enzymes and greater
relative abundances of Gram-negative and Gram-positive
bacteria. Later in decomposition, externally supplied N
slowed decomposition, increasing the fraction of slowly
decomposing litter, reducing lignin-degrading enzyme
activity, and relative abundances of Gram-negative and
Gram-positive bacteria.
Hobbie et al.
(2012)
4-50
-------
Table 4-8 (Continued): Decomposition.
Process/
Type of
Deposition
Addition
Indicator
Ecosystem
Region
kg/ha/yr
kg/ha/yr
Effect of Deposition
Reference
Decomposition
Deciduous
Catskill
9.0
50 mg N/ha/yr
N Addition: N addition caused decrease in mineralization
Lovett et al.
forests
Mountains of
N (NH4NO3)
and nitrification and an increase in forest floor C pools and
(2013)
southeastern
C:N, indicating that N addition increased C sequestration in
New York, U.S.
the organic horizons of the soil, most significantly in
hemlock plots.
Enzymes Northern Upper Michigan 15 to 20 30 N Addition: Caused slower organic matter decay and Freedman et
Decomposition hardwood forest altered microbial community composition and function. al. (2013)
Observed a decrease in the abundance and richness of
key protein-coding genes in archaea and bacteria
responsible for N fixation, ammonification, denitrification,
and assimilatory NO3" reduction; the same was true for
bacterial genes mediating nitrification and dissimilatory
NO3" reduction.
N Addition: (1) significantly altered the composition of Eisenlord et
actinobacterial and fungal genes mediating plant and al. (2013)
fungal cell wall depolymerization; (2) significantly
decreased the richness and diversity of genes involved in
the depolymerization of starch (-12%), hemicellulose
(-16%), cellulose (-16%), chitin (-15%), and lignin (-16%);
and (3) resulted in small changes in community
composition (25% difference in fungi; 18% in
actinobacteria).
Enzymes
Northern
Upper Michigan
15 to 20
30 as six equal
N Addition: Observed that atmospheric N deposition
Freedman
Decomposition
hardwood forest
applications of
increases saprotrophic bacterial laccase-like multicopper
and Zak
NaNC>3 pellets
delivered to the
forest floor over
the growing
season
oxidases (LMCOs). These results suggest a plausible
mechanism by which anthropogenic N deposition has
reduced decomposition, increased soil C storage, and
accelerated phenolic DOC production.
(2014)
Decomposition
Forest
Switzerland
Not
7 (NH4NO3)
N Addition: Promoted the production of new fungal
Grieoentroa
fungal residue
specified
70 (NH4NO3)
residues but slowed the decomposition of old residues in
et al. (2014)
forest soil fractions. Preservation of old microbial residues
could be due to decreased N limitation of microorganisms
and therefore a reduced dependence on organic N
sources.
Enzymes Sugar maple Michigan 5.8 to 7.3 30
Decomposition forests
4-51
-------
Table 4-8 (Continued): Decomposition.
Process/
Indicator
Type of
Ecosystem
Region
Deposition
kg/ha/yr
Addition
kg/ha/yr
Effect of Deposition
Reference
Enzyme activity Northern Bear Brook
hardwood and Watershed
softwood forest (BBWM), ME,
U.S.
25.5
N Addition: After 22 yr of N addition, N enrichment had Mineau et al.
little effect on microbial enzyme activity in terrestrial (2014)
compartments, even across varying degrees of organic
matter recalcitrance.
Soil respiration Multiple biomes Global
Varied
Meta-analysis: N addition significantly increased soil
respiration by 2.0% across all biomes but decreased
respiration by 1.44% in forests and increased it by 7.84 and
12.4% in grasslands and croplands, respectively (p < 0.05).
The response ratios of soil respiration to N addition were
positively correlated with mean annual temperature (MAT),
most significant when MAT was less than 15°C. N addition
largely altered root and microbial biomass and soil C
content, which are likely the mechanisms behind the
altered soil respiration.
Zhou et al.
(2014a)
Decomposition Grassland
Minnesota,
Nebraska, Iowa,
Kansas;
Colorado
3.1 to 18
100
N Addition: Decreased microbial respiration of OM by as Riggs et al.
much as 29% relative to control plots, and consequently, (2015)
decreased C loss from this pool.
Decomposition Forest
Great Lakes
region
6.8 to 11.8 30 as NaNQ3
N Addition: Fine root biochemistry was less responsive
than leaf litter to long-term simulated N deposition. Fine
roots were the dominant source of difficult-to-decompose
plant C fractions entering the soil.
When combined with litter production, simulated N
deposition increased N flux through leaf litter by an average
of 29% but did not affect fine root N flux.
Xia et al.
(2015)
Enzyme activity Subalpine forest
Loch Vale
watershed
(LVWS) is in
Rocky Mountain
National
Park (RMNP) on
the eastern edge
of the Colorado
Front Range.
3-4 kg 25 kg N Addition: There were no changes in the C degrading
N/ha/yr NhUNOs/ha/yr enzyme activity in response to fertilization, while the N
(wet) (1996-present) degrading enzyme activity was enhanced with elevated
activity of leucine aminopeptidase (LAP) and marginally
significant increase in N-acetyl-p-glucosaminidase (NAG)
with N fertilization.
Boot et al.
(2016)
4-52
-------
Table 4-8 (Continued): Decomposition.
Process/
Type of
Deposition
Addition
Indicator
Ecosystem
Region
kg/ha/yr
kg/ha/yr
Effect of Deposition
Reference
Enzyme activity
Northern
Michigan
5.9-7.4 kg
30 kg
N Addition: This long-term experiment found N deposition
Freedman et
hardwood forest
N/ha/yr (not
NaNC>3/ha/yr
decreased the activity reduced the activity of extracellular
al. (2016)
specified if
(beginning in
enzymes mediating plant cell wall decay.
wet, dry, or
1994)
total)
Soil respiration
Northern
White Mountain
8 kg N/ha/yr
30 kg N/
N Addition: The greatest reduction in soil respiration on N
Kana et al.
hardwood forest
National Forest,
(wet + dry)
ha/yr
and N + P fertilized plots occurred on the sites with lowest
(2016)
NH; Bartlett,
(20th
(NH4NO3),
pretreatment soil N mineralization and litterfall N flux.
Hubbard Brook,
century
10 kg P/ha/yr
Nutrient additions did not significantly affect either fine root
and Jeffers Brook
average at
(NahhPCM), or
turnover (minirhizotrons) or microbial respiration (laboratory
Forests
HBEF)
N + P
incubations).
(treatments in
year 2011)
Decomposition
Restored
Prairie Invasion
-1/3 of N
5 g N/m2/yr
N Addition: Increasing rainfall variability and N addition
Schuster
tallgrass prairie
and Climate
added
urea. "Seasonal
can stimulate litter decomposition in tall grass prairie.
(2016)
Experiment
maximum
(PRICLE)
5-day
Loveland, CO
cumulative
rainfall"
increased by 33
(2012) and 9%
(2013)
Decomposition
All ecosystems
Global
Not
Not specified
Meta-analysis: 198 peer-reviewed journal articles found N
Yue et al.
specified
addition did not significantly alter litter decomposition, soil
(2016)
respiration (except for wetlands, +28.26%), and microbial
respiration (except for forests, +9.08).
4-53
-------
Table 4-8 (Continued): Decomposition.
Process/ Type of Deposition Addition
Indicator Ecosystem Region kg/ha/yr kg/ha/yr
Decomposition Mixed hardwood Chronic Nitrogen 50-150
forest dominated Amendment
by black and red
oak (Quercus
velutina and Q.
rubra)
Effect of Deposition Reference
Changes in litter decay by was generally lower in the N van Diepen et
treatment microbes compared to control microbes for the al. (2017)
same species, a response not readily reversed when N
microbial isolates were grown in low N environments.
Changes in fungal behaviors accompany and perhaps drive
previously observed N induced shifts in fungal diversity,
community composition, and litter decay dynamics.
Study (CNAS)
located at the
Harvard Forest
Long-Term 57
Ecological
Research (LTER)
site in
Petersham, MA,
USA
Decomposition Forest Scotland 14-16 kg 1.18 g 15N in Litter decomposition is a larger source of N for trees than Nair et al.
N/ha/yr 4m plot simulated N deposition. (2017)
BBWM = Bear Brook Watershed; C = carbon; DOC = dissolved organic carbon; ha = hectare; kg = kilograms; LMCO = Laccase-like multicopper oxidase; mg = milligrams; N = nitrogen;
15N = nitrogen-15; NaN03 = sodium nitrate; NH4N03 = ammonium nitrate; (NH4)2S04 = ammonium sulfate; N03" = nitrate; 15N03" = nitrogen-15-labeled nitrate; OM = organic matter;
S = sulfur; SOM = soil organic matter; yr = year.
4-54
-------
The body of knowledge published since the 2008 ISA indicates the effects of N on
decomposition rates are inconsistent among studies. New field work supported the
concept that N deposition suppresses decomposition (Riggs et al.. 2015; Zak et al.. 2008).
Likewise, Kang et al. (2016). in a large-scale N addition study of 13 northern hardwood
forests in the U.S., supported the findings that N addition decreases forest respiration and
further identified the greatest reduction in soil respiration on N fertilized plots occurred
on the sites with lowest pretreatment soil N mineralization and litterfall N flux (Kang et
al.. 2016). In contrast, meta-analysis that evaluated the central tendencies of N addition
on total soil respiration in forest biomes reported inconsistent results, including an
increase (Yue et al.. 2016). decrease (Zhou et al.. 2014a). and no effect (Liu and Greaver.
2010). Meta-analyses that looked for the central tendencies of terrestrial soil respiration
to N addition also have inconsistent results; two meta-analyses found that N did not alter
total soil respiration in terrestrial soils (Yue et al.. 2016; Liu and Greaver. 2010). another
identified an increase (Zhou et al.. 2014a). and only one identified a decrease (Janssens et
al.. 2010). The different results in the meta-analyses reflect slightly different selection
criteria the authors used to determine which studies to include in the analyses. Among
biomes, differences in soil respiration may result largely from stimulation of autotrophic
respiration by N addition to croplands and grasslands compared with no significant
change for forests, and a simultaneous decline in heterotrophic/microbial respiration in
most biomes rLiu and Greaver (2010); Zhou et al. (2014a). with the exception of
croplands, tropical forests, and boreal forests].
One proposed mechanism for reduced microbial decay under increased N deposition is a
shift in the species composition of the microbial community with the consequence of
decreased lignin decomposition. Lignin is an organic polymer, particularly important in
cell wall formation of vascular plants. Under N addition, the microbial community would
shift from basidiomycete fungal activity, some of which oxidize lignin in plant detritus
and polyphenols in SOM to CO2, to more metabolism by bacteria and ascomycete fungi
(cellulose degraders), which only partially oxidize these organic substrates (Freedman
and Zak. 2014; Zak et al.. 2011). There is new evidence to support this theory in
northeastern hardwood forests in the U.S. where N enrichment increases the species
richness and diversity of ascomycetes generally [cellulose degraders; Morrison et al.
(2016)1. with evidence that observed changes in decay abilities were not readily reversed
when N isolates were grown in control environments, indicating that the fungal
community may not recover quickly following the cessation of N enrichment Ivan
Diepen et al. (2017); see Appendix 61.
There is new evidence to support that lignolytic enzyme activity decreases under N
addition (Freedman et al.. 2016; Keeler et al.. 2009; Manning et al.. 2008). This trend is
also supported by a meta-analysis by Whittinghill et al. (2012). who found that N
4-55
-------
addition decreased lignin decomposition rates by 30% and increased cellulose
decomposition by 9%. Fine roots are likely the dominant source of
difficult-to-decompose plant carbon fractions entering the soil, and this pattern appears to
be widespread in boreal and temperate forests (Xia et al.. 2015).
There are cases in which N addition causes no change in enzyme activity of microbes; for
example, the hardwood and softwood forests growing in the Bear Brook Watershed in
Maine (Mincau et al.. 2014) and C degrading enzymes in a subalpine forest (Boot et al..
2016). Another study in the Catskill Mountains of New York State found that patterns in
microbial community structure and function (detected by enzyme activity) were more
strongly influenced by the tree species present than by fertilization (Weand et al.. 2010).
Microbial decomposition response to N addition may also change through time. The 2008
ISA documented an N addition study lasting longer than 2 years that indicated that a shift
can occur from stimulation to depression of decomposition over time (Knorr et al.. 2005).
In a study looking at changes through time of the organic layer, Hobbie et al. (2012)
suggested N deposition in forest ecosystems may decrease mean residence times of active
fractions in fresh litter, while increasing those of more slowly decomposing fractions,
including more processed litter. In boreal forests, Talbot and Treseder (2012) found a
transition over time from competition among decomposers to high cellulase activity and
suppressed lignin loss under N fertilization. The trend suggests that, in N limited systems,
N fertilization may alter decomposer community structure by favoring a shift toward
cellulose- and mineral-N users.
There is also new evidence that experimental N deposition significantly decreased the
richness and diversity of microbial genes involved in the depolymerization of starch
(12%), hemicellulose (16%), cellulose (16%), chitin (15%), and lignin (16%) (Eisenlord
et al.. 2013).
4.3.8 Nitrogen Mineralization
Mineralization refers to processes that release carbon as CO2 and nutrients in inorganic
form. Nitrogen mineralization is the process by which organic N is converted to
plant-available inorganic forms. The 2008 ISA documented that the rate of mineralization
may be influenced by numerous factors including C:N of soil organic matter, soil pH, and
the microbial community. N mineralization has been shown to increase with increasing N
addition (Aberetal.. 1998). often up to 1.6 times the control (Gundersen et al.. 1998).
New publications (Table 4-9) support that soil N mineralization increases with N addition
across terrestrial ecosystems (Lu et al.. 2011a). specifically in temperate forests (Nave et
4-56
-------
al.. 2009a). likely due to increases in dissolved organic nitrogen (DON), the size of the N
pool in the soil, and decreases in C:N ratios. The forest floor responds differently than the
deeper mineral soil layer I Figure 4-7; Nave et al. (2009a); Lovett et al. (2013)1.
In two forested headwater catchments in Ontario, Casson et al. (2014a) found that N
mineralization and nitrification rates were similar, indicating that almost all mineralized
N was converted to NO3 (ranging from 71 to 99%) in catchments they studied. While
Bade et al. (2015) found in old-growth spruce forests, lower N mineralization occurred in
the more open patches than the closed ones. Possible reasons were reduced litter supply
and lower canopy N interception in gaps in this forest under exposure to high N
deposition. Further studies in other old-growth forests are needed to better understand the
mechanisms causing long-term change in N cycling with forest development.
In desert shrublands, Rao et al. (2009) found that N deposition may increase production
and/or alter litter C:N ratios that increase soil C. There was an inverse relationship
between the C:N ratio and total N mineralized, yet a positive relationship between
organic C and total N with mineralization. These results indicate that microbial activity in
low productivity arid land soils is primarily limited by C and secondarily limited by N.
4-57
-------
Table 4-9 Nitrogen mineralization.
Ambient N/S
Process
Type of
Deposition
N/S Addition
Endpoint
Ecosystem
Region
kg/ha/yr
kg/ha/yr
Effect of Deposition
Reference
Mineralization
Semiarid
Southern
Up to 35-45
Two
N is mineralized at a faster rate from grass litter
Sirulnik et al.
California,
forN
applications
enriched in N as a result of N deposition.
(2007a)
U.S.
per year of
30 kg N
Mineralization
Semiarid
Southern
Up to 35-45
Two
When significantly lower microbial N was reported in
Sirulnik et al.
California,
forN
applications
in fertilized plots it corresponded to faster net N
(2007b)
U.S.
per year of
mineralization. When higher microbial N was
30 kg N
observed in fertilized plots it corresponded to net
immobilization (though not significantly faster), thus
corroborating evidence that the microbial community
was taking up more N in fertilized plots. These rates
may have corresponded to C availability, which was
not measured.
Ca Gradient: The exchangeable Ca coupled with Page and Mitchell
soil moisture, soil organic matter, and ambient (2008)
temperature accounted for 61% of the variability in
extractable inorganic N across 11 sites. The
influence of Ca on soil inorganic N may be through
interactions between soil Ca concentrations and
species composition, which in turn affect the quality
of litter available for N mineralization.
Decomposition
Enzyme
activities
Spruce forest
Germany 8.5
N Exclusion: Some N cycling enzymes increased
activities, whereas others decreased under reduced
N treatment.
Enowashu et al.
(2009)
N mineralization
Desert
Joshua Tree 2.7 to 14.4
National Park
Calculated soil N from deposition was directly
correlated with measured soil C and N and
decreasing C:N ratios.
Rao et al. (2009)
Mineralization Mixed hardwood
soil [N] stands
soil [Ca]
Adirondack Not specified None
Mountains,
NY, U.S.
4-58
-------
Table 4-9 (Continued): Nitrogen mineralization.
Process
Endpoint
Type of
Ecosystem
Region
Ambient N/S
Deposition
kg/ha/yr
N/S Addition
kg/ha/yr
Effect of Deposition
Reference
N mineralization
Temperate
forests.
Northeastern
U.S.
Varies
Varies
Meta-analysis: Overall, N inputs increased soil C
(+7.7%) and N mineralization (+62%), while
decreasing C:N (-4.9%).
Nave et al. (2009a)
N mineralization
Forest
50 (6 yr of N
addition)
N Addition: There was a significant decline in
potential N mineralization and nitrification rates in
the mineral horizon but not in the forest floor.
Lovett et al. (2013)
N mineralization
Terrestrial
ecosystems
Varies
Varies
Meta-analysis: included 206 papers on responses
of ecosystem N cycle caused by N addition. N
addition increased mineralization rates.
Lu et al. (2011a)
N mineralization
Douglas fir forest
Oregon coast
range
2.0
None
Aboveground N uptake by plants increased with net
N mineralization, peaking at 35 kg N/ha/yr.
Perakis and
Sinkhorn (2011)
Net N
Mixed hardwood
Muskoka-
NO3-
Not applicable In all seasons, rates of nitrification were similar to
Casson et al.
mineralization
forested
Haliburton
deposition
rates of total mineralization, indicating that almost all
(2014a)
and nitrification
headwater
district of
2.78 ± 1.22;
mineralized N is converted to NO3" in both soil types
rates/NH4+ N,
catchments
south-central
NH4+
(ranging from 71 to 99%).
soil NO3",
Ontario,
deposition
stream NO3"
Canada
3.90 ± 1.39
N
Natural
Harz National
27 (open
Not applicable Net N mineralization (and ammonification) rates Bade et al. (2015)
mineralization,
(unmanaged)
Park in
areas)
were higher in the closed stands of the optimum and
ammonification,
old-growth
central
47 (closed
overmature stages than in the more open decay and
nitrification/N03"
Norway spruce
forest
Germany
forest)
regeneration stages. Only a small proportion of
NH4+ was oxidized to NO3" in the acidic soils.
Net litter
mineralization
Forest
Scotland
14-16 kg
N/ha/yr
1.18 g 15N
per 4m plot
Three times as much 15N was retained in the O and
A soil layers when N was derived from litter
decomposition than from mineral N additions.
Nairetal. (2017)
C = carbon; Ca = calcium; ha = hectare; kg = kilogram; N = nitrogen; NH4+ = ammonium; N03 = nitrate; yr = year.
4-59
-------
so"c C/N NmIn
Forest floor
Overall
i- N-fix
T3
2 N-fert
o
- (47)
(7)
- (25)
o— (28)
-{13)
-I 1 T 1—
- (87) D
" (115)
- (35)
—° (38)
"• (22)
-*> (16)
"j* (30)
-d-<61)
I I I
(86)
(8)
B
(45)
(33)
(121)
E
(50)
(25)
(46)
y/-
(81)
(2)
(37)
(42)
*-//¦ r"
vy—
(80)
(51)
—A
(24)
(5)
-T-V—
4-60
-------
4.3.9
Dissolved Organic Carbon Leaching
The acid-base characteristics of dissolved organic matter (DOM) are an important part of
understanding the recovery potential for soils, lakes, and streams impaired by acid
deposition. The composition of DOM includes fulvic and humic acids, carboxylic acids,
and amino acids; dissolved organic carbon (DOC) includes these organic acids. Here we
focus on the DOC component of DOM. The many carboxylic groups of DOC make it
chemically interact like a weak acid; therefore, DOC content may affect pH levels. In
addition, Fakhraei and Driscoll (2015) emphasized the importance of predicting
accurately the acid-base properties of recovering surface waters because, for example, the
acidic components of DOC act as hosts for binding trace metals like toxic Al (for
additional discussion on Al and DOM see Appendix 4.3.5).
In recent years, the DOC of many lakes and streams has risen, with the likely source
being the soils in the adjacent terrestrial watershed. However, the mechanism causing the
observed increase is unclear. The increase may be due to a combination of soil recovery
from acidification, changes in climate (e.g., temperature and precipitation), and N
deposition, among other mechanisms [for reviews see Kalbitz et al. (2000); Evans et al.
(2005)1. New studies in the literature have investigated soil DOC response to
acidification (N + S deposition) and/or looked at the effects of N addition; these studies
will be the focus of the following discussion (Table 4-10). For a discussion of DOC in
surface water see Appendix 7.
4-61
-------
Table 4-10 Terrestrial dissolved organic carbon (DOC) leaching.
Ambient N/S
Type of
Deposition
N/S Addition
Process Ecosystem
Region
kg/ha/yr
kg/ha/yr
Effect of Deposition
Reference
DOC Bog, heathland,
Synthesis of
Ranged from
Ranged from 10
Synthesis: N had inconsistent effect on
Evans et al. (2008)
grassland,
17 addition
5-16 N dep
to 150 N
DOC; however, the form of N applied
broadleaf and
experiments in
(chemical forms
indicates nonacidifying forms of N tend to
coniferous forest
the northeastern
for N varied)
increase DOC concentrations.
U.S. and
northern Europe
DOC Temperate 410 observations Not specified
mixed and globally
conifer forests,
boreal forests,
grasslands,
tropical forests,
arctic, wetlands,
desert, and
tundra
DOC Norway spruce Southern Not specified Not specified Model: The Stockholm Humic Model was Lofqren et al. (2010)
(Picea abies) Sweden used to model DOC solubility in soil water
forests and predicted that DOC trends could vary
between positive and negative depending
on changes in pH, ionic strength, and soil Al
pools.
10 to 650 Meta-analysis: N addition increases short- Liu and Greaver (2010)
term belowground C storage by increasing
C content of the organic layer. N addition;
increased DOC concentration (+18%); and
increased C content of the organic soil layer
(+17%) but not the mineral soil layer.
4-62
-------
Table 4-10 (Continued): Terrestrial dissolved organic carbon (DOC) leaching.
Process
Type of
Ecosystem
Region
Ambient N/S
Deposition
kg/ha/yr
N/S Addition
kg/ha/yr
Effect of Deposition
Reference
DOC
Beech forest
stands on
calcareous soils
Swiss Jura
mountain range
Not specified 5.5 (NH4NO3)
N Addition: Leaching of DOC from the litter
layer was not affected by N additions, but
DOC fluxes from the mineral soils at 5- and
10-cm depth were significantly reduced by
17%. 13C indicated that litter-derived C
contributed less than 15% of the DOC flux
from the mineral soil, with N additions not
affecting this fraction. Hence, the
suppressed DOC fluxes from the mineral
soil at higher N inputs can be attributed to
reduced mobilization of nonlitter derived
"older" DOC.
Haqedorn et al. (2012)
DOC
None—
theoretical
None—
theoretical
Not specified Not specified
N Addition: Increases mineralization
causing the pool of bioavailable DOC to
decrease. Consequently, relatively less
bioavailable DOC remains for NHV
assimilation and immobilization in microbial
biomass, leaving more NhV for nitrifiers. As
a result, internal NO3" production increased.
The higher bacterial demand for DOC under
elevated availability of N and electron
acceptors comes into conflict with
increasing chemical suppression of DOC
solubility and bioavailability in the
progressively acidifying soils and finally
results in the C limitation of microbial
metabolism.
Kopacek et al. (2013)
DOC Northern Upper Michigan 15 to 20 kg
hardwood forest N/ha/yr
30 (as six equal
applications of
NaN03 pellets
delivered to the
forest floor over
the growing
season)
N Addition: N addition accelerated
phenolic DOC production.
Freedman and Zak (2014)
4-63
-------
Table 4-10 (Continued): Terrestrial dissolved organic carbon (DOC) leaching.
Process
Type of
Ecosystem
Region
Ambient N/S
Deposition
kg/ha/yr
N/S Addition
kg/ha/yr
Effect of Deposition
Reference
DOC
Forest and
streams
Hubbard Brook
Experimental
Forest, NH
Variable
None
Field observation: Concentrations of DOC
showed statistically significant declines in
the Oa soil solutions of all three elevation
zones, and a subset of the Bs soil solutions
over the period of 1984-2011.
Fuss etal. (2015)
DCO Lakes Adirondack Variable None Field observation: increasing DOC in 29 of Driscoll et al. (2016)
Long-Term 48 lakes as lake pH increased.
Monitoring
(ATLM) Program
lakes at
Huntington
Forest and
Whiteface
Mountain,
Adirondack
Mountains, NY
DOC Forest and U.K. Mean 1993 to None
grassland 2010
gradients of
deposition:
S: 44 to
86 meq/m2/yr;
N: 40 to
90 meq/m2/yr;
CI: 94 to
306 meq/m2/yr
Modeling: Using the MADOC model, the Sawicka et al. (2017)
acidifying effect of S deposition was the
predominant control on the observed soil
water DOC trends. The relative importance
of S and N loading depended on soil
sensitivity to acidification, and on N
limitation.
13C = carbon-13; C = carbon; DOC = dissolved organic carbon; ha = hectare; kg = kilogram; N = nitrogen; NaN03 = sodium nitrate; NH4+ = ammonium; NH4N03 = ammonium nitrate;
N03" = nitrate; S = sulfur; yr = year.
4-64
-------
A number of studies suggest that recovery from acidification is driving the elevated
levels of DOC. An inverse relationship between mineral and organic acid export to
surface waters from soils was first proposed in the 1980s (Krug and Frink. 1983). Support
for this theory is found in a range of laboratory experiments in which increases in both
acidity and ionic strength (associated with a high S042 loading) have been shown to
reduce soil solution DOC (Kalbitz et al.. 2000). Evans et al. (2008) addressed whether
elevated N deposition or recovery from acidification associated with decreasing S cause
increased DOC loss from upland soil. Their analysis of a large experimental data set
including 12 sites in the U.S. and northern Europe showed that although the response to
N addition is inconsistent, DOC concentrations responded predictably to the chemical
form of N added. DOC concentrations increased with NaNOs additions or gaseous NH3
exposure and decreased with most NH/ salt additions. The authors cite the effect of the
chemical form of N on acidity as a plausible mechanism and further conclude that their
evaluation does not provide clear support for the role of N deposition as the sole driver of
rising DOC, but is consistent with an acidity-change mechanism. Evans et al. (2008) also
stated their finding is consistent with findings based on long-term monitoring data, that
DOC increases in northern European and North American surface waters are
substantially attributable to regional decreases in acidifying, primarily S, deposition
(Monteith et al.. 2007; Evans et al.. 2005). New studies also report that N addition has
caused inconsistent changes to DOC, with N addition causing both increases (Driscoll et
al.. 2016; Liu and Greaver. 2010) and decreases (Fuss et al.. 2015; Hagedorn et al..
2012).
In the Adirondack Long-term Monitoring Program, Driscoll et al. (2016) measured
increasing DOC in 29 of 48 monitored lakes as lake ANC and pH increased. In contrast,
decreases in DOC concentrations were observed in the soil solution from the Oa soil
horizon at all sites, the Bs soil horizon soil solution of one site, and the stream water
measured at the base of the watershed. These observations were attributed to recovery
from acidification in the HBEF, NH (Fuss et al.. 2015). Fuss et al. (2015) considered
these results surprising and therefore reviewed European and U.S. research literature,
found studies with similar observed trends, and identified a number of potential factors
leading to this difference. While surface water DOC generally emanates from soil water
DOC (with the possible exception of snowmelt), soil ionic strength and soil Al pools, as
well as soil depth, may also influence DOC in soil solution (Lofgren et al.. 2010).
Decreasing ionic strength can decrease DOC; higher levels of organic aluminum
complexes in soil can increase DOC solubility; and DOC in forest floor soil water may
decrease at a faster rate than in mineral soil water. The authors did note that HBEF's rate
of decreasing DOC concentrations appeared to be diminishing in the later 15 years of
reporting.
4-65
-------
Although most studies are on forests, Sawicka et al. (2017) used the MADOC model for
three forested and three grassland and heath sites and observed that the acidifying effect
of S deposition was the predominant control on the observed soil water DOC trends
(Sawicka et al.. 2017). The relative importance of S and N loading depended on soil
sensitivity to acidification and on N limitation. In all N limited soils investigated, the
modeled DOC increases over the monitoring period were dominated by the effects of
recovery from acidification (higher DOC solubility), but the effects of N enrichment
driving higher DOC production may have been important in the longer term. In contrast,
reductions in nonmarine chloride deposition and the effects of long-term warming
appeared to have been relatively unimportant.
Other work has further explored the mechanisms for how N addition (including nutrient
as well as acidification effects) affects DOC (Freedman and Zak. 2014; Kopacek et al..
2013). Freedman and Zak (2014) reported that atmospheric N deposition led to less
microbial biodiversity and favored bacterial species in the forest floor that lead to
reduced decomposition, increased soil C storage, and accelerated phenolic DOC
production (Figure 4-8). Kopacek et al. (2013) observed that N addition, together with
S042 deposition, increases the availability of electron acceptors for soil microbial
processes. The chemical and microbial responses include an increase in bacterial
mineralization creating DOC, increased N mineralization where N availability increases
(with an increase in bacterial uptake of DOC), chemical suppression of DOC solubility
and bioavailability in the progressively acidifying soils, and finally, the C limitation of
microbial metabolism. Bacterial assimilation of NO3 , which depends on the
stoichiometric NO;, to DOC ratio in the substrate, may initially increase with increasing
NO;, concentrations, but then decrease due to a lower pool of bioavailable DOC.
4-66
-------
Bacterial Metabolism of Lignln arxJ Humlcs
LMCO-fiartBflng Bacteria a. Acfflnbacfaite
N = nitrogen; LMCO = Laccase-like multicopper oxidase.
Source: Freedman and Zak (20141.
Figure 4-8 Conceptual diagram of positive (solid arrows) and negative
(dashed arrows) fluxes in nitrogen pools (squares) and carbon
pools (ovals) and the biological processes (no border) that are
affected by experimental nitrogen deposition.
4.3.10 Belowground Carbon Pools
In general, N enrichment influences C flux, C partitioning, and the amount of C
sequestered by ecosystems. A number of studies have suggested that C sequestration
increases with increasing N supply based on the changes in aboveground net primary
production [ANPP; LeBauer and Treseder (2008); Xia and Wan (2008)1. However, about
half of the C fixed annually by terrestrial vegetation is allocated to belowground pools.
Many studies have shown that the belowground C cycle does not always mirror the
aboveground cycle. For example, elevated CO2 and N both have been shown to increase
aboveground biomass production (LeBauer and Treseder. 2008). However, increases in
aboveground plant production and greater aboveground litter inputs do not necessarily
increase mineral soil C storage (Talhelm et al.. 2009; Gielen et al.. 2005; Lichter et al..
2005). and increases in soil C do not necessarily result from greater aboveground litter
inputs (Pregitzer et al.. 2008). This disparity indicates that it is inappropriate to
extrapolate from aboveground responses to belowground processes. Aboveground plant
biomass, once dropped from the canopy, is one of the major contributors to soil organic
4-67
-------
matter accumulation (Sullivan et al.. 2007a). ANPP generally increases with increasing N
supply (Le Bauer and Treseder. 2008; Xia and Wan. 2008). However, this increase in C
input may not cause an increase in belowground ecosystem C storage. This is because the
output of C from the soil may be stimulated by N induced alteration of plant tissue
chemistry or the ratio of root: shoot, each of which will change the rate of decomposition
(see Appendix 4.3.7). Since the 2008 ISA was published there have been several new
meta-analyses on the effects of N addition on belowground carbon pools (Janssens et al..
2010; Liu and Greaver. 2010). each integrating information from numerous pools to
improve the understanding of how N addition alters the carbon cycle belowground.
Additional studies are summarized in Table 4-11.
Liu and Greaver (2010) synthesized data from multiple terrestrial ecosystems to quantify
the response of belowground C cycling under N addition (Figure 4-9). N addition
increased aboveground litter inputs (+20%), but fine root litter inputs were unchanged. N
addition inhibited microbial activity, as indicated by a reduction in microbial respiration
(-8%) and microbial biomass carbon (-20%). Although soil respiration was not altered
by N addition, dissolved organic carbon concentration increased (+18%), suggesting C
leaching loss may increase. N addition increased the C content of the soil organic horizon
(+17%) but not the mineral soil. The increase in organic horizon C was attributed to both
increased litter input and decreased decomposition [inferred from the lower microbial
respiration rates (Liu and Greaver. 2010)1.
Field studies provide additional support for the finding that soil carbon responds
differently to N addition depending on the organic (vs. mineral) content of the soil, with
organic content tending to decrease with increasing soil depth. In four northern hardwood
forests spread across Michigan that received experimental N deposition (additional
30 kg N/ha/yr) for 10 years, Pregitzer et al. (2008) and Zak et al. (2008) reported
significant increases in soil C, particularly within the soil organic horizon. These
increases occurred despite no increase in aboveground litter production or root turnover.
A ' "NO;, -labeling experiment revealed that N accumulated in SOM by first flowing
through soil microorganisms to plants, which then shed the leaves back to the detrital
layer. In a meta-analysis including data from 72 north temperate forests, Nave et al.
(2009a) found N addition increased soil C (+7.7%). The forest floor responded differently
than the deeper mineral soil layer (5- to 100-cm depth) as increased soil C storage
occurred only in the mineral soil (+12.2%), in contrast to the findings of Liu and Greaver
(2010) and Pregitzer et al. (2008).
4-68
-------
ANPP = aboveground net primary production; DOC = dissolved organic carbon; ER = ecosystem respiration; GPP = gross primary
productivity; MBC = microbial biomass carbon; n.s. = nonsignificant; Raboveground = aboveground respiration; RautotroPhic = soil
autotrophic respiration; Rmicr0biai = microbial respiration; RsoH = total soil respiration.
t data from LeBauer and Treseder (20081; * data from Xia and Wan (20081; * data from Treseder (20041.
Source: Liu and Greaver (20101.
Figure 4-9 Estimation of the changes in carbon budget of terrestrial
ecosystem under nitrogen addition.
A new study has evaluated the consistency of N cycling in different ecosystems by
synthesizing data across five biomes in California. The results indicated that, across
biomes, N concentration in soil has a strong positive correlation to SOC (Yang et al..
2017). There are also new studies indicating N deposition increases SOM accumulation
without altering the biochemical composition (Zak et al.. 2017). New studies on how the
response of belowground soil C to N is modified by temperature and precipitation has
been published since the 2008 ISA (Ni et al.. 2017; Greaver et al.. 2016) and is
summarized in Appendix 4.7.
4-69
-------
Table 4-11 Belowground carbon pools.
Ambient N/S
Type of Deposition N/S Addition
Process Ecosystem Region kg/ha/yr kg/ha/yr Effect of Deposition Reference
SOC Soil Great Britain 0-2 g N/m2/yr None Model: Using model N14CP for data Tipping et al.
from -2,000 N limited field sites, it (2017)
was predicted that, N deposition
increased NPP between the
years 1750 and 2010, increasing
via detritus SOC by 1.2 kg C/m2
(-10%). (The authors assumed
-30% error of estimated values.)
SOC
Five biome classes
California
0.6-18.4 kg
Grassland soil:
Synthesis: Across biomes, total N
Yana et al. (2017)
in California: desert,
N/ha/yr
76.4 kg N/ha/yr as
concentration was strongly
grassland,
NPK (29:3:4);
correlated to SOC (R2 = 0.88;
shrubland, forest,
desert soil: 60 kg
log(y) = [0.81 * log(x)] - 1.10).
wetland
N/ha/yr as NH4NO3;
shrubland soil:
60 kg N/ha/yr as
NH4NO3
SOC
Hardwood forest
Jilin province,
23 kg N/ha/yr
0 kg N/ha/yr, 25 kg
N Addition: This was a 6-yr
Chen et al. (2017)
birch (Betula
China
N/ha/yr, and 50 kg
fertilization study. The authors
platyphylla) and
N/ha/yr
observed that fertilization
aspen (Populus
decreased the fraction of
davidiana)
macroaggregrates (2-8 mm) and
increased the fraction of
0.053-2 mm aggregates.
(P = 0.01). They concluded that N
deposition (as simulated with
fertilization) can increase formation
of micro- and macroaggregates
within macroaggregate soil and
thus stabilize C.
4-70
-------
Table 4-11 (Continued): Belowground carbon pools.
Process
Type of
Ecosystem
Region
Ambient N/S
Deposition
kg/ha/yr
N/S Addition
kg/ha/yr
Effect of Deposition
Reference
Soil organic matter
Sugar maple (Acer
saccharum)-dom\-
nated northern
hardwood forests
North-south
geographic range
of sugar
maple-dominated
upper Great Lakes
and eastern North
America
6.8-11.8 kg
N/ha/yr
(wet + dry)
across sites
30 kg N03"-N/ha/yr
(NaNOs)
N Addition: SOM accumulated as
soil particulate organic matter
occlusion at N addition sites.
Zaketal. (2017)
Abundance of
carboxyl, aryl,
O/N-alkyl, and alkyl
C in mineral soil
(biochemical
composition of
forest floor
[Oe/Oi-horizon
~4 cm] and organic
matter)
Sugar maple (Acer North-south
saccharum)-dom\-
nated northern
hardwood forests
geographic range
of sugar
maple-dominated
upper Great Lakes
and eastern North
America
6.8-11.8 kg
N/ha/yr
(wet + dry)
across sites
30 kg N03"-N/ha/yr
(NaNOs)
N Addition: Abundance of carboxyl Zak et al. (2017)
and aryl C in forest floor differed
across sites but the abundance of
O/N-alkyl and alkyl C did not
(P = 0.10-0.80). No difference was
observed in the proportion of aryl,
O/N-alkyl, or alkyl C in mineral soil
between ambient deposition and N
addition sites. (Particulate organic
matter's increased occlusion of
organics may shift fungal and
bacterial activity toward partial
oxidation of organics without
changing biochemistry of organic
matter.)
4-71
-------
4.3.11
New Biogeochemical Indicators
In this section, we report on new biogeochemical indicators reported in the literature
since the 2008 ISA. Recent literature offering potential new indicators of biogeochemical
change due to N and/or S deposition included fungal-to-bacterial ratio (Hogbcrg et al..
2013) and syntaxonomic associations I Wamclink et al. (2011); Table 4-121.
Hogberg et al. (2013) proposed soil microbial community indices as predictors of soil
solution chemistry and N leaching in Picea abies spruce forests in southern Sweden.
Stands with low concentrations of NOs" and Al3+ had higher fungi :bacteria ratios
compared with stands with higher concentrations of NOa" and Al. They identified three
promising microbial community indices as indicators of N leaching from forests; the soil
fungi:bacteria ratio was the most important.
In a study in the Netherlands, Wamelink et al. (2011) examined whether the abiotic
ranges of syntaxonomic units (associations) in terms of pH and NO;, concentration can
be estimated and then, in principle, used to estimate critical loads for acid and N
deposition. They used splines to estimate abiotic ranges of syntaxonomic units based on
measured soil pH and NO;, concentration and vegetation releves. They acknowledge it is
not yet possible to directly estimate ranges for syntaxa for pH and NO; on a large scale
using this approach; however, indirectly estimated soil pH and NO; concentrations are
sufficiently available to derive ranges for many associations.
4-72
-------
Table 4-12 New biogeochemistry indicators.
Endpoint
Nutrient Description/Direct Effect of
Enrichment Acidification Deposition Addition Soil/Water Endpoint on Biological
Indicator Indicator kg/ha/yr kg/ha/yr Effect
Reference
N leaching
X
Throughfall N
includes wet and
dry inputs and
ranged from 2.7 to
19
20 (NH4NO3) Microbial community composition in
the organic layer of spruce forests
and soil solution chemistry below the
rooting zone was highly correlated.
Stands with low concentrations of
NO3" and Al were fungi dominated
and had a higher fungi:bacteria ratio
compared with stands with high
concentrations of these leachates.
Leaching stands had higher
abundance of Gram-positive
bacteria.
Hoqberq et al.
(2013)
Vegetation
health
X
X
Not specified
None
(2011)
Acknowledge the approach is not yet Wamelink et al.
possible to directly estimate ranges
for syntaxa for pH and NO3" on a
large scale; however, indirectly
estimated soil pH and NO3"
concentrations are sufficiently
available to derive ranges for many
associations.
Al = aluminum; ha = hectare; kg = kilogram; N = nitrogen; NH4N03 = ammonium nitrate; N03 = nitrate; yr = year.
4-73
-------
4.3.12
Differential Effects of Reduced and Oxidized Nitrogen
Some biogeochemical process responses to N deposition have the potential to vary
depending on whether the dominant form of deposited N was oxidized or reduced. We
have focused on presenting meta-analyses as they synthesize the central tendencies of
many studies. In some cases, the results for a given endpoint are not consistent among
meta-analyses. Table 4-13 summarizes the studies that look at the most common forms
ofN.
Additional studies include the observation by Evans et al. (2008) that DOC
concentrations responded predictably to the chemical form of N used for manipulation,
increasing with NaNCh additions or gaseous NH3 exposure, and decreasing with most
NH4+ salt additions (Appendix 4.3.9). The authors cite the effect of the chemical form of
N on acidity as a plausible mechanism. Ramirez et al. (2010a) investigated whether soil
microbes in three distinct soils (from aspen, pine, and grassland ecosystems) respond
differently to six different reduced and oxidized N species, including NH4NO3, (NEb^CO
(urea), KNO3, NH4CI, (NH4)2S04, and Ca(NC>3)2 (also discussed in Appendix 6). The
authors found that all inorganic forms of N fertilizer significantly decreased microbial
CO2 production but that organic N (urea) decreased the respiration rate in forest soil only
by 27% (aspen) and 11% (pine) and increased the grassland microbial respiration rate by
20%. They added that the soil pH change resulting from N addition did not appear to
influence the observed decrease in the soils" microbial respiration rate.
Jovan et al. (2012) monitored tree trunk pH during their research on eutrophic lichen
abundance in the Los Angeles air basin, oxidized forms of N (particularly HNO3)
dominate dry deposition. They noted that NH3 deposition would normally raise bark pH
but in the arid, hot Mediterranean climate of the study area, high trunk pH dust raises
bark pH.
4-74
-------
Table 4-13 The effects of different forms of inorganic nitrogen on biogeochemical processes and indicators
according to meta-analyses. See Table 6-1 for the effects of different forms of inorganic nitrogen on
biological endpoints.
Process/Indicator
NO3-
NH4NO3
NH4+
Urea
UAN
Reference
Total soil C
NS
NS
NS
NS
NS
Liu and Greaver (2010)
Total soil C
NS
T
NS
NS
-
Yueetal. (2016)
Soil DOC
T
T
NS
NS
-
Yueetal. (2016)
Soil DOC
T
NS
NS
T
-
Liu and Greaver (2010)
Organic layer C
NS
T
-
T
-
Liu and Greaver (2010)
Soil respiration
NS
NS
NS
NS
NS
Yueetal. (2016)
Soil respiration3
NS
NS
T
NS
-
Liu and Greaver (2010)
CH4 emission
NS
NS
NS
NS
NS
Liu and Greaver (2009)
CH4 uptake
NS
Liu and Greaver (2009)
N2O
T
T
T
T
T
Liu and Greaver (2009)
N Recovery
80%
85%
53%
-
-
Templer et al. (2012)
aNot significant effect of the N mean for this endpoint; however, there was a significant difference among N forms,
t indicates significant increase.
I indicates significant decrease.
- indicates not reported.
4-75
-------
4.4
Soil Monitoring and Databases
There are several new studies using long-term monitoring data sets in the U.S. and
Europe (Table 4-14). New studies in the U.S. include an analysis of 45 years of
biogeochemical monitoring data (Yanai et al.. 2013) and sulfur accumulation (Mitchell
and Likens. 2011) at the HBEF, NH. At the Niwot Ridge LTER site, CO (Lieb et al..
2011). identify the effects of a decade of simulated N deposition in the southern Rocky
Mountains.
Yanai et al. (2013) evaluated 45 years of biogeochemical monitoring data at the HBEF,
NH. Since 1992, the ecosystem shifted to a net N sink ~8 kg N/ha/yr. There are several
possible explanations: (1) gaseous N fluxes from the ecosystem in response to
denitrification, (2) a budget discrepancy in the net error of the other measured and
estimated stocks and fluxes, and/or (3) N accumulation in an unidentified ecosystem
compartment.
As discussed in Appendix 4.3.3. Mitchell and Likens (2011) examined the sulfur
accumulation observed in over four decades of continuous long-term records for four
watersheds in HBEF, NH and found that as S deposition declined, soil moisture became a
more powerful control on S release from soils than did deposition.
At the southern Rocky Mountains Niwot Ridge LTER site, CO, Lieb et al. (2011) found a
decade of simulated N deposition to alpine ecosystems caused ongoing changes in
diversity and soil biogeochemistry, including lower soil acid buffering capacity,
decreased concentrations of exchangeable Mg2+, and increased concentrations of the
potentially toxic cations Mn2+ and Al3+. Their results suggested an N deposition threshold
for the onset of acidification at this site of 28 kg N/ha/yr.
4-76
-------
Table 4-14
Biogeochemistry monitoring and databases.
Process/
Deposition
Indicator
Type of Ecosystem
Region
(kg/ha/yr)
Addition (kg/ha/yr) Effects
Reference
Base cation (Be)
Alpine soils
Niwot Ridge in the
6 to 8 (10 yr)
8, 28, 48, and 68 Addition: Changes in diversity,
Lieb et al. (2011)
release
southern Rocky
(NH4NO3) lower soil acid buffering capacity,
soil [Al]
Mountains, CO,
decreased concentrations of Mg2+,
soil [Mn]
U.S.
and increased concentrations of
soil pH
the potentially toxic cations Mn2+
and Al3+. Results suggested an N
deposition threshold of
28 kg N/ha/yr.
Monitoring: Over four decades of Mitchell and
data were used to evaluate S Likens (2011)
budgets. Current declining inputs
of atmospheric S and the higher
outputs of SO42" in drainage
waters relative to precipitation
inputs are driven by the S stored
in the soil. Climatic change will
potentially increase SO42"
mobilization and hence could slow
the resultant recovery from
acidification.
Monitoring: Since 1992, the Yanai et al. (2013)
ecosystem shifted to a net N sink,
either storing or denitrifying
~8 kg N/ha/yr.
Sulfate leaching Northeastern forest HBEF, NH, U.S. Not specified None
[S0421
N accumulation C:N Northern hardwood HBEF, NH, U.S. ~7 None
soil [N] forest (since 1992)
4-77
-------
Table 4-14 (Continued): Biogeochemistry monitoring and databases.
Process/
Indicator
Type of Ecosystem
Region
Deposition
(kg/ha/yr)
Addition (kg/ha/yr)
Effects
Reference
NO3" leaching C:N
(lake)
C:N (mineral soil layer)
TOC:TN
Boreal soils and
lakes
Sweden
<3 to 17
None
Gradient: Significant relation
found between C:N ratios of the
organic soil layer and the ones of
lake waters. Evidence found of N
deposition having depressed the
C:N ratios of lake waters more
than the ones of organic soil
layers. Clear sudden increase
seen in NO3" leaching in regions
where N deposition exceeded
7.5 kg/ha/yr.
Khalili etal. (2010)
Oxidation of organic S
in humic soils due to S
dep
Ratio of reduced to
oxidized organic S
Grassland
Rothamsted Park
Grass Experiment,
Herfordshire,
England
Not specified
None
Monitoring: Analysis of the
effects of atmospheric SO2
emissions since the late 1800s
found acidification led to a
depletion of exchangeable Ca and
Mg and an 8x increase in
exchangeable Al.
Lehmann et al.
(2008)
Be, N, and SO42" flux
soil solution [Al]
soil solution Be
soil solution [Inorganic
N]
soil solution [SO42 ]
molar Bc:AI in soil
solution
Forest, including
beech and Norway
spruce
Switzerland, Jura
Mountains in
southern Alps. The
forests were not
managed during
the whole
observation period
Between 2000 and None
2007, mean
(-0.02 to 1.99
kmol/ha/yr)
Monitoring: A decade
(1995-2007) of monitoring data
indicates acidifying deposition
significantly decreased at three
out of the nine study sites due to a
decrease in total N deposition. In
the soil solution, no trend in
concentrations and fluxes of Be,
SO42", and inorganic N were
found at most soil depths at five
out of the seven sites, suggesting
that the soil solution reacted very
little to the changes in
atmospheric deposition.
Pannatier et al.
(2011)
4-78
-------
Table 4-14 (Continued): Biogeochemistry monitoring and databases.
Process/ Deposition
Indicator Type of Ecosystem Region (kg/ha/yr) Addition (kg/ha/yr) Effects Reference
pH, ANC, base cations, Forest and streams HBEF, NH Variable None Field observation: Soil solution Fuss et al. (2015)
Ali, Alo monitoring 1984-2011 at HBEF
showed that: pH and ANC did not
change significantly in the
Oa-horizon but ANC increased in
Bs-horizon. Total base cations
decreased in Bs-horizon. Al0
decreased and. Ali decreased in
the 2 high-elevation
subwatersheds.
Al = aluminum; Al, = inorganic aluminum; Al0 = organic aluminum; Be = base cation; C = carbon; Ca = calcium; ha = hectare; HBEF = Hubbard Brook Experimental Forest;
kg = kilogram; kmol = kilomole; Mg = magnesium; Mn = manganese; N = nitrogen; NH4N03 = ammonium nitrate; N03" = nitrate; S = sulfur; S02 = sulfur dioxide; S042" = sulfate;
TN = total nitrogen; TOC = total organic carbon; yr = year.
4-79
-------
There are new studies evaluating long-term monitoring in Europe, including those
looking at S dynamics in England (Lehmann et al.. 2008) and N and S dynamics in
Switzerland (Pannatier et al.. 2011). In Sweden, Khalili et al. (2010) examined N and C
interactions between boreal soils and lakes. Lehmann et al. (2008) found that soil
acidification from the input of oxidized organic S proceeded much more rapidly than did
recovery after reductions of atmospheric emissions and deposition. Pannatier et al. (2011)
examined monitoring data at Swiss Long-Term Forest Ecosystem Research sites. The
results suggested that the fluxes of Be, SO42 , and inorganic N in soil solution reacted
very little to the changes in atmospheric deposition. A stronger reduction in base cations
compared to Al3+ was detected at two sites, possibly indicating that acidification of the
soil solution was proceeding faster at these sites than the other sites. In Sweden, Khalili et
al. (2010) examined samples collected in selected years between 1993 and 2005.
Although they found a significant relation between C:N ratios of the organic soil layer
and lake waters, the large-scale variations in soil C content were not directly linked to C
concentrations in lake waters. Soil N seems to have leached in small amounts from the
soils directly into the lakes in the form of NO3 . NO;, leaching showed a clear and
sudden increase in regions where N deposition exceeded 7.5 kg/ha/yr.
4.5 Models
The 2008 ISA described the most commonly applied soil biogeochemistry models in the
U.S. used to track N and/or S deposition (Section A.3 of the 2008 ISA). The focus of
Appendix 4.5 is to update available information on several key models currently being
used in the U.S. to assess the effects of S and N deposition on terrestrial ecosystem soil
biogeochemistry. Steady-state models include steady-state simple mass-balance (SMB).
Dynamic models include the Very Simple Dynamic (VSD) soil acidification model,
MAGIC, PnET/BGC, and DayCent-Chem. One important input to these models is
estimating base cation weathering (BCw), and there are new updates on two methods to
estimate this parameter: Soil Texture Approximations (STA) and PROFILE.
4-80
-------
4.5.1 Updates to Key Previously Identified Models
4.5.1.1 Soil Texture Approximation (STA) and PROFILE:
Estimating Base Cation Weathering
Base cation weathering (BCw) rate is one of the most influential yet difficult to estimate
parameters in the calculation of critical acid loads of N and S deposition for terrestrial
systems. As discussed in Appendix 4.3. acidifying deposition causes base cation leaching
from soil. The supply of base cations is largely replenished by base cation weathering of
minerals from rock within the ecosystems. Obtaining accurate estimates of weathering
rates is difficult because weathering is a process that occurs over very long periods of
time. There are some new studies on estimating BCw, including a study on the clay
correlation-substrate method and PROFILE (Koseva et al.. 2010). a new application of
PROFILE in the U.S. (Phelan et al.. 2014). and an evaluation of uncertainty in estimating
BCw (Futtcr et al.. 2012).
The clay correlation-substrate method, also called clay-based Soil Texture
Approximation (STA), is an empirical steady-state model that has been used to estimate
BCw rates for forest ecosystems in the U.S. because it is simple and has low data
requirements. The STA method has an empirical function that was first developed for
European soils and later adapted to soils in Canada and the U.S. The method estimates
total base cation weathering rates (BCw, sum of Ca2+, Mg2+, Na+, and K+), based on the
relationship between established weathering classes and clay content of soils for different
acidity classes—acidic, intermediate, and basic. An alternate steady-state model,
PROFILE (Sverdrup and Warfvinge. 1993). may offer an improved method to estimate
BCw rates. It is a transferable, process-based model that simulates the weathering rates of
groups of minerals. PROFILE was developed in Sweden and is a mechanistic,
steady-state kinetics model that calculates the weathering of Ca2+, Mg2+, K+, and Na+ in
each horizon within a soil profile based on mineral specific chemical dissolution rates
and site and soil conditions. Because PROFILE is a steady-state model, weathering rates
calculated by PROFILE can be used with the SMB critical acid load model. One of the
main limitations that has discouraged the wider adoption of PROFILE for BCw rate
determinations in the U.S. is the large data requirements of the model (Koseva et al..
2010).
Koseva et al. (2010) reported on their evaluation and revision of the STA model for use
in Canadian forests. The authors compared the performance of the STA model to
PROFILE weathering estimates for soils at 75 sites in Canada. The relative ability of the
STA and PROFILE models to provide reasonable weathering rates was evaluated using
4-81
-------
base cation mass balances at a subset of sites in Ontario (n = 19). Mineral weathering
rates calculated with the STA method at 75 sites in Canada (6-367 eq/ha/yr) were up to
38 times lower (7 times on average) than rates estimated with PROFILE
(143-2,119 eq/ha/yr,/? < 0.0001). Despite deviations from the 1:1 line, weathering
estimates obtained using the STA method were significantly correlated using PROFILE
weathering estimates (p < 0.05). The authors concluded that the "revised" STA model
they used may be more widely applicable in Canada, but not necessarily suited to all
regions in the U.S. The uncertainty of the model is largely unknown, and the three
equations of the clay correlation-substrate method require recalibration or revision when
transferred to new locations. In addition, the STA equations were derived for young soils
that developed following the Late Wisconsin glaciation (Koseva et al.. 2010); the clay
correlation-substrate model, which is based on clay content and parent material acidity,
may not be suitable for older, more weathered soils that were not affected by the most
recent glaciation and which cover the majority of the U.S. (U.S. EPA. 2009c).
PROFILE requires over 26 time series or site parameters as model inputs. The U.S.
Geological Survey (USGS) recently completed a soil geochemical and mineralogical
survey of the U.S. as part of the North American Soil Geochemical Landscapes Project.
This project, hereafter referred to as the USGS Landscapes Project, included mineralogy
analyses conducted on soil samples collected from 4,871 evenly spaced sites across the
U.S. (Smith et al.. 2013). These new data allow the PROFILE model to be applied at a
larger scale. Phelan et al. (2014) evaluated PROFILE using national data sets as a method
to estimate BCw rates for forests in the U.S., focusing on Pennsylvania as the first test
state. The model was successfully applied at 51 forested sites across Pennsylvania.
Weathering rates ranged from 11.9 to 924.5 meq/m2/yr and were consistent with soil
properties and regional geology. The authors suggest that the method be applied to other
locations to further evaluate the performance of the model.
A data set of published BCw rates was evaluated by Futter et al. (2012). The data set
included 394 individual silicate mineral weathering rate estimates from 82 poorly
buffered, silicate-mineral-dominated locations across the world where at least 3 published
estimates of BCw were available. The researchers found uncertainty for the input data
high relative to the estimated contribution of model parameter uncertainty of BCw
weathering rate (meq/m2/yr) to overall variability related to PROFILE and MAGIC,
ranging from 1.7 to 2.1 for the two models, respectively. The two models have been most
widely used to make assessments of weathering rates for environmental decision making,
and each represents the landscape differently. PROFILE provides estimates for shallow
one-dimensional soil, whereas MAGIC integrates weathering processes for the whole
catchment, including deep soil weathering. Therefore, lower weathering rate estimates
are shown from PROFILE in comparison to MAGIC, which considers deep soil
4-82
-------
weathering to constitute a potentially important source of Be in catchment weathering
(Futter et al.. 2012).
Whitfield et al. (2018) evaluated major sources of uncertainty associated with using
PROFILE for upland forests in the continental U.S. Mineral stoichiometry was not an
important influence on BCw estimates (uncertainty < 1%). Characterizing B-horizon
mineralogy by averaging A- and C-horizons was found to be a minor (< 5%) contributor
to uncertainty in some areas, but where mineralogy is known to vary with depth the
uncertainty can be large. Estimating mineral-specific surface areas had a strong influence
on estimated BCw, however the greatest uncertainty in BCw estimates, was due to the
particle size class-based method used to estimate the total specific surface area upon
which weathering reactions can take place.
4.5.1.2 Steady-State Mass Balance
The 2008 ISA documented a model to assess CLs for acidification in forest soils based on
simple mass-balance equations [SMBE; McNultv et al. (2007)1. This study estimated
critical acid load and exceedance in soils at a 1-km2 spatial resolution across the U.S. A
second publication discussed the uncertainties associated with this model and
national-scale assessment (Li and McNultv. 2007). The authors quantified uncertainty
under natural variability in 17 model parameters and determined the relative
contributions of each in predicting critical loads. The results indicated that uncertainty in
the CLs came primarily from components of base cation weathering (BCw; 49%) and
acid neutralizing capacity (46%), whereas the most critical parameters were BCw base
rate (62%), soil depth (20%), and soil temperature (11%). The authors concluded that
improvements in estimates of these factors are crucial to reducing uncertainty and
successfully scaling up SMBE for national assessments. This work remains the best
national-scale estimate of terrestrial soil acidification in the U.S. (see Appendix 4.6). A
new study by Posch et al. (2011) is a regional application of SMB models, while several
other studies using SMB have been published to determine acidification critical loads
based on critical limits of Bc:Al and ANC (Phelan et al.. 2014; Duarte et al.. 2013; Jung
et al.. 2013; Whitfield and Watmough. 2012; Forsius et al.. 2010; McNultv and Boggs.
2010; Nasr et al.. 2010). The biological implications of these critical loads are discussed
in Appendix 5 and a summary of SMB CLs in the U.S. is presented in Appendix 4.6.
Posch et al. (2011) published an article outlining SMB models for use in N deposition on
ecosystem biodiversity. This approach has the well-established benefit of easy regional
applicability, while incorporating specified critical chemical criteria to protect specified
receptors. Rather than indicating an upper limit for deposition (i.e., critical load), linked
4-83
-------
nutrient nitrogen and acidity chemical criteria for plant occurrence produce an ""optimal"
nitrogen and sulfur deposition envelope. This method is similar to the methods developed
in the 2012 Policy Assessment for the Review of the secondary NAAQS for Oxides of N
and S.
4.5.1.3 ForSAFE and ForSAFE-VEG
The ForSAFE model (Wallman et al.. 2005) is the biogeochemical simulator platform
that simulates the cycles of carbon, nitrogen, base cations (Be), and water in a forest
ecosystem, including simulation of soil chemistry, tree growth, and soil organic matter
accumulation or depletion. ForSAFE requires site-specific inputs of the physical
properties of the soil (including mineralogy, hydrological parameters, density, depth, and
stratification), tree type, and time series of atmospheric deposition and climatic data
(temperature, light, and precipitation). The model gives monthly estimates of weathering
rates, soil moisture, soil solution concentrations, uptake fluxes of N and Be, litterfall,
decomposition and mineralization, as well as photosynthesis and growth rates.
ForSAFE-VEG is a composite model, in which the VEG module (Sverdrup et al.. 2007)
reads a set of five drivers (soil solution pH, Be concentration, N concentration, ground
level light, soil moisture) from ForSAFE, including air temperature, and uses them to
estimate the relative abundance of a set of indicator plants at the site. The result is a
model chain that can link changes in atmospheric deposition, climatic conditions, and
land use to responses in the biogeochemistry and plant community composition at the site
level, both historically and in the future. A new study by Belvazid et al. (201 la) revealed
limitations in the model simulation of N concentrations in soil solution. The authors
concluded that the biogeochemical model platform must be improved to simulate N
processes more accurately before it is used to calculate CL for N deposition. The model
overestimated the actual N concentrations in the soil solution. Several new applications
of ForSAFE-VEG have been published (Sverdrup et al.. 2012; Belvazid et al.. 201 la),
and the results are discussed in Appendix 5 and Appendix 6.
4.5.1.4 Model of Acidification of Groundwater in Catchment
(MAGIC)
The Model of Acidification of Groundwater in Catchment [MAGIC; Cosby et al.
(1985a); Cosby et al. (1985b); Cosby et al. (1985c)l is one of the most well-known
dynamic models of aquatic and terrestrial acidification. It is a lumped-parameter model of
soil and surface water acidification in response to atmospheric deposition based on
process-level information about acidification. "Lumped-parameter' refers to the extent
4-84
-------
that spatially distributed physical and chemical processes in the catchment are averaged
or lumped together without affecting the model's reproduction of catchment response.
Process-level information refers to how the model characterizes acidification into (1) a
section in which the concentrations of major ions are assumed to be governed by
simultaneous reactions involving S042 adsorption, cation exchange,
dissolution-precipitation-speciation of aluminum, and dissolution-speciation of inorganic
carbon and (2) a mass balance section in which the flux of major ions to and from the soil
is assumed to be controlled by atmospheric inputs, chemical weathering, net uptake and
loss in biomass, and losses to runoff. One strength of MAGIC is the size of the pool of
exchangeable base cations in the soil. As the fluxes to and from this pool change over
time due to changes in atmospheric deposition, the chemical equilibria between soil and
soil solution shift to give changes in surface water chemistry. The degree and rate of
change of surface water acidity thus depend both on flux factors and the inherent
characteristics of the affected soils. The data requirements to run dynamic models like
MAGIC are considerable. The equations that characterize the chemical composition of
soil water in MAGIC contain 33 variables and 21 parameters. Data required to conduct
dynamic modeling are not as available in as many places as the data required to conduct
steady-state modeling.
Oulehle et al. (2012) presented a new formulation of the acidification model MAGIC that
uses decomposer dynamics to link N cycling to C turnover in soils. In comparisons with
earlier versions, the new formulation more accurately simulates observed short-term
changes in NO;, leaching, as well as long-term retention of N in soils. The authors state
that the new formulation gives a more realistic simulation of observed changes in N
leaching. The new formulation also provides a reasonable simulation of the long-term
changes in C and N pools (and C:N ratio) with SOM.
MAGIC has recently been used to calibrate BCw at 140 locations throughout the
southern Appalachian Mountains, where input data were sufficient for running the
dynamic model (Povak et al.. 2014). Results were then extrapolated to the region.
McDonnell et al. (2014b) used these calibrated regionalized BCw values as inputs for
regional steady-state CL modeling.
4.5.1.5 Photosynthesis and Evapotranspiration—
Biogeochemical (PnET-BGC)
The Photosynthesis and Evapotranspiration—Biogeochemical (PnET-BGC) model is an
integrated forest-soil-water model that has been used to assess the effects of air pollution
and land disturbances on forest and aquatic ecosystems (Gbondo-Tugbawa et al.. 2001).
The model was developed by linking two submodels: PnET-carbon and nitrogen
4-85
-------
[PnET-CN; Aber et al. (1997)1 and BGC (Gbondo-Tugbawa et al.. 2001). The main
processes in the model include tree photosynthesis, growth and productivity, litter
production and decay, mineralization of organic matter, immobilization of nitrogen,
nitrification (Aberetal.. 1997). vegetation and organic matter interactions of major
elements, abiotic soil processes, solution speciation, and surface water processes
(Gbondo-Tugbawa et al.. 2001). The hydrologic algorithms used in PnET-BGC are
summarized by Aber and Federer (1992) and Chen and Driscoll (2005). PnET-BGC has
the capability of using multiple soil layers (Chen and Driscoll. 2005) to model seasonal
variations in soil parameters and chemistry. Applications and conceptual advancements
to the model published since 2008 are summarized in Table 4-15.
4.5.1.6 DayCent-Chem
DayCent-Chem links two widely accepted and tested models, one of daily
biogeochemistry for forest, grassland, cropland, and savanna systems, DayCent (Parton et
al.. 1998). and the other of soil and water geochemical equilibrium, PHREEQC
(Parkhurst and Appelo. 1999). The linked DayCent/PHREEQC model was created to
capture the biogeochemical responses to atmospheric deposition and to explicitly
consider those biogeochemical influences on soil and surface water chemistry. The linked
model expands on DayCent"s ability to simulate N, P, S, and C ecosystem dynamics by
incorporating the reactions of many other chemical species in surface water.
4-86
-------
Table 4-15 Photosynthesis and Evapotranspiration—Biogeochemical (PnET-BGC) and DayCent.
Region
Deposition
Model
(Dynamic or
Steady State)
Approach/Observation
Reference
HBEF, NH
526 eq S/ha/yr;
327 eq N03"/ha/yr;
-522 eq
NOs" + NHs/ha/yr
PnET-BGC
Application: Combination of multiple deposition scenarios of S, NO3", and
base cation deposition (0 to 100% decrease in 20% increments after 2008), as
well as current climate and climate change scenarios to year 2100.
Wu and Driscoll
(2010)
HBEF, NH
Not specified
PnET-BGC
Conceptual advancement: PnET-BGC modified to include CO2.
Pourmokhtarian et al.
(2012)
San Bernadino
Mtns., CA
Two sites:
(1) 8.8 kg N/ha/yr;
(2) 70 kg N/ha/yr
DayCent
Conceptual advancement: DayCent modified to include O3 effects.
Bvtnerowicz et al.
(2013)
Adirondack
Long-term
Monitoring
Program (ALTM;
128 lakes)
Mean wet + dry dep.
2009-2011:
S = 20-34 meq/m2/yr;
N = 24-33 meq/m2/yr
PnET-BGC
Application: Controlling S deposition is more effective in promoting acidic lake
recovery than controlling S + N deposition.
Reducing S dep. 60% beyond 2011 level is predicted to restore 28% of
impaired lakes to ANC 20 peq/L > by 2050 and 60% of lakes by 2200. An ANC
of 11 peq/L can be achieved to 53% of lakes by 2050.
Fakhraei et al. (2014)
Adirondack Mtns.
Region, NY
S042" = 290.3-365.7
eq/ha/yr;
NOs" = 172.5-233.5
eq/ha/yr
Application: PnET-BGC was used to evaluate biophysical factors that affect
CLs and TLs of acidity for the Constable Pond watershed, as an example of a
chronically acidic drainage lake in the Adirondack region of New York, U.S.
These factors included a range of future scenarios of decreases in atmospheric
nitrate, ammonium and sulfate deposition from present to 2200.
Zhou et al. (2015c)
Great Smoky
Mtns. National
Park
6.8-27.8 kg S/ha/yr
6.1-16.6 kg N/ha/yr
(2004-2008)
PnET-BGC
Application: Simultaneous reduction in SO42" and NO3" deposition is more
effective at increasing stream ANC than SO42" alone. NO3" leaching continues
as N deposition decreases. Stream recovery is delayed as NO3" facilitates
desorption of legacy SO42" that is adsorbed to acid soils.
Zhou et al. (2015b)
4-87
-------
Table 4-15 (Continued): Photosynthesis and Evapotranspiration—Biogeochemical (PnET-BGC) and DayCent.
Region
Deposition
Model
(Dynamic or
Steady State)
Approach/Observation
Reference
Great Smoky
Mtns. National
Park
3.1 kg S/ha/yr
5.1 kg N/ha/yr
PnET-BGC
Application: Due to soil SO42" adsorption capacity, reducing NHV deposition
would be more effective in stream recovery than reducing NOx and SO2.
Fakhraei et al. (2016)
4-88
-------
4.5.2
New Models (Published since 2008)
4.5.2.1 SMARTml
SMARTml is an assemblage of surface complexation models (SCMs), which are used in
predicting dynamics in soil chemistry without any site-specific calibration. In the first
published application of SMARTml to a spruce forest site in Germany, model results
matched observations well overall (Bontcn et al.. 2011). Simulations deviated from
observations only for soil layers or parameters for which insufficient information was
available. These positive results demonstrate the potential to further apply SCMs in
dynamic modeling. Current results only refer to a single site, and the testing of SCMs for
more cases with differences in soil types, depositions, and environmental conditions are
needed to better understand the strengths of this model.
4.5.2.2 Very Simple Dynamic (VSD) and VSD+
The Very Simple Dynamic (VSD) soil acidification model is the simplest extension of
the steady-state SMB model into a dynamic model. It does this by including cation
exchange and time-dependent N immobilization (accumulation). The VSD model is
designed for sites with little available data and for applications on a large regional or
continental scale. The model has a short execution time that allows rapid scenario
analyses and the calculation of target loads (i.e., deposition targets), which result in a
desired chemical condition in the soil (solution) in a specified year. Posch and Reinds
(2009) developed a version of the VSD for steady-state critical load applications at the
regional scale. Although simpler than other widely used dynamic models (such as
MAGIC and SAFE), VSD contains the basic physical and chemical relationships
common to all these models. However, the model's simplicity also means that some
processes have either been left out altogether (e.g., SO42 sorption) and others strongly
simplified (e.g., N cycling processes). As a consequence, VSD is not best suited for sites
where SO42 adsorption is important. Furthermore, the simple description of N processes
does not allow simulating decreasing soil N pools (and increasing C :N ratios) under
reduced N inputs.
Bonten et al. (2015) describe an extension of the VSD model, called VSD+, which
includes an explicit description of C and N turnover. The VSD+ model includes an
explicit description of organic C and N turnover, whereas in the VSD model, N
4-89
-------
immobilization depends on the N availability and C sequestration mostly depends on the
N immobilization rate, which is controlled by user-prescribed C:N ratios. The VSD+
model also includes S adsorption on soils. The authors apply VSD+ to three forest stands,
which differ in N deposition and soil C:N ratios. Results suggest that at some locations
VSD+ can accurately predict trends and absolute values of NO, and NH44" concentrations
in soil and stream waters, soil C :N ratios and pH.
4.5.2.3 Soil Organic Matter
The Soil Organic Matter (SOM) model was developed as an alternative to the
decomposition module of the PnET model for application in northeastern forests. It relies
on empirical representations of litter decomposition and soil C turnover rates, and
explicitly represents multiple soil horizons. Tonitto et al. (2014) examined the effect of N
addition on SOM dynamics. Their model simulations suggested that ambient atmospheric
N deposition at the forest has led to an increase in cumulative 0-, A-, and B-horizons C
stocks of 211 g C/m2 (3.9 kg C/kg N) and 114 g C/m2 (2.1 kg C/kg N) for hardwood and
pine standards, respectively. They concluded that the model proved largely able to
simulate soil C and N dynamics at their study site under control conditions and that field
observations, mechanistic experiments, and model simulations suggest that the addition
of N to forest ecosystems could have a substantial effect on forest soil C accumulation
via suppression of organic matter decomposition.
4.5.2.4 ORCHIDEE—Carbon-Nitrogen
ORCHIDEE—Carbon-Nitrogen (O-CN) is a terrestrial biosphere model which has been
developed from the land surface model ORCHIDEE and describes the N and C fluxes
and stocks of vegetation and SOM for 10 natural plant functional types, as well as C3 and
C4 croplands at a half-hourly timescale. Zaehle (2013) determined that the estimate C
sequestration from the process-based O-CN model is lower than was found in earlier
studies based on simple biogeochemical models and upscaling of field-based
measurement, which have estimated C sequestration based on N deposition estimates as
0.4-0.7 Pg C/yr in 1990. The O-CN results were within the range of previous model
simulations with the current generation of C-N cycle models (0.2-0.6 Tg N/yr). The
authors believe the study provides an advance over previous assessments because it relies
on a "process-based ecosystem model that integrates the key C-N cycle interactions and
their coupling to biogeophysical processes, while considering the impacts of atmospheric
(climate, CO2) and land cover changes." Key uncertainties in the modeling include (1) the
response of canopy-level photosynthesis to N additions, (2) changes in the allocation
4-90
-------
patterns (root:shoot ratio), and (3) the competition of plants and soil microbes for the
added (or reduced) amount of N.
4.5.2.5 Dynamic Model N14C
Dynamic model N14C is a plant-soil N and C cycling model that simulates terrestrial
ecosystem responses to atmospheric N deposition (Tipping et al.. 2012). The model
includes four plant functional types: broadleaved and coniferous trees, herbs, and dwarf
shrubs. It simulates net primary production (NPP); C and N pools; leaching of DOC,
DON, and inorganic N; denitrification; and the radiocarbon contents of organic matter on
an annual time step. The model simulates annual plant growth and turnover and soil C
and N cycling and is reasonably successful at reproducing average results. However, for
individual sites, there are no significant correlations for C pools or C :N ratios and only
weak relationships for N pools and inorganic N leaching. Inorganic leaching has
traditionally been considered one of the main indicators of N saturation (however, see
new studies in Appendix 4.3.2) and therefore an important goal of N14C is to simulate its
response to N enrichment.
4.5.2.6 Dynamic Simulation Model of Ecosystem Nitrogen
Perakis and Sinkhorn (2011) reported S15N constraints on long-term N balances in
temperate forests using a dynamic simulation model of ecosystem N and S15N. Their
model evaluated which combination of N input and loss pathways could produce a range
of high ecosystem N contents characteristic of forests in the Oregon Coast Range, U.S.
Ecosystem S15N displayed a curvilinear relationship with ecosystem N content and
largely reflected mineral soil, which accounted for 96-98% of total ecosystem N. Model
simulations of ecosystem N balances parameterized with field rates of N leaching
required long-term average N inputs that exceed atmospheric deposition and asymbiotic
and epiphytic N2-fixation, and that were consistent with cycles of post-fire N2-fixation by
early successional red alder. Soil water 5'"NO; patterns suggested a shift in relative N
losses from denitrification to NO;, leaching as N accumulated, and simulations identified
NO;, leaching as the primary N loss pathway that constrains maximum N accumulation.
4.5.3 Comparative Analyses
Tominaga et al. (2009) used HBEF located in NH, as the setting to evaluate the
performance of three uncalibrated process-oriented models. They performed a Monte
4-91
-------
Carlo multiple-model evaluation framework of the dynamic models MAGIC, SAFE, and
VSD. The greatest differences in model outputs were attributed to the cation exchange
submodel, with Gapon exchange-based models retaining more base cations on the
exchange complex and releasing less into solution, resulting in lower soil solution ANC
values. Given the same deposition scenario, the three models (without calibration)
simulate changes in soil and soil solution chemistry differently, but the basic patterns
were similar.
Bonten et al. (2015) compared common dynamic models used to evaluate soils
(Table 4-16). Comparisons are summarized in Table 4-17.
4-92
-------
Table 4-16 Overview of properties of four dvnamic soil chemistrv models as characterized in Bonten et al.
(2015).
Features
VSD
MAGIC
ForSAFE
SMARTml
Number of soil layers
1
1-3
1-20
1-15 soil horizons
Temporal resolution
Annual
Annual/monthly
Monthly
Variable (day-yr)
Be weathering
External input
Calibrated, or
external-input
Submodel
External input
Forest nutrient uptake
External input
External input
Submodel
Submodel
Runoff-to-precipitation ratio
External input
External input
Fixed ratio
External input
Sulfate adsorption
Not included
Langmuir isotherm
Yes
At Fe/AI hydroxides using a 2pK-DDL model
N immobilization
Fractional, fixed, or
external input
Fractional, fixed, or
external input
Submodel
Submodel
Nitrification
100%
Fractional
Submodel
Submodel
Denitrification
Fractional, fixed
Fractional
None or
submodel
Submodel
Soil N build-up controlled by C:N ratio
Yes
Yes
No
Submodel
CO2 degassing in surface water
Yes
Yes
Yes
No surface water
AI(OH)3 precipitation in stream
Yes
Yes
Yes
No surface water
Lumped base cations (Ca2+, Mg2+, and K+)
i Yes
No
Yes
No
DOC dissociation model
Oliver or simple
monoprotic
T riprotic
Oliver
NICA-Donnan
4-93
-------
Table 4-16 (Continued): Overview of properties of four dynamic soil chemistry models as characterized in Bonten
et al. (2015).
Features
VSD
MAGIC
ForSAFE
SMARTml
Cation exchange
Gaines-Thomas or
Gapon
Gaines-Thomas
Gapon
OM: NICA-Donnan clay: Donnan gel hydroxides: 2pk
DDL
Number of cation exchange equations
2
4
1
Not relevant
Ions in soil solution charge balance
12
28
16
Full speciation of ions in the soil solution
First appearance in the literature
2009
1985
1993
2011
4-94
-------
Table 4-17
Model comparison.
Soil Process/
Indicator
Type of Ecosystem
Region
Deposition
kg/ha/yr
Modeled Observation
Reference
Soil S
Forest
Three Swiss forest
monitoring sites
Not specified
In a comparison of four models VSD, MAGIC,
ForSAFE, and SMARTml, ForSAFE-modeled S
concentrations tended to be underestimated by
the model.
Bonten et al. (2015)
Soil pH
Forest
Three Swiss forest
monitoring sites
Not specified
In a comparison of four models VSD, MAGIC,
ForSAFE, and SMARTml, ForSAFE results
indicated upper soil layers can have higher
modeled pH than observed.
Bonten et al. (2015)
Soil ANC and pH
Scots pine plantation
Long-term forest
monitoring site in
U.K. (Peak District
of northern
England)
Not specified
In a comparison of four models VSD, MAGIC,
ForSAFE, and SMARTml, MAGIC successfully
reproduced the trend in acidity (ANC and pH) in
soil solution over the majority of the (ca. 1990s)
monitoring period.
Bonten et al. (2015)
Soil SO42"
Spruce forest
Gardsjon. Sweden
Not specified
In a comparison of four models VSD, MAGIC,
ForSAFE, and SMARTml, ForSAFE modeling
results showed SO42" was slightly
underestimated compared to measured SO42"
concentrations in soil solutions at 5-, 10-, and
20-cm depth (1990-2000).
Bonten et al. (2015)
Soil SO42"
Spruce forest
Bechtel,
Switzerland
Not specified
In a comparison of four models VSD, MAGIC,
ForSAFE, and SMARTml, ForSAFE-modeled
soil SO42" was reasonably well modeled at
20-cm depth but clearly below measured values
at 100-cm depth (1990-2005).
Bonten et al. (2015)
4-95
-------
Table 4-17 (Continued): Model comparison.
Soil Process/ Deposition
Indicator Type of Ecosystem Region kg/ha/yr Modeled Observation Reference
Ion Forest Three Swiss forest Not specified Comparison of four models VSD, MAGIC, Bonten et al. (2015)
concentrations, Al monitoring sites ForSAFE, and SMARTml. VSD reproduced the
ion concentrations similar to measured values,
however the modeled Al concentrations were
substantially lower than measured ones. VSD
only calculates free Al3+, whereas
measurements also include other Al-species as
Al-hydroxides and Al complexed by dissolved
organic matter and fluoride.
ForSAFE results compared to observations. In Bonten et al. (2015)
Sweden, nutrient base cations were slightly
underestimated compared to measured values
in soil solution at 5-, 10-, and 20-cm depth
(1990-2000), while Na+ in soil solution was
relatively well modeled at 5-, 10-, and 20-cm
depth but less well in deeper soil. In
Switzerland, base cations modeling results were
"within the range of the measured values with a
slight underestimation at 20-cm depth"
(1990-2005). In Germany, Ca2+ agreed with
measured values in soil solution (1990-2005).
Al
Spruce forest
Bechtel,
Switzerland
Not specified
ForSAFE-modeled Al3+ was "in the range of the
measured soil solution values at 20-cm depth"
(1990-2005).
Bonten et al. (2015)
Al, Ca, and
protons
Norway spruce forest
Long-term
monitoring site in
Germany
Not specified
SMARTml results showed that modeled Al3+,
Ca2+, and protons agree with measured values
observations in soil solution (1990-2005).
Bonten et al. (2015)
Base cations Spruce forest
(nutrients
Ca + Mg + K; Na;
and Ca)
Gardsjon, Sweden; Not specified
Bechtel,
Switzerland; and
long-term
monitoring site in
Germany
4-96
-------
Table 4-17 (Continued): Model comparison.
Soil Process/
Indicator
Type of Ecosystem
Region
Deposition
kg/ha/yr
Modeled Observation
Reference
Base saturation
Forest
Great Smoky
Mountains National
Park
5.1 kg N/ha/yr
(36.5 mmolc/m2/yr)
PnET-BGC used to model 30 stream
watersheds during 1981-2014 when SO42" and
NO3" deposition decreased (81 and 53%, resp.).
Hindcast modeling (beginning ca. 1850) showed
decreased soil base saturation from 17.8%
(preindustrial median) to 12.6% (current
median).
Fakhraei et al. (2016)
: kilogram; HBEF = Hubbard Brook Experimental Forest; S = sulfur; yr = year.
4-97
-------
4.6
National-Scale Sensitivity
The 2008 ISA documented that by the end of the 1980s, after nearly a decade of intensive
research performed under the original National Acid Precipitation Assessment Program
(NAPAP) research program, the regions of the U.S. with acid-sensitive waters and
ecosystems were well recognized. These acid-sensitive ecosystems are generally located
in upland mountainous terrain in the eastern U.S. and are underlain by weathering-
resistant bedrock, such as granite or quartzite sandstone. The 2008 ISA documented maps
of the U.S. that showed areas sensitive to acidification. However, similar maps for areas
sensitive to the eutrophication effects of nitrogen were not available. Strong evidence was
documented demonstrating that biogeochemical sensitivity to deposition-driven
eutrophication and acidification is the result of historical loading, geologic/soil conditions
(e.g., mineral weathering and S adsorption), and nonanthropogenic sources of N and S
loading to the system. The 2008 ISA documented that there was no single deposition
level applicable to all ecosystems in the U.S. that will describe the onset of eutrophication
or acidification. Since the 2008 ISA, there are new publications commenting on recovery
of terrestrial ecosystems at either the national scale (NAPAP. 2011) or in specific regions
(Lawrence et al.. 2015a; McDonnell et al.. 2013; Elliott et al.. 2008). One new paper
evaluates national-scale terrestrial critical loads for nitrate leaching (Pardo et al.. 2011b).
but work on national-scale soil acidification published in 2007 remains the most recent
national assessment of this effect (McNultv et al.. 2007).
4.6.1 Acidification Recovery
It is important to note that different chemical pools within the soil may recover from
declining N and S atmospheric deposition at different rates. For example, soil solution
CaAl ratio or SO42 or NO, concentration are faster responding indicators than total N.
Indicators that are linked to slow pools (such as soil percentage base saturations or soil
C-to-N ratios) will have slow response times with regard to changes in atmospheric
deposition. An indicator such as ANC which is influenced by both fast (solution) and
slow (soil) pools has an intermediate response time. In addition to how indicators convey
rates of recovery in different biogeochemical pools, recovery can be documented by
empirical evidence and projected by models of recovery trajectories.
The most recent national-scale assessment of recovery from acidifying deposition was the
National Acid Precipitation Assessment Program (NAPAP) report to Congress (NAPAP.
2011). NAPAP is a cooperative federal program first authorized in 1980 to coordinate
acid rain research and report the findings to Congress. The NAPAP member agencies are
4-98
-------
the U.S. EPA, the U.S. Department of Energy, the U.S. Department of Agriculture, the
U.S. Department of Interior, the National Aeronautics and Space Administration, and the
National Oceanic and Atmospheric Administration. The report concluded that few studies
have evaluated terrestrial ecosystem health relative to acidification effects over time, and
soils in the most acid-sensitive regions for which there is more data continue to acidify.
New evidence since the publication of the NAPAP (2011) is summarized in Table 4-18.
A new study supports the beginning of recovery from soil acidification in the
northeastern U.S. (Fuss et al.. 2015; Lawrence et al.. 2015a). Fuss et al. (2015) studied
HBEF in NH and found indicators of recovery of acidification in soil solution
measurements taken from 1984-2011. At the same site, Phelan et al. (2016) conducted a
modeling study that showed current reductions in deposition generally halted further
damage to soils and plants and resulted in no or only partial recovery.
In the southern Appalachian Mountains, modeling studies suggest current stress and
recovery likely to take decades even under scenarios of large reductions in S deposition
(McDonnell et al.. 2013; Elliott et al.. 2008). Rice et al. (2014) calculated for many
forested, unglaciated watersheds from Pennsylvania to Georgia will begin releasing SO42
over the next two decades (Appendix 4.3.3). Unglaciated soils, like those that occur in the
southeastern U.S., accumulate S that is slowly released from soil pools into drainage
water, a process that temporarily delays ecosystem recovery in response to decreases in S
deposition (Appendix 4.3.3).
The acid-base characteristics of DOC are an important part of understanding the recovery
potential for soils, lakes, and streams impaired by acid deposition; these effects are
discussed in Appendix 4.3.9.
4-99
-------
Table 4-18
Recovery.
Empirical or
Region
Type of Ecosystem
Model
Approach/Observation
Reference
Southern Forest soil NuCM Model: Modeled three S deposition simulations: current, 50% decrease, Elliott et al. (2008)
Appalachian and 100% increase, at Joyce Kilmer (JK), Shining Rock (SR), and Linville
Mountains, NC Gorge (LG) wilderness areas. Low Ca:AI ratios results suggest that
forests at SR and LG are significantly stressed under current conditions.
The soil SO42" retention is low, perhaps contributing to the high degree of
acidification. The soils are very acidic and low in weatherable minerals.
Even with large reductions in SO42" and associated acid deposition, it
may take decades before these systems recover from depletion of
exchangeable Ca, Mg, and K.
Southern 65 streams and their MAGIC model Model: A study used the MAGIC model to evaluate soil Be status. Future McDonnell et al.
Appalachian watersheds S deposition reduction scenarios (6, 58,65, and 78% reduction), and (2013)
Mountains, U.S. various changes in timber harvest, temperature, and precipitation were
modeled. Each of the scenario projections indicated that median year
2100 soil exchangeable Ca will be at least 20% lower than preindustrial
values. The simulations suggested that substantial mass loss of soil Be
has already occurred since preindustrial times. Soil Be pools in the study
region are expected to remain significantly below preindustrial conditions
for more than 100 yr, regardless of changes in climate, S deposition, or
timber harvest.
Great Smoky Forest soils and PnET-BGC Application: Simultaneous reduction in SO42" and NO3" deposition is Zhou et al. (2015b)
Mountains streams more effective at increasing stream ANC than SO42" alone. NO3" leaching
National Park, continues as N deposition decreases (this is due to unmanaged forests).
U.S. Stream recovery is delayed as NO3" facilitates desorption of legacy SO42"
that is adsorbed to acid soils.
Great Smoky Forest soils and PnET-BGC Model: Modeling of 30 stream watersheds characterized by decreased Fakhraei et al. (2016)
Mountains streams SO42" and NO3" deposition during 1981-2014 (81 and 53%) showed that
National Park, stream recovery has been limited and delayed due to the high sulfate
U.S. adsorption capacity of soils in the park resulting in a long lag time for
recovery of soil chemistry.
4-100
-------
Table 4-18 (Continued): Recovery.
Empirical or
Region Type of Ecosystem Model Approach/Observation Reference
Eastern Canada Forest and streams Recovery Field observation: 27 sites exposed to reductions in wet SO42" Lawrence et al.
and the deposition of 5.7-76%, over intervals of 8-24 yr. Results are decreased (2015a)
northeastern U.S. exchangeable Al in the O-horizon and increases in pH in the O and
B-horizons at most sites. Among all sites, reductions in SO42" deposition
were positively correlated with ratios of base saturation and negatively
correlated with exchangeable Al ratios in the O-horizon. However, base
saturation in the B-horizon decreased at one-third of the sites, with no
increases.
HBEF, NH Forest and streams Ali Recovery Field observation: Slowed losses of base cations from soil and Fuss et al. (2015)
decreased mobilization of dissolved inorganic aluminum were observed.
Stream water pH at the watershed outlet increased at a rate of
0.01 units/yr and the acid neutralizing capacity (ANC) gained
0.88 peq/L/yr. Dissolved organic carbon generally decreased in stream
water and soil solutions. Both baseline and chronic acidification
(measured during snowmelt) are recovering at this site.
HBEF, NH, and Forests soils and Model Model: Only when future deposition to 2100 was returned to preindustrial Phelan et al. (2016)
Bear Brook plants ForSAFE-VEG levels was recovery of soil and plant community to 1900 conditions
Watershed, ME projected. Policy-based reductions in deposition generally halted further
damage to soils and plants and resulted in no or only partial recovery.
Adirondack Forests soils and PnET-BGC Model: Controlling S deposition is more effective at acidic lake recovery Fakhraei et al. (2014)
Long-term lakes than controlling S + N deposition. Reducing N deposition is less effective
Monitoring because the resulting increase in soil pH leads to soil desorption of SO42"
Program (ALTM; and other anions.
128 lakes) Reducing S dep. 60% beyond 2011 level is predicted to restore 28% of
impaired lakes to ANC 20 peq/L > by 2050 and 60% of lakes by 2200. An
ANC of 1 peq/L can be achieved to 53% of lakes by 2050.
4-101
-------
Table 4-18 (Continued): Recovery.
Empirical or
Region Type of Ecosystem Model Approach/Observation Reference
Adirondack Mtns. Forests soils and PnET-BGC Model: Decreasing SO42 deposition is 4.6x more effective than NO3 Zhou et al. (2015c)
Region, NY lakes decrease for years 2040-2050.
Decreases in NO3" deposition is more effective at increasing lake ANC
than an equal decrease in NhU"1" deposition.
Due to the higher mass transfer coefficient for in-lake retention of NO3",
decreasing NO3" deposition decreases NO3" leaching but lowers ANC
production.
As SO42" and NO3" deposition decrease, a significant lake DOC
concentration increase has been observed. However, PnET-BGC does
not consider the decrease in soil organic matter partitioning as acid
deposition decreases and soil pH increases.
MAGIC = Model for Acidification of Groundwater in Catchments; NuCM = Nutrient Cycling Model; S042 = sulfate; Ca = calcium; Mg = magnesium; K = Potassium; S = sulfur;
Be = base cation.
4-102
-------
4.6.2
Critical Loads
4.6.2.1 Soil Acidification
The 2008 ISA included McNultv et al. (2007) a national assessment of soil critical loads
for acidification calculated by a SMBE technique (see Section 4.5.1.8 of the 2008 ISA).
The uncertainty associated with the SMBE technique was discussed in a second
publication (Li and McNultv. 2007). In general, the SMBE predicted that 26% of U.S.
forest soils have critical loads less than 1,000 eq/ha/yr. Low critical loads outside of New
England, New York, and the Appalachian Mountain region are not necessarily
problematic to forest health because acid deposition is much lower across most of the
U.S. compared to these areas. Unfortunately, mountainous terrain comprises much of the
area with very low forest soil critical loads. Mountain forests receive some of the highest
local rates of acidic deposition.
No new publications are identified in this review on the subject of national-scale
sensitivity; however, there are new reports on regional sensitivity. Figure 4-10 is a map of
soil CLs presented by McNultv et al. (2007) and updated with newer SMB modeling,
where available (McDonnell et al.. 2014b; Phelan et al.. 2014; Duarte et al.. 2013;
Sullivan et al.. 201 lb; Sullivan et al.. 201 la). Duarte et al. (2013) is a new evaluation of
critical loads for terrestrial acidification in New England, U.S. The steady-state mass
balance method is applied at over 4,000 plots. The acceptable ANC leaching rate was
calculated based on the critical chemical criteria of no change in base saturation. Over
80% of the critical loads were between 850 and 2,015 eq/ha/yr. Phelan et al. (2014)
evaluated PROFILE using national data sets as a method to estimate BCw rates for
forests in the U.S., focusing on Pennsylvania as the first test state. The model paired with
national data sets was successfully applied at 51 forested sites across Pennsylvania. The
soil critical loads were evaluated using a Be Al value of 10.0. The CL values ranged from
4 to 10,503 eq/ha/yr. Sullivan et al. (201 la) used the dynamic model MAGIC to model
terrestrial soil acidification in the Adirondacks, NY. Simulations were based on one
driver of acidic deposition (S) and included evaluation of CLs for soil solution molar
Bc:Al and CaAl ratio, two critical threshold levels (1 and 10), and two endpoint years of
model simulation (2050 and 2100). Statistically selected lakes (n = 44) were modeled to
represent a population of 1,320 sites. Nearly all (>93%) had a very low TL (<25 eq/ha/yr
for the year 2100) when the protection threshold was set to BC Al = 10, and the majority
(>60%) had a high TL of (>100 eq/ha/yr for the year 2100) to achieve BcAL = 1.
Sullivan et al. (201 lb) calculated surface water CLs for 66 sites in the Blue Ridge
4-103
-------
mountains (North Carolina, Tennessee, and South Carolina). CL were reported for
surface waters; however, the MAGIC model was parameterized for all 66 sites including
terrestrial geochemistry. McDonnell et al. (2014b) studied acidification in the southern
Appalachians from northern Georgia to southern Pennsylvania and from eastern
Kentucky and Tennessee to central Virginia and western North Carolina. Although soil
critical loads were not reported in the publication, soil solution data are reported in
Figure 4-10.
Forest Ecosystems Critical Loads for Acidity
6,001-8,800
States
No Data
eq = equivalent; ha = hectare; yr = year.
(A.) McNultv et al. (2007) critical loads are mapped at 1 -km2 grids (center map). For uncertainty, see Li and McNultv (20071.
(B.) Duarte et al. (20131 critical loads are mapped at 4-km2 grids; (C. and D.) Phelan et al. (20141 critical loads are mapped for each
sampling site (Pennsylvania). McDonnell et al. (2014b1: Sullivan et al. (2011 bl: Sullivan et al. (2011a) critical loads are mapped as a
single point at the center point of the watershed (New York and North Carolina).
Source: http://nadp.slh.wisc.edu/committees/clad.
Figure 4-10 Forest ecosystem critical loads for soil acidity related to base
cation soil indicators.
4-104
-------
In the case of semiarid ecosystems of southern California, which are not highly sensitive
to acidification, soils in high N deposition areas of chaparral and forested areas in the Los
Angeles basin have acidified significantly (Pardo et al.. 2011b).
4.6.2.2 Nitrate Leaching
Pardo et al. (2011b) documented the threshold N deposition value which caused
increased NO;, leaching from forest ecosystems into surface water was 8 to
25 kg N/ha/yr. This information is summarized by ecoregion (Omernick Level 1) in
Figure 4-11. At 4 kg N/ha/yr, increasing NO;, was reported in the organic horizon in the
Colorado Front Range, which suggests incipient N saturation. In the northeastern U.S., N
budgets from 83 forested watersheds show that N retention averages 76% of incoming
atmospheric-N deposition and decreases from 90% retention for sites receiving
7 kg N/ha/yr to 60% retention for sites receiving 11 kg N/ha/yr (Aber et al.. 2003). The
highest critical loads were reported for Mediterranean California mixed-conifer forests.
Since the publication of Pardo et al. (2011b). new studies have been published on nitrate
leaching in the U.S. (see Appendix 4.3.2); one critical load is from Bowman et al. (2014).
who identify 10 kg N/ha/yr as the deposition level associated with nitrate leaching in
RMNP. New evidence from Europe by Pise et al. (2009) shows approximately 95% of
forests receiving less than 8 kg N/ha/yr have leaching of less than 1 kg N/ha/yr. In
Sweden, Khalili et al. (2010) showed a clear sudden increase in NO3 leaching in regions
where N deposition exceeded 7.5 kg/ha/yr.
4-105
-------
Empirical CL of N (kg ha y)
4-17 Northwest Forested Mountains
| 8 Northern Forests; Eastern Temperate Forests
J 10-17 Mediterranean California
10-26 Great Plains
CL = critical load; ha = hectare; kg = kilogram; N = nitrogen; yr = year.
The range of critical loads based on increased nitrate leaching for each ecoregion. The hatch marks indicate increasing levels of
uncertainty: no hatch marks for the most certain "reliable" category and single hatching for the "fairly reliable" category. White areas
lack data for critical loads determination for nitrate leaching.
Source: Pardo et al. (2011 bl.
Figure 4-11 Map of critical loads for nitrate leaching by ecoregion in the U.S.
4.7 Modification of Terrestrial Soil Response to Nitrogen (N)
Biogeochemical responses to N deposition can be modified by many environmental
factors including phosphorus, disturbance, stand age and climatic shifts in temperature
and precipitation. Here we provide a very brief overview of these topics. Appendix 13
provides an overview of modification of ecosystem response to N driven by climate,
whereas this section describes a brief summary of how climate modifies terrestrial soil
response to N.
Uncertainty
| Reliable
[\^j Fairly Reliable
Expert Judgment
4-106
-------
4.7.1
Disturbance and Stand Age Effects on Nitrogen Retention
The 2008 ISA reported that varying degrees of N assimilation, leaching, and microbial
transformation often reflect differences in N status among treatment sites. These
variations have most often been attributed to disturbance history, dating back a century or
more (Goodale and Aber. 2001). Sites that have undergone disturbances that cause loss of
soil N, such as logging, fire, and agriculture, tend to be most effective at retaining
atmospheric and experimental inputs of N. Fire causes substantial N losses from
ecosystems. Timber harvest contributes to nutrient removal from the ecosystem via
biomass export and acceleration of leaching losses (Mann et al.. 1988; Bormann et al..
1968). In particular, logging contributes to loss of N and Ca2+ from the soil (Lattv et al..
2004; Tritton et al.. 1987). Fire has been shown to alter N retention by its effects on
vegetation, leading to less interception of deposition, and soil characteristics; the effects
on these processes may be seen for decades after fire occurs (Kahl et al.. 2007). N
retention capability often decreases with stand age, suggesting that older forests are more
susceptible than younger forests to becoming N saturated (Hedinetal.. 1995).
One new study has been published since 2008 on how disturbance affects N retention
(Vourlitis and Pasquini. 2008). Fire did little to alter patterns of soil N enrichment from
atmospheric N deposition; however, periodic fires have important implications for the
structure and function of chaparral shrublands and their propensity to become N saturated
under current and future N deposition and fire regimes.
4.7.2 Nitrogen and Phosphorus Interactions
Mechanisms driving terrestrial N versus P (phosphorus) limitations differ greatly since
their sources and biogeochemical dynamics differ; therefore, there is variability across
ecosystems (Vitousek et al.. 2010). Organic N can be decomposed by a variety of
enzymes to mineralize N from substrate; however, organic P relies on phosphatase
enzymes to mineralize it, an enzymatic process that is independent of C respiration and N
mineralization (Marklein et al.. 2016).
Literature on N deposition's effect on N and P dynamics focuses on N and P cycling and
the influence of N on phosphatase enzyme activity. Phosphatase is an enzyme released
into the soil by organisms. It cleaves or hydrolyzes phosphorous (phosphoric acid) from
substrate (soil, roots) to yield available phosphate for uptake (Marklein et al.. 2016). The
chemical structure of phosphatase includes N, therefore Vitousek et al. (2010) explained
that N fertilization increases P cycling because it enables organisms to increase their
extracellular phosphatase and, in turn, release more phosphate from soil organic matter.
4-107
-------
There are some conflicting ideas of how N addition effects P in soils. N addition can
acidify soil, therefore phosphatase activity may be suppressed by this mechanism,
decreasing P release to soil (Wang et al.. 2016b). In contrast, N addition can decrease the
adsorption capacity of soil for P ions, thus increasing P availability and uptake by plants
for productivity. It is also well known that N addition can increase litter decomposition,
which may also increase P release/cycling rates. An increased supply of P can delay the
onset of P limitation. It is important though to acknowledge that the type ofN applied,
application rate, and ecosystem type may influence the phosphatase activity (Song et al..
2017).
4.7.3 Climate Modification of Acidification Effects on Soil
For soil acidification, potential future changes in the quantity and temporal distribution of
precipitation and temperature (and their interactions) is expected to alter the wet-dry
cycles that govern the timing and amount of acidic inputs in precipitation, microbial
transformation in the soil, and the flush of acid anions from soils to surface waters.
There are two recent papers on the relationship between precipitation and sulfur in U.S.
watersheds. Rice et al. (2014) indicated that hydrology, and specifically runoff, is an
important controller of sulfate recovery in watersheds because drainage flushes the
accumulated sulfur from the soil. As precipitation and runoff patterns change with a
changing climate, this important process will be affected. Changing hydrology in the
Northeast is well documented and underway, not just a future effect. Similarly, Mitchell
et al. (2011) observed that following decades of changes in stream sulfate concentrations
and fluxes that have been driven by atmospheric deposition, variation in stream sulfate is
now being controlled by variations in precipitation inputs. Increased variation in
precipitation will increase wetting and drying cycles that promote mineralization of
sulfate from soil and subsequent methylation.
In general, if acid anions build up in soil during periods of drought, the eventual flushing
likely causes a more potent acidification event (Moslev. 2015; Whitehead et al.. 2009). If
the acidification event occurs during a time when sensitive biota or lifestages of biota are
present, acidification may cause more adversity to these populations (Kowalik et al..
2007). Increases in storm frequency associated with global climate change (Collins et al..
2013) could increase the frequency and severity of acidification driven by high levels of
sea salt deposition in coastal regions (Wright and Schindler. 1995). Although the
mechanisms of interaction are unclear, increases in DOC concentrations in aquatic
ecosystems across Europe and the U.S. have been linked to acidification, N cycling, and
4-108
-------
climate change, with important implications for water quality and ecosystem function
rEvans et al. (2008); see Appendix 4.3.91.
Warmer temperatures increase decomposition and nitrification. Nitrification will also
increase with increased N supply caused by increased weathering or decomposition
(Booth et al.. 2005). The process of nitrification generates protons that increase the rate
of nitrate and base cation leaching to drainage waters (Murdoch et al.. 1998). The
combined increase of NO3 leaching and loss of base cations has the potential to magnify
acidification in forest soils (Fernandez et al.. 2003). Soil weathering is typically the key
buffer to acidic deposition (Li and McNultv. 2007). and while weathering is increased by
both soil temperature and soil moisture (Gwiazda and Broecker. 1994). it is unclear
whether any future change in the magnitude of temperature and precipitation will be
enough to alter base cation supply or influence the acid-base balance of sensitive
ecosystems. Furthermore, it is unclear whether increased supply of N in soils from either
deposition, increased decomposition, or increased nitrogen fixation will negate the
ameliorative effect of enhanced weathering. Some studies show that climate change will
mitigate acidification through increased weathering (Kopacek et al.. 2017; Belvazid et al..
2011a). while others show that climate change will aggravate acidification although
increased nitrification outpacing enhanced weathering (Wu and Driscoll. 2010). In
general, increased temperature and precipitation will likely enhance inputs of buffering
agents from weathering and deposition, but also increase inputs of acidifying agents from
deposition and enhanced N cycling. The relative sensitivity of these opposing processes
to a given change in climate remains unresolved.
Ecological effects of atmospheric nitrogen (N) and sulfur (S) deposition on two
hardwood forest sites in the eastern United States were simulated in the context of a
changing climate using the dynamic coupled biogeochemical/ecological model chain
ForSAFE-VEG. The main driver of ecological effects was soil solution N concentration
(Mcdonnell et al.. 2018a).
4.7.4 Climate Modification of Nitrogen-Driven Eutrophication in Soil
The following is a brief summary of how temperature, snow, and precipitation affect soil
response to N. In addition, information is summarized on N effects on the flux of three
GHGs (CO2, CH4, and N2O) from terrestrial ecosystems.
Snow interacts with soil biogeochemistry in a number of ways, mostly by temperature
shifts associated with snowpack and soil moisture flux during snowmelt. The highest
flow and lowest ANC periods of the year tend to coincide with times of high snowmelt.
This is when acidification is likely to occur. Snowpack insulates soils from frigid winter
4-109
-------
air temperatures. With a warming climate, soils are expected to become colder and
experience more winter soil freeze-thaw cycles as snow cover continues to decline.
A number of experiments have evaluated snow effects on soil N; these include studies on
the timing, frequency, and severity (snow depth). Contosta et al. (2017) found evidence
that in winters with more snow, a large pulse of NO, may move from soils to streams
when temperatures are too low/flows too high to allow for biological uptake. Groffman et
al. (2001) conducted a snow manipulation experiment at HBEF and found mild late-
season freezing increases soil NO;, levels by physical disruption (increased fine root
mortality) causing reduced plant N uptake and reduced competition for inorganic N,
allowing soil NO;, levels to increase even with no increase in net mineralization or
nitrification. Later studies at the same site have shown that soil N mineralization rates
were more strongly related to soil volumetric water content than to root biomass, snow or
soil frost, or winter soil temperature (Sorensen et al.. 2016). and the more freeze/thaw
cycles anticipated with climate warming supported lower rates of N mineralization at
HBEF (Duran et al.. 2016). The effects of snow depth on soil N have been synthesized in
a meta-analysis by Li et al. (2016c). who evaluated the central tendencies from
41 publications on 12 variables related to soil N. Results are that snow addition
(increasing depth) significantly increased foliar N (+4.5%) and microbial N (MBN,
+35.9%); however, snow removal did not significantly change under snow removal
manipulation. Snow removal manipulation significantly decreased soil N2O efflux
(-34.1%) and nitrification (-24.8%). However, altered snow depth did not significantly
affect soil dissolved organic N (DON), total inorganic N, nitrate, and N leaching.
Because increasing snowpack depth promoted MBN, the unchanged, net N
mineralization and soil ammonium content were probably due to limitation of the soil N
availability and other soil abiotic factors rather than soil microbes.
Greaver et al. (2016) created a synthesis of meta-analyses to study the individual effects
of either N, temperature, or precipitation on different pools and processes related to
ecosystem C pools and fluxes (Figure 13-2). Nineteen different pools and processes were
included in this analysis.
Initial findings are consistent with the biokinetic effects of warming; long-term data and
meta-analyses show that soil respiration, including decomposition and microbial
respiration, is stimulated by increasing temperature (Lu et al.. 2013; Bond-Lambertv and
Thomson. 2010; Rustad et al.. 2001). Most empirical studies show rising temperature
stimulates N release by mineralization (Churkina et al.. 2010). which may be driven more
by temperature effects on moisture (Emmett et al.. 2004). In some dynamic land models,
the additional N from mineralization will stimulate C uptake by plants even more than
current N deposition (Burd et al.. 2016). At the same time, increased N from
4-110
-------
mineralization may cause N induced inhibition of decomposition, a feedback mechanism
that might decrease the amount of N released and which is currently considered by a few
models (Gerber et al.. 2010). The mechanisms causing N driven reduction in
decomposition are not well understood but are thought to result from changes in
microbial community composition and the production of decomposition enzymes, as well
as possible changes in the character and degradability of soil organic matter (Conant et
al.. 2011; Janssens et al.. 2010). Climate change could also affect decomposition rates by
altering both available soil moisture and microscale connectivity among microorganisms,
water, and nutrients within the soil matrix, which, in turn, may alter microbial processes
(Xiang et al.. 2008). Although there is no consensus about how dissolved organic carbon
(DOC) in surface water is regulated overall, increasing N and temperature increase DOC
concentrations (Laudon et al.. 2012). While few meta-analyses have examined
precipitation effects on the soil C cycle, precipitation tends to increase the root C pool
(Figure 13-2).
Ni et al. (2017) created a meta-analysis to evaluate the interactions between warming and
N deposition and found the interaction greatly increased the soil C input (+49%)
compared with the single factor of either warming (+5%) or N deposition (+20%). Soil C
loss was not significantly affected by the interaction of N and warming likely because
increases in decomposition due to warming are offset by the decreases by N addition.
N addition alters fluxes of GHGs by regulating plant and microbial activities that are
directly associated with GHG production and consumption processes (Figure 4-12). It is
important to consider net ecosystem flux because consumption may be offset by
production. For example, N addition stimulates plant growth in most ecosystems
(LeBauer and Treseder. 2008). and may in turn increase C sequestration in plant biomass.
On the other hand, maintenance respiration is positively correlated with tissue N content
(Reich et al.. 2008). and litter with higher N content also decomposes faster. Therefore,
increased leaf N content under elevated N may result in higher C loss by increasing both
autotrophic and heterotrophic respiration. Like CO2, ecosystems may consume and
produce CH4 and N2O, and the balance determines their net release.
Liu and Greaver (2009) conducted a meta-analysis of 313 observations across 109 studies
to evaluate the effect of N addition on the flux of three GHGs: CO2, CFL, and N2O. The
results indicated that N addition increased ecosystem carbon content of forests by 6%,
marginally increased soil organic carbon of agricultural systems by 2% but had no
significant effect on net ecosystem CO2 exchange for nonforest natural ecosystems.
Across all ecosystems, N addition increased CFL emission by 97%, reduced CFL uptake
by 38%, and increased N2O emission by 216%. Most often, N addition is considered to
increase forest C sequestration without consideration of N stimulation of GHG
4-111
-------
production in other ecosystems. However, this study indicated that although N addition
increased the global terrestrial C sink, the GHG benefits of an atmospheric CO2 reduction
could be largely offset (53-76%) by N stimulation of global CH4 and N2O emission from
multiple ecosystems.
CO
n2o
A
N input
Denitrifying bacteria (+)
N,
-n2o*-no
\
NH/ —~NHiONO —NOj"
/
NiO
Nitrifying bacteria (+) Aerobic
C cycle
N cycle
SOC
DOC
DIN/DON
+ Positive feedback
- Negative feedback
ANPP = aboveground net primary productivity; BNPP = beiowground net primary productivity; SOC = soil organic carbon;
DOC = dissolved organic carbon; DIN = dissolved inorganic nitrogen; DON = dissolved organic nitrogen.
Source: Liu and Greaver (20091.
Figure 4-12 The potential mechanisms that regulate the responses of carbon
dioxide (CO2), methane (CH4), and nitrous oxide (N2O) production
and consumption to elevated nitrogen (N).
4-112
-------
4.8 Summary
4.8.1 Sources
The effects of N and S deposition on soil biogeochemistry cause cascading effects on
biological species. The biological effects are discussed in Appendix 5 and Appendix 6.
Since 2008, there have been a number of estimates of sulfate, oxidized nitrogen, and
reduced nitrogen from atmospheric deposition. The most recent estimates are
summarized in Appendix 2. and maps showing the geographic distribution of deposition
are presented for total acidifying deposition (Figure 2-12). total N deposition
(Figure 2-13). and total S deposition (Figure 2-31). Maps depicting how the relative
contribution of oxidized and reduced N species varies across the U.S. are presented in in
Figure 2-14.
In the 2008 ISA, atmospheric deposition was identified as the main source of
anthropogenic N, relative to other N sources, to nonmanaged terrestrial ecosystems. This
has been confirmed by new studies on N sources to U.S. lands and waterways, which find
human-mediated N inputs are spatially heterogeneous across the country, ranging from
<1.0 to 34.6-fold the rate of background N input across the conterminous U.S. Synthetic
N fertilizer and atmospheric N deposition are the largest and second-largest overall
human-mediated N sources (the single largest sources in 41 and 33% of HUC-8s,
respectively). There is no new information published on nondeposition sources of S in
terrestrial ecosystems. S inputs from emissions to the atmosphere are discussed in
Appendix 2.
4.8.2 Soil Processes and Indicators
Soil N enrichment and soil acidification are occuring in sensitive ecosystems across the
U.S. at recent levels of deposition. Soil acidification is a natural process that can be
accelerated by N and/or S deposition. A number of soil geochemical processes are
associated with acidification (Appendix 4.3 and Table 4-19). Base cations counterbalance
acid anions. Base cations are added to a watershed by weathering and atmospheric
deposition and are removed by leaching and (perhaps temporarily) by uptake into
growing vegetation. Acidifying deposition can deplete exchangeable base cation pools in
soils where acid deposition is high relative to base cation input. The limited mobility of
anions associated with naturally derived acidity (organic acids and carbonic acid)
4-113
-------
controls the rate of base cation leaching from soil where rates of atmospheric deposition
of S and N are low. In addition to increasing ion concentration in soil, inputs of S and N
in acidifying deposition provide anions that are highly mobile, further accelerating the
rates of base cation leaching. In addition, the deposition of reduced forms of N
(e.g., NHX) can stimulate nitrification, which is the microbial oxidation of NH4+ to NO;, .
This oxidation process acidifies soils because 2 moles of hydrogen ion (H+) are released
per mole of NH4+ converted to NO;, . Therefore, both chemically reduced and oxidized
forms of N and S deposition contribute to terrestrial acidification. Several useful
indicators of soil acidification (Table 4-19) include indicator thresholds related to
biological responses, the biological basis of which are discussed in Appendix 5.
It is clear from Table 4-19 that some of the same processes and indicators associated with
acidification are also associated with N enrichment. The N enrichment of terrestrial soils
occurs in response to the input of exogenous N. The new studies on N deposition
contribution to total N loading in terrestrial U.S. ecosystems are discussed in
Appendix 4.2.
The 2008 ISA documented that most N, often more than 85% of the total ecosystem N, is
stored in the soil in temperate forest ecosystems. There is new evidence that soil litter is
the largest N pool in grasslands, shrublands, and wetlands (Appendix 4.3.1). The ability
of atmospheric deposition to increase soil N is indicated by a positive correlation between
atmospheric deposition levels and total N concentration in the organic and mineral soil
horizon. Soil N accumulation is linked to increased N leaching and decreased N retention
(Appendix 4.3.2). Thresholds of N deposition that are associated with the onset of
elevated NO; leaching are discussed in Appendix 4.6.2.2.
Recent work suggests that N leaching from soil can precede the complete saturation of
the biotic and abiotic processes that retain N (Appendix 4.3.2). Lovett and Goodale
(2011) proposed a model of N saturation in which NO; leaching can occur even if the
ecosystem N retention capacity has not yet been saturated, as is observed at many sites
and supported by several new studies. Although N leaching may occur prior to saturation,
N budgets from 83 forested watersheds in the northeastern U.S. show that N retention
averages 76% of incoming N deposition. Nitrogen retention decreases from 90%
retention for sites receiving 7 kg N/ha/yr to 60% retention for sites receiving
11 kg N/ha/yr, suggesting a range of response in which leaching is exacerbated in this
region (Aber et al.. 2003).
The 2008 ISA documented that N enrichment alters rates of microbial transformation of
chemicals in the soil, such as increased rates of nitrification and denitrification, and
altered rates of decomposition. The addition of N to terrestrial ecosystems can increase
nitrification, an important process for soil acidification and the production of NO; .
4-114
-------
Nitrification is often stimulated in soils with a C:N ratio below approximately 20 to 25,
and N deposition often causes the ratio to fall below this threshold. The NO;, created by
nitrification may be leached or denitrified. In terrestrial ecosystems, denitrification of
NO;, mainly occurs in saturated zones. Several syntheses have been published since 2008
evaluating N addition effects on denitrification and nitrification in terrestrial ecosystems
(Appendix 4.3.6). A new meta-analysis shows N addition significantly increases
denitrification from many types of ecosystems (coniferous forest, deciduous forest,
tropical forest, wetland, grassland), except heathlands.
4-115
-------
Table 4-19 Summary of key soil geochemical processes and indicators associated with eutrophication and
acidification.
Endpoint
N-Driven Nutrient
Enrichment
Acidification
The Effect of Deposition
PROCESS
N saturation
X
X
New empirical evidence suggests revising the N saturation concept, because
observed NO3" leaching can occur even if the ecosystem N retention capacity
has not yet been saturated.
Soil N accumulation
X
X
New meta-analysis across ecosystem types confirms inorganic soil NO3"
concentration increases with N addition. New gradient study confirms that N
concentration increases with N deposition. New addition study confirms
increased soil accumulation.
NO3" leaching
X
X
New meta-analysis confirms leaching increases with N addition.
New regional-scale gradient analyses: <8 kg N/ha/yr onset of leaching
<1 kg N/ha/yr in European forests. Second study suggests <7.5 kg N/ha/yr
increases leaching in Swedish forests.
New addition study: 9 to 14 kg N/ha/yr increase in inorganic N concentrations.
New USFS critical loads for the onset of leaching: 8 to 10 kg N/ha/yr in eastern
and western U.S., 17 kg N/ha/yr in the Sierra Nevada and San Bernardino
Mountains.
S accumulation and
adsorption
X
Some soils (notably in many watersheds in the SE U.S.) have the capacity to
adsorb substantial quantities of S, with essentially no acidification of drainage
water. Nevertheless, S adsorption capacity is finite and under continual high S
deposition loading, the adsorptive capacity of soil will eventually be depleted.
New studies of 27 watersheds in the SE indicates most will begin releasing
S042"in the next two decades, and northeastern watersheds show a net loss of
S from soils now in response to decreased levels of atmospheric S deposition.
SO42" leaching
X
Atmospheric S deposition generally increases leaching of SO42" to surface
waters. The amount of deposition that causes the onset of leaching varies
across the landscape.
4-116
-------
Table 4-19 (Continued): Summary of key soil geochemical processes and indicators associated with
eutrophication and acidification.
N-Driven Nutrient
Endpoint Enrichment Acidification
Base cation X
release/depletion
The Effect of Deposition
Base cation (Ca, Mg, K, Na) release occurs in response to the input of acid
anions (SO42" and NO3") from deposition. This proportional change in base
cations relative to acid anions from N and S deposition is called the F-factor.
New studies confirm base cation depletion continues to occur in the Rocky
Mountains (threshold 28 kg N/ha/yr) and in U.K. grasslands, while in a
northeastern forest, 17 yr of N addition did not cause further depletion. A
meta-analysis indicates cation depletion occurs early after increased deposition
of acid anions, which tapers off with time.
Al mobilization X <15 to 20% soil base saturation is the threshold for inorganic Al mobilization
from soil. This is an extremely important effect of acidifying deposition because
inorganic monomeric Al, including Al3+ and various hydroxide species, is toxic to
biota (Appendix 5). Inorganic Al is minimally soluble at pH about 6.0, but
solubility increases steeply at pH values below about 5.5.
There have been no new studies on Al mobilization.
Nitrification X X Nitrification releases 2 mol hydrogen ion (H+) per mol NHV converted to NO3",
acidifying soils. As soil inorganic N accumulates, net nitrification rates often
increase, and NO3" can leach from the ecosystem.
New N gradient and meta-analysis studies confirm N addition increases
nitrification.
Denitrification X Denitrification is the microbial reduction of NO3" to nitrite (NO2"), nitric oxide
(NO), the greenhouse gas nitrous oxide (N2O), and N2. It occurs under
anaerobic conditions. In Europe, soil switched from a source to a sink after two
decades of N deposition exclusion.
New meta-analysis confirms N addition increases denitrification rates.
DOC leaching X X Acidity of some surface waters is partly regulated by DOC concentrations. Soil
acidification can suppress the natural production of DOC. In recent years, the
DOC concentration of some lakes and streams has risen, with adjacent
terrestrial ecosystems as the DOC source. However, the mechanism for this
increase is unclear. It may be due to soil recovery from acidification or N
deposition effects on decomposition.
New studies include a meta-analysis and field addition studies.
4-117
-------
Table 4-19 (Continued): Summary of key soil geochemical processes and indicators associated with
eutrophication and acidification.
Endpoint
N-Driven Nutrient
Enrichment
Acidification
The Effect of Deposition
Decomposition/
mineralization
X
X
Decomposition rates correlate with ratios of C:N, lignin:N, or lignin:cellulose in
litter. The addition of N can stimulate the decomposition of labile compounds
that degrade during the initial stages of decomposition, but added N can
suppress the decomposition of more recalcitrant material. Evidence for this is
widespread in forests but has not yet been well documented in grasslands and
other ecosystems.
There are new addition studies and meta-analysis to better understand the
mechanisms and response trends.
INDICATOR
Soil [N]
X
X
Increasing N deposition increases concentration of N in soil and reflects soil N
accumulation that may be assimilated by organisms or mobilized via leaching.
Soil C:N ratio
X
X
Increasing N deposition decreases the C:N in plant tissue, causing cascading
effects during decomposition on microbial transformation, especially nitrification,
leading to increased NO3" leaching.
THRESHOLD: <20 to 25 associated with elevated rates of nitrification and
elevated risk of nitrate leaching in the U.S. and <25 to 30 for increased NO3"
leaching in Europe.
Soil base saturation
X
Increasing N + S deposition decreases the soil pool of exchangeable base
cations.
THRESHOLD: <15 to 20% is the value at which exchange ion chemistry is
dominated by inorqanic Al and may cause iniurv to veqetation (see Appendix 5).
Soil Bc:AI ratio
X
Increasing N + S deposition decreases the soil pool of exchangeable base
cations, often decreasing the Ca:AI ratio.
THRESHOLD: Ca:AI <1.0 to 10 causes physiological stress and decreased
qrowth and survival of sensitive plant species (see Appendix 5).
Fungi:bacteria ratio
X
New indicator since 2008 ISA: Increasing N deposition decreases the fungi-to-
bacteria ratio and causes a transition from N to C limitation among soil food
webs.
Al = aluminum; Be = base cation; C = carbon; Ca = calcium; DOC = dissolved organic carbon; H+ = hydrogen ion; ha = hectare; K = potassium; kg =kilogram; Mg = magnesium;
N = nitrogen; N2 = molecular nitrogen; N20 = nitrous oxide; NE = northeast; NH4+ = ammonium; NO = nitric oxide; N02" = nitrite; N03" = nitrate; S = sulfur; SE = southeast; yr = year.
4-118
-------
The effects of N on decomposition, which is the breakdown of organic matter, is an
active area of research (Appendix 4.3.7). Within the soil microbial community, bacteria
and fungi are the primary decomposers of organic matter. Both microbial community
composition (see Appendix 6) and microbial enzyme activity can respond dynamically to
shifts in inorganic nutrient and substrate availability, reflecting the nutrient and energy
limitation of the microbial community. Litter decay rates are also well established to
correlate with ratios of C:N, lignin:N, or ligninxcllulosc in litter. These chemical traits
have been shown to account for over 73% of the variation in litter decomposition rates
worldwide. Based on these observations, it could be assumed that added N would
stimulate decomposition and the loss of C from soil pools. There is now widespread
evidence, however, that the stimulatory effects of N on decomposition are limited to the
early stages of mass loss, when more labile compounds are degraded. During the later
stages of decomposition when the rates of mass loss are slow and controlled by the
degradation of recalcitrant compounds, the addition of N slows decomposition. This
slowing of decomposition appears to be a consequence of decreases in the production of
some extracellular enzymes by fungi. It is clear that N additions generally decrease
respiration from soil heterotrophs, but this may not decrease total soil respiration because
heterotrophic respiration accounts for only a portion of soil CO2 efflux.
Several new meta-analyses have been published since 2008 on N addition effects on
belowground carbon cycling (Appendix 4.3.10). About half of C fixed annually by
terrestrial vegetation is allocated to belowground C pools; therefore, it is important to
understand how N affects belowground C to better understand changes in plant
physiology, plant growth, and ecosystem C cycling (Appendix 6). Many studies have
shown that changes in the belowground C cycle do not always match the aboveground C
cycle. Therefore, it is inappropriate to extrapolate from aboveground responses to
belowground processes.
4.8.3 Monitoring
Several new studies use long-term monitoring data sets. Yanai et al. (2013) evaluated
45 years of biogeochemical monitoring data at the HBEF, NH. From 1966 to 1977, more
N was accumulating in living biomass than was input by atmospheric deposition; the
missing N source was attributed to N fixation. Since 1992, biomass accumulation has
been negligible, and the ecosystem shifted to a net missing N sink of ~8 kg N/ha/yr,
which the authors hypothesized may not be the result of N retention, but may be due to
gaseous N fluxes in response to N deposition or uncertainty the amount of N in the
mineral soil. Mitchell and Likens (2011) examined sulfur accumulation observed in over
four decades of continuous long-term record for four watersheds, showing that the
4-119
-------
biogeochemical control of annual S042 export in stream water draining from forested
watersheds has shifted from control by atmospheric S deposition to soil moisture driven
by climate (reported in Appendix 4.3.3). At the Niwot Ridge LTER, CO, Lieb et al.
(2011) found that a decade of simulated N deposition to alpine ecosystems caused
ongoing changes in diversity and soil biogeochemistry, including lower soil acid
buffering capacity, decreased concentrations of Mg2+, and increased concentrations of the
potentially toxic cations Mn2+ and Al3+. Their results suggested an N deposition threshold
of 28 kg N/ha/yr.
There are new studies on long-term monitoring in Europe, including investigations of S
dynamics in England (Lehmann et al.. 2008). N and S dynamics in Switzerland
(Pannatier et al.. 2011). and N and C interactions between boreal soils and lakes in
Sweden (Khalili et al.. 2010). Notably, Khalili et al. (2010) found a significant
relationship between C:N ratios of the organic soil layer and C in lake waters, and a clear
sudden increase in NO;, leaching in regions where N deposition exceeded 7.5 kg N/ha/yr.
4.8.4 Models
The most commonly used ecosystem models in the U.S. were described in the 2008 ISA.
Here we focus on the several primary models that are currently being used in the U.S. to
assess the effects of S and N deposition on terrestrial ecosystems. One important input to
these models are estimates of base cation weathering (BCw). There are new updates on
two methods to estimate this parameter: Soil Texture Approximations (STA) and
PROFILE. Steady-state models include steady-state mass balance equations (SMBE),
while dynamic models include the VSD and VSD+, MAGIC, NuCM, PnET/BGC, and
DayCent-Chem.
Base cation weathering rate is one of the most influential yet difficult to estimate
parameters in the calculation of critical acid loads of nitrogen (N) and sulfur (S)
deposition for terrestrial systems. Commonly used models include STA, a simple
empirical steady-state model with low data requirements, and PROFILE, a mechanistic,
steady-state kinetics model that requires 26 input parameters. There is a new study on
estimating BCw. Kosevaet al. (2010) confirmed that the STA equations were derived for
young soils that developed following the Late Wisconsin glaciation and may not be
suitable for older, more weathered soils that were not affected by the most recent
glaciation and which cover the majority of the U.S. (U.S. EPA. 2009c). The larger data
set required to run PROFILE prohibits applying the model to many parts of the U.S.
However, a recent soil geochemical data set compiled by the USGS has made it possible
to apply PROFILE in the U.S. The initial application in 51 forested sites across
4-120
-------
Pennsylvania indicates this technique may be appropriate for applications nationwide
(Phelan et al.. 2014). The uncertainty associated with the calculation of BCw has recently
been evaluated by Futter et al. (2012). who estimated that the contribution of model
parameter uncertainty to overall variability related to PROFILE and MAGIC (a dynamic
model that typically uses PROFILE for initial estimates of BCw) was relatively small.
Estimates of BCw are input parameters in soil acidification models. The simplest
acidification models are the steady-state mass balance models. The 2008 ISA
documented a model based on simple mass-balance equations (SMBE) to assess critical
loads for acidification in U.S. forest soils (McNultv et al.. 2007). A second publication
discussed the uncertainties associated with this model and national-scale assessment (Li
and McNultv. 2007). The results indicated that uncertainty in the critical load came
primarily from components of BCw (49%) and acid neutralizing capacity (46%), whereas
the most critical parameters were BCw base rate (62%), soil depth (20%), and soil
temperature (11%). The authors concluded that improvements in estimates of these
factors are crucial to reducing uncertainty and successfully scaling up SMBE for national
assessments. Posch et al. (2011) discussed a regional application of SMB models.
Dynamic models are typically applied to much smaller spatial areas than steady-state
models because they require more data and parameterization. The VSD model is the
simplest extension of the steady-state SMB model into a dynamic model that is designed
for sites with few data available and applications on a large regional or continental scale.
Posch and Reinds (2009) developed a version of the VSD for steady-state critical load
applications at the regional scale. The ForSAFE model Wallman et al. (2005) simulates
the cycles of carbon, nitrogen, base cations (Be), and water in a forest ecosystem, while
simultaneously simulating soil chemistry, tree growth, and soil organic matter
accumulation or depletion. ForSAFE-VEG (Sverdrup et al.. 2007) is a composite model
that can link changes in atmospheric deposition, climatic conditions, and land use to
responses in the biogeochemistry and plant community composition at the site level, both
historically and into the future. Belvazid et al. (201 la) stated that the biogeochemical
model platform must be improved to simulate N processes more accurately before being
applied to calculate critical loads. The model overestimated the actual N concentrations
in the soil solution. Several new applications of ForSAFE-VEG were published rBelvazid
et al. (2011a); Sverdrup et al. (2012); McDonnell et al. (2013). and the results are
discussed in Appendix 5 and Appendix 61.
MAGIC (Cosby et al.. 1985a; Cosby et al.. 1985c; Cosby et al.. 1985b) is one of the most
well-known dynamic models of aquatic acidification. An update to MAGIC by Oulehle et
al. (2012) gives a more realistic simulation of observed changes in N leaching. The new
formulation also provides a reasonable simulation of the long-term changes in C and N
4-121
-------
pools (and C:N ratio) with SOM. PnET-BGC is an integrated forest-soil-water model that
has been used to assess the effects of air pollution and land disturbances on forest and
aquatic ecosystems; it uses a new conceptual advancement that incorporates CO2
dynamics into the model. DayCent-Chem links two widely accepted and tested models,
one of daily biogeochemistry for forest, grassland, cropland, and savanna systems,
DayCent (Parton et al.. 1998). and the other of soil and water geochemical equilibrium,
PHREEQC (Parkhurst and Appelo. 1999). There is a new conceptual advancement for
DayCent which now includes O3 (Bvtncrow icz et al.. 2013). The Nutrient Cycling Model
(NuCM) was developed to synthesize current understanding of nutrient cycling in forests
and to predict how forests respond to changing S and N atmospheric deposition rates. No
revisions to these models have been identified.
4.8.5 National-Scale Sensitivity
The 2008 ISA documented that by the end of the 1980s, after nearly a decade of intensive
research performed under the original NAPAP, the regions of the U.S. with many
acid-sensitive waters and ecosystems were well recognized. These acid-sensitive
ecosystems are generally located in upland mountainous terrain in the eastern U.S. and
are underlain by weathering-resistant bedrock like granite or quartzite sandstone.
However, similar work characterizing areas sensitive to the eutrophication effects of
nitrogen is not available. Typically, all terrestrial ecosystems are sensitive to N
deposition.
There is strong evidence to show that biogeochemical sensitivity to N-driven
eutrophication and acidification from atmospheric deposition is the result of historical
loading, geologic/soil conditions (e.g., mineral weathering and S adsorption), and
nonanthropogenic sources of N and S loading to the system. No single deposition level is
applicable to all ecosystems in the U.S. that will describe the onset of eutrophication or
acidification. Ecosystem sensitivity is heterogeneous.
New publications have commented on recovery of terrestrial ecosystems from
acidification at the national scale (NAPAP. 2011). specifically in the Northeast (Phelan et
al.. 2016; Fuss et al.. 2015; Lawrence et al.. 2015a). and the lack of recovery in the
southern Appalachian Mountains (Fakhraei et al.. 2016; Zhou et al.. 2015b; McDonnell et
al.. 2013; Elliott et al.. 2008). Work on national-scale soil acidification published in 2007
remains the most recent national assessment of this effect (McNultv et al.. 2007);
however, Figure 4-10 is a map of soil CLs that updates (McNultv et al.. 2007) with newly
available modeling (McDonnell et al.. 2014b; Phelan et al.. 2014; Duarte et al.. 2013;
Sullivan et al.. 201 lb; Sullivan et al.. 201 la).
4-122
-------
New critical loads for nitrate leaching in forest ecosystems reported by Pardo et al.
(2011b) ranges from 8 to 25 kg N/ha/yr across ecoregions (Omernick Level 1) in the U.S.
(Figure 4-11). At 4 kg N/ha/yr, increasing NO;, was reported in the organic horizon in
the Colorado Front Range, which suggests incipient N saturation. In the northeastern
U.S., N budgets from 83 forested watersheds show that N retention averages 76% of
incoming atmospheric-N deposition and decreases from 90% retention for sites receiving
7 kg N/ha/yr to 60% retention for sites receiving 11 kg N/ha/yr (Aber et al.. 2003). The
highest critical loads were reported for Mediterranean California mixed-conifer forests.
Since the publication of Pardo et al. (2011b). new studies have been published on nitrate
leaching (see Appendix 4.3.2). Evidence from Europe by Pise et al. (2009) showed
approximately 95% of forests receiving less than 8 kg N/ha/yr have leaching of less than
1 kg N/ha/yr. In Sweden, Khalili et al. (2010) showed a clear sudden increase in NO;,
leaching in regions where N deposition exceeded 7.5 kg/ha/yr.
4.8.6 Climate Modification of Soil Response to Nitrogen Addition
Soil biogeochemistry response to N deposition can be modified by climatic shifts in
temperature and precipitation. For soil acidification, potential future changes in the
quantity and temporal distribution of precipitation and temperature (and their
interactions) is expected to alter the wet-dry cycles that govern the timing and amount of
acidic inputs in precipitation, microbial transformation in the soil, and the flush of acid
anions from soils to surface waters. In general, increased temperature and precipitation
will likely enhance inputs of buffering agents from weathering and deposition, but also
increase inputs of acidifying agents from deposition and enhanced N cycling. The relative
sensitivity of these opposing processes to a given change in climate remains unresolved.
Snow interacts with soil biogeochemistry in a number of ways, mostly by temperature
shifts associated with snowpack and soil moisture flux during snowmelt (Duran et al..
2016; Sorensen et al.. 2016; Groffman et al.. 2001). The highest flow and lowest ANC
periods of the year tend to coincide with times of high snowmelt. This is when
acidification is likely to occur. Snowpack insulates soils from frigid winter air
temperatures. With a warming climate, soils are expected to become colder and
experience more winter soil freeze-thaw cycles as snow cover continues to decline. There
have been a number of experiments evaluating how snow effects soil N; these include
studies on the timing, frequency, and severity [snow depth; Li et al. (2016c)l.
Greaver et al. (2016) created a synthesis of meta-analyses to study the effects of either N,
temperature, or precipitation on different pools and processes related to ecosystem C
responses to these variables and to gain some insight as to the direction and magnitude of
4-123
-------
the change observed (Figure 13-2). Nineteen different pools and processes were included
in this analysis. Liu and Greaver (2009) conducted a meta-analysis of 313 observations
across 109 studies of terrestrial and wetland ecosystems to evaluate the effect of N
addition on the flux of three GHGs (CO2, CH4, and N2O). This study indicated that
although N addition increased the global terrestrial C sink, the GHG benefits of an
atmospheric CO2 reduction could be largely offset (53-76%) by N stimulation of global
CH4 and N2O emission from multiple ecosystems. The authors noted that N addition
alters fluxes of GHGs by regulating plant and microbial activities that are directly
associated with GHG production and consumption processes. It is therefore extremely
important to consider total ecosystem flux of GHGs because consumption may be offset
by production, even within the same ecosystem
4-124
-------
APPENDIX 5 BIOLOGICAL EFFECTS OF
TERRESTRIAL ACIDIFICATION
This appendix characterizes the biological effects of acidifying deposition of nitrogen (N)
and sulfur (S) to terrestrial ecosystems. Appendix 5.2 discusses effects on trees and
forests (Appendix 5.2.1). understory plants and grasslands (Appendix 5.2.2). lichens
(Appendix 5.2.3). soil biota (Appendix 5.2.4). and fauna (Appendix 5.2.5). The
characteristics, distribution and extent of sensitive ecosystems is presented in
Appendix 5.3. Next, Appendix 5.4 presents the application of terrestrial acidification
models. Finally, levels of deposition at which effects are manifested is discussed in
Appendix 5.5. including a discussion of critical loads (Appendix 5.5.3). Climate
modification of acidifying deposition effects is discussed in Appendix 5.6. A summary
section with causal determinations based on a synthesis of new information and previous
evidence of biological effects of terrestrial acidification is presented in Appendix 5.7.
5.1 Introduction
Changes in biogeochemical processes caused by acidifying deposition of N and S to
terrestrial systems (Appendix 4) are linked to changes in terrestrial biota and have
significant ramifications for biological functioning of these ecosystems. In the 2008 ISA
for Oxides of Nitrogen and Oxides of Sulfur—Ecological Criteria (hereafter referred to as
the 2008 ISA), the evidence was sufficient to infer a causal relationship between
acidifying N and S deposition and changes in terrestrial biota. The strongest evidence for
a causal relationship comes from studies of terrestrial systems exposed to elevated levels
of acidifying deposition that show decreased plant health, reduced plant vigor, and/or loss
of terrestrial biodiversity. In multiple studies, consistent and coherent evidence has
shown that acidifying deposition can affect terrestrial ecosystems by causing direct
effects on plant foliage and indirect effects associated with changes in soil chemistry
(Section 3.2.2.3 of 2008 ISA). Biological indicators with reported responses to acidifying
deposition and conditions created by acidifying deposition include forest trees, shrubs,
lichens, grasslands, and Arctic and alpine tundra.
Acidifying deposition can affect terrestrial ecosystems via direct effects on plant foliage
and indirect effects associated with changes in soil chemistry (Figure 5-1). Biological
effects of acidification on terrestrial ecosystems are generally attributable to aluminum
(Al) toxicity and decreased ability of plant roots to take up base cations (especially
calcium [Ca]) and water from the soil (Cronan and Grigal. 1995). Acidifying deposition
to acid-sensitive soils can cause soil acidification, increased mobilization of Al from soil
5-1
-------
to drainage water, and depletion of the pool of exchangeable base cations in the soil (see
Appendix 4.3.4 and Appendix 4.3.5 for descriptions of these processes). Effects on the
soil and direct effects of acidifying deposition on foliage can influence the response of
plants to climatic stresses such as drought and cold temperature. The effects of acidifying
deposition can also influence the sensitivity of plants to other stresses, including insect
pests and disease (Joslin et al.. 1992).
Since publication of the 2008 ISA, the overarching understanding of terrestrial
acidification has not appreciably changed. More recent research has confirmed and
strengthened this understanding and provided more quantitative information, especially
across the regional landscape. This appendix highlights findings from the literature after
the completion of the 2008 ISA. A number of studies have evaluated the relationships
between soil chemistry indicators of acidification and ecosystem biological endpoints
(Table 5-6). Much of the new literature reviewed in Table 5-6 concerns natural variability
in soil pH, Ca concentrations and studies that have used elevated N and S have been
noted in that table under the "N and S Deposition/Additions" column. Soil chemistry
indicators examined in recent literature include exchangeable base cations, soil pH,
exchangeable acidity (hydrogen ions [H+] and Al), exchangeable Ca:Al ratio, base
saturation, and Al concentrations. Biological endpoints evaluated included physiological
responses of trees and other vegetation, lichens, soil biota, and fauna. Trees and other
vegetation included sugar maple (Acer saccharum), red spruce (Piceci rubens), yellow
birch (Betala ctlleghctniensis), American beech (Fagus grctndifolict), American basswood
(Tilla ctmericctna), black cherry (Primus serotina), eastern hophornbeam (Ostrva
virginiana), white ash (Fraxinus ctmericctna), hickories (Cctryct spp.), northern red oak
(Quercus rubra), and forest understory, grassland, and alpine plant species. Table 5-1
provides a summary of the soil chemistry indicator-biological endpoint relationships that
have been evaluated in the literature since the 2008 ISA. The 2008 ISA only considered
changes in vegetation in the analysis of how acid deposition affected terrestrial
ecosystems. More recent research has quantified effects on fauna (e.g., birds, snails) and
soil biota (Appendix 5.2.4; Appendix 5.2.5).
Together with the information available in the 2008 ISA, this body of evidence is
sufficient to infer a causal relationship between acidifying N and S deposition and
the alteration of the physiology and growth of terrestrial organisms and the
productivity of terrestrial ecosystems. Further, the body of evidence is sufficient to
infer a causal relationship between acidifying N and S deposition and the alteration
of species richness, community composition, and biodiversity in terrestrial
ecosystems.
5-2
-------
Transpiration
H20
FUNCTION
Membrane integrity
Stomatal regulation
Enzyme activation
Carbohydrate metabolism
Cold hardiness
Defense/chemical-physical
I
GROWTH
Cell division
Cell wall synthesis
Stress tolerance
I
STRUCTURE
Canopy integrity
Leaf form
Wood quality
Tree height
Root distribution
H'J
Deposition
Cytoplasm
Cell Wall
Cal
Leaching
Throughfall
. Root xylem and cortex
>tion
s In
ta
iCa, Al
VV
Soil
Solution
Ca, Al
Leaching
Physiological Processes
Ca Supply Rate
«
Biogeochemical Processes
Al = aluminum; Ca = calcium; H = hydrogen; H20 = water; S = sulfur.
Source: Fenn et al. (20061.
Figure 5-1 Diagram based on Fenn et al. (2006) showing indicators of forest
physiological function, growth, and structure that are linked to
biogeochemical cycles through processes that control rates of
calcium supply. Calcium affects plant physiological processes
that influence growth rates and the capacity of plants to resist
environmental stresses, such as extremes of temperature,
drought, insects, and diseases. Therefore, acidifying deposition,
which can deplete soil calcium or interfere with calcium uptake
through mobilization of soil aluminum, can affect forest health.
5-3
-------
Table 5-1 Relationships between soil chemistry indicators and biological endpoints that have been evaluated
in the literature since the 2008 Integrated Science Assessment.
Taxa
Exchangeable Base
Cations
PH
Exchangeable
Acidity
Exchangeable Ca:AI
Ratio
Base Saturation
Exchangeable Al or
Al Concentrations
Sugar maple
Beier et al. (2012);
Bilodeau-Gauthier et
al. (2011):
Long et al. (2009);
Page and Mitchell
(2008);
Sullivan et al. (2013):
Cleavitt et al. (2014);
Duchesne and
Ouimet (2009);
Pitel and Yanai
(2014)
Long et al. (2009);
McEathron et al.
(2013); Sullivan et
al. (2013);
Miller and
Watmough (2009)
Bilodeau-Gauthier
et al. (2011)
Bilodeau-Gauthier Bilodeau-Gauthier
et al. (2011);
Sullivan et al.
(2013)
et al. (2011)
Yellow birch
McEathron et al.
(2013)
American beech
Page and Mitchell
(2008);
Duchesne and
Ouimet (2009)
American basswood
Page and Mitchell
(2008);
Beier et al. (2012)
- - - -
Black cherry
-
Long et al. (2009)
Eastern hophornbeam
Beier et al. (2012)
-
Hickories
-
Elias et al. (2009)
5-4
-------
Table 5-1 (Continued): Relationships between soil chemistry indicators and biological endpoints that have been
evaluated in the literature since the 2008 Integrated Science Assessment.
Taxa
Exchangeable Base
Cations
PH
Exchangeable
Acidity
Exchangeable Ca:AI
Ratio
Exchangeable Al or
Base Saturation Al Concentrations
Northern red oak
-
-
-
-
Elias et al. (2009)
Forest understory plant
species
Horslev et al. (2008)
Horslev et al. (2008)
Horslev et al. (2008)
-
Horslev et al. (2008)
Grassland plant species
-
Pannek et al. (2015)
-
-
-
Lichens
-
Cleavitt et al.
(2011a)
-
-
-
Soil biota
Sridevi et al. (2012):
Ohta et al. (2014)
Chen et al. (2013):
Rousk et al. (2009):
Gilliam et al.
(2011b):
Sridevi et al. (2012)
Chen etal. (2013):
Rousk et al. (2009)
Fauna
Beier et al. (2012):
Pabian and
Brittinqham (2012)
Pabian and
Brittinqham (2012)
-
-
¦
Al = aluminum; Ca = calcium.
Note: Only soil chemistry indicators with a reported significant relationship (positive or negative) with a biological endpoint are indicated in this table. See Table 5-6 for a listing of soil
chemical indicators included in each study.
5-5
-------
5.2
Effects on Terrestrial Organisms and Ecosystems
5.2.1 Trees and Forests
Both coniferous and deciduous forests throughout the eastern U.S. have been
experiencing gradual losses of base cation nutrients from the soil due to accelerated
leaching from acidifying deposition. This change in base cation nutrient availability can
reduce the quality of forest nutrition over the long term in sensitive areas of low base
cation soils. Evidence suggests that red spruce and sugar maple in some areas in the
eastern U.S. have experienced declining health as a consequence of acidifying deposition.
The 2008 ISA reported several indicators of stress to forest trees (Table 3-3 in the 2008
ISA), including the percentage of dieback of canopy trees, dead tree basal area (as a
percentage), crown vigor index, and fine twig dieback. Biological effects of acidification
on terrestrial ecosystems were generally attributed to A1 toxicity, decreased ability of
plant roots to take up nutrient cations (due to leaching of base cations from soil and
interference with uptake), and elevated leaching of Ca2+ from conifer needles. The Ca:Al
ratio in soil solution is a chemical indicator of the negative impacts of soil acidification
on terrestrial vegetation (Cronan and Grigal. 1995). As tree species have shown similar
sensitivities to Ca:Al and the molar ratio of base cation (Bc):Al in soil solution (with Be
representing Ca2+, Mg2+, and K+), the Bc:Al ratio was used to represent the Ca: A1 ratio,
and it is the most commonly used indicator in the simple mass balance (SMB) model to
estimate critical acid loads in the European Union, Canada, and the U.S. (McNultv et al..
2007; Ouimet et al.. 2006; Spranger et al.. 2004).
Sverdrup and Warfvinge (1993). in a meta-analysis of laboratory, greenhouse, and field
studies, reported that the critical soil solution Bc:Al ratios for a large variety of tree
species ranged from 0.09 to 20.0, although the Be Al ratio range reported for tree species
native to North America was more restricted at 0.09 to 2.0. This range is similar to that
described by Cronan and Grigal (1995) for Ca:Al. In their meta-analysis of studies
examining sensitivities to the soil solution Ca: Al ratio, plant toxicity or nutrient
antagonism was reported to occur at Ca:Al ratios ranging from 0.2 to 2.5. The
meta-analyses conducted by Sverdrup and Warfvinge (1993) explored the relationships
between Bc:Al ratios in soil solution and the growth of different tree species. They
reported the Bc:Al ratios at which growth was reduced by 20% relative to controls. A
Bc:Al ratio of 1.0 is often applied to protect forested systems of Europe, particularly
conifers (Spranger et al.. 2004). and a Be Al ratio of 10.0 has been identified for forests
5-6
-------
in North America, to protect deciduous forests (McNultv et al.. 2007; Ouimet et al..
2006). The higher ratio (i.e., 10.0 vs. 1.0) provides a greater level of protection for a
wider range of species and various biotic and abiotic conditions.
However, when using soil solution Bc:Al or Ca:Al ratios as the chemical indicators or
criterion in the estimation of terrestrial acidification critical loads, several uncertainties
must be considered. Many of these considerations have been addressed in the reviews
conducted by Cronan and Grigal (1995) and Vanguelova et al. (2007). Although
recognized as one of the best indicators of soil acidity risk in forest ecosystems, soil
Ca:Al and Bc:Al ratios are variable and can change with soil conditions temporarily and
spatially. Estimates of soil base cations and Al concentrations are also influenced by
sampling and laboratory analysis methods, and there is still uncertainty regarding which
forms of Al are phytotoxic. Soil solution Al occurs in many different ionic and
complexed forms, depending on the soil pH and concentrations of soil ligands. In
addition, uncertainties regarding the sensitivity of the biological receptor should also be
considered. Critical Bc:Al and Ca:Al ratios are often based on seedling studies in
controlled environments, and the relationships are less consistent for trees growing in the
forest. Environmental and biological conditions, such as differences in tree age, soil
horizon chemistry experienced by the roots, and root mycorrhizae are important to
consider when comparing laboratory and field-based research studies. Recognizing these
different sources of variability, Cronan and Grigal (1995) recommended applying a ±50%
uncertainty to their critical soil Ca:Al ratio of 1.0. Critical fine root Ca:Al ratios have also
been suggested as indicators of stress in acidic forest soils (Vanguelova et al.. 2007;
Cronan and Grigal. 1995). Fine root Ca and Al indicate what is absorbed by the tree, and
there are often strong correlations between fine root and soil solution Ca:Al ratios.
Similarly, studies have shown relationships between fine root Ca:Al ratio and metrics of
root and above-ground tree health and productivity. However, similar to critical soil
solution Ca:Al and Bc:Al ratios, fine root Ca:Al ratios also have sources of uncertainty
that can be attributed to factors including soil water and chemistry variability, analysis
methods and timing, fine root age, and differences in tree physiology. In addition, U.S.
EPA is unaware of a method that uses fine root Ca:Al ratio as the chemical indicator
within SMB calculations of critical loads.
The tree species most commonly studied in North America and used in the 2008 ISA to
assess the impacts of acidification due to total nitrogen and sulfur deposition include red
spruce and sugar maple, although other tree species like flowering dogwood (Cornus
floridd) can also be sensitive to conditions created by acidifying deposition. New
information regarding the effects of acidifying deposition on the commonly studied sugar
maple and red spruce, as well as other tree, understory, and grassland species is
summarized below.
5-7
-------
5.2.1.1 Sugar Maple
Among broadleaf tree species in the northeastern U.S., sugar maple is the species most
commonly associated with adverse acidification-related effects resulting from S and N
deposition, although other base cation accumulating hardwoods may also be at risk
(Driscoll et al.. 2001b). Sugar maple is distributed from Missouri and Minnesota in the
northcentral U.S. eastward to Maine and the central Appalachian Mountain region, and
the species is a dominant component of the northern hardwood forest type (Braun. 2001).
Within this range, soil acidification is widespread throughout the northeastern U.S. and
within many portions of the Appalachian Mountains (Warbv et al.. 2009).
The 2008 ISA reported that acidifying deposition, in combination with other stressors, is
a likely contributor to the decline of sugar maple trees (Section 3.2.2.3 of 2008 ISA).
Sugar maple decline has been noted to occur in some portions of the eastern U.S., on base
cation-poor soils developed from parent material dominated by sandstone or other base
cation-poor substrates. Sugar maple appears to be particularly sensitive to reduced Ca and
magnesium (Mg) availability caused by acidifying deposition. Based on the results from
a compilation of laboratory studies, sugar maple growth can be reduced by approximately
20% at a Bc:Al soil solution ratio of 0.6 (Sverdrup and Warfvingc. 1993).
The more recent literature on sugar maple is consistent with the 2008 ISA. In these
studies, sugar maple was sensitive to soil conditions and chemistries associated with
acidifying deposition (Appendix 5.2.1.1.1). Soil chemical indicators that were evaluated
included exchangeable base cations, soil pH, exchangeable acidity (H+ and Al),
exchangeable CaAl ratios, base saturation, and Al concentrations (Al form unspecified).
Measured sugar maple responses included changes in basal area, growth, regeneration
success, and foliar nutrient concentrations and chemistry (Table 5-6). In addition, several
studies evaluated physiological mechanisms through Ca and Al addition that could
explain the response of sugar maple to changes in soil chemistry induced by acidifying
deposition (Table 5-2).
5-8
-------
Table 5-2 Summary of calcium addition studies in North America.
Reference
Region
Species
Additions
(Ca and/or Al)
Description
Battles et al.
New
Sugar maple,
Approximately
Ca additions resulted in the recovery of tree biomass increments, higher aboveground NPP,
(2014)
Hampshire
American
1,000 kg Ca/ha
and increased photosynthetic surface area. Sugar maple exhibited the largest cumulative
beech, yellow
(applied in 1999)
change in biomass, while American beech showed a negative cumulative change in
birch, red
biomass.
spruce, and
balsam fir
Bovce et al. New Red spruce Total of 38 g Ca/m2 Trends toward greater foliar Ca and Ca:AI ratios and lower Al concentrations across the
(2013) Hampshire, and balsam fir and/or 10.8 g Al/m2 treatment gradient. Ca availability appeared to enhance the ability of red spruce and balsam
Vermont fir to repair oxidative stress damage.
Cleavitt et
al. (2011b)
New
Hampshire
Sugar maple
Approximately
1,000 kg Ca/ha
(applied in 1999)
Masting events were not influenced by the Ca treatment. Seeds from the Ca treated sites
had lower concentrations of Al, K, and Mg and significantly higher concentrations of Si (but
no differences in Ca concentrations). Seeds from the Ca treated sites had higher
percentage of seedling emergence and higher seedling survival for the first 3 yr.
Halman et
al. (2008)
New
Hampshire
Red spruce
Approximately
1,000 kg Ca/ha
(applied in 1999)
Foliar Ca and total sugar concentrations were significantly higher in trees in the Ca addition
watershed. Cold tolerance of foliage was significantly greater in trees in the Ca addition
watershed.
Halman et
al. (2013)
New
Hampshire
Sugar maple
Total of 38 g Ca/m2
and/or 10.8 g Al/m2
Ca additions were found to increase C allocation to sugar maple growth. Al additions
increased root Al concentrations, root cell membrane disturbance, and ascorbate
peroxidase and glutathione reductase activity, and reduced foliar reflush following frost
injury and the number of viable seeds.
Halman et
al. (2015)
New
Hampshire
Sugar maple
and American
beech
Annual application of
CaCh (2.5 g/m2) and
AlCb (0.9 g/m2) from
1995-1999;
Wollastonite (38 g/m2)
was applied in 1999 a
single dose
The Al and Ca treatments did not affect beech foliar chemistry. Ca additions resulted in
significantly higher Ca concentrations in both dominant and nondominant sugar maple trees.
By 2008, the growth of American beech was higher than that of sugar maple on the control
plots and Al treated plots, and nondominant sugar maple growth was greater than that of
American beech and dominant sugar maple on the Ca treated plots. Increases in tree
mortality on the Al treated plots may have released surviving American beech and increased
their growth.
5-9
-------
Table 5-2 (Continued): Summary of calcium addition studies in North America.
Reference
Region
Species
Additions
(Ca and/or Al)
Description
Minocha et
al. (2010)
New
Hampshire
Sugar maple,
yellow birch,
and American
beech
Approximately
1,000 kg Ca/ha
(applied in 1999)
Foliar soluble Ca increased significantly in all species at mid and high elevations at Ca
supplemented watershed. Sensitivity to Ca limitation increased with elevation.
Moore et al.
(2012)
Quebec,
Canada
Sugar maple
8 dolomite lime
addition rates—0, 0.5,
1, 2, 5, 10, 20, and
50 megatonnes/ha
(applied in 1994)
Foliar Ca and Mg concentrations were found to be higher for treated trees relative to the
control trees. Mean crown dieback of trees decreased and seedling density increased with
the lime rate. The stem basal area increment for maple trees and proportion of the sugar
maple seedlings was increased with lime treatment.
Schabera et
al. (2011)
New
Hampshire
Red spruce
Approximately
1,000 kg Ca/ha
(applied in 1999)
Trees from the Ca addition watershed had higher estimated levels of foliar biomass. Ca
addition increased the stress tolerance of red spruce foliage during the cold season,
resulting in greater crown mass.
Smith et al.
(2009)
New York,
Vermont
Red spruce
Total of 160 kg Ca/yr
(1992-1995)
Greater amounts of Ca were found in the wood from the high-Ca location than from the
low-Ca location. Ca concentration was greater in wood formed in the 1970s than for wood
formed in the 1980s, and Ca treatments resulted in increased concentration of Ca in both
the 1970s and 1980s decadal bands of wood. Foliar concentrations of Ca oxalate were also
higher on the high-Ca site.
Sridevi et al.
(2012)
New
Hampshire
Soil microbes
(bacteria)
Approximately
1,000 kg Ca/ha
(applied in 1999)
Bacterial community structure in the Ca treated and nontreated reference soils was found to
be significantly different, with differences among communities being more pronounced in the
mineral soils.
Al = aluminum; AICI3 = aluminum chloride; C = carbon; Ca = calcium; CaCI2 = calcium chloride; g = gram; ha = hectare; K = potassium; kg = kilograms; m = meter; Mg = magnesium;
NPP = net primary production; Si = silicon; yr = year.
5-10
-------
5.2.1.1.1 Soil Chemical Indicators for Sugar Maple
Soil Exchangeable Base Cations
Sugar maple basal area and growth were responsive to soil exchangeable base cation
concentrations in several studies in the northeastern U.S. In the Adirondack Mountains of
New York, Beier et al. (2012) reported sugar maple basal area was positively correlated
to forest floor and mineral soil (B-horizon) exchangeable soil Ca on sites that ranged in
soil Ca from 1.83 to 53.89 cmolc/kg (in Oa-horizon) and 0.28 to 7.73 cmolc/kg in the
B-horizon (see Figure 3-2 in the 2008 ISA for a description of soil horizons). Similar
results were reported in Quebec, Canada by Bilodeau-Gauthier et al. (2011) who found
that sugar maple basal area growth was positively correlated to concentrations of base
cations (Ca, potassium [K], and Mg) in wood and mineral soil (B-horizon), and Long et
al. (2009) who reported that sugar maple basal area increment (BAI) was positively
correlated with exchangeable Ca and Mg in the upper B-horizon in the northeastern U.S.
Page and Mitchell (2008) also found that the relative basal area of sugar maple was
positively correlated with mineral soil (0-10 cm) exchangeable Ca in the Adirondack
Mountains. In the same region, Sullivan et al. (2013) determined that sugar maple canopy
vigor was positively correlated with soil exchangeable Ca and Mg, and mean growth
rates (measured as BAI) were positively correlated with exchangeable Ca at the
watershed level.
Relationships between soil exchangeable base cations and sugar maple regeneration (the
growth and abundance of seedlings and saplings) have also been demonstrated. Cleavitt
et al. (2014) evaluated sugar maple seedling survival and cause of death across 22 sites in
New Hampshire that varied in soil Ca and topographic position. Soil Ca concentration
exhibited a ninefold change across the study sites and was positively correlated to
increases in sugar maple abundance and initial seedling densities. However, soil Ca
concentration was not a significant predictor of lst-year mortality, nor was it a factor that
distinguished among the three main site types. In a study in Quebec, Canada, Duchesne
and Ouimet (2009) explored relationships between soil chemistry and sugar maple in the
sapling stratum and found that the basal area of sugar maple was positively correlated
with soil exchangeable Ca and Mg.
Pitel and Yanai (2014) evaluated the abiotic and biotic factors influencing the mortality
of dominant and codominant sugar maple trees in 47 stands in Massachusetts, Vermont,
and New York that had experienced defoliation by native forest tent caterpillars
(Malacosoma disstria) between 2002 and 2007. A total of 54 predictive variables
(defoliation year, stand and site characteristics, and soil chemistry variables) were
5-11
-------
evaluated through multiple linear regression followed by stepwise regression to produce
the best predictive models of sugar maple mortality from 2006 to 2008. Mortality was
highest in stands with the greatest amount of crown dieback the previous year, and
drought, cold winter temperatures, concave micro-relief, and soil base cation availability
were significant predictors of mortality. Concentrations of exchangeable Ca, Mg, and K
in the upper B soil horizon were inversely correlated with sugar maple mortality, with
exchangeable K showing the strongest relationship. Sites with above average sugar maple
mortality (>3 or 4%) were found to occur on soils with low concentrations of
exchangeable Ca (0.31-0.46 cmolc/kg), Mg (0.06-0.10 cmolc/kg), and K
(0.03-0.05 cmoL/kg). There was also an interaction between defoliation and soil base
cation availability: stands defoliated in 2005 that had low soil (A-horizon) Mg saturation
suffered higher rates of mortality.
Soil Exchangeable Acidity and pH
Bilodeau-Gauthier et al. (2011) found that sugar maple tree growth was negatively
correlated to soil exchangeable acidity (H+ and exchangeable Al), and through a
multifactor analysis, showed that tree age and soil exchangeable Al accounted for 51% of
the variation in sugar maple BAI. Positive correlations between mineral soil pH and sugar
maple basal area growth have also been reported by Long et al. (2009) and McEathron et
al. (2013). Sullivan et al. (2013) found that sugar maple canopy vigor was positively
correlated with soil pH.
One study reported a relationship between sugar maple foliar chemistry and
exchangeable soil chemistry. Miller and Watmough (2009). in their evaluation of
hardwood plots along air-pollution (N, S, and ozone), soil-acidity, and climate gradients
in Ontario, Canada, showed that foliar Ca and Mg content was positively correlated and
foliar manganese (Mn) content was negatively correlated with soil A-horizon pH.
Soil Base Saturation
An investigation of every relationship between sugar maple BAI and soil variables
revealed that base saturation was the best predictor of BAI (nonlinear) and explained
43% of variance (Bilodeau-Gauthier et al.. 2011). Similarly, Sullivan et al. (2013) found
that soil base saturation was related to sugar maple regeneration and growth. Plots with
lower soil base saturation did not have sugar maple regeneration; the proportion of sugar
maple seedlings dropped substantially at base saturation levels less than 20%
(Figure 5-2). Mean growth rates were positively correlated with soil base saturation at the
watershed level.
5-12
-------
c
o
r
o
Q.
O
O)
C
0
a>
CO
(/)
0.8 -
0.6 -
0.4
0.2 -
0
0
wi
CO I
o«
CMl
COl
CQ,
%
¦<9
5
o
20
40
60
80
100
BS in Upper B Horizon
BS = base saturation; SM = sugar maple.
Note: Plots were rank ordered based on soil base saturation, and a 5-plot rolling average was applied to both the soil base
saturation and the seedling proportion data. Reference lines are added at base saturation values of 12 and 20%, indicating break
points in the sugar maple seedling proportion-response function.
Source: Reprinted with permission from Sullivan, TJ; Lawrence, GB; Bailey, SW; McDonnell, TC; Beier, CM; Weathers, KC,
McPherson, GT; Bishop, DA (2013) Effects of acidic deposition and soli acidification on sugar maple trees in the Adirondack
Mountains, New York. Environmental Science and Technology 47(22): 12687-12694. Copyright (2020) American Chemical Society.
Figure 5-2 Relationship between the proportion of seedlings that were sugar
maple and soil base saturation in the upper B-horizon.
Soil Exchangeable Calcium:Aluminum Ratio
The relationship between sugar maple BAI and a soil exchangeable Ca:Al threshold of
<0.03 was evaluated by Long et al. (2009) through an analysis of a regional network of
76 forest sites that spanned Pennsylvania, New Hampshire, New York, and Vermont.
However, a repeated-measures analysis did not reveal statistically significant effects of
the Ca:Al soil ratio threshold of <0.03 on BAI.
5.2.1.1.2 Physiological Mechanisms for Sugar Maple
A number of studies were also conducted to determine the potential mechanisms
underlying the sensitivity of sugar maple to the soil conditions created by acidifying
deposition. These studies evaluated the influence of Ca additions on sugar maple
physiology, growth, seeds, and seedlings (Table 5-2). In Hubbard Brook Experimental
5-13
-------
Forest (HBEF), NH, one study involved the addition of a total of approximately 1,000 kg
Ca/ha as pelletized wollastonite (CaSiCh) in October 1999 to a watershed (Battles ct al..
2014). This treatment resulted in increases in: soil pH (in Oie-horizon) from 4.29 to 5.45;
base saturation from 9.6 to 78.6% (in Oie layer), 56% (in Oa-horizon), and 14.3% (in
upper mineral soil); and soil solution Ca:Al (inorganic monomeric Al) from 1.6 to 6.4 (in
Oa-horizon) and from 0.97 to 3.8 (in the mineral soil). As part of this study, Battles et al.
(2014) evaluated the effects of the Ca additions on tree growth and found that sugar
maple in the treated watershed exhibited the largest increase in cumulative biomass
relative to other species. Another study at the same experiment reported that sugar maple
seedlings were 50% larger in the treated watershed and mycorrhizal colonization of
seedlings was much higher in the treated watershed (22.47% of root length) as compared
with the reference watershed [4.4%; Juice et al. (2006)1. Mycorrhizal colonization also
increased in mature sugar maple in the treated watershed (56%) compared to the
reference watershed [35%; Juice et al. (2006)1. In contrast to sugar maple, American
beech showed a negative cumulative change in biomass. Minocha et al. (2010) examined
the effects of the Ca addition treatment on foliar chemistry and found that sugar maple
had increased foliar concentrations of Ca, which were accompanied by increases in
soluble phosphorous (P), chlorophyll, and two amino acids (i.e., glutamate and glycine).
The authors also reported decreases in known metabolic indicators of physiological stress
(i.e., arginine and y-aminobutyric acid, as well as putrescine, a diamine). Sugar maple
also exhibited a decrease in foliar putrescine at higher elevations in the watershed,
indicating possible remediation from Ca deficiency. Cleavitt et al. (201 lb) evaluated
sugar maple seed production, seed chemistry, seedling growth (lifestage), and seedling
survival on the Ca treated watershed and found that seeds from the Ca treated sites had
lower concentrations of Al, K, and Mg and significantly higher concentrations of silicon
(Si) than seeds from the reference watershed. The percentage of seedling emergence was
also higher, average lifestage was more significantly advanced, and seedling survival was
greater on the Ca treated watershed.
In a second study (NuPert Study) conducted at HBEF, Halman et al. (2013) evaluated the
impacts of Ca (380 kg/ha) and Al (108 kg/ha) additions on sugar maple physiology. The
Ca additions increased the proportion of carbon allocated to sugar maple growth. In
contrast, Al additions increased root Al concentrations and root cell membrane
disturbance, increased ascorbate peroxidase (APX) and glutathione reductase (GR)
activity, reduced foliar reflush following frost injury, and reduced the number of viable
seeds. Ascorbate peroxidase and GR are antioxidant enzymes that target reactive oxygen
species generated by environmental stresses, such as photo-oxidative stress damage at
low temperatures. Elevated APX and GR activity are indicators of increased oxidative
stress. Several years later, Halman et al. (2015) re-evaluated the NuPert study to
determine the long-term and contrasting responses of sugar maple and American beech to
5-14
-------
the Ca and A1 treatments following a major ice storm in 1998. The A1 and Ca treatments
did not affect beech foliar chemistry. However, Ca additions significantly increased Ca
concentrations in both dominant and nondominant sugar maple trees, relative to the trees
treated with Al. Interestingly, for all treatments, foliar A1 concentrations were higher in
the nondominant than dominant sugar maple trees, and the opposite was found for foliar
Ca concentrations. By 2008 (11 years after the ice storm), the growth of American beech
was higher than that of sugar maple on the control plots and Al treated plots, and
nondominant sugar maple growth was greater than that of American beech and dominant
sugar maple on the Ca treated plots. These differential growth responses emerged within
2 to 11 years following the ice storm, depending on the treatment and species. Although
plots were mainly composed of sugar maple, American beech experienced the greatest
growth on Al treated plots. Increases in overstory tree mortality on the Al treated plots
may have increased light availability, released surviving American beech, and increased
beech growth.
Moore et al. (2012) evaluated soil chemistry and sugar maple status 15 years after
treatment with dolomite lime (0 to 50 megatonnes/ha) to a hardwood forest in Quebec,
Canada, and found results similar to those in HBEF. Foliar Ca and Mg concentrations
were higher for treated trees relative to the control trees, and mean crown dieback of trees
decreased quadratically with the lime addition rate, from 39% for the control trees to a
value of 1 to 3% for the lime rates of 5 megatonnes/ha and higher. In addition, sugar
maple BAI for the limed trees was nearly double that of the nonlimed trees. The lime
application was also beneficial to the sugar maple regeneration. The overall sugar maple
seedling density increased with the lime rate, doubling in the 50 megatonnes/ha
(32 seedlings/m2) compared with the controls (16 seedlings/m2). The proportion of the
sugar maple seedlings to all of the other species increased quadratically from 22% in
controls to more than 55% in the 5 to 50 megatonnes/ha treatments. In contrast, the
proportion of competitive species decreased quadratically with the lime rate, including
American beech, for which the proportion in the treated plots (24%) was nearly half the
proportion observed in the controls (46%).
5.2.1.2 Red Spruce
Red spruce is a conifer that occurs mainly in the northeastern U.S. and at scattered
high-elevation sites (e.g., mountain and ridge tops) in the Appalachian Mountains. Red
spruce dieback or decline has been observed across high elevation landscapes of the
northeastern, and to a lesser extent, southeastern U.S. At high elevations in the
Adirondack Mountains in New York and the Green Mountains in Vermont, more than
50% of the canopy red spruce trees died during the 1970s and 1980s. In the White
5-15
-------
Mountains in New Hampshire, about 25% of the canopy spruce died during that same
period (DcHaves et al.. 1999). Dieback of red spruce has also been observed in mixed
hardwood-conifer stands at relatively low elevations in the western Adirondack
Mountains, an area that receives high inputs of acidifying deposition (Shortlc et al..
1997); acidifying deposition has been implicated as a causal factor (DcHaves et al..
1999).
The 2008 ISA reported that changes in soil chemistry (e.g., depletion of soil base cations,
Al toxicity to tree roots, leaching of base cations into drainage water; see Appendix 4.3.4
and Appendix 4.3.5) have contributed to high mortality rates and decreased growth of red
spruce trees in some areas of the eastern U.S. over the past three decades. Studies
evaluating the physiologic basis behind the responses of red spruce to acidifying
deposition have attributed the reduced vigor of the species to increased sensitivity to frost
injury and cold temperatures. The frequency of freezing injury to red spruce needles had
increased during a period in the latter half of the 20th century that coincided with
increased emissions of S and N oxides and increased acidifying deposition (DeHaves et
al.. 1999).
Since the 2008 ISA, research evaluating the sensitivity of red spruce to soil indicators of
acidifying deposition have mainly focused on the potential mechanisms underlying the
sensitivity of the species to acidic soil conditions. These studies have examined the
physiological response of red spruce to Ca additions (Table 5-2).
In the HBEF study that involved the addition of approximately 1,000 kg Ca/ha, Halman
et al. (2008) and Schaberg et al. (2011) evaluated the response of red spruce foliar
chemistry. Halman et al. (2008) reported significantly higher foliar Ca and total sugar
(fructose and glucose) in the Ca addition watershed than in trees in the reference
watershed. Foliar APX activity was also higher in trees in the Ca addition watershed
during winter. Cold tolerance of foliage was significantly greater in trees in the Ca
addition watershed than in trees in the reference watershed. Schaberg et al. (2011)
measured concentrations of foliar polyamines and free amino acids, foliar chlorophyll,
and sapwood area (as a proxy for foliar biomass) in Ca treated and nontreated trees.
Foliar polyamines (putrescine and spermidine) and free amino acids (alanine, gamma
aminobutyric acid [GABA]) are putative stress protection compounds that may directly
protect or provide other benefits to foliage that increase stress tolerance. The Ca additions
increased November concentrations of alanine, GABA, putrescine, and spermidine
relative to foliage from the reference watershed. Consistent with increased stress
protection indicated by the elevated polyamine and free amino acids, foliage from the Ca
addition watershed had higher total chlorophyll concentrations than foliage from the
reference watershed. In contrast, foliage from the reference watershed had significantly
5-16
-------
higher alanine:glutamic acid ratios, which have been attributed to cold sensitivity or
damage in other species. In addition to concentration-based differences in foliar
compounds, trees from the Ca addition watershed had higher estimated levels of foliar
biomass than trees from the reference watershed. These findings suggested that Ca
addition increased the stress tolerance and productive capacity of red spruce foliage
during the cold season and resulted in greater crown mass compared to trees growing on
nontreated soils.
Similar findings were reported by Bovce et al. (2013). who examined the influence of Ca
and Al on the physiology of red spruce and balsam fir (Abies balsamea) in three different
locations that varied in soil nutritional status. Processes known to be Ca sensitive (root
and foliar cation concentrations, chlorophyll fluorescence, soluble sugar concentrations,
and the activities of APX and GR in current-year foliage) were measured, and the results
from the study suggest that Ca availability enhanced the ability of red spruce and balsam
fir to repair oxidative stress damage, including photo-oxidation.
Concentrations of Ca in sapwood and accumulations of Ca oxalate in foliage have been
used as markers of environmental change due to acidic deposition or forest management
practices. Smith et al. (2009) compared the effects of different Ca fertilization treatments
(approximately 178 kg Ca/ha/yr during 1992-1995) on Ca concentrations in wood and
Ca and oxalate (Ox) concentrations in red spruce foliage at two locations with different
initial concentrations of Ca in the soil (6.4 cmolc/kg vs. 13.7 cmolc/kg). Greater amounts
of Ca were found in the wood from the high-Ca location than from the low-Ca location.
Foliar concentrations of Ca oxalate were higher on the high-Ca site than on the low-Ca
site and increased in response to Ca additions.
5.2.1.3 Other Tree Species
In the 2008 ISA, there was some information regarding the effects of acidification on
dogwood (Cornns) species. Loss of base cations, specifically Ca2+, had been implicated
in increased susceptibility of flowering dogwood to dogwood anthracnose, its most
destructive disease. Susceptibility to the disease and disease severity in stands appeared
dependent on several factors, including acid deposition and various edaphic
characteristics and meteorological conditions. Studies pointed to greater vulnerability of
dogwoods to anthracnose under simulated acid rain treatments (Britton et al.. 1996) and
under Ca2+ deficiency rHolzmueller et al. (2007); Section 3.2.2.3 of 2008 ISA],
Since the 2008 ISA, no further research on dogwood has been published. However, other
tree species have been evaluated in studies relating soil chemistry to tree physiology,
including red maple, white oak, yellow birch, white ash, American beech, black cherry,
5-17
-------
northern red oak, hickories, American basswood, and eastern hophornbeam. These
species were reported to vary in their sensitivities to soil conditions associated with
acidifying deposition (Table 5-6).
In a recent study in West Virginia, Thomas et al. (2013) used dendroisotopic techniques
to show the recovery of eastern redcedar (Juniperus virginiana) trees from decades of S
pollution using an analysis of a tree ring chronology from 1909 to 2008. Growth,
measured as BAI, in eastern redcedar old-growth stands (118-480 years old) increased
significantly since 1970. A multivariate correlation analysis using historical climate
variables, atmospheric CO2 concentrations, and U.S. SO2 and NOx emissions estimates
showed that the growth of cedar trees over the 100-year chronology is explained best by
increases in atmospheric CO2 and NOx emissions, and decreases in SO2 emissions.
Through carbon isotope (13C) analyses, the researchers were able to show that the stomata
of cedar may be more sensitive to SO2 emissions than to increasing atmospheric CO2. A
breakpoint in the 100-year S13C tree ring chronology occurred around 1980, as SO2
emissions declined, indicating a gradual increase in stomatal conductance and a
concurrent increase in photosynthesis related to decreasing SO2 emissions and increasing
atmospheric CO2. After 1990, calculated stomatal conductance increased more than
photosynthesis. These patterns in physiology led to changing trends in intrinsic water use
efficiency (/WUE; the ratio of photosynthesis to water loss through stomatal
conductance). The calculated /'WUE generally increased from the 1940s to 1990 mainly
because of SO2 emission effects on stomatal conductance, and then /WUE began to
decrease after 1990 because SO2 emissions no longer constrained stomatal function.
Through S isotope analysis (tree ring S34S), the study showed a synchronous change in
the sources of S used at the whole-tree level that indicated a reduced anthropogenic
influence. The increase in growth and the S13C and S34S trends in the tree ring chronology
of these Juniperus trees provide evidence for a distinct physiological response to changes
in atmospheric SO2 emissions since 1980. The authors attributed the changes since 1980
to an indirect effect of decreases in acid deposition. However, the exact mechanism is
unclear because the 100-year chronology could only be correlated to estimated SO2
emissions because acidifying deposition measurements were not available. Other
researchers have pointed out that the trees in the Thomas et al. (2013) study were
growing on a limestone outcrop that could be well buffered from soil acidification
(Schaberg et al.. 2014). Further, the rapid recovery of tree growth could also point to a
direct effect of gaseous SO2 rather than an indirect effect of soil acidification, which
would have a longer response time to decreases in emissions. See Appendix 3.2 for
further discussion of direct gaseous SO2 effects on vegetation.
5-18
-------
5.2.1.3.1
Soil Chemical Indicators for Other Tree Species
Soil Exchangeable Base Cations
Several studies reported on the relationships between species-specific basal area and
growth and soil exchangeable base cations. In the Adirondack Mountains, McEathron et
al. (2013) evaluated the relationships between basal area and soil chemistry and reported
that yellow birch basal area was positively correlated with mineral soil exchangeable Ca.
Similarly, Page and Mitchell (2008) compared the basal areas of American basswood,
American beech, and white ash with mineral soil (0-10 cm in depth) exchangeable Ca
concentrations and found exchangeable Ca had a positive correlation with relative basal
area of American basswood, a negative correlation with the relative basal area of
American beech, and no correlation with relative basal area of white ash. The positive
relationship between American basswood and soil exchangeable Ca is consistent with
other studies. Beier et al. (2012). in their evaluation of vegetation communities across
northern hardwood sites that ranged in exchangeable Ca concentrations from 1.83 to
53.89 cmoL/ha (in Oa-horizon) and 0.28 to 7.73 cmolc/ha (in B-horizon), reported that
American basswood and eastern hophornbeam were only found on the sites with the
highest exchangeable Ca concentrations. Similarly, negative relationships between
American beech and soil exchangeable Ca have been reported elsewhere. Duchesne and
Ouimet (2009) found that basal area of American beech in the sapling stratum in sugar
maple-dominated forests was negatively correlated with exchangeable Ca and Mg. Soil
concentrations of Ca and Mg on sites colonized by American beech were, on average,
roughly half those on sites where American beech was absent, thereby suggesting that
beech effectively colonizes sites with low base status. These findings are also consistent
with the responses of American beech to the Ca addition studies described earlier (Battles
et al.. 2014; Moore et al.. 2012).
Soil Base Saturation
One study of 30 forest plots within Monongahela National Forest, WV reported on the
relationship between soil base saturation and tree abundance. Elias et al. (2009) evaluated
relationships between tree growth parameters and soil indicators of acidifying deposition
and found that hickories were the only species (out of four within-species comparisons)
to have significantly lower numbers on sites with base saturation below 20% (A-horizon)
and 2.5% (B-horizon). They also found that the percentage of dead northern red oak was
highest on sites with Al concentrations (A-horizon) above 43 cmolc/kg. The authors
reported that sites in the year 2000 with subsurface base saturation above 10% had more
unique species.
5-19
-------
Soil Exchangeable Calcium:Aluminum Ratio
A study of 76 hardwood stands across the northeastern U.S. noted the relationship
between exchangeable Ca:Al ratios and BAI. Long et al. (2009) found that, unlike sugar
maple, black cherry BAI was greater in stands with exchangeable Ca:Al ratios
(B-horizon) below the <0.03 threshold adopted by the study.
5.2.2 Forest Understory and Grassland Species
The 2008 ISA did not specifically assess the effect of acidification on forest understory
species due to the lack of studies. Also, the 2008 ISA reported that grasslands were
thought to be less sensitive to acidification than woodlands (Kochv and Wilson. 2001;
Blake et al.. 1999). and grasslands with calcareous soils will be less sensitive than those
with acidic soils (Bobbink et al.. 1998).
Since the 2008 ISA, several studies have evaluated the relationships between soil
chemistry indicators of acidification and forest understory and grassland species
(Table 5-6).
5.2.2.1 Soil Chemical Indicators for Forest Understory and
Grassland Species
Soil pH, Exchangeable Base Cations, and Exchangeable Acidity
One study examined the relationship between soil chemistry indicators of acidification
and forest understory species. Horslev et al. (2008) evaluated 35 soil chemistry, stand,
and climatic variables as predictors of understory plant species composition in northern
hardwood stands, and found that a base cation-acid cation nutrient gradient accounted for
71.9% (in New Hampshire and Vermont) and 63% (in Pennsylvania and New York) of
the variation in the nonmetric multidimensional scaling ordination analyses of plant
community composition. Soil Ca, Mg, and pH formed the base end and Al, Mn, K, soil
acidity, and organic matter represented the acid end of the gradient. Based on results from
McNemars" exact test, a total of 50 of the 234 understory species were associated with
the base end of the base cation-acid cation nutrient gradient. These species have value as
indicators of sites at the high end of the base cation nutrient gradient in northern
hardwoods, sites that would be suitable for acid-sensitive species such as sugar maple.
In the U.K., Stevens et al. (2010b) used data from a national survey to evaluate the
species richness of 68 grasslands along an N deposition gradient. Ellenberg R
5-20
-------
(reaction-soil pH) and N (soil nutrient) scores and an index of soil acidity preference
were used as the metrics to characterize the plant diversity responses to the soil chemistry
conditions created by N deposition. Although the study did not find any significant
relationships between the Ellenberg and soil acidity index values and N deposition, there
was evidence that soil acidification was contributing to changes in species diversity and
community composition. Soil acidification may have led to decreased nutrient
availability and increased A1 solubility, thus, masking the effects of increased soil N
availability (see also Appendix 6.3.5 for N discussion).
A similar analysis along an N deposition gradient (2.4 to 43.5 kg N/ha/yr) was conducted
by Pannek et al. (2015) using a vegetation data set from 153 seminatural acidic
grasslands in northwestern Europe. Species frequency in response to N and other factors
including soil (0-10 cm) P, pH, NH44", and NO;, and geographical, climatic, and
management factors were evaluated. A second set of data from acidic grasslands in
Germany (392 plots) and the Netherlands (144 plots) containing plots from different time
periods were also included in the analyses to determine whether the results of the spatial
gradient approach coincided with temporal changes in the abundance of species. Out of
44 species included in the study, 16 were found to be affected by N deposition, with 12 of
them exhibiting a decreased abundance response. Increasing soil pH and P influenced 24
and 14 species, respectively, predominantly increasing abundance. Change of species
over time was unrelated to their responses to pH. However, species were found to
significantly decline over time in both Germany and the Netherlands in response to N
deposition, soil P, and NO3 . These results show the influence of N deposition in a
decline for many plant species in seminatural acidic grasslands, although impacts of N
deposition on pH did not appear to be the mechanism of plant responses (see also
Appendix 6.2.5 for N discussion).
5.2.3 Lichens
In the 2008 ISA it appeared that lichen populations were affected in areas with acidifying
deposition (Davies et al.. 2007). However, it was not clear whether effects were due to
direct effects of SO2, N effects, or acidifying deposition. See Appendix 6.3.7 for a
discussion of N deposition on lichens.
We have not identified any new studies published since the 2008 ISA that evaluated the
relationships between conditions created by acidifying deposition and lichen physiology.
However, in Acadia National Park, Cleavitt et al. (201 la) evaluated the interactions
among N and S deposition, tree type, and epiphytic lichen and bryophyte diversity,
biomass, and abundance to (1) document any differences in the depositional
5-21
-------
environments (throughfall and cloud water) of red spruce and red maple (Acer rub rum)
trees, (2) relate deposition differences to both chemistry of the bark and corresponding
epiphyte biomass and composition on the bark, and (3) describe any species-specific
relationships that emerged between the epiphytes and bark chemistry (Table 5-6).
Throughfall N and S deposition were higher under spruce trees than under maple trees,
and average S concentrations in cloud water were negatively correlated with bark Ca and
bark Mg. Epiphytic lichen richness was higher on maple trees than spruce trees, and tree
species differed in the number of rare epiphytic species. Several pollution-sensitive
epiphytes were restricted to the maples: the cyanolichens Leptogium cyctnescens
(bipartite) and Lobaria pulmonaria (tripartite), the large ruffle lichen Parmotrema
crinitum, and the moss Zvgodon viridissimus var. rupestris. Cyanolichens only occurred
on maple bark and did not occur on any bark with a pH below 5.02. Apparent overlap
between the bark chemistry of spruce and maple, particularly for samples from higher on
maple boles, suggest a reduction in the area of chemically suitable substratum for
epiphytes.
5.2.4 Soil Biota
Soil biota were not specifically addressed in the 2008 ISA. Since the 2008 ISA, several
studies have evaluated the relationships between soil chemistry indicators of acidification
and soil biota (Table 5-6). The effects of elevated N on soil biota is addressed throughout
Appendix 6 (see Appendix 6.2.3. Appendix 6.2.4. Appendix 6.2.5. and Appendix 6.3.5).
5.2.4.1 Soil Chemical Indicators for Soil Biota
Soil pH and Exchangeable Calcium:Aluminum Ratio
Bardhan et al. (2012) evaluated the relationships between soil chemistry and bacterial
diversity in 30 plots along a soil chemistry and deposition gradient in high-elevation
spruce-fir forests in the Great Smoky Mountain National Park (GSMNP). Modeled S
deposition ranged from 6 to 41 kg S/ha/yr, measured soil pH ranged from 3.0 to 4.6, and
measured CEC ranged from 1.3 to 23.1 cmolc/kg (soil samples pooled from the 0-, A-,
and B-horizons). However, bacterial diversity and community composition did not
change along the gradients of S deposition, soil pH, or exchangeable CaAl ratio.
5-22
-------
Soil pH and Aluminum Concentrations
Chen et al. (2013) evaluated the impacts of S additions on soil microbial and nematode
communities in a grassland system of Inner Mongolia. Seven treatments of 28 to 166 mol
H+/ha were applied in three doses in the form of sulfuric acid (H2SO4) from 2009 to 2010.
The authors found that the proportions of soil bacteria, fungi, and nematodes were altered
by the treatments, and the responses were related to soil pH and Al concentrations.
Increases in fungal fatty acids (49% increase) and fungi:bacteria ratio (120% increase)
and decreases in total and bacterial fatty acids (40-47% decrease) relative to the controls
were attributed to soil pH and Al concentrations. High Al concentrations (51 to
83 mg/kg) were associated with decreases in total fatty acids and bacterial fatty acids and
increases in fungal fatty acids. These results are consistent with a study of an agricultural
soil gradient that showed that at soil pH below 4.5, microbes (bacteria and fungi) are
decreased due to increased Al solubility (Rousk et al.. 2009). and that fungi are often
more dominant in acidic soils. Chen et al. (2013) also reported that acid additions also
impacted the soil nematode community by initially increasing the total number of soil
nematodes and then altering nematode community composition: bacterivorous and
fungivorous nematodes increased, but plant-feeding and omnivorous and carnivorous
nematodes decreased. The shifts in the nematode community were attributed to decreased
soil pH and changes in soil moisture.
Soil pH and Exchangeable Calcium
One study also investigated the influence of Ca addition (-1,000 kg Ca/ha/yr in 1999) on
the soil bacterial community in a northeastern hardwood forest (Sridevi et al.. 2012). The
study detected 1,756 taxa spanning 42 phyla, 53 classes, 127 orders, and 154 families in
the Ca treated and reference watersheds. Bacterial community structure was significantly
different between the Ca treated and nontreated reference soils, with differences among
communities being more pronounced in the mineral soils. Calcium additions resulted in
significant changes in bacterial community composition in the organic and in the mineral
soil horizons. The numbers of detectable taxa in families such as Acidobacteriaceae,
Comamonadaceae, and Pseudomonadaceae were lower in the Ca amended soils, while
Flavobacteriaceae and Geobacteraceae were higher. Analyses of relationships between
soil chemistry and the bacterial communities indicated that only exchangeable Ca, pH,
and P were significantly correlated with bacterial community structure.
5-23
-------
Soil pH
Gilliam et al. (201 lb) evaluated the microbial community across a soil weathering
gradient in a northern hardwood forest in West Virginia. Microbial community
composition, characterized through phospholipid fatty acid (PLFA) analysis, varied
among sites. Fungi were dominant at the most weathered, low NO, -production site,
while Gram-negative bacteria were significantly higher at the less weathered, moderate
and high NO, -production sites. Accordingly, the fungi:bacteria ratio increased in the
direction of the low NO, -production plots in ordination space. Correlations between the
soil parameters and PLFA results suggest that low soil pH and NO;, concentrations
supported fungal dominance, although other important factors including differences in
plant community and clay and organic matter content may have also influenced the soil
microbial community.
Soil Exchangeable Calcium
A study in Japan evaluated the relationship between forest type and soil invertebrate
communities in evergreen broad-leaved forests versus Japanese cedar (Cryptomerici
japonica) plantation forests (Ohta et al.. 2014). Exchangeable Ca was found to be
significantly higher in soil from the Japanese cedar plantations than the evergreen forest.
The invertebrate community composition also differed significantly between the two
forest types and was best explained by exchangeable Ca concentrations. Two major taxa
of soil crustaceans, Talitridae and Ligidium jctponiciim, were only found in the Japanese
cedar plantations. In contrast, millipedes (Paradoxosomatidae) and beetles were relatively
abundant in the evergreen plots.
5.2.5 Fauna
Fauna were not specifically assessed in the 2008 ISA. Since the 2008 ISA, two studies
were found that evaluated the relationships between soil chemistry indicators of
acidifying deposition and fauna (Table 5-6).
5.2.5.1 Soil Chemical Indicators for Fauna
Soil Exchangeable Calcium
Beier et al. (2012) characterized the variation in gastropod, salamander, and vegetation
communities among northern hardwood forests attributable to soil exchangeable Ca. The
5-24
-------
sites represented the variability in soil Ca in the Adirondack Mountains, ranging from
1.83 to 53.89 cmoL/kg (Oa-horizon) and 0.28 to 7.73 cmolc/kg (B-horizon). Snail
community richness and the abundance and live biomass of red-backed salamanders
(Plethodon cinerens) were all positively correlated with soil Ca. Land snail species
richness and abundance were positively correlated with Oa-horizon Ca and negatively
correlated to SO42 deposition and site elevation (and NO;, deposition for snail
abundance). Salamander communities were dominated by mountain dusky salamanders
(Desmognathns ochrophcteus) at Ca poor sites, with composition continuously shifting
toward dominance by red-backed salamanders as Ca availability increased. Several
known calciphilic species of snails (Pcircivitrea multidentata, Gcistrocoptci pentodon, and
Eucomdus polygyrus) were found only at the highest Ca sites. Some of the observations
(e.g., decreasing snail abundance) were confounded by a strong positive correlation
between elevation and estimated acid deposition. However, although the underlying
mechanisms require further study, these findings indicate that Ca availability, which is
shaped by geology and acidic deposition inputs, influences northern hardwood forest
ecosystems at multiple trophic levels.
Soil Exchangeable Calcium and pH
Pabian and Brittingham (2012) determined the relationships between soil (Oa-horizon)
chemistry and forest bird community composition, abundance, and diversity, and
evaluated potential mechanisms responsible for the relationships in oak and red
maple-dominated forests in Pennsylvania. Mean soil (Oa-horizon) exchangeable Ca for
the 14 forest sites ranged from 5.28 to 23.5 meq/100 g and pH ranged from 3.6 to 5.1.
Bird community composition (species richness and species abundances) varied with soil
Ca and pH, with 10 bird species having the highest abundances in forests with high-Ca
soils, and 5 species having the highest abundances with low-Ca soils. Five species were
classified as ""generalists" because they had high abundances and were present at all
forest sites. Bird species associated with low-Ca soils were associated with high densities
of mountain laurel (Kalmia latifolia) and five tree species whose basal areas were
explained by low soil pH and Ca. Bird species associated with high-Ca soils were
associated with high densities of saplings and high basal areas of acid-sensitive tree
species (17 species whose basal areas were, in part, explained by high Ca and soil pH).
All environmental and soil variables explained 37.8% of the variation in bird species
abundance data, with environmental variables explaining 36.0%, soils explaining 0.5%,
and 1.3% being explained by both. Most (67%) of the variation in the bird abundance
data explained by soils was also explained by the vegetation and invertebrate variables,
thereby supporting the hypothesis that the environmental variables were responsible for
the soil-bird relationships.
5-25
-------
5.3 Characteristics, Distribution, and Extent of Sensitive
Ecosystems
In the 2008 ISA, it was known that geology (particularly surficial geology) is the
principal factor governing the sensitivity of terrestrial and aquatic ecosystems to
acidification from S and N deposition. Geologic formations with low base cation supply
(e.g., sandstone, quartzite), due mainly to low weathering, generally underlie the
watersheds of acid-sensitive lakes and streams. Bedrock geology has been considered in
numerous acidification studies (Sullivan et al.. 2007b; Vertucci and Eilers. 1993; Stauffer
andWittchen. 1991; Stauffer. 1990; Bricker and Rice. 1989). Other factors contribute to
the sensitivity of soils and surface waters to acidifying deposition, including topography,
soil chemistry, land use, and hydrologic flowpath.
Forests of the Adirondack Mountains of New York, Green Mountains of Vermont, White
Mountains of New Hampshire, the Allegheny Plateau of Pennsylvania, and mountain top
and ridge forest ecosystems in the southern Appalachians are the region's most sensitive
to terrestrial acidification from atmospheric deposition (Section 3.2.4.2 of 2008 ISA).
Recent decreases in acid deposition had been linked to improvements in surface water
chemistry (Appendix 7). However, there remains widespread measurements of ongoing
depletion of exchangeable base cations in forest soils in the northeastern U.S.
(Appendix 4.3.4).
At the time of the 2008 ISA, there had been no systematic national survey to determine
the extent and distribution of terrestrial ecosystem sensitivity to the effects of acidifying
deposition. However, one preliminary national evaluation by McNultv et al. (2007) used
a simple mass balance model and available national databases to estimate forest soil
critical acidifying deposition loads (for wet and dry deposition of S and N) and
exceedances. They found that approximately 15% of forest soils in the U.S. receive
acidifying deposition that exceeds the estimated critical load of wet and dry deposition of
S and N by more than 250 eq/ha/yr (McNultv et al.. 2007). The areas where exceedances
reach this level could be considered to represent those areas that are likely most sensitive
to continued high levels of acidifying deposition.
Since the 2008 ISA, a series of studies have evaluated the characteristics, distribution,
and/or sensitivity of ecosystems to acidifying N and S deposition. For descriptions of
studies that characterized ecosystem sensitivity through critical load and critical load
exceedance determinations, see Appendix 5.5.
5-26
-------
5.4 Application of Terrestrial Acidification Models
The models that were used in recent studies to evaluate terrestrial acidification and its
components included the simple mass balance [SMB; Forsius et al. (2010); McNultv and
Boggs (2010); Nasr et al. (2010); Whitfield and Watmough (2012); Duarte et al. (2013);
Jung et al. (2013); Phelan et al. (2014)1. soil texture approximation [STA; Whitfield et al.
(2010b)l. MAGIC (Sullivan et al.. 201 la; Whitfield et al.. 2010a; Whitfield et al.. 2009).
ForSAFE-VEG (McDonnell et al.. 2014a; Sverdrup et al.. 2012). and empirical models
outlined by Spranger et al. (2004) and Krzvzanowski (2011). See Appendix 4.5 for a
review of models.
Phelan et al. (2014) applied the PROFILE model (see Appendix 4.5.1.1) to estimate BCw
to support SMB critical load estimates for 51 hardwood forest sites in Pennsylvania. The
rates of BCw ranged from 119 to 9,245 eq/ha/yr and were consistent with soil properties
and regional geology. Critical loads ranged from 4 to 10,503 eq/ha/yr. The PROFILE
model estimates by Phelan et al. (2014) were three times larger than those reported for
the same sites by McNultv et al. (2007) who used the clay correlation-substrate method
and SMB models to estimate BCw rates and critical loads, respectively. These PROFILE
model results suggest that the hardwood sites in Pennsylvania may not be as sensitive to
acidifying deposition as previously estimated by McNultv et al. (2007). It should be
noted that BCw rates were not measured for these areas and may be a result of the
different (empirical vs. mechanistic) approaches of each model. As the Phelan et al.
(2014) study only tested and compared the PROFILE model in Pennsylvania, the
researchers recommended applying PROFILE in different regions and ecosystems in the
U.S. to gain a better understanding of the model performance and the degree to which
BCw rates estimated by PROFILE differ from those estimated using the clay
correlation-substrate model.
5.5 Levels of Deposition at Which Effects Are Manifested
Since the 2008 ISA, several studies have evaluated the relationships between N and S
deposition and the growth and physiology of terrestrial organisms and ecosystem
function. In addition, numerous studies have used estimates of critical load exceedances
caused by historic, current, and future N and S deposition levels to characterize the
impacts of acidifying deposition. In this section, these relationships are described as
(1) impacts of elevated N and S deposition, (2) impacts of ambient levels ofN and S
deposition, and (3) critical loads and exceedances.
5-27
-------
5.5.1
Impacts of Elevated Nitrogen and Sulfur Deposition
A number of studies published since the 2008 ISA have evaluated the impacts of elevated
N and S deposition on trees, herbaceous species, and soil biota (Table 5-3). "Elevated" N
and S deposition in this section refers to additions of N and S above ambient atmospheric
deposition during the time of the specified study. As detailed in Appendix 6. N deposition
may lead to enrichment effects as well as acidification.
In Bear Brook Watershed, ME, Bethers et al. (2009) investigated the effects of
chronically elevated N and S deposition (i.e., 25.2 kg N/ha/yr and 28.8 kg S/ha/yr
additions since 1989) on growth, foliar nutrients, and photosynthetic capacity of sugar
maple saplings. Sugar maple saplings in the treated watershed had higher foliar Al
(+56%), N (+15%), P (+10%), and K (+15%) and lower foliar Ca (-25%) compared to
the nontreated watershed, presumably through influences on soil chemistry. The treated
saplings also had lower photosynthetic capacity, higher N:P ratios, negative correlations
between leaf N and electron transport capacity, and reduced carboxylation capacity,
which suggest nutrient imbalances induced by the elevated N and S deposition. However,
sapling growth was unaffected by the treatments. In another study conducted in Quebec,
Canada, Moore and Houle (2013) observed similar results in their evaluation of the
effects of 8 years of NH4NO3 additions (26 and 85 kg N/ha/yr applied from 2001-2008)
on sugar maple physiology and soil chemistry. Foliar Ca in the high N treatment
decreased by 79% compared to the control and reached 0.24%, the lowest foliar Ca
concentration ever reported for sugar maple. The treatments did not significantly alter
dieback rate or basal area growth. These results corresponded to changes in soil
chemistry; the treatments significantly decreased exchangeable Ca, Mg, Mn, and K in at
least one of the top organic soil layers. The largest differences were observed for
exchangeable Ca between the control and the high N treatment, with the L and the H
layers of the soil organic horizon showing exchangeable Ca decreases of 29 and 72%,
respectively. These results suggest that increased N deposition can strongly affect Ca
nutrition of sugar maple at sites with low base cation saturation; however, effects on tree
growth have not been documented.
In West Virginia, Jensen et al. (2014) examined long-term impacts of relatively high N
and S additions (22 years of 169 kg/ha/yr (NH^SO-O on black cherry and yellow poplar
(Liriodendron tulipifera) bole wood Ca, Mg, and Mn concentrations; tree growth; and
basal area. Bole wood Ca concentrations were lower and Mn concentrations higher in
both species on the treated versus nontreated watershed. Growth responses, measured
through relative growth rates of cored trees and changes in basal area, were not
conclusively affected by the treatment.
5-28
-------
Jung and Chang (2012) evaluated the impacts of 4 years of elevated N and S deposition
(30 kg N/ha/yr, 30 kg S/ha/yr, and 30 kg N/yr + 30 kg S/ha/yr additions from 2006 to
2009) on a forest composed of trembling aspen (Populus tremuloides), white spruce
(Piceci glauca), balsam fir, black spruce (Piceci mariana), and paper birch (Betiila
papyrifera) in the Athabasca oil sands region of Alberta, Canada. None of the treatments
influenced the growth of the understory plants or soil microbial biomass; understory and
microbial community composition were not measured. However, N increased tree growth
in the N and N + S treatments, indicating N limitation. Nitrogen and S additions
decreased soil exchangeable Ca2+ and Mg2+, and these decreases were attributed to a
combination of increased tree uptake to support greater growth and increased leaching
with SO42 .
Guv et al. (2013) conducted a greenhouse experiment to assess the potential effects of
acidic deposition on the root system morphology of three endemic species that grow on
sand dunes in the Athabasca oil sands region of Alberta, Canada: Armeria maritima,
Deschctmpsia mackenzieana, and Stellaria arenicola. Plants were exposed to three pH
treatments (pH 5.6, 5.0, and 4.2) together with additions of 2.61 to 4.67 mg/L of SO42
and 1.95 to 3.51 mg/L NO;, for 55-60 days. There were no statistically significant
differences in plant survival, root length, root surface area, or root tip numbers between
acid treatments. Therefore, current rates of acidifying N and S deposition are not likely a
threat to these species.
Since the 2008 ISA, three studies were identified on the impacts of N and/or S deposition
on soil biota. Payne et al. (2010) examined the impacts of elevated SO42 deposition on
the microbial community in a Scottish peatland and showed that additions of 95 kg
SO42 /ha/yr applied over 18 months decreased the concentrations and percentages of live
amoebae. Abundances of Trinemct line are, Corythion dubium, and Euglypha rotunda
were also significantly reduced. These results suggest the potential importance of SO42
deposition in influencing testate amoebae communities in the peatland soils. In another
study, Chen et al. (2013) evaluated the impacts of S additions on soil microbial and
nematode community composition in an Inner Mongolian grassland in China. Seven
treatments of 28 to 166 mol H+/ha were applied in three doses in the form of sulfuric acid
(H2SO4) from 2009 to 2010. Fungal fatty acids were increased by 49% and fungi:bacteria
ratio increased by up to 120% by the acid additions. Total and bacterial fatty acids were
decreased by the S treatments by up to 47 and 40%, respectively. Acid additions also
impacted the soil nematode community by initially increasing the total number of soil
nematodes, then altering community composition; bacterivorous and fungivorous
nematodes increased, while plant-feeding and omnivorous and carnivorous nematodes
decreased. Hu et al. (2013) examined the impacts of N and S additions on soil microbial
biomass and function in a boreal mixed-wood forest in the Athabasca oil sands region of
5-29
-------
Alberta, Canada. Five years of N and S additions (30 kg N/ha/yr and/or 30 kg S/ha/yr)
did not influence soil microbial biomass C and N. However, activities of some
extracellular enzymes in the soil were decreased by the treatments, with greater
(3-glucosidase activity in the N + S treatment than in the S treatment and decreased soil
arylsulfatase activity in the S addition treatment. Thus, the additions of N and/or S
strongly affected soil microbial community functions and enzymatic activities without
changing soil microbial biomass in this boreal forest.
5-30
-------
Table 5-3 Impacts of acidifying nitrogen and sulfur deposition.
Reference
Ecosystem Type/Region
Species
Nitrogen and Sulfur
Deposition/Additions
Description
Bardhan et al. (2012)
Forest
Great Smoky Mountain
National Park, TN, NC
Soil microbes (bacteria)
6 to 41 kg S/ha/yr (modeled)
Soil analyses indicated only minor
differences in bacterial diversity among
sites; the bacterial community did not
change along the gradients of S
deposition.
Bethers et al. (2009)
Forest
Bear Brook watershed,
ME
Sugar maple
Ambient: 8.4 kg N/ha/yr and 14.4 kg
S/ha/yr; elevated: 25.2 kg N/ha/yr
and 28.8 kg S/ha/yr
Treated watershed had a 56% increase in
foliar Al, a 25% reduction in foliar Ca, N
(+15%), P (+10%), and K (+15%). The
treated saplings had lower photosynthetic
capacity, high N:P ratios, and negative
correlations between leaf N and electron
transport capacity, which may indicate
nutrient imbalance.
Boot et al. (2016)
Loch Vale watershed in
Rocky Mountain National
Park, CO
Soil microbes
17 yr of 25 kg N/ha/yr (as NH4NO3)
addition
Long-term fertilization resulted in increased
soil acidity and reduced soil C. Soil
microbial biomass in the fertilized soils was
also lower (22%), the microbial community
was altered through reductions in vesicular
arbuscular mycorrhizae and saprotrophic
fungi, and activity of N degrading microbial
enzymes was decreased.
Chen et al. (2013)
Grassland
Mongolia, China
Soil microbes, bacteria,
fungi, and nematodes
Sulfuric acid (0, 2.76, 5.52, 8.28,
11.04, 13.8, and 16.56 mol H+/m2)
Fungal fatty acids were increased by 49%
and fungi:bacteria ratio increased by up to
120% by the acid additions relative to the
controls. Total and bacterial fatty acids
were decreased by the S treatments by up
to 47 and 40%, respectively. In the
treatments, bacterivorous and fungivorous
nematodes increased, while herbivorous
and omnivorous + carnivorous nematodes
decreased.
5-31
-------
Table 5-3 (Continued): Impacts of acidifying nitrogen and sulfur deposition.
Reference
Ecosystem Type/Region
Species
Nitrogen and Sulfur
Deposition/Additions
Description
Cleavitt et al. (2011a)
Forest
Acadia National Park, ME
Epiphytic lichens
12 to 18 kg S/ha/yr
Throughfall chemistry influenced bark pH
and the suitability of tree boles as habitat
for lichen; epiphytic lichen species richness
and presence of pollution-sensitive
epiphytes were greater on red maple trees,
which have a higher pH in the bark relative
to red spruce trees.
Dietze and Moorcroft
(2011)
Forest
Eastern and central
portions of the U.S.
267 tree species
organized into 10 PFTs
6 to 16 kg/ha/yr as NO3"; 4 to
30 kg/ha/yr as S042" (NADP wet)
Mean sensitivity by covariate and PFTs
showed that overall tree mortality was
most sensitive to atmospheric pollutants,
with acid deposition (SO42") showing the
highest sensitivity and N deposition the
third highest sensitivity. Stand DBH
showed the second highest sensitivity.
Duarte et al. (2013)
Forest
Northeastern U.S.
21 tree species
256 to 920 eq/ha/yr of N; 242 to
1,154 eq/ha/yr of S (modeled)
Statistically significant negative
correlations between critical load
exceedance and growth (17 species), and
crown density (4 species) were
determined. Positive correlations between
critical load exceedance and declining
vigor (4 species), crown dieback
(4 species), and crown transparency
(7 species) were determined. Species that
were most negatively affected by N and S
deposition included balsam fir, red spruce,
quaking aspen, and paper birch.
Jensen et al. (2014)
Forest
Fernow Experimental
Forest, WV
Yellow poplar and black
cherry
35.5 kg N/ha/yr; 40.5 kg S/ha/yr as
(NH4)2S04 (addition)
Bole wood Ca concentrations were lower
and Mn concentrations were higher in both
species on the treated watershed. Growth
responses were not conclusive and
appeared to differ by species.
5-32
-------
Table 5-3 (Continued): Impacts of acidifying nitrogen and sulfur deposition.
Reference
Ecosystem Type/Region
Species
Nitrogen and Sulfur
Deposition/Additions
Description
Jung and Chang (2012) Forest
Alberta, Canada
Trembling aspen, white
spruce, balsam fir,
black spruce, and
paper birch
30 kg N/ha/yr, 30 kg S/ha/yr, and
30 kg N/yr + 30 kg S/ha/yr (additions)
Nitrogen addition increased tree growth in
the N and N + S treatments. None of the
treatments affected understory growth or
soil microbial biomass. Annual leaching
losses of SO42" were increased by S and
S + N additions. Leaching of base cations
showed a similar trend to SO42" leaching.
Miller and Watmough
(2009)
Forest
Southern Ontario,
Canada
Sugar maple
9 to 12.8 kg N/ha/yr; 7.6 to 14.8 kg
S/ha/yr (ambient)
Sugar maple foliar S and N contents were
positively correlated with modeled N and S
deposition. Foliose lichen species richness
was negatively correlated with modeled air
pollution levels (S deposition, N deposition,
and atmospheric ozone).
Pannek et al. (2015)
Acidic grassland
Atlantic biogeographic
region of Europe
44 grassland species
within the Violion
caninae alliance
2.4 to 43.5 kg N/ha/yr (ambient)
Out of 44 species studied, 16 were
affected by N deposition, with 12 of them
exhibiting a negative response. Increasing
soil pH and P influenced 24 and
14 species, respectively, predominantly
positively. Species that were negatively
affected by high N deposition and/or high
soil P also showed a negative temporal
trend characterized by short stature and
slow growth.
Pavne et al. (2010)
Peatland
Scotland
Soil microbes (testate
amoebae)
Ambient: 5 kg SO42 /ha/yr; elevated:
95 kg S042"/ha/yr
Analysis showed that the SO42" treatment
reduced the concentrations and
percentages of live amoebae, suggesting a
less active community as a result of the
treatment. In addition, abundances of
Trinema lineare, Corythion dubium, and
Euglypha rotunda were significantly
reduced.
5-33
-------
Table 5-3 (Continued): Impacts of acidifying nitrogen and sulfur deposition.
Reference
Ecosystem Type/Region
Species
Nitrogen and Sulfur
Deposition/Additions
Description
Quesnel and Cote
(2009)
Forest
Boreal region of Canada
White spruce
NA
Meta-analysis of foliar results (nutrient
concentrations) from 23 white spruce
forests (natural and nonfertilized plantation
forests) revealed that N deficiencies
occurred in less than 10% of the sites.
Base cation deficiencies identified were
attributed to removals with harvest, higher
requirements of white spruces for K and
Ca, and nutrient deficiencies/imbalances
induced by N additions/deposition.
Rose etal. (2016)
Open uplands, open
lowlands, and woodlands
U.K.
Vascular plants
NA
Changes in vegetation from 1993-2012 at
a network of plots (504) within
12 Environmental Change Network (ECN)
sites. Significant increases in species
richness were detected at the network
level. Increases in acid-sensitive species
and comparatively little change in
acid-tolerant species were noted. Changes
are consistent with increases in pH
observed and attributed to the large
reductions in acid deposition. Increases in
species diversity were also attributed to
wetter summers and a reduction in soil N
availability at some of the upland locations.
Soule (2011)
Forest
Grandfather Mountain,
NC
Red spruce
NA
Radial growth rates of red spruce
increased through time, and growth rates
were significantly correlated to temperature
(positively), days with precipitation
(negatively), atmospheric CO2 (positively),
and emissions of SOx and NOx
(negatively).
Sullivan etal. (2013)
Forest/Adirondack
Mountains, NY
Sugar maple
750 to 1,120 eq/ha/yr as N + S
(NADP wet, CASTNET dry)
Found that plots without sugar maple
seedlings had higher rates of atmospheric
N + S deposition.
5-34
-------
Table 5-3 (Continued): Impacts of acidifying nitrogen and sulfur deposition.
Nitrogen and Sulfur
Reference Ecosystem Type/Region Species Deposition/Additions Description
Thomas et al. (2013) Forest Eastern redcedar NA Dendroisotopic techniques showed the
Appalachian Mountains, recovery of eastern redcedar trees from
yyy decades of S pollution. Analysis provided
evidence for a distinct physiological
response to changes in atmospheric SO2
emissions since 1980.
Annual mean and cumulative N deposition
were strongly correlated with decreases in
lichen species richness, decreases in N
sensitive species, and poorer thallus
condition. Cumulative dry deposition of S
had the best fit to decreases in thallus
condition, poorer community-based S
index values, and the absence of many S
sensitive species.
Al = aluminum; C = carbon; Ca = calcium; CASTNET = Clean Air Status and Trends Network; C02 = carbon dioxide; DBH = diameter at breast height; eq = equivalents;
H+ = hydrogen ion; ha = hectare; K = potassium; kg = kilogram; L = liter; m = meter; mg = milligram; MN = manganese; mol = mole; N = nitrogen; NA = not applicable;
Na2S04 = sodium sulfate; NADP = National Acid Deposition Program; NH4NO3 = ammonium nitrate; (NH4)2S04 = ammonium sulfate; N03" = nitrate; NOx = the sum of nitric oxide and
nitrogen dioxide; P = phosphorus; PFT = plant functional types; S = sulfur; S02 = sulfur dioxide; S042" = sulfate; SOx = sulfur oxides; yr = year.
Cleavitt et al. (2015) Forest Lichens 3 to 8 kg N/ha/yr and 4.5 to 5.2 kg
VT, NH, ME S/ha/yr
5-35
-------
5.5.2
Impacts of Ambient Deposition
Five studies that were published since the 2008 ISA have been identified that
documented the impacts of ambient N and S deposition on terrestrial biological
endpoints, including tree species, tree species grouped into plant functional types (PFTs),
lichens, and soil biota (Table 5-3). "Ambient" N and S deposition in this section refers to
the levels of atmospheric N and S deposition experienced during the time of the specified
study.
Dietze and Moorcroft (2011) conducted a large-scale analysis of central and eastern U.S.
that investigated 13 covariates (including climate, air pollutants [N deposition, acid
deposition, and ozone averaged 1994-2005], topography, and stand characteristics) as
predictors of individual tree mortality. Tree species (267 species) were grouped into
10 PFTs based on hardwoods/softwoods, latitude, successional phase (early, mid, and late
successional), and hydrological soil status (i.e., hydric). The researchers found that tree
mortality was most sensitive to stand characteristics and air pollutants. Nine PFTs had
decreased mortality with increased N deposition, and only the northern mid successional
hardwoods showing the opposite pattern (i.e., increased mortality with increased N
deposition). Seven PFTs showed large increases in mortality with acid deposition (as
SO42 ). with only late successional conifers showing the inverse (i.e., a weak decline in
mortality with increased acid deposition). Mean sensitivity by covariate and PFT showed
that overall tree mortality was most sensitive to atmospheric pollutants, with acid
deposition (SO42 ) showing the highest sensitivity and N deposition the third highest
sensitivity. Individual tree DBH showed the second highest sensitivity.
Duarte et al. (2013) evaluated the responses of tree species in 4,057 forest plots in the
northeastern U.S. to the exceedance of critical loads of N and S deposition. Modeled S
and N deposition ranged from 242 to 1,154 eq/ha/yr and 256 to 920 eq/ha/yr,
respectively. Critical loads were exceeded in 45% (calculated using midpoint weathering
rates) of the plots. See Appendix 5.5.3 for a greater description of critical loads and
exceedances from this study. Results from Spearman's rank correlation analyses showed
that the growth of 17 species and crown density of 4 species were negatively correlated
with critical load exceedance (Table 5-4). Positive correlations between critical load
exceedance and declining vigor (three species), crown dieback (four species), and crown
transparency (seven species) were also found (Table 5-5). Crown dieback was considered
the most reliable indicator of forest health. Based on this metric, balsam fir, red spruce,
quaking aspen, red maple, and paper birch were identified as the species most negatively
impacted by N and S deposition.
5-36
-------
Table 5-4 Results of Spearman's rank correlation analysis comparing growth
versus critical load exceedance by species for the forested plots in
the northeastern U.S. [from Duarte et al. (2013)1. Modeled sulfur and
nitrogen deposition on the plots ranged from 242 to 1,154 eq/ha/yr
and 256 to 920 eq/ha/yr, respectively. Correlations shown here are
significant at a = 0.05.
Species
Correlation Coefficient
Sample Size (n)
Black spruce (Picea mariana)
-0.44
41
Balsam fir (Abies balsamea)
-0.33
499
Chestnut oak (Quercus prinus)
-0.28
238
Paper birch (Betula papyrifera)
-0.27
574
Scarlet oak (Quercus coccinea)
-0.26
170
Bigtooth aspen (Populus grandidentata)
-0.25
280
Eastern white pine (Pinus alba)
-0.19
1,464
Black oak (Quercus velutina)
-0.18
257
White ash (Fraxinus americana)
-0.18
1,256
Sweet birch (Betula lenta)
-0.12
460
Yellow birch (Betula alleghaniensis)
-0.12
669
Northern red oak (Quercus rubra)
-0.11
1,533
Red maple (Acer rubrum)
-0.11
3,861
Red spruce (Picea rubens)
-0.09
616
American beech (Fagus grandifolia)
-0.08
1,449
Black cherry (Prunus serotina)
-0.07
1,026
Eastern hemlock (Tsuga canadensis)
-0.07
1,055
Pignut hickory (Carya glabra)
0.44
31
Norway spruce (Picea abies)
0.38
148
Yellow poplar (Liriodendron tulipifera)
0.35
47
White spruce (Picea glauca)
0.24
97
eq = equivalents; ha = hectare; yr = year.
Source: Duarte et al. (20131.
5-37
-------
Table 5-5 Results of Spearman's rank correlation analyses comparing tree vigor,
crown density and dieback, and canopy transparency versus critical load
exceedance by species for the forested plots in the northeastern U.S.
[from Duarte et al. (2013)1. Modeled sulfur and nitrogen deposition on the
plots ranged from 242 to 1,154 eq/ha/yr and 256 to 920 eq/ha/yr,
respectively. Correlations shown here are significant at a = 0.05.
Vigor
Crown Density
Crown Dieback
Canopy T ransparency
Species
Correlation
Coefficient
Sample
Size (n)
Correlation Sample
Coefficient Size (n)
Correlation Sample
Coefficient Size (n)
Correlation
Coefficient
Sample
Size (n)
American
beech
0.13
433
0.12 502
0.30
502
Ash
0.22
189
Balsam
fir
-0.23 161
0.56 161
0.18
161
Black
cherry
0.33
47
Eastern
hemlock
0.36 82
-0.24 148
Fir
-0.36
33
Northern
red oak
0.27
208
Paper
birch
-0.29 125
0.13 300
Quaking
aspen
-0.56 61
0.44 61
0.40
61
Red
maple
-0.16 337
0.10 767
0.19
767
Red
spruce
0.20 115
0.43 168
0.61
168
Sugar
maple
0.06
3,408
-0.04 3,466
0.12
3,467
White
ash
-0.32
49
eq = equivalents; ha = hectare; yr = year.
Source: Duarte et al. (20131.
5-38
-------
Sullivan et al. (2013) evaluated relationships between deposition and sugar maple
regeneration in the Adirondack Mountains. Total N and S deposition ranged between 750
and 1,120 eq/ha/yr, and plots that received higher rates of N (median of approximately
560 eq/ha/yr), S (median of approximately 480 eq/ha/yr), and N + S (median of
approximately 1,040 eq/ha/yr) deposition had no sugar maple seedlings.
Soule (2011) sampled red spruce trees within a high elevation red spruce-Fraser fir (Abies
fraseri) forest on Grandfather Mountain, NC between 2006 and 2008, and used
dendrochronology to evaluate the influences of climate (precipitation, temperature, and
CO2 concentrations) and emissions of SOx and NOx on growth. Radial growth rates of
red spruce increased through time and were positively correlated to temperature and
atmospheric CO2 but were negatively correlated to emissions of SOx and NOx and
number of days with precipitation in a year.
Miller and Watmough (2009) evaluated hardwood plots along air pollution (N, S, and
ozone), soil acidity, and climate gradients in Ontario, Canada, and found that foliose
lichen species richness was negatively correlated with modeled air pollution levels (S
deposition, N deposition, and atmospheric ozone). In the same study, however, no
relationship was seen in canopy condition and ground vegetation richness and diversity.
Bardhan et al. (2012) evaluated bacterial diversity along a soil and S deposition gradient
in high elevation spruce-fir forests in GSMNP. Modeled S deposition ranged from 6 to
41 kg S/ha/yr. However, neither bacterial diversity nor community composition changed
along the gradient of S deposition. The bacterial community on all sites was dominated
by members of the phyla Actinobacteria, Acidobacteria, Planctomycetes, Proteobacteria,
and Chloroflexi. Species from these phyla are often found in highly acidic environments
such as acid-mine drainage and sphagnum bogs. Therefore, these analyses suggested that
despite reductions in acid deposition, the soil conditions in these GSMNP sites were still
acidic and had not yet reached a threshold suitable for nonacidophilic bacterial
communities.
Pannek et al. (2015) used a data set from seminatural grasslands found on acidic soils
along an N deposition gradient (2.4 to 43.5 kg N/ha/yr) in northwestern Europe to
examine the response of species frequency to N and other factors including soil
(0-10 cm) P, pH, NH4+, and NO;, and geographical, climatic, and management factors.
Out of 44 studied species, 16 were affected by N deposition, with 12 of them exhibiting a
negative response (see also Appendix 6.2.5 and Appendix 6.3.5).
5-39
-------
5.5.3
Critical Loads and Exceedances
Since the 2008 ISA, critical load evaluations have been conducted in the northeastern
forests of the U.S., hardwood forests in Pennsylvania, forests in western Canada, and
arctic, subarctic, alpine, subalpine, and boreal systems. Studies also simulated target
loads and critical loads and exceedances into the future under different deposition and
climate scenarios and evaluated the sensitivity of critical load estimates to varying
environmental conditions and forest management and uncertainties in critical load
estimates. See Appendix 4.6.2.1 for more information on critical loads based on soil
acidification and Appendix 6.5 for information on N critical loads in terrestrial systems.
Duarte et al. (2013) calculated critical loads of N and S for 4,057 forested plots in the
northeastern U.S. using the steady-state SMB model (see also Appendix 4.5.1.2) and soil
solution Be:A1 of 10.0 as the chemical indicator and threshold. In addition, the study
evaluated the influence of three different BCw rates on critical loads: minimum,
midpoint, and maximum. These BCw rates were determined using the clay
correlation-substrate method and the range of U.S. Department of Agriculture (USDA)
Natural Resources Conservation Service (NRCS) Soil Survey Geographic Database
(SSURGO) values for each soil series. Critical loads using the midpoint weathering rates
were found to range from 11 to 6,540 eq/ha/yr (over 80% within the range of
850-2,050 eq/ha/yr), and in comparisons with deposition estimated using the ClimCalc
model, were exceeded in 98% (calculated using minimum weathering rates), 45%
(calculated using midpoint weathering rates), and 15% (calculated using maximum
weathering rates) of plots. Similarly, Phelan et al. (2014) calculated critical loads of N
and S deposition for 51 hardwood forests in Pennsylvania using the SMB model, soil
solution Be:A1 of 10.0 as the chemical indicator, and the PROFILE model to estimate
BCw rates. They found that critical loads ranged from 4 to 10,503 eq/ha/yr, and that
critical loads at 53% of the sites were exceeded by the 2002 N and S deposition.
Krzvzanowski (2011) modeled deposition and soil acidification critical loads in
northeastern British Columbia, Canada using empirical methods described in Sprangcr et
al. (2004). Slow weathering of shale- and mudstone-derived feldspars, micas, and quartz
placed the soils of the study area in the "high-sensitivity" class of critical loads. Soil
acidification critical load was estimated to be 200 eq/ha/yr. Neither S nor N deposition
exceeded this critical load. The combined S and N deposition at the dominant point of
reception (i.e., near Blueberry River First Nation) was estimated to be 73.5 eq/ha/yr.
Estimates of S and N critical loads and exceedances for upland soils were generated for
the Georgia Basin in British Columbia, Canada using the SMB model and zero base
cation depletion as the chemical criterion and threshold (Nasr et al.. 2010). The objective
of the study was to evaluate critical load and exceedance calculations in the context of
5-40
-------
sustainable forest and soil management policies by using "no further changes in soil base
saturation" as the critical load criterion. Critical loads were found to range between 140
and 4,000 eq/ha/yr and were generally lowest on ridge tops and increased towards the
valleys. Critical load exceedance ranged from 13% of the basin based on wet deposition
fluxes to 32% under modeled total (wet and dry) deposition. With continued N and S
deposition, significant portions of the basin were predicted to experience
exceedance-enhanced base cation depletion rates greater than 200 eq/ha/yr.
Forsius et al. (2010) determined critical loads and exceedances for terrestrial ecosystems
in the arctic and subarctic regions (with latitudes north of 60°) for 1990, 2000, 2010, and
2020 using the SMB model and two different chemical indicators (soil solution Al:Bc of
1.0 and ANCle of 0.0). The Al:Bc indicator was hypothesized to protect against fine root
damage, and the ANCle indicator was hypothesized to preserve existing soil base cation
pools. The 2020 deposition scenario was based on the maximum feasible reduction
(MFR) emissions for 2020. Critical load estimates were generally comparable among the
three main regions (northern Europe, Russia, and North America) and were influenced by
the chemical indicator and associated threshold. The Al:Bc of 1.0 was less stringent than
ANCle of 0.0, with the median critical loads for the two indicators being 700 and
300 eq/ha/yr, respectively. In North America, the lowest critical loads occurred in eastern
Canada above latitudes of 60°. In general, low critical loads were found in areas with low
weathering rates associated with coarse soils on acidic parent material. Critical loads in
North America (above latitudes of 60°) were not exceeded by any of the deposition levels
for any of the years: The lowest critical load was 130 eq/ha/yr, while the maximum rates
of N and S deposition in North America were estimated to be 30 to 40 eq/ha/yr.
Simkin et al. (2016) quantified the effect of N deposition on species richness in the
continental U.S. using data from over 15,000 plots. N deposition had a strong effect on
species richness, but this effect differed between closed canopy and open ecosystems
(i.e., forests and nonforest). In nonforested ecosystems (grasslands, deserts, shrublands,
subalpine ecosystems), there was a positive relationship between N deposition and
herbaceous species richness at low rates of N deposition, then a decrease in species
richness with higher rates of N deposition over a threshold of 8.7 kg N/ha/yr. This study
focused N enrichment, but highlighted interactions with soil acidification. Vegetation on
acidic soils was more susceptible to species loss under elevated N. See Appendix 6.2 and
Appendix 6.5.3 for more information on this study.
5-41
-------
5.5.3.1 Target and Future Critical Loads/Critical Load
Exceedances
Target loads are generally defined as the deposition that is determined to provide a
selected level of protection for or recovery of an ecosystem in a given future year. Target
loads of S deposition were calculated for 44 watersheds and extrapolated to
I,320 acid-sensitive watersheds in the Adirondacks using MAGIC (see also
Appendix 4.5.1.4) and different soil chemical indicator and thresholds of base saturation
(5 and 10%), soil solution Bc:Al, and Ca:Al [1.0 and 10.0; Sullivan et al. (201 la)l. In a
comparison of target loads (out to years 2050 and 2100) and the 2002 deposition, only
II.6 to 13.5% of the watersheds were simulated to be in exceedance of target loads to
protect soil base saturation to 5%. For target loads to protect soils to a base saturation of
10%, 79.7 to 87.5% of the watersheds were in exceedance. For soil solution BcAl, 7.8
and 98.1% of watersheds were exceeded by the 2002 deposition for target loads to protect
soil solution ratios of 1.0 and 10.0, respectively. For soil solution Ca:Al, 44.1 to 58.2% of
watersheds experienced exceedances of target loads associated with a soil solution ratio
of 1.0, and 98.1% of watersheds with target loads to protect Ca:Al ratios of 10.0 were
exceeded. Further investigations revealed that 58.2, 85.7, and 93.6% of watersheds could
not obtain threshold values of 10% base saturation, Bc:Al = 10, or CaAl = 10,
respectively, even if acidic deposition was held at zero, thereby demonstrating that these
chemical indicator threshold values were not useful for target load calculations using
MAGIC in the Adirondack Mountains.
Sverdrup et al. (2012) and McDonnell et al. (2014a) used the ForSAFE-VEG model (see
also Appendix 4.5.1.3) to evaluate potential long-term effects of climate change and
atmospheric N deposition on alpine/subalpine ecosystems. Critical loads in both studies
were defined as the amount of N deposition to protect against a change in plant
biodiversity of 5 to 10 Mondrian (M) units (i.e., 5-10% change in plant species cover).
Sverdrup et al. (2012) focused on a "generalized" alpine/subalpine site in the Rocky
Mountain National Park, and simulated plant responses to a future climate (IPCC
scenario A2) and four levels of N and S deposition (preindustrial background, Clean Air
Act [CAA] controls, no CAA controls, and no CAA controls + high deposition). Soil
base saturation decrease to less than 1% and soil solution Bc:Al were predicted to be less
than 10.0 after year 2100, with the futures of no CAA emission controls and no CAA
controls + high deposition (indicating soil acidification). Future plant species coverages
were predicted to change in successively greater amounts in response to the altered
climate, CAA emission controls, no CAA emission controls, and no CAA emission
controls + high deposition. Critical loads of N deposition were calculated (based on a
change of 5 M) to be 1 kg N/ha/yr. Critical loads related to S deposition were not
discussed. All future N deposition scenarios (except preindustrial background N) were
5-42
-------
simulated to result in critical load exceedance. McDonnell et al. (2014a) found similar
results in their study that evaluated the response of plants on a site in the Loch Vale
watershed, CO to 100 different scenario combinations ofN deposition, precipitation, and
temperature. The estimated critical load of N to protect against future (average of
years 2010-2100) change in biodiversity (10 M) was estimated to be between 1.9 and
3.5 kg N/ha/yr, depending on the temperature increase scenario. Current deposition levels
were found to exceed the critical load. Future increases in temperature were forecasted to
substantially impact plant community composition beyond the predicted changes in
response to N alone; N deposition was forecasted to result in approximately a 10- to
25-M change by 2100, whereas a +4.6°C increase in temperature was forecasted to result
in approximately a 38- to 48-M change by 2100. In both studies, plant community
response appeared to be attributed to N enrichment or eutrophication, as McDonnell et al.
(2014a) stated that the critical load adopted in their study was much lower than critical
loads to protect against NO;, leaching and soil acidification determined by other studies.
5.5.3.2 Sensitivity of Critical Load Estimates to Forest
Management and Environmental Stresses
Influences of Forest Management
Removal of N and base cations with tree harvesting can be included in the calculation of
critical loads. Several recent studies evaluated the amounts of base cations removed
through different harvesting practices and determined that base cation removal can be
substantial and contribute to site acidification.
Duchesne and Houle (2008) determined base cation budgets in a managed boreal balsam
fir forest in Quebec, Canada according to six different scenarios, including two
harvesting scenarios (whole-tree and stem-only harvesting), and three scenarios of
mineral weathering. Whole-tree harvesting was found to remove twice as much Ca (1,358
vs. 664 mol/ha) and K (483 vs. 200 mol/ha) as stem only. In contrast to Ca and Mg,
immobilization of K within tree biomass (69 mol/ha/yr) was the main pathway of K
losses from the soil exchangeable reservoir, being five times higher than losses via soil
leaching (14 mol/ha/yr). The amounts of K contained within the aboveground biomass
and the exchangeable soil reservoir were 3.3 and 4.2 kmol/ha, respectively. Whole-tree
harvesting was estimated to remove 44% of the K that was readily available for cycling
in the short term.
Similar results were found by Iwald et al. (2013) in their evaluation of the removal of
base cations with the harvesting of tree stumps and logging residues for biofuels in
5-43
-------
Sweden. Their study evaluated three levels of harvesting intensity that varied in the
amounts of stump and logging residues left on site (0-40%). Results from the study
indicated that harvesting of stumps and logging residues constituted 13 to 24% and 27 to
45% of total (stumps + stem wood + logging residues) base cation extraction,
respectively, depending on harvest intensity and tree species. The higher acidifying effect
of logging residue removal was explained by the higher contents of base cations in
needles and branches compared to stem wood. In a comparison between total net cation
extraction by tree harvesting and maximum levels of current acid deposition, the
acidifying effect of pine harvesting was found to be 57 to 108% of that of acid
deposition, the acidifying effect of spruce harvesting was 114 to 263%, and the acidifying
effect of birch harvesting was 60 to 171%.
In the northeastern U.S., Lucas et al. (2014) evaluated base cation extractions under three
different management scenarios that varied in management intensity and conservation
focus. The MaxGrowth management option, which involved site scarification, intensive
fertilization (amonium nitrate), and short rotation lengths (60-100 years), resulted in
roughly 50 to 100% greater base cation removal than that associated with the other three
management scenarios. Additionally, removal of treetops, branches, and stumps
consistently resulted in removal rates of Ca, Mg, and K three to four times greater than
conventional stem-only harvests.
Johnson et al. (2015a) used mass balance calculations to evaluate the impact of
harvesting on ecosystem balances of Ca, Mg, K, and acidification of forest soils at
40 sites in Ireland. Three harvesting scenarios were evaluated: stem-only harvest (SOH),
stem plus branch harvest (SBH), and whole-tree harvest (WTH). Mass balances for Ca,
Mg, and K were determined based on the difference between long-term inputs
(atmospheric deposition plus mineral weathering) and outputs (biomass removal plus
leaching losses). Soil acidification was calculated using a simplified acidity balance of
inputs (base cation and Na deposition and base cation weathering) and outputs (base
cation uptake, S deposition, and CF deposition). Under the SOH and SBH management,
inputs of Ca, Mg, and K were predicted to be sufficient to meet outputs. Atmospheric
deposition was the most important source of Ca and Mg input. For K, inputs from soil
weathering were as important as deposition. Under the WTH scenario, Ca output
exceeded input at 19 of the 40 study sites. However, the differences were small relative to
the sizes of the Ca pools; at the 19 sites, exchangeable pools could support Ca removal
with WTH for a median of 220 years. Mg and K removal with WTH was balanced by
inputs through deposition and soil weathering. In contrast, for soil acidification budgets,
base cation removal with all scenarios of harvesting was much greater than that generated
by soil weathering, suggesting that soils will become acidified over the long term.
However, there was considerable uncertainty around the calculation of base cation fluxes.
5-44
-------
For Ca balances, confidence intervals spanned positive and negative values at so many
sites that it was not possible to predict the balance of Ca budgets. In addition, uncertainty
in flux calculations was particularly important for K, as soil exchangeable pools were
small and could be depleted within one or two rotations.
Zetterberg et al. (2014) evaluated the long-term effects of whole-tree harvest (WTH) on
soil and stream water acidity in three forested catchments dominated by Norway spruce
in Sweden. Potential influences of varying the amounts of logging residues, Ca
concentrations in tree biomass, and site productivity on the model predictions were also
evaluated through a sensitivity analysis. The MAGIC model was used to simulate
changes in forest soil exchangeable Ca pools and stream water ANC from 1850 to 2100,
with WTH occurring in 2020. Large depletions in soil Ca supply and a reversal of the
positive trend in stream ANC were predicted for all three catchments sites after WTH.
However, the magnitude of impact on stream ANC varied depending on site and the
concentrations of mobile strong acid anions. Varying the tree biomass Ca concentrations
was found to have the largest impacts on modeled soil and stream chemistry. Site
productivity was the second most influential variable, and changing the amount of
harvest residues left on site only marginally affected soil exchangeable Ca and stream
water ANC. The results from this study suggest that future research should concentrate
on minimizing uncertainties in tree biomass Ca concentrations and performing studies on
biological feedback mechanisms that can increase Ca availability in the soils.
Influences of Environmental Stresses
Climate and other environmental stresses may also directly or indirectly alter ecosystem
parameters that are used in the SMB model to determine critical loads. The potential
influences of climate were examined by McNultv and Boggs (2010) in their case-study
evaluation of red spruce stands in western North Carolina that were experiencing low
versus high rates of mortality associated with a pine beetle outbreak. There were positive
relationships between site fertility (forest floor and soil measurements, foliar N
concentrations, and Mg:N ratios) and red spruce mortality. Annual basal area growth of
red spruce on more fertile plots was more sensitive to drought than on less fertile sites.
Based on these observations, critical loads of acidity could change as a result of episodic
stress (e.g., drought, insect infestations, frost injury, cold tolerance, etc.), and increased
growth due to higher fertility and/or nutrient imbalances caused by acidifying deposition
may make red spruce more susceptible to environmental stresses. Critical loads of N and
S, therefore, might be lower when environmental stresses are present.
5-45
-------
5.5.3.3 Uncertainties in Critical Load Estimates
Reinds and de Vries (2010) evaluated the uncertainties in critical and target loads of S
and N for 182 European forest soil plots using the Very Simple Dynamic (VSD) model
(see also Appendix 4.5.2.2). Target loads were defined as the deposition that leads to a
desired chemical state of the ecosystem in a given future year. The VSD model was
calibrated using Bayesian prior probability functions for model parameters based on
literature data, data from 200 Dutch forest sites and from simulated denitrification rates
from a detailed ecosystem model, which improved the model fit to observed soil
measurements. Critical loads determined by the calibrated model varied by chemical
criterion. Minimum critical loads of N ranged between 181 (5th percentile) and
1,606 eq/ha/yr (95th percentile), maximum critical loads of N ranged between 502
(5th percentile) and 31,247 eq/ha/yr (95th percentile), and maximum critical loads of S
ranged between 4 (5th percentile) and 9,670 eq/ha/yr (95th percentile), with the critical
loads based on Al:Bc = 1 being higher than those determined using the ANC = 0 eq/ha/yr
criterion. Uncertainty analysis also showed that the main drivers of uncertainty were
largely dependent on the chemical criterion used in the critical and target load
calculations. Base cation weathering, deposition, and the parameters describing the H-Al
equilibrium in the soil solution were the main sources of uncertainty in the estimates of
maximum critical loads for S (Clmax[S]) based on the Al:Bc criterion of 1.0, and
uncertainty in Clmax(S) based on ANC was completely determined by base cation inputs.
The denitrification fraction was the most important source of uncertainty for the
maximum critical loads of N (Clmax[N]). Calibration of VSD reduced the levels of
uncertainty for all critical loads and criteria. After calibration, the coefficient of variation
(CV) for Clmax(S) was below 0.4 for almost all plots, and target loads were not needed
in any of the simulations for 40% of the plots. According to the noncalibrated model,
there was a positive probability for the need of a target load for almost all plots.
5.6 Climate Modification of Ecosystem Response
The effect of acidifying deposition on terrestrial ecosystems can be modified by climate
shifts in temperature and precipitation. Appendix 13 provides an overview on this topic
and Appendix 4.7.1. provides information on climate modification of soil acidification.
Warmer temperatures increase decomposition and nitrification. Nitrification will also
increase with increased N supply caused by increased weathering or decomposition
(Booth et al.. 2005). The process of nitrification generates protons that increase the rate
of nitrate and base cation leaching to drainage waters (Murdoch et al.. 1998). The
combined increase of NO3 leaching and loss of base cations has the potential to magnify
5-46
-------
acidification in forest soils (Fernandez et al.. 2003). Soil weathering is typically the key
buffer to acidic deposition (Li and McNultv. 2007). and while weathering is increased by
both soil temperature and soil moisture (Gwiazda and Broecker. 1994). it is unclear
whether any future change in the magnitude of temperature and precipitation will be
enough to alter base cation supply or influence the acid-base balance of sensitive
ecosystems. Furthermore, it is unclear whether increased supply of N in soils from either
deposition, increased decomposition, or increased nitrogen fixation may negate the
ameliorative effect of enhanced weathering. Some studies show that climate change will
mitigate acidification with increased weathering (Belvazid et al.. 201 la), while others
show that climate change will aggravate acidification although increased nitrification
outpacing enhanced weathering (Wu and Driscoll. 2010). In general, increased
temperature and precipitation will likely enhance inputs of buffering agents from
weathering and deposition, but also increase inputs of acidifying agents from deposition
and enhanced N cycling. The relative sensitivity of these opposing processes to a given
change in climate remains unresolved.
The mountainous regions of the Eastern U.S. are especially interesting to study because
acidifying deposition and climate change interact in these areas. Recently, Wason et al.
(2017) studied the responses of red spruce and balsam fir to acidic deposition and trends
in climate on Whiteface Mountain in New York using tree ring analysis in forests plots
along an elevational gradient. They found that both species increased growth with
increased precipitation and pH. Red spruce growth appeared to increase growth with a
warming climate and balsam fir did not. Despite the changes in growth due to
precipitation chemistry and climate, the authors did not detect changes in the distribution
in the spruce-fir forest and perhaps this is a longer-term process. The study demonstrated
the complexity of forest response as multiple environmental factors change as these
forests recover form acidifying deposition.
Koo et al. (2014) used the Annual Radial Model (ARIUM) to investigate projected
climate change effects and changing air pollution on red spruce growth in the Great
Smoky Mountain national park. The model estimated that high elevation (<1,700 m) red
spruce growth would decline 10.8% when a climate change interacted with a 10%
increase in air pollution. However, growth increased by 8.4% when air pollution
decreased 10% in air pollution with climate change. In contrast, low elevation red spruce
growth decreased with future climate change with decreased, increased and no change in
air pollution.
5-47
-------
5.7 Summary
Since publication of the 2008 ISA, the overarching understanding of terrestrial
acidification has not appreciably changed. More recent research has confirmed and
strengthened this understanding and provided more quantitative information, especially
across the regional landscape. A number of studies have evaluated the relationships
between soil chemistry indicators of acidification and ecosystem biological endpoints
(see Table 5-6). Soil chemistry indicators examined in recent literature include
exchangeable base cations, soil pH, exchangeable acidity (H+ and Al), exchangeable
Be A1 ratio, base saturation, and Al concentrations. Biological endpoints included in the
evaluations consisted of physiological and community responses of trees and other
vegetation, lichens, soil biota, and fauna.
5.7.1 Physiology and Growth
The physiological effects of acidification on terrestrial ecosystems in the U.S. were well
characterized at the time of the 2008 ISA. Consistent and coherent evidence from
multiple species and studies in 2008 showed that the biological effects of acidification on
terrestrial ecosystems were generally attributable to physiological impairment caused by
Al toxicity and decreased ability of plant roots to take up base cations (Section 3.2.2.3 of
the 2008 ISA). Acid deposition can also leach membrane-associated stores of Ca from
young red spruce needles, which affects membrane stability and freezing tolerance. New
information since the 2008 ISA has supported these conclusions (Appendix 5.2).
including further studies on the impact of acidifying deposition on sensitive tree species,
such as sugar maple and red spruce. Much of the new evidence for the negative effects of
acidifying deposition on these species comes from Ca addition experiments, in which the
addition of Ca has alleviated many of the negative plant physiological and growth effects.
Consistent with the findings of the 2008 ISA, the body of evidence is sufficient to infer
a causal relationship between acidifying N and S deposition and the alteration of the
physiology and growth of terrestrial organisms and the productivity of terrestrial
ecosystems.
In the 2008 ISA, acidifying deposition, in combination with other stressors, was found to
be a likely contributor to physiological effects that led to the decline of sugar maple trees
living at higher elevation in some portions of the eastern U.S. that have geologies
dominated by sandstone or other base-poor substrate, and that have base-poor soils.
Studies since the 2008 ISA support these findings (see Appendix 5.2.1.1). For example,
recent field studies demonstrated relationships between soil chemical indicator threshold
values and tree responses. Substantial declines in sugar maple regeneration were found at
5-48
-------
soil base saturation levels <20%, which is consistent the range reported in the 2008 ISA.
In new studies, sugar maple showed positive growth and regeneration responses to
increasing exchangeable base cations, base saturation and soil pH, and negative
relationships with increasing exchangeable Al. Additional studies indicated that growth,
regeneration and physiological responses of sugar maple to soil conditions created by
acidifying N and S deposition were reversed or ameliorated by Ca additions. The
responses to Ca additions included increased growth and regeneration, decreased crown
dieback, increased foliar chlorophyll, and decreases in foliar metabolic indicators of
physiological stress.
The 2008 ISA reported that changes in soil chemistry, such as depletion of soil base
cations, increasing Al concentration, and leaching of base cations into drainage water,
have contributed to physiological stress, high mortality rates, and decreasing growth
trends of red spruce trees. New information since the 2008 ISA from Ca addition studies
supports the conclusion that depletion of base cations contributed to these effects in trees
(Appendix 5.2.1.2). Foliar biomass and physiological responses of red spruce to soil
conditions created by acidifying N and S deposition were reversed or ameliorated by Ca
additions. The responses included higher foliar antioxidant activity in the winter,
significantly greater foliar cold tolerance, and higher levels of foliar metabolic
compounds that indicate an increased tolerance of environmental stresses. (Schaberg et
al.. 2011; Halman et al.. 2008).
In the 2008 ISA, there was a limited amount of information on acidification effects on
flowering dogwood. Since the 2008 ISA, no additional information on dogwood has been
published; however, other tree species that have been evaluated in studies relating soil
chemistry to tree physiology include yellow birch, white ash, American beech, black
cherry, northern red oak, hickories, American basswood, and eastern hophornbeam.
These species were reported to vary in their sensitivities to soil conditions associated with
acidifying deposition. Data are insufficient to draw general conclusions for other species.
New information was also published on the recovery of eastern redcedar since the 1980s
as SO2 emissions declined. However, it is unclear whether this recovery was from acid
deposition, direct SO2 effects, or a combination of each (Appendix 5.2.1.3).
New studies since the 2008 ISA have also added new information about acidifying
deposition on forest understory vegetation, grasslands, lichen communities, and higher
trophic-level organisms (snails and salamanders). These studies are not as numerous as
those from the decades of tree research; however, results of these studies support the
conclusions of the 2008 ISA regarding the effects of acid deposition on terrestrial
ecosystems.
5-49
-------
5.7.2
Biodiversity
Consistent with the findings of the 2008 ISA, the body of evidence is sufficient to infer
a causal relationship between acidifying N and S deposition and the alteration of
species richness, community composition, and biodiversity in terrestrial ecosystems.
The physiological effects of acidifying deposition can result in changes in species
composition and biodiversity whereby sensitive species are replaced by more tolerant
species. For example, sugar maple was found to have greater growth and seedling
colonization with increasing soil cation availability, and American beech was more
prevalent on soils with lower levels of base cations—locations where sugar maple does
less well (Appendix 5.2.1.3.1). As noted above, studies have found increased
regeneration of sugar maple with Ca additions and less regeneration with increasing
exchangeable Al. Soil acid-base chemistry was found to be a predictor of understory
species composition. Fifty understory species were associated with the basic end of a soil
pH gradient, and these species could have value as indicators of sites with high base
cation status and potentially suitable habitat for acid-sensitive species like sugar maple.
In another set of studies, soil acid-base chemistry predicted and was correlated with soil
biota diversity and community composition. Proportions of soil bacteria, fungi, and
nematodes were found to be correlated to soil pH and Al concentrations. Fungi and
nematodes were more abundant in acidic soils. Ca additions resulted in a change in soil
bacterial community composition, and the bacterial community structure was found to be
significantly correlated with soil exchangeable Ca, pH, and P (Appendix 5.2.4).
5.7.3 National-Scale Sensitivity and Critical Loads
Sensitivity of soils to acidifying deposition is discussed in detail in Appendix 4 and
summarized in Section IS.5.1. In general, surficial geology is the principal factor
governing the sensitivity of terrestrial ecosystems soil to acidification from S and N
deposition. Other factors contribute to the sensitivity of soils to acidifying deposition,
including topography, soil chemistry, and land use. Forests of the Adirondack Mountains
of New York, Green Mountains of Vermont, White Mountains of New Hampshire, the
Allegheny Plateau of Pennsylvania, and mountain top and ridge forest ecosystems in the
southern Appalachians are the regions most sensitive to terrestrial acidification from
atmospheric deposition (Section 3.2.4.2 of the 2008 ISA). Sensitive ecosystems can also
be characterized by presence of acid-sensitive soils and plant species (Appendix 5.3).
Models used to determine critical loads of acidifying deposition included: SMB, STA,
MAGIC, ForSAFE-VEG, and empirical models. Several models and extrapolation
methods to estimate BCw rates were also investigated. The PROFILE model was
5-50
-------
evaluated as a model to estimate soil BCw rates to support estimates of SMB critical
loads in the U.S. In general, recently published models used soil solution Bc:Al of 10.0 as
an indicator to estimate critical loads in North America. These models are described in
more detail in Appendix 4.5.
Sensitivities of ecosystems to ambient N and S deposition were also characterized
through critical loads and exceedances (Appendix 5.5). Calculated critical loads for forest
sites based on the soil solution Bc:Al of 10.0 in the northeastern U.S. ranged from 11 to
6,540 eq/ha/yr (eq quantifies the supply of H+ ions available for acid-base reaction,
allowing the acidifying effects of N and S deposition to be combined into the same unit),
and 15-98% of these critical loads were exceeded by N and S deposition. In this region,
correlation analyses showed that the growth of 17 species and crown density of 4 species
were negatively correlated with critical load exceedance. In Pennsylvania, critical loads
based on the soil solution Bc:Al of 10.0 for hardwood forests ranged from 4 to
10,503 eq/ha/yr and were exceeded by the 2002 N and S deposition in 53% of the plots.
In comparison, critical loads for terrestrial ecosystems in the arctic and subarctic regions
of North America were not exceeded by estimated deposition in the years 1900, 2000,
2010, and 2020. For these high latitude ecosystems, the lowest critical load was
130 eq/ha/yr, while the maximum N and S deposition was 30 to 40 eq/ha/yr.
In western Canada, critical loads ranged from 40 to 4,000 eq/ha/yr depending on the
study and location. Several studies evaluated the influence of BCw rates, soil chemical
indicators and thresholds, N retention, tree species-specific base cation uptake, and/or
bulk (i.e., wet) versus total deposition on critical load estimates. All of these parameters
were found to influence critical load and exceedance determinations, thereby
demonstrating the uncertainties and sensitivities associated with critical load estimates.
5-51
-------
Table 5-6 Mode of action for acidifying nitrogen and sulfur deposition.
Reference
Ecosystem
Type/Region
Species
N and S
Deposition/
Additions
Soil Indicator
Description
Bardhan et al.
(2012)
Forest
Great Smoky
Mountain
National Park,
TN, NC
Soil
microbes
(bacteria)
6 to 41 kg
Soil pH and
S/ha/yr
exchangeable
modeled
Ca and Al, CEC
deposition
and base
classes
saturation (O-,
(Weathers et
A-, and
al.. 2006)
B-horizons)
14.57 to
Exchangeable
19.7 kg/ha/yr as
Ca and pH (Oa-
wet NO3"; 17.44
and B-horizons)
to
29.09 kg/ha/yr
as wet SO42",
modeled (Ito et
al.. 2002)
1990-1999
Only minor differences in bacterial diversity among sites; the bacterial
community did not change along the gradients of S deposition, soil pH,
or exchangeable Ca:AI ratio. High elevation sites remain acidic and
have not yet reached a threshold suitable for nonacidophilic bacterial
communities.
Beier et al. (2012)
Forest
Adirondack
Mountains, NY
(12 upland
hardwood
forests)
NA
Increasing trends in snail community richness and abundance, live
biomass of red-backed salamanders, and canopy tree basal area with
increasing soil Ca. Land snail species richness and abundance were
positively correlated with Oa-horizon Ca and negatively correlated to
SO42" deposition. Salamander communities changed continuously
along the Ca gradient. Several known calciphilic species of snails and
plants were found only at the highest Ca sites. The proportion of basal
area attributed to standing dead trees decreased significantly with
Oa-horizon Ca.
Bilodeau-Gauthier
et al. (2011)
Forest
Quebec lakes
network,
Quebec,
Canada
(6 watersheds)
Sugar
maple
NA
Soil pH, base Tree growth was positively correlated to concentrations of base cations
saturation, (Ca, K, and Mg) in wood and soil, and negatively correlated to
exchangeable concentrations of acidic metals in wood (Al, Mn, and Cd) and soil (H+
Ca:AI ratio, and exchangeable Al). Percentage base saturation was the best
exchangeable predictor of BAI (nonlinear) and explained 43% of variance.
Al, Ca, Mg, Mn, Multifactorial relationships indicated that tree age and soil
and K (forest exchangeable Al accounted for 51% and tree age and log of the ratio of
floor and 0 to base cations (Ca + Mg + K):Al in the soluble (water and acid soluble
15 cm of wood extracts) fractions accounted for 46% of the variation in sugar
B-horizon) maple BAI.
5-52
-------
Table 5-6 (Continued): Mode of action for acidifying nitrogen and sulfur deposition.
Reference
Ecosystem
Type/Region
Species
N and S
Deposition/
Additions
Soil Indicator
Description
Bondi etal. (2016)
White, Green,
and Adirondack
Mountains
(34 forest sites)
Red-backed NA
salamander
(Plethodon
cinereus)
Soil pH Studied the relationships between soil pH (organic horizon) and the
(Oa-horizon—on abundance and health of red-backed salamanders. No associations
16 sites between soil pH and salamander metrics (abundance, body size, and
A-horizon). Soil body condition) were found, and the salamanders did not appear to
pH range was select habitats based on soil pH. The strongest driver of the abundance
2.73-5.54. of red-backed salamanders was the presence of dusky salamanders
(Desmognathus spp.).
Caietal. (2017a)
Semiarid
grassland
Inner Mongolia,
China
Grasses
(Agropyron
cristatum,
Stipa
krylovii),
forbs
(.Artemisia
frigida,
Potentilla
tanacetifolia,
Potentilla
bifurca)
0, 5, 10, and Soil pH and Al N additions resulted in decreased pH and increased soil available Fe,
15 g N/m2/yr) plant tissue Zn, Mn, and Cu concentrations, with water additions partially counteracting
and two water Cu, Mn, and Fe the impacts of N. Nitrogen additions caused higher foliar Mn, Cu, and
addition concentrations Zn and lower Fe concentrations, resulting in micronutrient imbalances,
treatments Similar to the soils, water additions partially offset the impacts of the N
additions on foliar chemistry.
Chen et al. (2013) Grassland
Mongolian
steppe, China
Soil Seven Soil pH and Fungal fatty acids were increased by 49% and fungi:bacteria ratio
microbes treatments of S extractable increased by up to 120% by the H2SO4 additions, relative to the
(bacteria, additions as cations (Al, Ca, controls. The H2SO4 treatments decreased total and bacterial fatty
fungi, and sulfuric acid (0, Mg, and NA; 0 acids by up to 47 and 40%, respectively. These responses were
nematodes) 2.76,5.52, to 15 cm) attributed to soil pH and Al3+concentrations. High Al3+concentrations
8.28, 11.04, (51 to 83 mg/kg) were associated with decreased total fatty acids and
13.8, and decreases in bacterial and increases in fungal fatty acids. Soil
16.56 mol nematode numbers were initially increased by the H2SO4 treatments
H+/m2) as three followed by changes in the nematode community. The shifts in the
additions nematode community were attributed to decreased soil pH and
(2009-2010) changes in soil moisture.
5-53
-------
Table 5-6 (Continued): Mode of action for acidifying nitrogen and sulfur deposition.
Reference
Ecosystem
Type/Region
Species
N and S
Deposition/
Additions
Soil Indicator
Description
Chen et al. (2016)
Alpine Soil
grasslands and microbes
NA
global
terrestrial
biomes
(bacteria,
fungi)
Soil pH
Studied the relative importance of site, biotic, and climatic factors on
soil microbial communities in two alpine grasslands in the Tibetan
Plateau and across global terrestrial biomes. Microbial communities
were characterized by phospholipid fatty acid (PFLA) and grouped as
total microbial biomass, fungal biomass, arbuscular mycorrhizal fungi
biomass, actinomycete biomass, and fungal:bacterial ratio. In the
Tibetan grasslands, all measures of soil microbial communities were
found to be negatively related to soil pH (soil pH ranged from -6-10)
and positively correlated with soil C:N ratio (C:N ratio range of 4-15).
The fungal:bacterial ratio showed a positive relationship with pH at soil
pHs that ranged from 7.5-10, and negative relationship with soil C:N.
Across the global biomes, total microbial biomass was also found to be
positively correlated with soil C:N and negatively correlated with soil
pH. Soil variables (alone) accounted for 43.4% of the variation in total
microbial biomass, while climatic and biotic variables (alone) only
accounted for 4.5 and 0.2% of variation, respectively.
Cleavitt et al. (2014)
Forest
Hubbard Brook
Experimental
Forest, NH
Sugar
maple
NA
Exchangeable Soil Ca concentration exhibited a 9x change across the study sites and
Ca (top 5 cm of was positively correlated to sugar maple abundance and initial seedling
B-horizon) densities. However, soil Ca concentration was not a significant
predictor of 1 st-year mortality, nor was it a factor that distinguished
among the three main site types.
Duchesne and
Ouimet (2009)
Forest
Southern
Quebec,
Canada
(426 monitoring
plots)
Sugar
maple and
American
beech
NA
Soil pH, The basal area of sugar maple in the sapling stratum was positively
exchangeable correlated with soil exchangeable Ca and Mg. Basal area of American
K, Ca, Mg, H, beech in the sapling stratum was negatively correlated with
and Al (upper exchangeable Ca and Mg. However, the basal area of sugar maple in
B-horizon) the sapling stratum was positively correlated with both the relative
basal area of dead sugar maple and sugar maple in the tree stratum.
The basal area of American beech was also positively correlated with
the relative basal area of American beech in the tree stratum.
5-54
-------
Table 5-6 (Continued): Mode of action for acidifying nitrogen and sulfur deposition.
Reference
Ecosystem
Type/Region
Species
N and S
Deposition/
Additions
Soil Indicator
Description
Elias et al. (2009) Forest
Hardwood
Monongahela anc' cor|ifer
National Forest, *ree sPecies
WV (FIA plots)
NA
Soil pH, base The study authors found that hickories were the only species to be in
saturation, significantly lower numbers on sites with base saturation below 20
exchangeable (A-horizon) and 2.5% (B-horizon). Percentage of dead northern red oak
Ca:AI (A- and was highest on sites with A-horizon Al concentrations above
B-horizons) 43 cmolc/kg of soil. Soil exchangeable Ca:AI and sum of base cations
in the B-horizon were highest in stands that experienced the lowest
species turnover rates (1989-2000). Periodic mean annual volume
increment (whole stand) was positively correlated with A-horizon base
saturation (range of 5 to 77%), Ca concentrations, and exchangeable
Ca:AI (range of 0.17 to 10.2) and B-horizon pH. The exchangeable
Ca:AI of the A-horizon accounted for over 30% of the variation in the
periodic mean annual volume increment.
Gilliam et al.
(2011b)
Forest
Fernow
Experimental
Forest, WV
Soil
microbes
(bacteria
and fungi)
NA
Soil pH (5 cm) Fungi were dominant at the most weathered, low NO3" site, while
Gram-negative bacteria were significantly higher at the less weathered,
moderate and high NO3" sites. Accordingly, the fungi:bacteria ratio
increased in the direction of the low NO3" plots in ordination space.
Correlations between the soil parameters and PLFA results suggest
that low soil pH and NO3" concentrations supported fungal dominance,
although other factors including differences in plant community and
clay and organic matter content may have also influenced the soil
microbial community.
Horslev et al. (2008)
Forest
Pennsylvania,
New York, New
Hampshire, and
Vermont
(86 northern
hardwood
stands)
234 forest
understory
species
NA
The study authors found that a base cation-acid cation nutrient gradient
accounted for 71.9% (in NH and VT) and 63% (in PA and NY) of the
variation in the nonmetric multidimensional scaling ordination analyses
of plant community composition. Sugar maple foliar Mg and Ca had the
strongest association with the base end of the gradient. Exchangeable
Al (in NH and VT) and foliar Mn (in PA and NY) were strongly
associated with the acid end of the gradient. A total of 50 of the
234 understory species were associated with the base end of the base
cation-acid cation nutrient gradient.
5-55
-------
Table 5-6 (Continued): Mode of action for acidifying nitrogen and sulfur deposition.
Reference
Ecosystem
Type/Region
Species
N and S
Deposition/
Additions
Soil Indicator
Description
Huang et al. (2017)
Forest
Eastern China
Soil
microbes
(Archaeal
diversity)
NA
Soil pH and Al Studied influence of Al additions on the archaeal diversity in red soils,
concentrations Three Al treatments of 0, 100, or 200 mg Al/kg soil (as AICI3 6H2O); pH
maintained at pretreatment level. Al additions were found to increase
the abundance but decrease the evenness of the Archaea.
Abundances of Crenarchaeota increased while those of Euryarchaeota
decreased in response to higher Al concentrations, suggesting that
Crenarchaeota is more tolerant of Al than is Euryarchaeota
Kunito et al. (2016) Forest
Japan
(27 sites)
Soil
microbes
NA
pH and Al Evaluated the relationship between soil chemistry (Al concentrations
(soluble and and pH) and soil microbial biomass and enzymes involved with C, N,
exchangeable) and P cycling (p-D-glucosidase, polyphenol oxidase, L-asparaginase,
concentrations acid phosphatase). The researchers found that p-D-glucosidase and
in 0-15 cm of polyphenol oxidase were reduced with higher amounts of soluble and
soil exchangeable Al, acid phosphatase shared an inverse relationship with
soil pH, and L-asparaginase increased as pH increased. Microbial
biomass was also found to decrease as organically bound Al and Fe
increased.
Li et al. (2016a)
Natural steppe
ecosystem
Inner Mongolia,
China
Soil
microbes
(bacteria)
Elevated N Soil pH
5-15 g N/m2/yr (0-15 cm)
as urea applied
in May and
June since
2005
Evaluated: (1) the impacts of N and water additions on soil microbial
community (0-15 cm of soil), (2) the linkages between the variation in
belowground bacterial community and aboveground vegetation
community, and (3) relationships between soil/plant factors and soil
microbial community. In general, bacterial alpha diversity was positively
correlated with total N and pH and negatively correlated with soil C:N
ratio and concentrations of NH4 and NO3. Beta diversity of the bacterial
community was significantly correlated with C/N ratio, inorganic N, and
pH. The relative abundances of Proteobacteria, Firmicutes, TM7, and
OD1 increased with N addition gradient. The relative abundances of
Proteobacteria Firmicutes, TM7, and OD1 were positively correlated
with inorganic N and negatively correlated with soil pH.
5-56
-------
Table 5-6 (Continued): Mode of action for acidifying nitrogen and sulfur deposition.
Reference
Ecosystem
Type/Region
Species
N and S
Deposition/
Additions
Soil Indicator
Description
Long et al. (2009) Forest
Sugar
Pennsylvania, maple and
New York, New black cherry
Hampshire, and
Vermont
(76 sites)
NA
Soil Ca:AI Exchangeable Ca, Mg, and pH in upper B-horizon were positively
(threshold of correlated with sugar maple BAI growth in 1987-1996. Generally, in
0.03; upper long-term trends (1937-1996), sugar maple in stands with
B-horizon) below-threshold amounts of foliar Ca or Mg and above-threshold
amounts of Mn had decreasing BAI trends, while stands with
above-threshold Ca and Mg and below-threshold Mn showed a leveling
off of the BAI. Black cherry consistently showed greater growth on
stands with below-threshold foliar Ca and Mg compared with
above-threshold stands. Black cherry BAI was also greater in stands
with below-threshold Ca:AI molar ratios in the upper B-horizon.
Lucash et al. (2012)
Forest
19 sites in
northeastern
U.S. (New York
and New
Hampshire)
Hardwood NA
and conifer
tree species
Exchangeable
Ca, Mg, and Al
(Oie, Oa 0 to
10 cm, 10 to
30 cm, and
30 cm to top of
C-horizon)
Concentrations of Ca and Mg in foliage were correlated with
exchangeable Ca and Mg concentrations in the upper mineral soil; for
most tree species they were also correlated to acid-extractable Ca and
Mg in the parent material (C-horizon). Foliar Al was insensitive to soil Al
concentrations.
Medeiros et al.
Greenhouse
Red maple
N and P
PH
Studied the influence of pH (simulated acid rain) on leaf, xylem, and
(2016)
study
and white
oak
treatments
(control and low
N/P—90% of
control N/P
levels), pH (4.5
and 6)
hydraulic trait coordination responses of 1-yr-old red maple and white
oak seedlings in a greenhouse study. The researchers found
interactions between nutrient levels and pH; low pH reduced the ability
of both species to adjust xylem traits and leaf water relations
(i.e., hydraulic acclimation) in response to changes in nutrient
availability.
McEathron et al.
Forest
Sugar
NA
Soil pH and
Evaluated the relationships between species-specific basal area and
(2013)
Ha-De-Ron-
maple, black
exchangeable
soil and stream water chemistry. Sugar maple basal area was
Dah Wilderness
cherry,
Ca and Al (0 to
positively correlated with mineral soil pH, and yellow birch basal area
Area in
American
10 cm mineral
was positively correlated with mineral soil exchangeable Ca. Sugar
Adirondack
beech, red
soil horizon)
maple basal area was also negatively correlated with stream water
Mountains, NY
maple, and
DOC.
(seven
yellow birch
subwatersheds)
5-57
-------
Table 5-6 (Continued): Mode of action for acidifying nitrogen and sulfur deposition.
Reference
Ecosystem
Type/Region
Species
N and S
Deposition/
Additions
Soil Indicator
Description
Mueller et al. (2016) Sugar maple
dominated
forests
Michigan
Soil
microbes
NA
Soil pH
Studied the relationships between soil bacterial and fungal community
activity and soil pH and C:N gradients in the litter from sugar maple
stands in Michigan. The proportion of active bacteria were found to
increase along the soil pH gradient, but decreased along the soil C:N
gradient. In contrast, no significant correlations were detected for the
fungal community. Similarly, no significant correlations were between
temperature and the fungal or bacterial communities.
Pabian and
Brittinaham (2012)
Forest
Pennsylvania
(14 forest sites)
NA NA
Exchangeable
Ca and pH
(Oa-horizon).
Mean soil
exchangeable
Ca and pH for
the 14 study
sites ranged
from 5.28 to
23.5 meq/100 g
and 3.6 to 5.1,
respectively.
Sugar NA
Exchangeable
maple,
Ca,
American
exchangeable
beech,
Ca:AI (forest
American
floor and upper
basswood,
[0 to 10 cm]
and white
mineral soil)
ash
Bird community composition (species richness and species
abundances) varied with soil Ca and pH, with 10 bird species having
the highest abundances in forests with high-Ca soils, and 5 species
having the highest abundances with low-Ca soils. Bird species
associated with low-Ca soils were associated with high densities of
acid-loving mountain laurel (Kalmia latifolia) and five tree species with
basal area explained by low soil pH and Ca, whereas bird species
associated with high-Ca soils were associated with high densities of
saplings and high basal area of acid-sensitive tree species.
Page and Mitchell
(2008)
Forest
Adirondack
Mountains, NY
(11 sites)
Evaluated the relationships between exchangeable soil Ca
concentrations and tree basal area. There were no observed trends
relating total basal area to mineral soil (0 to 10 cm) exchangeable Ca
concentrations; however, the relative basal areas of sugar maple and
American basswood were positively correlated with mineral soil
exchangeable Ca, and relative basal area of American beech was
negatively correlated. Relative basal area of white ash was not
corelated to soil exchangeable Ca.
Perakis et al. (2013)
Forest
Oregon
(coastal range)
Douglas fir NA
(plantations)
None
Nitrate leaching (at 20 and 100 cm) increased, soil pH declined (from
5.8 to 4.2), and exchangeable soil Ca, Mg, and K decreased (10x
declines) along the soil N gradient. Exchangeable Ca and Mg (in both 0
to 20 cm and 0 to 100 cm) and K (0 to 20 cm) declined with increasing
nitrate leaching. Mean soil profile pH declined logarithmically with
nitrate leaching at 20 and 100 cm. The sum of exchangeable Ca, Mg,
and K was positively correlated with soil pH at 20 and 100 cm.
Aboveground tree biomass contained an increasing percentage of total
ecosystem Ca, Mg, and K as soil N increased.
5-58
-------
Table 5-6 (Continued): Mode of action for acidifying nitrogen and sulfur deposition.
Reference
Ecosystem
Type/Region
Species
N and S
Deposition/
Additions
Soil Indicator
Description
Pitel and Yanai
(2014)
Forest
Massachu-
setts, Vermont,
and New York
Sugar
maple
NA
Soil pH, Evaluated the mortality of dominant and codominant sugar maple in
exchangeable 47 stands that had experienced defoliation by native forest tent
Ca:AI, Al, and caterpillar (Malacosoma disstria). Mortality was found to be highest in
cations (Ca, Mg, stands with the greatest amount of crown dieback the previous year.
K), cation (Ca, Drought, cold winter temperatures, concave microrelief and soil base
Mg, K, and total) cation availability were also significant predictors of mortality,
saturation, and Concentrations of exchangeable Ca, Mg, and K in the upper B soil
effective CEC horizon were inversely correlated with sugar maple mortality, with
(A- and upper exchangeable K having the strongest relationship with mortality. Site
B-horizons) with above-average sugar maple mortality (>3 or 4%) occurred on soils
with low concentrations of exchangeable Ca (0.31 to 0.46 cmolc/kg),
Mg (0.06 to 0.10 cmolc/kg), and K (0.03 to 0.05 cmolc/kg). Stands
defoliated in 2005 that had low Mg saturation (A-horizon) suffered
higher rates of mortality, suggesting an interaction between low base
cations and defoliation events.
Sridevi et al. (2012)
Forest
Hubbard Brook
Experimental
Forest, NH
Soil
microbes
(bacteria)
NA
Soil pH, Ca additions of 1,000 kg Ca/ha applied in 1999. The bacterial
exchangeable community structure in the Ca treated and nontreated reference soils
cations, Al, Fe, was found to be significantly different, with differences among
P, Mn, Zn, communities being more pronounced in the mineral soils. Calcium
exchangeable additions resulted in a change in bacterial community composition of
acidity, and 23% in the organic and 22% in the mineral soil horizons. Numbers of
CEC (0 to detectable taxa in some families were lower in the Ca amended soils,
15 cm) while some families were higher. Analyses of relationships between soil
chemistry and the bacterial communities indicated that only Ca, pH,
and P were significantly correlated with bacterial community structure.
Stevens et al.
(2010b)
Grassland
U.K. (68 sites)
Grassland
species
6.2 to 36.3 kg
N/ha/yr
Centre for
Ecology and
Hydrology
(CEH) National
Atmospheric
Deposition
Model
Ellenburg R Data from a national survey were used to evaluate species richness of
(reaction-soil 68 U.K. grasslands along an N deposition gradient. The results suggest
pH) and N (soil that soil acidification (instead of eutrophication) was contributing to
nutrient) scores changes in species diversity and composition. Soil acidification may
have led to reduced nutrient availability and increased Al solubility
preventing the "fertilizing" effects of N addition from being apparent.
5-59
-------
Table 5-6 (Continued): Mode of action for acidifying nitrogen and sulfur deposition.
Reference
Ecosystem
Type/Region
Species
N and S
Deposition/
Additions
Soil Indicator
Description
Sullivan etal. (2013)
Forest
Adirondack
Mountains, NY
(50 plots in
20 small
watersheds
with sugar
maple
overstory)
Sugar
maple
75 to
112 meq/m2/yr
as N + S
NADP (wet
deposition);
CASTNET (dry
deposition)
2000-2004
(average)
Soil pH,
exchangeable
Ca,
exchangeable
Mg, and base
saturation
(forest floor, A-
and upper B- (0
to 10 cm)
horizons)
The study found that plots with lower soil base saturation did not have
sugar maple regeneration (these same plots also received higher N
and S deposition levels); proportion of sugar maple seedlings dropped
substantially at base saturation levels less than 20%. Canopy vigor was
positively correlated with soil pH and exchangeable Ca and Mg. Mean
growth rates (BAI) were positively correlated with exchangeable Ca
and base saturation at the watershed level.
Tian etal. (2016b)
Temperate
steppe
Inner Mongolia,
China
Grassland
species;
Stipa krylovii
and
Artemesia
frigida
Elevated N
9 yr of urea
additions at 0,
1, 2, 4, 8, 16,
32, and 64 g
N/m2/yr
Soil pH,
exchangeable
Mn and Al (three
soil depths:
0-10, 10-20,
20-30 cm)
Long-term N additions increased total above-ground plant biomass but
decreased species richness; N additions significantly reduced forb
species richness, while the diversity of grass species was not affected.
Soil chemistry was influenced by the N additions; soil pH was reduced
and concentrations of exchangeable Mn, ferric Fe, and Al were
increased. Foliar concentrations of Mn in both A. frigida and S. krylovii
were increased by N additions. A greenhouse study showed that the
biomass of A. frigida seedling shoots and roots were significantly
reduced with MnCh additions, but the treatments had no effect on S.
krylovii seedlings.
Tu etal. (2016)
Forest types
Soil NA
Soil pH
The soil diazotrophic community structure was found to differ
across North
diazotrophs
(0-10 cm)
significantly across the six forests; lower microbial spatial turnover and
America
(N2 fixing
microbes)
greater community diversity were found in rainforests relative to
temperate forests. In addition, community diversity was strongly
correlated with latitude, mean annual temperature, plant species
richness, and total annual precipitation. Diazotrophic community
diversity was weakly correlated with soil pH and moisture.
Yina etal. (2017)
Grassland
Inner Mongolia,
China
Soil
microbes
(ammonia-
oxidizing
bacteria and
Archaea)
Elevated S
H2SO4 additions
started in 2009;
0, 2.76, 5.52,
8.28, 11.04,
13.80,
16.56 mol H/m2
Soil pH and Study evaluated the impacts of soil pH on the abundance and structure
NH4+-N and of soil ammonia-oxidizing bacteria (AOB) and Archaea (AOA)
NC>3"-N communities. Decreasing soil pH was found to be correlated with
decreases in AOB abundance (pH range of 5.0-7.3) and increases in
AOA (pH range of 5.3-7.3) abundance, except at the lowest pH (pH
5.0), which negatively influenced AOA abundance. Soil acidification did
not significantly influence AOA or AOB community composition.
5-60
-------
Table 5-6 (Continued): Mode of action for acidifying nitrogen and sulfur deposition.
Reference
Ecosystem
Type/Region
Species
N and S
Deposition/
Additions
Soil Indicator
Description
Yuan etal. (2016)
Alpine tundra Soil bacteria Elevated N
Niwot Ridge, 20yrofNand
Colorado P additions—0
Rocky and 20 g
Mountain N/m2/yr and 0
National Park and 2 g P/m2/yr
(1993-1995), 0
and 10 g
N/m2/yr and 2 g
P/m2/yr
(1996-1997),
same
application
rates every 2 yr
(1998-2009),
10 g N/m2/yr
was added as
Ca(N03)2 to
address soil
acidification
Soil pH (0-10 Soil bacterial communities differed by plant community type, with
cm) bacterial alpha diversity being significantly correlated with plant
richness and production of forbs. N additions also influenced soil
bacteria, with bacterial communities treated with N (and N + P) being
significantly different than those in the control and P treatments.
Chloroflexi and Bacteroidetes responded positively to the additions of
N, while Acidobacteria and Verrucomicrobia responded negatively to N.
The N additions resulted in reduced soil pH; from wet to dry systems,
the relative importance of N additions on soil pH increased. Of the soil
variables, pH shared the strongest correlations with plant and bacterial
diversity metrics. Structural equation modeling showed that the indirect
effects (as opposed to direct effects) of N additions—changes in soil
pH and plant communities—were the strongest determinants of soil
bacterial community responses.
Zena etal. (2016)
Temperate
steppe
Inner Mongolia,
China
Soil
microbial
communities
Elevated N
6 yr of NH4NO3
additions at 0,
60, 120 and
240 kg N/ha/yr
Soil pH (0-10 Plant community biomass was found to significantly increase, species
cm and 10-20 richness decrease, and N concentrations increase with N additions,
cm) Seven bacterial phyla (mainly rare phyla) in the 0-10-cm soil depth
were significantly changed with N additions, while only 1 phylum in the
10-20-cm depth was affected by the additions of N. Hierarchical
structural equation modeling revealed that changes in bacterial
community composition were due to changes in soil pH and plant
composition, while shifts in bacterial richness were attributed to NH4+
concentrations.
5-61
-------
Table 5-6 (Continued): Mode of action for acidifying nitrogen and sulfur deposition.
Reference
Ecosystem
Type/Region
Species
N and S
Deposition/
Additions
Soil Indicator
Description
Zhang et al. (2016b)
Temperate
steppe
Inner Mongolia,
China
57 vascular
plant
species
Elevated N
6 yr of NH4NO3
additions at 0,
1, 2, 3, 5, 10,
15, 20, and
50 g N/m2/yr
(2x and 12x per
year)
Soil pH (0-10 N additions, at both frequencies, significantly decreased the number of
cm) new species gained and increased the number of old species lost.
However, the number of new species gained was lower on the low
frequency of N addition plots compared to the high frequency, while the
number of old species lost was not affected by N addition frequencies.
There was a negative correlation between the cumulative gain of new
species and soil pH, NH4 concentrations and community biomass
accumulation, while cumulative loss of old species was positively
correlated with these variables.
Basto et al. (2015b) Grassland
Peak District
National Park,
U.K.
found to be correlated with decreases in total and grass seed
abundance, declines in the persistence of H. pulchrum seed, and
declines in damage to C. rotundifolia seed. Seed germination was not
influenced by pH. In soil with pH higher than 5.6, indirect effects of pH
(through increased fungal pathogens) appeared to decrease the
persistence of the seed of all three grassland species. This study
suggested that: acidic soils are associated with increased seed
persistence; the longevity and size of grassland seed banks decline as
soil pH increases; and that pH indirectly influences seed persistence.
261 NA Soil pH Studied seed bank and seed germination, viability, and damage
grassland (gradient with (through seed burial experiment conducted with Scabiosa columbaria,
species; soil pH range of Hypericum pulchrum, and Campanula rotundifolia) along a natural pH
seed bank 3.5 to 6.5) gradient from acidic to calcareous grasslands. Increasing soil pH was
Al = aluminum; BAI = basal area increment; Ca = calcium; CASTNET = Clean Air Status and Trends Network; Cd = cadmium; CEC = cation exchange capacity; CEH = Centre for
Ecology and Hydrology; cm = centimeter; DOC = dissolved organic compound; Fe = iron; FIA = Forest Inventory and Analysis; g = grams; H+ = hydrogen; ha = hectare;
K = potassium; kg = kilograms; m = meter; meq = milliequivalent; mg = milligrams; Mg = magnesium; Mn = manganese; mol = mole; N = nitrogen; NA = not applicable;
NADP = National Acid Deposition Program; P = phosphorus; PLFA = phospholipid fatty acid: N03" = nitrate; S = sulfur; S042" = sulfate; yr = year; Zn = zinc.
5-62
-------
APPENDIX 6 TERRESTRIAL ECOSYSTEMS:
NITROGEN ENRICHMENT EFFECTS
ON ECOLOGICAL PROCESSES
Appendix 5 examined the biological effects of terrestrial acidification from nitrogen (N)
and sulfur (S) deposition. This appendix characterizes the biological effects of terrestrial
N enrichment that can be caused by atmospheric N deposition. Following an introduction
(Appendix 6.1). this appendix is composed of five major sections: effects on physiology,
growth, and productivity (Appendix 6.2); changes in biodiversity and community
composition (Appendix 6.3); climate modifications of N enrichment effects
(Appendix 6.4); critical loads (Appendix 6.5); and a summary (Appendix 6.6). The first
two sections begin with an introduction (Appendix 6.2.1 and Appendix 6.3.1) reviewing
the previous causal determination and presenting the current causal determination.
Following this introduction, an overview is presented of the mechanisms operating across
ecosystems to link N enrichment to biological change (Appendix 6.2.2 and
Appendix 6.3.2). Appendix 6.2. Appendix 6.3. and Appendix 6.5 are further divided into
subsections based on ecosystem type (e.g., forests, grasslands, etc.) or functional group
(e.g., lichens, trees, herbaceous plants). Finally, Appendix 6.6 provides a summary of the
new information generated since the 2008 ISA for Oxides of Nitrogen and
Sulfur—Ecological Criteria (hereafter referred to as the "2008 ISA").
6.1 Introduction
Nitrogen (N) is a key element required by all organisms in order to build amino acids and
nucleic acids, the basic biochemical subunits needed to synthesize the proteins, enzymes,
RNA, and DNA sustaining all biological processes. By the second half of the 19th
century, before these biochemical pathways had been identified, it was already
understood that N was a component of plant and animal tissues and a supply of N was
essential for plant growth (Galloway et al.. 2004). Indeed, the ability of added N to
stimulate plant growth had been recognized by science (and commerce) for over a
century (Galloway et al.. 2004; Galloway and Cowling. 2002) prior to the 2008 ISA. By
2008, it was already clear that N availability broadly limited productivity in terrestrial
ecosystems.
By 2008, it was also clear N availability could alter the biodiversity of terrestrial
ecosystems. Broadly, the effects ofN deposition on the diversity of terrestrial ecosystems
stem from four mechanisms [sensu Bobbink et al. (2010)1: (1) eutrophication,
(2) acidification, (3) direct toxicity and damage, and (4) increased susceptibility to
6-1
-------
secondary stress. Multiple mechanisms likely operate simultaneously to alter diversity
within a particular community or ecosystem. As described in Appendix 5. N deposition
may lead to soil acidification, which can have negative effects on plants and
microorganisms. Direct toxicity and damage from N deposition often come from an
accumulation of NH44" in soils and plant tissues. Increased susceptibility to secondary
stressors includes greater impacts of pathogens (Bobbink et al.. 2010) and shifts in
herbivory as a result of altered tissue chemistry (Throop and Lerdau. 2004). Among these
four major mechanisms, eutrophication is perhaps the most complex because it can
change the physiology of individual organisms, alter the relative growth and abundance
of species, transform relationships between species, and indirectly affect the availability
of other essential resources such as light, water, and nutrients rSuding et al. (2005); Clark
et al. (2007); Hautier et al. (2009); see also Appendix 41. These can lead to biodiversity
shifts, including community compositional changes, the loss of species, and decline in
species richness.
Since the 2008 ISA, the effects of N deposition on terrestrial ecological processes and
biodiversity have continued to be widely studied. With the increasing volume of research,
a number of new studies have been conducted using meta-analysis to synthesize
published observations or using continental or global data sets to understand broad-scale
patterns. New studies have provided a more detailed understanding of how N influences
terrestrial ecosystem growth and productivity; community composition and biodiversity;
and sensitive organisms and ecosystems. Further, a large body of work has been
published on critical loads (CLs) for N since 2008. Together with the information
available in the 2008 ISA, this body of evidence is sufficient to infer a causal
relationship between N deposition and (1) the alteration of the physiology and
growth of terrestrial organisms and the productivity of terrestrial ecosystems; and
(2) the alteration of species richness, community composition, and biodiversity in
terrestrial ecosystems.
6.2 Linking Nitrogen Deposition to Changes in Physiology,
Growth, and Productivity in Terrestrial Ecosystems
6.2.1 Introduction
In the 2008 ISA, evidence was sufficient to infer a causal relationship between N
deposition and the alteration of the terrestrial carbon (C) and N biogeochemical cycles.
These effects included not only changes in soil C and N pools and fluxes (described in
6-2
-------
Appendix 4). but also significant alterations of plant and microbial growth and
physiology. The 2008 ISA built upon the conclusions of the 1993 Oxides of Nitrogen Air
Quality Criteria Document. By 1993, a series of hypotheses regarding the effects of
chronic N deposition on northern temperate forests had already been developed from
early field observations (Aber et al.. 1989). Briefly, N deposition was expected to
increase tree growth in most forests, but once plant demand for N had been satisfied
(biological and physical sinks had reached "N saturation"), further N deposition could
lead to nitrification, soil cation leaching, acidification, nutrient deficiencies, and
decreased tree growth rAber et al. (1989); described in Appendix 4 and Appendix 51. A
revised form of these N saturation hypotheses (Aber et al.. 1998) provided much of the
conceptual foundation in the 2008 ISA for understanding how N deposition influenced
plant physiology, growth, and ecosystem productivity.
The effects of N deposition on terrestrial ecological processes have been widely studied
since 2008. Research on N deposition has continued in North America and Europe, with
other areas receiving less attention [e.g., temperate forests in Asia and the Southern
Hemisphere (Gilliam. 2016)1. A significant new body of research has developed in the
boreal, arid, and subtropical ecosystems of Asia, particularly in China [e.g., Du and Fang
(2014); Du et al. (2014b); Du et al. (2014a); Sun et al. (2014); Zhang et al. (2015e)l. New
meta-analyses have provided a more detailed understanding of how added N affects
productivity responses in different biomes (LeBauer and Treseder. 2008). growth among
plant functional types (Xia and Wan. 2008). growth of individual plant parts (Xia and
Wan. 2008). root growth (Li et al.. 2015). and ecosystem C storage (Liu and Greaver.
2009). while other broad-scale analyses have examined changes in plant N concentrations
(Xia and Wan. 2008). microbial biomass (Treseder. 2008). and belowground C cycling
(Liu and Greaver. 2010). This expanded body of research has created a better
understanding of how N influences processes at molecular to global scales. Together with
the information available in the 2008 ISA, this body of evidence is sufficient to infer a
causal relationship between N deposition and the alteration of the physiology and growth
of terrestrial organisms and the productivity of terrestrial ecosystems.
6.2.2 Mechanisms Operating across Terrestrial Ecosystems
The 2008 ISA evaluated a large number of studies assessing how N deposition has
changed terrestrial C cycling and found an array of ecological responses. The most
extensive evidence of the effect of N deposition on C cycling was available for forests in
North America and Europe. In experimental N addition studies, moderate to high
additions of N led to either no significant change in tree growth rates or transient growth
increases (generally at deposition rates lower than 10 kg N/ha/yr), followed by increased
6-3
-------
mortality, especially at higher rates of N addition. This group of studies showed
coherence in effects and indicated the presence of a biological gradient in responses with
increasing N deposition.
Some of the broad-scale changes caused by N additions to terrestrial ecosystems have
been quantified by meta-analysis, a data synthesis tool that started to become common in
ecological research beginning in the late 1990s (Gurevitch et al.. 2001). Using this tool to
understand the effects of N additions on herbaceous plant communities, (Gough et al..
2000) found that N additions stimulated aboveground plant productivity by an average of
53%. In a broader synthesis, Elser et al. (2007) conducted a meta-analysis of N addition
effects on plant community productivity and observed average increases in productivity
of 20-30% in grasslands, forests, tundra, and wetlands. In addition to changes in plant
growth, Koricheva et al. (1998) identified via meta-analysis that added N could alter the
chemistry of plant tissues, including increasing tissue concentrations of N and free amino
acids, while decreasing concentrations of starch and C based secondary compounds
(important for defense against herbivores), including phenylpropanoid compounds. In a
meta-analysis of the belowground effects of N, Treseder (2004) found N additions
decreased both the abundance of mycorrhizal fungi and the percentage of plant roots
colonized by mycorrhizal fungi. In addition, multiple lines of evidence showed that N
deposition increases the performance of insect herbivores, and potentially, insect
populations (Throop and Lerdau. 2004). In a synthesis of 500 observations of the effect
of N on litter decomposition rates, Knorr et al. (2005) found that added N stimulated
decomposition at sites with low rates of ambient N deposition (<5 kg N/ha/yr), but
slowed decomposition at sites with moderate rates of N deposition [5-10 kg N/ha/yr;
Knorr et al. (2005)1 and that N additions at rates from 2 to 20 times ambient N deposition
inhibited decomposition by 8 to 16%.
These changes can cause a cascade of ecological consequences. For instance, the
inhibition of decomposition can increase soil C content, but the accumulation of plant
litter on the soil surface can inhibit the establishment of some plant species (Patterson et
al.. 2012; Cleavitt et al.. 201 lb; Xiong and Nilsson. 1999; Facelli and Pickett. 1991). In
the western U.S., Fenn et al. (2003a) suggested greater plant growth caused by N
deposition could increase plant litter accumulation, in turn increasing the susceptibility of
forests and other wildlands to severe wildfires. Nitrogen deposition also affects the
patterns of C allocation because most of the additional plant growth occurs aboveground.
This increases the shoot-to-root ratio, which can be detrimental to plants because of
decreased resistance to environmental stressors, such as drought and wind (Braun et al..
2003; Krupa. 2003; Minnich et al.. 1995; Fangmeier et al.. 1994).Thus, it was recognized
in the 2008 ISA that the effects of N deposition on biological and ecological processes in
6-4
-------
terrestrial ecosystems were pervasive, complex, and difficult to fully understand and
predict.
Although evidence showed plant productivity responses to N deposition as of the 2008
ISA, the extent to which N increased forest C content in C budget measurements
(quantified input and output of C to the ecosystem) was uncertain. A prominently
reported estimate of >600 kg C sequestered per kg of N deposited to forests, which was
based on a regional gradient technique (Magnani et al.. 2007). was widely criticized.
Subsequent reassessments of these data suggested that forest C sequestration was an
order of magnitude lower (Sutton et al.. 2008). Liu and Greaver (2009) found in a
meta-analysis that N additions ranging from 25.5 to 200 kg N/ha/yr increased forest
ecosystem C content (+6%).
In addition, fewer studies had examined the effects of N deposition on the eutrophication
of nonforested ecosystems. Mack et al. (2004) examined C and N pools in a long-term N
addition experiment at the Arctic Long-Term Ecological Research site near Toolik Lake,
AK. Plant growth increased as a result of N additions, but the N additions enhanced
decomposition of belowground C pools in deep soil layers more than it increased primary
production, leading to a substantial net loss of C from this ecosystem. More broadly, Liu
and Greaver (2009) conducted a meta-analysis of 16 observations from nine publications
to evaluate the relationship between N addition (16 to 320 kg N/ha/yr) and C
sequestration in nonforest ecosystems and did not find a significant effect on net
ecosystem exchange (NEE, kg C/ha/yr).
Xia and Wan (2008) identified nearly 1,600 observations of plant biomass growth in
response to N additions, excluding agricultural and horticultural species (Figure 6-1 A).
Overall, plant biomass increased by 54%, with a larger gain in seeded plants (+55%) than
spore plants (+21%; not shown). Among the seeded plants, grasses showed the largest
biomass response, followed by trees, forbs, and shrubs. Annual herbs (+92%) showed a
stronger response than perennial herbs (+56%). Overall, biomass responses to N
increased linearly with mean annual precipitation (MAP). The studies included in the
analysis had addition rates ranging from 10 kg N/ha/yr to 1,000 kg N/ha/yr. When the N
additions were divided into low (<100 kg N/ha/yr) and high (>100 kg N/ha/yr) groups,
there were only a few differences among functional groups. The response of woody
species was lower than for herbaceous species at the low addition rate (+25 vs. +51%),
but the groups had similar responses at the high N addition rate. Among herbaceous
plants, both legumes and nonlegumes had positive responses to low rates of N additions
(19 and 23%, respectively), while the high rate of N additions increased nonlegume
biomass (+43%) but did not significantly affect legume growth. The authors also
examined how these responses varied in response to the rate of N additions. The N
6-5
-------
addition studies were divided into 50 kg N/ha/yr increments from 0 to 300 kg N/ha/yr,
but aside from increasingly positive growth responses for trees and woody plants with
greater N additions, there were few apparent trends, and sample sizes were inconsistent.
In their analysis, Xia and Wan (2008) separated plant growth data into aboveground
(887 observations) and belowground (340 observations) components. Across almost all
functional groups, there were greater increases in biomass in aboveground than
belowground components (Figure 6-1 A; see also Appendix 4). In a subsequent
meta-analysis of plant C pools, Lu et al. (2011b) also found that N additions to plants in
nonagricultural systems stimulated shoot C by 28.5% (n = 146) and root C by 20%
(n = 77). These results indirectly suggest that plant C allocation changes in response to N
addition. These changes were also shown by a more direct analysis of a much smaller
meta-analysis data set (n = 15) in which Li et al. (2015) observed a significant 11%
decrease in the root:shoot biomass ratio. Together, these results provide further support to
the conclusion in the 2008 ISA that N additions can alter plant C allocation and result in
elevated shoot:root ratios. Further, Lu et al. (2011b) analyzed the response of some
individual plant parts to added N and observed a 1.6% increase in leaf C mass.
Belowground processes are important components of terrestrial ecosystems. As detailed
in Appendix 5. soils are often the largest ecosystem pools of both C and N; soil
respiration can be the largest ecosystem efflux of CO2; and soils can be important sources
and sinks for the greenhouse gases N2O and CH4. Biologically, the large C fluxes plants
allocate belowground for root exudation, the growth and maintenance of roots, and the
support of mycorrhizal fungi help support complex belowground food webs. Less
information is available about belowground responses to N additions than aboveground
responses. However, Li et al. (2015) recently conducted a meta-analysis of how root traits
respond to N additions. Notably, Li et al. (2015) found that although total root biomass
increased (+20%; n = 74), fine root biomass declined (-13%; Figure 6-2B). This
discrepancy is apparently accounted for by an increase in the biomass of coarse roots
(+57%; n = 7), which are typically structural and conductive tissues. Morphological traits
of fine roots, such as length (n = 25), diameter (n = 10), and specific root length (n = 22),
were unaffected by N additions. Fine root turnover rate was 21% higher (n = 12), but fine
root production was not significantly affected by the N additions [n = 22; Li et al.
(2015)1.
Aboveground net primary productivity (NPP), which measures aboveground plant growth
at the community scale, was less responsive on average (+29%) than individual plant
growth to N additions I LeBauer and Treseder (2008); Figure 6-IB and Figure 6-2A1.
Consistent with the plant functional group analysis conducted by Xia and Wan (2008).
LeBauer and Treseder (2008) observed a large stimulation of aboveground NPP in
6-6
-------
grass-dominated ecosystems such as temperate grasslands (+53%), tropical grasslands
(+26%), and tundra (+35%), whereas NPP increased only 19% in temperate forests.
There were also significant NPP increases in tropical forests and wetlands. Added N did
not significantly increase NPP in deserts, but the analysis only included three
observations.
More recently, Tian et al. (2016a) synthesized aboveground NPP data from studies that
had experimentally added N at more than one rate in order to examine how plant growth
responses per unit N changed with increasing levels of N. Their metric, termed "N
response efficiency" [100 * (ANPPtreatment - ANPPControi)/ANPPControi/N addition rate] did
not significantly differ across the three ecosystem types (wetlands, forests, grasslands),
with an average increase of 3 to 4% in NPP per g of added N. The N response efficiency
decreased with N addition rates above 50 kg N/ha/yr, consistent with a saturating
response to N.
Liu and Greaver (2009) had smaller data sets, with 16 observations of net ecosystem
exchange (NEE; all in nonforested systems) and 17 observations of ecosystem C content
(Figure 6-2). There were no significant effects of added N on NEE overall or within the
individual biomes included in the analysis (grasslands, wetlands, tundra). However,
added N did increase forest ecosystem C content by 6%. Belowground NPP responses
have not been synthesized, at least in part, because data are lacking (LeBauer and
Treseder. 2008).
In terms of altered physiology, there is also widespread evidence that N additions
increase plant tissue N concentrations. The meta-analysis by Xia and Wan (2008)
included changes in plant N concentrations overall, aboveground, and belowground.
Overall, N additions increased plant N concentrations by an average of 28.5% (Xia and
Wan. 2008). Although there was significant variation among plant functional groups, the
average increase in tissue N concentration among seeded plants was much more similar
across functional groups than the biomass growth response, ranging only from +24 to
+35% once legumes (+7%) and the two broader functional groups containing legumes
(forbs, +14%; perennial herbs: +22%) were excluded. Likewise, belowground and
aboveground plant tissues had a similar change in N concentration, in contrast to the
varying biomass growth responses to N (Xia and Wan. 2008). Li et al. (2015) also
observed that N additions significantly increased root N concentrations in a root trait
meta-analysis (+18%; n = 84). Because of these increased N concentrations, N additions
can significantly decrease plant C:N ratios (Yue et al.. 2017).
6-7
-------
NPP = net primary productivity.
Notes: (A) Mean change in plant biomass growth in response to N additions redrawn from Xia and Wan (20081. with blue bars
representing the overall response, orange bars representing aboveground growth, and grey bars representing belowground growth.
Vertical grey lines identify different ways the data were parsed (i.e., all seeded plants; woody vs. herbaceous plants, etc.). (B) Mean
change in aboveground net primary productivity data from LeBauer and Treseder (20081. Error bars represent the 95% confidence
interval. Numbers above the error bars indicate the number of observations included in the analysis.
Figure 6-1 Effects of nitrogen additions on plant growth and net primary
productivity.
6-8
-------
B
60
50
40
g30
a>
.ti 20
Ł10
0
-10
-20
-30
60
50
40
-r-30
*->
u
Ł 10
0
-10
-20
-30
Total
Pool
Below
Ground
Pools
37
Above
Ground
Biomass
Net Ecosystem
Flux Input and
Output
Fluxes
53
m
Soil
Organic SoilC DOC
Horizon
57
31
Mycorrhiza
Microbial
C
Total
Plant
Fluxes
126
16
No No
AnalysisAnalysis
No No
AnalysisAnalysis
Below
Ground
Fluxes
500
No
GPP R Above Below HlT"
Ground Ground jutocoph
NPP NPP
Decomposition
^3 Analysis ^7
•utoropfc
10^-'
hetr-oroph
NEE
DOC = dissolved organic carbon; NEE = net ecosystem exchange; NPP = net primary production; GPP = gross primary production;
Recosystem = ecosystem respiration; Recosystem autotroph = plant respiration; Rsoii = soil respiration; Rsomheterotroph = heterotroph soil
respiration.
Notes: mean effect sizes from meta-analyses of N addition experiments on ecosystem, plant, and soil pools ([a] top panel) and
fluxes ([b] bottom panel). Error bars represent 95% confidence intervals. Numbers above the bars are the sample sizes for each
analysis. Letter with each bar denote the data source. No Analysis denotes a pool or flux that has not yet been meta-analyzed.
Data sources: (A) Liu and Greaver (20091: (B) Xia and Wan (20081: (C) Lu et al. (2011 bl: (D) Li et al. (20151: (E) Liu and Greaver
(20101: (F) Treseder (20041: (G) LeBauer and Treseder (20081: (H) Knorretal. (20051.
Figure 6-2
Effects of added nitrogen on ecosystem carbon pools and fluxes.
6-9
-------
The 2008 ISA linked enhanced terrestrial productivity to increases in photosynthesis and
gross primary productivity. Gross primary productivity (GPP) can increase as a result of
either a greater amount of photosynthetic tissue (more light absorption) or a higher rate of
photosynthesis per leaf. There is a strong mechanistic and conceptual link between
greater foliar N concentrations, higher leaf concentrations of the C assimilating enzyme
rubisco (Evans. 1989). and greater maximum rates of leaf-level photosynthesis in
vascular plants (Wright et al.. 2004). Consequently, the increase in foliar N caused by N
additions has been linked to increases in the leaf-level rate of photosynthesis for decades
[e.g., Teskev et al. (1994) and references therein]. Alternately, evidence cited in the 2008
ISA suggests that much of the increased foliar N observed when N is added may be
physiologically inactive because it manifests as an increase in storage compounds such as
free amino acids (Bauer et al.. 2004). While increases in photosynthesis in response to N
additions have been observed in trees, grasses, and shrubs, these increases have not been
consistent [e.g., Gulmon and Chu (1981); Laitha and Whitford (1989); Newman et al.
(2003); Chen et al. (2005b); Elvir et al. (2006); Talhelm et al. (2011); Pivovaroff et al.
(2016)1 and there does not appear to be a meta-analysis or other synthesis on the response
of leaf-level photosynthesis or GPP (Figure 6-2B) to N additions. Similarly, there is a
strong fundamental relationship between tissue N concentration and respiration rates in
plants (Reich et al.. 2008; Reich et al.. 2006; Ryan et al.. 1996). such that this relationship
is used to model respiration rates [e.g., Amthor (2000); Hanson et al. (2004)1. However,
there is evidence that this relationship can breakdown in N addition studies
[e.g., Schaberg et al. (1997); Drake et al. (2008); Burton et al. (2012)1. and there are
currently no broad analyses on the effects of N additions on ecosystem or plant-scale
autotrophic respiration (Figure 6-2B).
The 2008 ISA noted some observations of decreased microbial biomass as a result of
added N, particularly for mycorrhizal fungi ITreseder (2004); Figure 6-2B1. Since 2008,
however, it has become increasingly clear that N deposition can greatly impact microbial
communities, often including a decrease in microbial biomass. Meta-analyses conducted
across all ecosystem types have found that N additions can decrease microbial biomass
(Treseder. 2008). microbial biomass C (Liu and Greaver. 2010). and microbial biomass N
(Lu et al.. 2011b) and that the effects of added N on microbial biomass increase with the
duration of N additions and the total amount of N added (Treseder. 2008). These changes
are noteworthy given the diverse role of microorganisms in nutrient cycling, greenhouse
gas fluxes, and other ecosystem services. Although N additions frequently decrease
microbial biomass, the results are not always consistent (Liu and Greaver. 2010;
Treseder. 2008). The effects of N were not significant at the level of individual microbial
domains (bacteria, fungi), although there were trends toward increasing negative effects
of N additions on fungal biomass as duration and cumulative N load increased (Treseder.
2008). In a meta-analysis, Carey et al. (2016) observed that N additions increased the
6-10
-------
abundance of ammonia-oxidizing bacteria across nonagricultural ecosystems, but did not
affect the abundance of ammonia-oxidizing archaea. Recently, Li et al. (2015) provided a
new meta-analysis of N addition effects on mycorrhizal colonization of fine root tips and
observed an overall 17% decline (n = 21). While this reinforces the conclusions of an
earlier meta-analysis by Treseder (2004). there are important qualifiers to this finding: the
sample only included observations from forests and grasslands in temperate and
subtropical climates; significant decreases were only observed in temperate and forested
ecosystems, after more than 3 years of N additions, or at N addition rates
>100 kg N/ha/yr (Li et al.. 2015).
Changes in mycorrhizae in turn can affect plant productivity. Mycorrhizal fungi provide
benefits to plants, yet also carry costs. In exchange for nutrients and water from the
fungus, the plant provides C from photosynthesis (Hogberg et al.. 2010; Rillig. 2004). In
many cases, this tradeoff works to the plant's advantage. For instance, Van der Heiiden et
al. (1998) observed higher biomass for most of the individual plant species in their study
when inoculated with arbuscular mycorrhizae versus when mycorrhizae were absent.
They suggested the mycorrhizal acquisition of a limiting nutrient, in this case phosphorus
(P), for the plant in exchange for C explained this finding. However, under conditions in
which the nutrient limitation of the plant is relieved (e.g., high N availability),
mycorrhizae may no longer be as needed by the plant for nutrient acquisition, yet
concomitantly still impose a C cost. In this case, the tradeoff is no longer as beneficial to
the plant. With increased N availability, plants that make fewer mycorrhizal associations
can benefit both from the physiological advantages accompanying an alleviation of N
limitation (e.g., increases in photosynthesis, changes in plant chemistry) and from
incurring a lower C cost for mycorrhizae (Johnson et al.. 2008). Indeed, shifts in C
allocation away from mycorrhizae may be a mechanism to support increased plant
productivity in situations where plants do not show gains in photosynthesis [e.g., Talhelm
et al. (2011)1.
There are numerous mechanisms through which N deposition could impact microbial
biomass, including changes in soil chemistry, changes in the rates of aboveground and
belowground plant C inputs (including litter production, root exudates, and C supplied to
mycorrhizal fungi), and changes in plant tissue chemistry (Treseder. 2008). Nitrogen
additions can change the chemistry of litter inputs by altering the tissue composition of
plant species (Throop and Lerdau. 2004) or by changing the composition of plant
communities contributing to the ecosystem flux of plant litter [e.g., Manning et al.
(2008)1. Within a particular species, N additions can cause changes in both of the
dominant forms of plant litter: leaf litter and fine roots [e.g., Xia et al. (2015)1. These
changes can be relatively direct, such as increases in tissue concentrations of inorganic
and organic forms of N [e.g., Koricheva et al. (1998); Bauer et al. (2004)1. While leaf
6-11
-------
litter N concentrations are not as widely measured as green leafN, increases in leaf litter
N concentrations due to added N are also likely to be widespread because increased N
availability often either decreases or has no effect on the fraction of N plants absorb from
leaves during senescence (Aerts. 1996). Other changes in litter chemistry, such as
increases in condensed tannins, soluble phenolics, and nonstructural carbohydrates (Xia
et al.. 2015). likely involve more complex biological mechanisms. In a meta-analysis of
plant biochemistry data, Liu et al. (2016a) observed different effects of added N on trees
and herbs. For the live tissues of trees, N additions decreased concentrations of lignin,
cellulose, and nonstructural carbohydrates, and increased the concentration of protein. In
the live tissues of herbs, added N decreased the concentrations of nonstructural
carbohydrates and hemicellulose, but increased lignin, cellulose, and protein
concentrations. These changes have implications for herbivores and detritivores, as well
plant growth dynamics.
In a litter decomposition meta-analysis, Knorr et al. (2005) observed that the effects of
added N also differed based on the length of the study and the initial litter lignin content.
Added N stimulated decomposition in studies lasting less than 2 years, but reduced
decomposition in studies lasting more than 2 years. High-lignin litter decomposed more
slowly than low-lignin litter. Because lignin concentrations tend to increase during litter
decomposition, both of these results were considered to be consistent with earlier
evidence that increased N supplies stimulate microbial activity responsible for the
decomposition of labile litter constituents, while suppressing the microbial production of
extracellular enzymes responsible for the degradation of lignin rKnorr et al. (2005); see
also Appendix 41. In a meta-analysis of biochemistry data for decomposing litter, Liu et
al. (2016a) observed that N additions increased the concentration of lignin in
decomposing litter for herbs and shrubs, but did not have a significant effect on
decomposing tree litter. In comparison, N additions decreased cellulose concentrations in
decomposing tree litter, but did not significantly affect cellulose concentrations in
decomposing herb litter. In another meta-analysis, Liu and Greaver (2010) found that
although N additions increased aboveground litter production (+20%; n = 37; not shown)
across all biomes and N addition rates, there were no significant overall changes in either
total soil respiration (Figure 6-2B) or heterotrophic soil respiration (Figure 6-2B). In fact,
both forms of soil respiration tended to decline. Currently, no similar cross-biome
analysis has been conducted for changes in autotrophic soil respiration (respiration from
roots and mycorrhizal fungi), but Janssens et al. (2010) noted in a meta-analysis that
autotrophic respiration was suppressed by N additions in forests.
Because some biogeochemical processes involve specific chemical forms of N
(e.g., denitrification, ammonium toxicity; see Appendix 4). there is the potential that
biological responses to N deposition (or N addition) could depend on whether the
6-12
-------
dominant form of deposited N is oxidized (NOy) or reduced (NHX). A number of studies
have specifically addressed this issue, either by conducting experiments directly testing
additions of different forms of N or indirectly through syntheses comparing the effects of
NOy and NHx in different experiments. Different responses to individual forms of N
have been observed for some endpoints, such as increases in dissolved organic C,
decreases in ecosystem N retention, increases in soil N2O emissions, plant growth, and
defense against pathogens rLiu and Greaver (2009); Liu and Greaver (2010); Templer et
al. (2012); Verhoeven et al. (2011); Yue et al. (2016); Mur et al. (2016); see also
Table 4-131. By contrast, other studies have failed to observe a difference between the
effects of N forms. One direct test occurred in the Front Range of the Rocky Mountains
in Colorado, where Ramirez et al. (2010a) investigated whether soil microbes respond
differently to additions of NH44" versus NO;, and found that the total amount of N added
was correlated with a decrease in soil respiration, not the form of the N (NH44" vs. NO;, ).
A more comprehensive understanding is available by reviewing the results of
meta-analyses comparing the responses of N addition experiments conducted with
different forms of N (Table 6-1). With notable exceptions, most often differences in the
effect of the form of N were not observed in these meta-analyses. Moreover, studies
finding differences tended to occur where sample sizes were small [e.g., belowground C
pools in Yue et al. (2016)1.
6-13
-------
Table 6-1 The effects of different forms of inorganic nitrogen on biological
endpoints according to meta-analyses.
See Table 4-13 for the effects of different forms of inorganic nitrogen on
biogeochemical processes and indicators.
Reference
Endpoint
Effect of NOy vs. NHx Forms
Yue et al. (2016)
Aboveground plant C pool
Not significant
LeBauer and Treseder (2008)
Aboveground plant productivity
Not significant
Yue et al. (2016)
Aboveground plant productivity
Not significant
Treseder (2008)
Bacteria biomass
Not significant
Yue et al. (2016)
Belowground plant C pool
Increase with NHV
Liu and Greaver (2009)
Ecosystem C content
Not significant
Liu and Greaver (2010)
Fine root litter production
Not significant
Treseder (2008)
Fungal biomass
Not significant
Liu and Greaver (2010)
Leaf litter production
Insufficient data
Yue et al. (2016)
Leaf litter production
Insufficient data
Knorr et al. (2005)
Litter decomposition
Not significant
Yue et al. (2016)
Litter decomposition
Not significant
Treseder (2008)
Microbial biomass
Not significant
Liu and Greaver (2010)
Microbial biomass C
Decrease with NhV
Yue et al. (2016)
Microbial biomass C
Not significant
Treseder (2004)
Mycorrhizal abundance
Not significant
C = carbon; 15N = tracer isotope of nitrogen; N20 = nitrous oxide; NH4+ = ammonium; NH4N03 = ammonium nitrate; NHX = sum of
reduced forms of N; N03" = nitrate; NOY = the sum of oxidized forms of nitrogen.
Notes: References ordered by endpoint. Only statistically significant differences between the effects of forms of N listed as
increases or decreases.
Before the 2008 ISA, neither terrestrial N cycling nor anthropogenic N deposition had
been widely incorporated into Earth systems models (ESMs) used to understand and
forecast global climate and biogeochemical cycling. Thornton et al. (2007) made the first
effort to understand how both coupled C and N cycling and anthropogenic N deposition
would impact ESM predictions for terrestrial C uptake by inserting a new land
biogeochemistry model into a coupled climate system model. In the resulting model
6-14
-------
output, N limitation greatly decreased the amount of terrestrial C uptake predicted from
future increases in atmospheric CO2 and decreased the sensitivity of terrestrial C
sequestration to increases in temperature and precipitation (Thornton et al.. 2007).
Including N deposition within the model directly increased terrestrial C uptake and also
indirectly increased terrestrial C uptake by removing some of the N limitation predicted
to occur under increased atmospheric CO2 (Thornton et al.. 2007). Subsequently, efforts
to model biogeochemical processes at regional and global scales have expanded (Thomas
et al.. 2015; Zaehle and Dalmonech. 2011). Although the integration of terrestrial N
cycling into ESMs remains relatively weak, more ESMs are incorporating the coupling of
terrestrial C and N cycling as overall model development and sophistication advances
(Thomas et al.. 2015; Arora et al.. 2013; Zaehle and Dalmonech. 2011; Bonan and Levis.
2010; Gerber et al.. 2010). Consistent with the findings of Thornton et al. (2007).
inclusion of coupled C-N cycling in ESMs has two primary effects: decreasing the
stimulatory effects of elevated atmospheric CO2 on terrestrial productivity and decreasing
the sensitivity of terrestrial C sequestration to climate warming because increased soil N
mineralization stimulates plant productivity (Arora et al.. 2013; Zaehle and Dalmonech.
2011; Zhang et al.. 201 lb; Arneth et al.. 2010; Bonan and Levis. 2010; Gerber et al..
2010; Yang et al.. 2010; Zaehle et al.. 2010; Thornton et al.. 2009). However, the ESMs
that do include basic terrestrial C-N coupling lack more recently identified interactions
such as plant organic N uptake, soil priming (root exudation), and the suppression of litter
decomposition at high soil N availability, all of which could increase terrestrial C uptake
to varying extents (Thomas et al.. 2015; Zaehle and Dalmonech. 2011). In addition, the
interactions between and the net effects of N, precipitation, and temperature on
ecosystem C response in soils are unknown in many cases (see Appendix 13.1.2.1).
In the few ESMs that have directly included the effects of N deposition, the additional N
increased terrestrial C uptake and increased the extent to which elevated atmospheric CO2
stimulates terrestrial C uptake (Devaraiu et al.. 2016; Bonan and Levis. 2010; Yang et al..
2010). In addition, N deposition in the Northern Hemisphere (particularly in the U.S.)
apparently compounded the increase in C sequestration caused by the regrowth of
secondary forests during the late 20th century following agricultural abandonment and
timber harvest (Gerber et al.. 2013; Yang et al.. 2010). However, ESMs that do not
include potential N saturation may overestimate the effect of N deposition on terrestrial C
uptake in regions experiencing high N deposition rates (Lu et al.. 2016). In addition, it
should be noted that although N deposition and the overall anthropogenic production of
reactive N increases terrestrial C sequestration, they are not the only influence of
anthropogenic N on global climate. It is difficult to quantify the overall climate impact of
anthropogenic N [e.g., Pinder et al. (2013)1 because reactive N can change the planetary
albedo by enhancing aerosol formation, stimulate the production of biogenic greenhouse
gases, alter the production and destruction of methane and tropospheric ozone in the
6-15
-------
atmosphere, and indirectly decrease terrestrial productivity due to the phytotoxic effects
of ozone (Lu and Tian. 2013; Pinder et al.. 2013; Pinder et al.. 2012; Arneth et al.. 2010;
Zaehle et al.. 2010). Many of these effects are discussed further in the atmospheric
chemistry and terrestrial biogeochemistry portions of this ISA (see Appendix 2 and
Appendix 4).
In addition to the global-scale analyses conducted using ESMs, biogeochemical process
models have been used to assess the impact of N deposition on terrestrial productivity
and C sequestration at national and regional scales. Tian et al. (2012) used the Dynamic
Land Ecosystem Model (DLEM) to model the influence of climate, tropospheric ozone,
fertilizer use, land use/land cover change, atmospheric CO2, and N deposition on
terrestrial C storage over the 20th century in the southeastern U.S. (encompassing Texas
and Oklahoma to Florida and Virginia). Terrestrial C storage increased from 1951-2007,
with the model identifying atmospheric CO2 and N deposition as the environmental
factors responsible for the increase in C sequestration (Tian et al.. 2012). In China, a
series of papers has been published using biogeochemical process models to identify how
this same set of environmental factors influenced terrestrial C cycling in that country.
Rates of N deposition in China grew rapidly during the late 20th century and were
considerably higher than in the U.S. by the early 21st century, with nationwide average
rates of approximately 20 kg N/ha/yr and rates in southeast China averaging 35-40 kg
N/ha/yr (Lu and Tian. 2013; Lu et al.. 2012). Tian et al. (2011) applied the Terrestrial
Ecosystem Model and the DLEM model to China using data from 1961-2005. In both
models, China was a C sink during this time period, with the combination of N deposition
and agricultural fertilizer use accounting for 61% of the increase in C sequestration.
Notably, the responsiveness of terrestrial C sequestration to N deposition has declined
since the 1980s as N deposition in China increased, providing evidence that terrestrial
ecosystems are becoming less N limited in China (Tian et al.. 2011). Lu et al. (2012)
conducted a similar analysis, using DLEM to understand how multifactor environmental
change influenced terrestrial C sequestration in China from 1901-2005. Like Tian et al.
(2011). Lu et al. (2012) observed that N deposition increased terrestrial C sequestration in
China throughout much of the late 20th century, but that the responsiveness of terrestrial
C sequestration to N deposition has declined since the 1980s. Moreover, Lu et al. (2012)
reported that all areas of China, aside from some shrublands and portions of western
China, are becoming N saturated.
6-16
-------
6.2.3
Forests
6.2.3.1 Aboveground Processes
The 2008 ISA noted a wide range of forest productivity responses to added N. Responses
to low levels of N were often positive because N availability tends to limit growth in
terrestrial ecosystems (LcBaucr and Treseder. 2008). However, forest productivity
responses to higher rates of N addition were neutral or negative [e.g., MagiUetaL
(2004)1. The effects of N deposition were variable across species. Conifer species,
particularly at high elevations, were more likely to exhibit negative growth responses or
mortality in response to added N. Conifer species were less likely to demonstrate
increased growth in response to additional N and more often exhibited decreased growth
and increased mortality [e.g., McNultv et al. (2005); Beier et al. (1998); Boxman et al.
(1998a) |. Differences between broadleaf and conifer species were especially clear in
long-term N addition experiments: Elvir et al. (2003) observed increased sugar maple
(Acer sacchamm) basal area growth in response to long-term (NH^SC^ (25 kg N/ha/yr
for 10 years) additions, but red spruce (Picea rubens) growth was unchanged. At Harvard
Forest, oak (Quercus velutina, O. rubra) increased growth in response to chronic N
additions (50 or 150 kg N/ha/yr for 15 years), while red pine (Pinus resinosct) growth
slowed and mortality increased (Magill et al.. 2004). Most empirical observations of the
effects on atmospheric N deposition on forest productivity came from chronic N addition
experiments in temperate forests in the U.S. [e.g., Aber et al. (1995); McNultv et al.
(1996); Elvir et al. (2003); Magill et al. (2004); Pregitzer et al. (2008)1 and temperate and
boreal forests in Europe (Hvvonen et al.. 2008; Hogberg et al.. 2006; Beier etal.. 1998;
Boxman et al.. 1998a). Empirical analyses of the effects of atmospheric N deposition on
forest productivity in the U.S. were lacking.
Research published since 2008 has reinforced many of the ideas in the 2008 ISA. There is
considerable evidence from deposition gradient studies, forest modeling, and long-term N
addition experiments that N deposition broadly stimulates tree growth and the
productivity of forested ecosystems, but that effects vary by species. Using forest
inventory data collected between 2000 and 2016, Horn et al. (2018) found N deposition
coincided with a small overall increase in tree growth and mortality. The particular
response though varied greatly between species. The study analyzed results for 94 tree
species, but ultimately focused on a subset of 71. Of the 71 species, 39 exhibited a
significant relationship between growth and N deposition. Twenty of the 39 had
increasing growth with greater N deposition across the full depositional range
experienced by that species. Seventeen displayed increasing growth at lower levels of N
6-17
-------
deposition but decreasing growth at higher levels. Two species (yellow birch [Betula
alleghaniensis\ and eastern hemlock [Tsiigct canadensis]) declined in growth consistently
with increasing N deposition. Mortality also varied between species, with 3 species
increasing in survival, 25 species increasing in survival at lower N deposition before
decreasing in survival at higher levels, and 6 species consistently declining in survival
(Horn et al.. 2018).
Similarly, other studies have shown variation in tree growth and mortality with N
deposition. Thomas et al. (2010) assessed tree species response to N deposition using
forest inventory data from the early 1980s through the mid-1990s. Of the 23 species
surveyed, N deposition accelerated growth in 11 species, including 3 of the 4 most
abundant species (red maple [Acer nibrum], sugar maple, and northern red oak | One reus
rubra]). Negative effects on growth were seen in three species, all of which were
evergreen conifers (red pine [Pinus resinosa], red spruce, northern white cedar [Thuja
occidental-is]). All five of the arbuscular mycorrhizal tree species included in the analysis
exhibited increased growth. Eight species exhibited higher mortality rates with increasing
N deposition in the Thomas et al. (2010) analysis, notably several oak species, including
northern red oak. Only three species showed increased survival. Additionally, Xia and
Wan (2008) observed positive effects of added N on growth for both broadleaf and
coniferous trees in a meta-analysis, with broadleaf trees (+73%) more responsive than
conifers (+37%).
Examining forest stand-level responses, Hember et al. (2017) concluded N deposition had
increased forest stand growth in one of the five Canadian ecozones (Montane Cordillera),
decreased it in another (Boreal Plain), and did not significantly affect stand growth in the
remaining three (Pacific Maritime, Boreal Shield, and Atlantic Maritime). They similarly
found differences in forest stand mortality, with N deposition decreasing mortality in
three of the five ecozones (Pacific Maritime, Montane Cordillera, and Boreal Plain) and
no significant effect in the remaining two. In an analysis of forest inventory data from
across the entire eastern U.S. from the 1970s through early 2000s, Dietze and Moorcroft
(2011). as noted in Appendix 5. found N deposition was linked to decreased tree
mortality in 9 of 10 plant functional types and increased mortality only in the northern
midsuccessional hardwoods functional type. Across a smaller gradient within the
Adirondacks, Bedison and McNeil (2009) found a significant positive overall effect of N
deposition on tree growth from 1984-2004, but positive growth effects only for red
maple, balsam fir (Abies balsamea), and red spruce at the species level. Thus, it appears
from these inventory analyses that while tree growth has generally been enhanced by N
deposition over the last several decades, individual species have exhibited variable
responses to N deposition in mortality and growth.
6-18
-------
Variation has also been shown in tree responses to experimental N additions. Overstory
sugar maple trees increased growth in response to added N in Michigan [PregitzeretaL
(2008); N added at 30 kg N/ha/yr] and Maine rElvir et al. (2003); 25 kg N/ha/yr], but
mature sugar maple and red maple did not respond in the Catskills rLovett et al. (2013);
50 kg N/ha/yr], Northern red oak increased growth at Harvard Forest rFrev et al. (2014);
50 and 150 kg N/ha/yr] but showed no growth response at two sites in New York state
[Wallace et al. (2007); 75 kg N/ha/yr, Lovett et al. (2013); 50 kg N/ha/yr], with increased
mortality at one of the New York sites (Wallace et al.. 2007). Red pine at Harvard Forest
exhibited decreased growth and higher mortality in response to chronic N additions rFrev
et al. (2014); 50 and 150 kg N/ha/yr], while red spruce showed no growth response in
Maine rElvir et al. (2003); 25 kg N/ha/yr], A 13-year study in a young forest in West
Virginia included the two species with the most positive growth responses in the Thomas
et al. (2010) analysis and found that N addition (35 kg N/ha/yr of [NH^SO-O generally
decreased growth of black cherry (Primus serotina) and tulip poplar (Liriodendron
tulipifera), although these changes were not statistically significant (May et al.. 2005).
Notably, N additions in these studies often exceeded levels observed in forest inventory
studies (Horn et al.. 2018; Thomas et al.. 2010). For instance, Horn et al. (2018) found
growth in black cherry increased with N deposition up to just over 15 kgN/ha/yr, before
declining with increases in deposition beyond that. Thus, the N addition of 35 kg N/ha/yr
in (May et al.. 2005) far exceeded the point of maximum growth for black cherry. Tulip
poplar consistently increased in growth with N deposition according to Horn et al.
(2018). but, even in this case, the 35 kg N/ha/yr added by May et al. (2005) exceeded the
maximum amount of deposition (almost 34 kg N/ha/yr) experienced by tulip poplar in the
Horn et al. (2018) study. This could help explain some of the differences in tree growth
and mortality responses between experimental N addition and forest inventory studies.
Other studies have also observed similarly mixed results on tree growth and mortality in
U.S. forests. At the same sites in Michigan where long-term N additions increased growth
of mature sugar maple (Pregitzer et al.. 2008). N additions decreased the growth and
survival of sugar maple saplings (Talhelm et al.. 2013; Patterson et al.. 2012). Notably,
this negative effect occurred without increase in overstory leaf area that would reduce
light availability or a decrease in soil pH (Talhelm et al.. 2013). More recently, Ibanez et
al. (2016) observed that although small overstory trees (5-10 cm diameter at breast
height) at these sites were growing faster, they were also experiencing increased
mortality. As with sugar maple saplings in Michigan, N deposition appears to be
decreasing growth among northern red oak saplings in the Chicago area, potentially
contributing to the inability of this species to regenerate in that region (Bassirirad et al..
2015). In Vermont, 2 years of N addition at 150 kg N/ha/yr increased growth in four
hardwood species (including sugar maple and northern red oak), decreased growth in one
6-19
-------
conifer, and had no effect on two other hardwoods (Finzi. 2009). Allen et al. (2010)
reported higher mortality and a decrease in ectomycorrhizal root tips (Table 6-2). with N
addition (100 kg N/ha/yr) in ectomycorrhizal pinon pine in New Mexico, while no
change in the mortality of arbuscular mycorrhizal juniper. Allison et al. (2010) observed
a 2.5-fold increase in aboveground net primary productivity in response to several years
of N additions (100 kg N/ha/yr) in central Alaska in a recently burned boreal forest. At
two mixed conifer forests in the Sierra Nevada, 2 years ofN additions (12 or
24 kg N/ha/yr) had a positive effect on herb community biomass at one site
(24 kg N/ha/yr) and a negative effect at the other site (12 kg N/ha/yr); shrub biomass was
unaffected (Hurteau and North. 2008). Modeling 50 years into the future based on these
results, a lower rate of N deposition (12 kg N/ha/yr) was expected to increase herb
biomass, while the high rate of N deposition (24 kg N/ha/yr) led only to a small increase
in shrub biomass. In combination with a wetter precipitation regime, both N deposition
rates were predicted to increase shrub and herb biomass (Hurteau et al.. 2009). In a
greenhouse study, N additions of up to 120 kg N/ha/yr had no effect on the N fixing tree
black locust (Robinict psendoacacia) when grown in a monoculture or in competition
with the sawtooth oak (Onerous acntissima), but N additions increased the height and
total biomass of the sawtooth oak when grown in competition with black locust (Luo et
al.. 2014).
6-20
-------
Table 6-2 Growth, productivity, and carbon cycle responses of ectomycorrhizal fungi to nitrogen added via
atmospheric deposition or experimental treatments.
Reference
Study
Location
Vegetation
Ambient
Deposition or
Addition
Nitrogen
Addition Rate
(kg N/ha/yr)
Duration
(yr)
Endpoint
Effect of Additional
Nitrogen
Hoqberq et al. (2010)
Sweden
Scots pine (Pinus
sylvestris)
Addition
100
2
13C labeling of EM
biomarker PLFA
1 8:2oj6,9
Decrease
Nasholmetal. (2013)
Sweden
Scots pine (Pinus
sylvestris)
Addition
100
0
(2 weeks)
13C labeling of EM
biomarker PLFA
1 8:2oj6,9
Not significant
Parrent and Vilaalvs
(2009)
North Carolina
Loblolly pine (Pinus
taeda)
Addition
112
1
Ectomycorrhizal 18S
RNA expression
Not significant
Hoqberq et al. (2011)
Sweden
Scots pine (Pinus
sylvestris)
Addition
120
20; 15-yr
recovery
Ectomycorrhizal
biomarker 18:2u>6,9
Not significant
Hoqberq et al. (2011)
Sweden
Scots pine (Pinus
sylvestris)
Addition
30, 70
35
Ectomycorrhizal
biomarker 18:2u>6,9
Decrease
Nasholmetal. (2013)
Sweden
Scots pine (Pinus
sylvestris)
Addition
100
0
(2 weeks)
Fine root chitin
concentration
Not significant
Ki0ller et al. (2012)
Denmark
Norway spruce
(Picea abies)
Ambient
27-43
n/a
Mycelium production
Decrease
Bahretal. (2013)
Sweden
Norway spruce
(Picea abies)
Ambient
0.9-24.6
n/a
Mycelium production
Decrease
Bahretal. (2015)
Sweden
Norway spruce
(Picea abies)
Addition
200 (once)
1
Mycelium production
Decrease
Hasselquist et al.
(2012)
Sweden
Scots pine (Pinus
sylvestris)
Addition
20, 100
6
Mycorrhizal respiration
Low dose: increase
Hiqh dose: decrease
6-21
-------
Table 6-2 (Continued): Growth, productivity, and carbon cycle responses of ectomycorrhizal fungi to nitrogen
added via atmospheric deposition or experimental treatments.
Reference
Study
Location
Vegetation
Ambient
Deposition or
Addition
Nitrogen
Addition Rate
(kg N/ha/yr)
Duration
(yr)
Endpoint
Effect of Additional
Nitrogen
Vallack et al. (2012)
Sweden
Scots pine (Pinus
sylvestris)
Addition
100
2
Mycorrhizal respiration
Decrease
Allen et al. (2010)
New Mexico
Pinon pine (Pinus
edulis) and
one-seeded juniper
(Juniperus
monospermum)
Addition
100
7+
Mycorrhizal root tip
colonization (by
ectomycorrhizal and
arbuscular
mycorrhizae)
Decrease in
ectomycorrhizal root
tips; no change in
arbuscular
mycorrhizal root tips
Pritchard et al. (2014)
North Carolina
Loblolly pine (Pinus
taeda)
Addition
112
6
Mycorrhizal root tip
production
Decrease
Pritchard et al. (2014)
North Carolina
Loblolly pine (Pinus
taeda)
Addition
112
6
Mycorrhizal root tip
survival (deep soil)
Increase
Kou et al. (2017)
China (sub-
tropical)
Slash pine (Pinus
eiiiottii)
Addition
40, 120
2
Mycorrhizal
(ectomycorrhizal)
survival
Increase at both low
and high additions,
except for deeper soil
and dichotomous
mycorrhizae
Parrent and Vilaalvs
(2009)
North Carolina
Loblolly pine (Pinus
taeda)
Addition
112
1
Root ammonium
transport gene
expression
Not significant
Garcia et al. (2008)
North Carolina
Loblolly pine (Pinus
taeda)
Addition
112
2
Root colonization (%)
Increase
Diaz et al. (2010)
Spain
Aleppo pine (Pinus
haiepensis)
Addition
35, 60, 120 mg/
plant
1
Root colonization (%)
Decrease
Ki0ller et al. (2012)
Denmark
Norway spruce
(Picea abies)
Ambient
27-43
n/a
Root colonization (%)
Decrease
Kou et al. (2015)
China
Slash pine (Pinus
eiiiottii)
Addition
40, 120
2
Root colonization (%)
Low dose: not
significant
Hiah dose: increase
6-22
-------
Table 6-2 (Continued): Growth, productivity, and carbon cycle responses of ectomycorrhizal fungi to nitrogen
added via atmospheric deposition or experimental treatments.
Reference
Study
Location
Vegetation
Ambient
Deposition or
Addition
Nitrogen
Addition Rate
(kg N/ha/yr)
Duration
(yr)
Endpoint
Effect of Additional
Nitrogen
Parrent and Vilqalvs
(2009)
North Carolina
Loblolly pine (Pinus
taeda)
Addition
112
1
Root monosaccharide
transport gene
expression
Not significant
Allison et al. (2008)
Alaska
Black spruce (Picea
mariana)
Addition
140
5
Sporocarp abundance
Decrease
Allen et al. (2010)
New Mexico
Pinon pine (Pinus
edulis) and one-
seeded juniper
(Juniperus
monospermum)
Addition
100
7+
Sporocarp abundance
Decrease
Gilletetal. (2010)
Switzerland
Norway spruce
(Picea abies)
Addition
150
12
Sporocarp abundance
Decrease
Hasselquist et al.
(2012)
Sweden
Scots pine (Pinus
sylvestris)
Addition
20, 100
6
Sporocarp abundance
Low dose: not
significant
Hiah dose: decrease
Hasselauist and
Hoabera (2014)
Sweden
Scots pine (Pinus
sylvestris)
Addition
110
20, 15-yr
recovery
Sporocarp abundance
Not significant
Hasselauist and
Hoabera (2014)
Sweden
Scots pine (Pinus
sylvestris)
Addition
20, 100
6
Sporocarp abundance
Low dose: not
significant
Hiah dose: decrease
Hasselauist and
Hoabera (2014)
Sweden
Scots pine (Pinus
sylvestris)
Addition
35, 70
40, 2-yr
recovery for
70-kg
treatment
Sporocarp abundance
Decrease
EM = ectomycorrhizal fungi; ha = hectare; kg = kilogram; N = nitrogen; n/a = not applicable; PLFA = phospholipid fatty acids; RNA = ribonucleic acid; yr = year.
Notes: single studies are reported more than once if multiple endpoints were measured. References ordered by endpoint. Only statistically significant effects are listed as increases
or decreases.
6-23
-------
A number of studies on the effects on N deposition or chronic N additions have been
conducted in temperate and boreal forests in Europe, and these studies have more
consistently found positive effects of N on tree growth [e.g., Solberg et al. (2009);
Ferretti et al. (2014); Nair et al. (2016)1. Ferretti et al. (2014) observed a positive
relationship between N deposition and tree growth across 26 forest plots in Italy. A
positive effect of N deposition on tree growth was also found on 135 Swiss long-term
observation plots for mature beech and spruce from the mid-1980s through late 2000s
(Braun et al.. 2010). Further N additions decreased tree growth at three of seven
experimental sites, yet these changes were linked to P deficiencies (Braun et al.. 2010).
Eastaugh et al. (2011) took a more complex approach to understand growth trends of
Norway spruce (Piceci ctbies) in the Austrian National Forest Inventory from 1960-2008.
Using the Biome-BGC model to isolate growth trends from variation in climate, the
authors found a positive relationship between tree growth and N deposition. Using data
from long-term N addition experiments in Norway spruce and Scots pine (Finns
sylvestris) forests throughout Sweden and Finland, Hvvonen et al. (2008) found increased
tree growth at 11 of 12 sites. The effects of added N on forest productivity appear to
persist for decades and carry over through major ecosystem disturbances. From et al.
(2015) studied young Norway spruce and Scots pine forests planted in the late 1990s after
the original forests were clear-cut. The original forests had received N additions of
150 kg N/ha/yr (as NH4NO3) either twice (in 1977 and 1985), once (1985), or never. Tree
height growth of the young forests was positively related to the amount of previous N
additions, with significantly greater height growth in the forests that had received two N
additions. Foliar N concentrations were also significantly higher in the forests that had
received N additions (From et al.. 2015). Biomass of understory shrubs and bryophytes
decreased in response to N additions (100 kg N/ha/yr for 6 years) on boreal forest islands
in Sweden, although one of three shrub species increased in growth (Wardle et al.. 2016).
The 2008 ISA tied greater tree growth and forest productivity to increases in foliar N,
photosynthesis, and gross primary productivity. In addition, foliar N was identified as a
sensitive indicator of changes in forest N availability (Aber et al.. 1989). Since the 2008
ISA, increases in foliar N continue to have been linked to N deposition in forests in the
U.S., Europe, and Asia in pollution gradient studies (Sardans et al.. 2016; Talhelm et al..
2012; Jones et al.. 2011; Cox et al.. 2010; TTnmonier et al.. 2010; Fenn et al.. 2008) and N
addition experiments (Du. 2017; Fusaro et al.. 2017; Gilliam et al.. 2016a; Wardle et al..
2016; Fowler et al.. 2015; Du and Fang. 2014; Lovett et al.. 2013; Talhelm et al.. 2013;
Lovett and Goodale. 2011; Talhelm et al.. 2011; Allen et al.. 2010; Braun et al.. 2010).
For instance, Cox et al. (2010) observed a 40% increase in foliar N (from 15 mg/g to
21 mg/g) in Scots pine forests in Germany and the U.K. along an N deposition gradient of
4.6 to 28 kg N/ha/yr. In addition to trees, forest bryophytes and lichens also show
increased tissue N concentrations when exposed to additional N (Maaroufi et al.. 2016;
6-24
-------
McMurrav et al.. 2015; McMurrav et al.. 2013; Gundale et al.. 2011; Fenn et al.. 2008;
Glavich and Geiser. 2008). Increased foliar N is not universally observed in N addition
studies [e.g., Lovett et al. (2013); Zhang et al. (2015c); Zhang et al. (2017)1 or along N
deposition gradients (Watmough and Meadows. 2014). but a meta-analysis found that N
additions generally increase foliar N in trees (Lu et al.. 2011b).
Although there are clear links between N deposition and increased foliar N and between
higher foliar N and increased photosynthesis, there is only limited evidence that chronic
N deposition directly increases leaf-level photosynthesis in forests. For instance, in
long-term simulated N deposition experiments in Massachusetts and Maine, only one of
four species exhibited increased photosynthesis with N additions (Elvir et al.. 2006;
Bauer et al.. 2004). Research on photosynthesis since the 2008 ISA has been similarly
mixed. Talhelm et al. (2011) did not observe a significant increase in either
photosynthesis or canopy leaf area in mature sugar maple trees at four sites in Michigan
in response to N additions (30 kg N/ha/yr as NaNCh for 14 years). Using a
canopy-applied N treatment at two Swiss forests, Wortman et al. (2012) found that N
additions improved photosynthetic processes in oak at one site, but did not significantly
influence photosynthetic parameters in spruce or beech at the other site. In Swedish
boreal forests, higher tissue N concentrations resulting from N additions (20 or
100 kg N/ha/yr for 4 years) increased the amount of 13CC>2 assimilated into shrub
(Vciccinium) foliage but decreased the amount of C assimilated by bryophytes
(Hasselquist et al.. 2016). At a broader scale, Fleischer et al. (2013) analyzed the effect of
N deposition on photosynthesis using observations from 80 forest eddy covariance C flux
measurement sites predominantly located in the eastern U.S. and western Europe. In this
data set, canopy photosynthetic capacity was positively correlated to N deposition for
conifer forests when N deposition was below an observed critical load of approximately
8 kg N/ha/yr (Fleischer et al.. 2013). Most of the stimulus provided by N deposition
occurred in boreal forests, with little to no influence of N deposition in broadleaf forests
or forests in the temperate climate zones (Fleischer et al.. 2013). Leonardi et al. (2012)
compiled a global data set of published tree ring 13C chronologies and found that the rate
of N deposition had a significant linear effect on intrinsic water-use efficiency (iWUE)
and 13C discrimination (A13C) in conifers and a quadratic effect in angiosperms. This
change in 13C discrimination suggests that the additional N either stimulated
photosynthesis or stimulated growth to the extent that the trees became more water
limited.
The apparent contradiction between the wide observations that N deposition can increase
both tree growth and foliar N concentrations and the inconsistent effects of N deposition
on leaf-level photosynthesis could be explained by two other physiological responses to
N. First, plants can store excess N in foliage as free amino acids and other
6-25
-------
physiologically inactive forms of N (Bauer et al.. 2004). Increases in foliar amino acid
concentrations, particularly arginine, have been observed in forests in chronic N addition
experiments (Bauer et al.. 2004) and along N deposition gradients (Braun et al.. 2010). In
these cases, the added foliar N would have no impact on photosynthesis. Second, plants
can respond to added N by altering C allocation. As noted in the 2008 ISA, higher N
availability often causes trees to allocate less photosynthate to roots and belowground
processes and more toward aboveground growth (Vicca et al.. 2012; Hogbcrg et al..
2010; Janssens et al.. 2010; Litton et al.. 2007; Aerts and Chapin. 1999; Minnich et al..
1995). Research since the 2008 ISA has provided further evidence that N deposition is
likely capable of causing such a shift in forests. In a greenhouse experiment, N additions
(2, 5, and 10 kg N/ha/yr) resulted in more biomass allocated to leaves than to roots at all
four N addition levels for oak and black locust trees. At the same Michigan sites where
Talhelm et al. (2011) observed that chronic N additions had no effect on leaf-level
photosynthesis or stand leaf area, the combination of increased aboveground growth, no
change in root production or respiration, and decreased mycorrhizal abundance led the
authors to conclude that a change in C allocation away from mycorrhizae was the likely
cause for enhanced tree growth. More broadly, Vicca et al. (2012) found that across
49 forest sites, higher site fertility increased both the fraction of biomass that was
allocated aboveground and the fraction of photosynthate that was allocated toward
biomass production. The increase in C allocated to biomass production was thought to
most likely result from less C being allocated to root symbionts such as mycorrhizae
(Vicca et al.. 2012). Likewise, Janssens et al. (2010) conducted a meta-analysis of
20 forest N addition experiments and found that N additions decreased the fraction of C
allocated to belowground processes but did not affect root biomass. As an example of
this, Hogberg et al. (2010) conducted a short-term (~2 hour) 13CC>2 canopy labeling
experiment in a young Scots pine forest in northern Sweden to determine how N
additions (two 100 kg N/ha additions over 2 years) altered the allocation of recent
photosynthate belowground. In the 3 weeks following the canopy labeling, the total soil
respiratory efflux of the 13C label was 62% lower in the plots that had received the N
additions. Other observed physiological responses to N include delayed bud burst (De
Barba et al.. 2016). decreased tree ring width, and increased xylem conduit density
(Borghetti et al.. 2017).
The effects of N addition on forest productivity have the potential to be moderated by
increases in herbivory, particularly among insects (Throop and Lerdau. 2004). For
instance, Andersen et al. (2010) observed that N additions increased foliar N and
photosynthetic rates within a tropical forest in Panama, but that tree growth rates were
unaffected because of increased herbivory. At Harvard Forest in Massachusetts, northern
red oak regeneration declined with N additions [50 and 100 kg N/ha/yr; Bogdziewicz et
al. (2017)1. Acorn production increased, but so did weevil infestation of the acorns, and
6-26
-------
germination rates declined. In a Pennsylvania common garden experiment with northern
red oak, saplings receiving N additions (200 kg N/ha/yr) tended to have greater herbivore
damage, but the effects of N addition on herbivory and the link between herbivory and
decreased growth both varied by the type of herbivore, the tree lineage (breeding family),
and the study location (Cha et al.. 2010). Similarly, Jones et al. (2008) and Jones et al.
(2011) observed that the most abundant herbivore of bracken fern (Pteridium aqiiilimim)
and a beetle herbivore of California black oak (Onerous kelloggii) were increased by N
addition at a heavily polluted site, but not at a drier, less polluted site. However, there
was a positive correlation between leaf NO, concentrations and the abundance of several
groups of black oak herbivores.
6.2.3.2 Belowground Processes
The 2008 ISA analysis of how belowground processes in forests reacted to N deposition
focused on fine root dynamics, aboveground litter inputs, decomposition, soil respiration,
and soil C. Johnson (2001) had found a significant increase in forest soil C in response to
N additions as part of a meta-analysis, but only a single long-term N addition field study
had observed a significant increase in forest soil C I Prcgitzcr et al. (2008); see
Appendix 41. For fine roots, Nadelhoffer (2000) hypothesized that N deposition would
decrease biomass, but stimulate turnover and production. The 2008 ISA found little
evidence with which to evaluate this hypothesis. There was more available research on N
effects on decomposition and soil respiration. As noted earlier, Knorr et al. (2005)
observed in a meta-analysis that N additions increased decomposition at sites receiving
low rates of ambient N deposition (<5 kg N/ha/yr), but N additions suppressed
decomposition at sites receiving moderate rates of N deposition (5-10 kg N/ha/yr).
Research completed since the 2008 ISA has advanced our understanding of how N
affects belowground processes in forests. In two meta-analyses, Liu and Greaver (2010)
did not find a consistent effect of N additions on forest fine root production, while Li et
al. (2015) observed a decrease (-13.5%) in forest fine root biomass. Broadly, some of the
results discussed in the 2008 ISA that appeared inconsistent at the time were likely a
reflection of how N additions interacted with other ecological processes. In young,
rapidly expanding forests, N addition may increase root biomass as a consequence of an
overall enhancement of plant growth (Janssens et al.. 2010). For instance, N additions
(150 kg N/ha/yr for 3 years) to a young bamboo forest increased fine root growth by
>30%, even though this forest already received high rates of atmospheric N deposition
[95 kg N/ha/yr of wet deposition; Tu et al. (2015)1. However, as noted in the
"Aboveground Processes" section (Appendix 6.2.3.1). increases in N deposition tend to
decrease the proportion of C allocated to roots relative to aboveground growth (Li et al..
6-27
-------
2015; Vicca et al.. 2012; Janssens et al.. 2010; Litton et al.. 2007; Minnich et al.. 1995).
Thus, in more mature forests, root biomass and production are not consistently affected
by N deposition. For instance, there were no effects of chronic N additions on fine root
biomass in four mature hardwood forests in Michigan [30 kg N/ha/yr for over 15 years;
Burton et al. (2012)1. and in two mature mixed oak (Onerous) stands at Harvard Forest
[50 or 150 kg N/ha/yr for over 20 years; Frev et al. (2014)1. Likewise, N addition
(50 kg N/ha/yr) did not cause significant changes in root biomass in a tropical forest in
Puerto Rico (Cusack et al.. 2010). Root turnover did not respond to N addition
(30 kg N/ha/yr for 3 years) across 13 successional (20-40 years old) and mature
(>90 years old) hardwood stands in central New Hampshire (Kang et al.. 2016). In
contrast with the mixed oak stands, fine root biomass decreased in two red pine stands at
Harvard Forest [50 or 150 kg N/ha/yr for over 10 years; Frev et al. (2014)1. and root
biomass also decreased with N additions (50 or 150 kg N/ha/yr for 3 years) in a
subtropical, broadleaf forest in China (Peng et al.. 2017).
Observations of how other belowground processes respond to N additions provide further
evidence of an increase in the ratio of tree C allocated to aboveground growth and
productivity versus the C allocated belowground. In a meta-analysis of forest soil
respiration responses to N addition, Janssens et al. (2010) noted that N additions
decreased soil respiration overall, with a portion of this effect caused by a decrease in
autotrophic respiration. Within forests, autotrophic respiration may make up more than
50% of total soil respiration (Hasselquist et al.. 2012). Much of this autotrophic
respiration is mycorrhizal respiration, with the C allocated to and respired by
mycorrhizae estimated to account for 9 to 34% of forest soil respiration (Hasselquist et
al.. 2012; Heinemever et al.. 2007).
Mycorrhizal fungi have long been observed to be sensitive to increased forest N
availability, particularly ectomycorrhizae because these fungi have direct roles in plant N
acquisition, and their production of aboveground fruiting bodies (sporocarps) makes it
easier to observe changes in abundance (Treseder. 2004; Lilleskov et al.. 2002; Wallenda
and Kottke. 1998). As noted in Appendix 4. when N supply increases, C allocation to
ectomycorrhizal fungi decreases, and their abundance and activity decline. In a
meta-analysis, Li et al. (2015) found that N additions decreased mycorrhizal colonization
of fine root tips by 19% (n = 12). In contrast to the widespread increases in aboveground
tree growth, studies of ectomycorrhizal growth and productivity responses to added N
have documented nearly universally negative or neutral effects on metrics such as
mycorrhizal root colonization, sporocarp abundance, and the abundance of the fungal
lipid biomarker 18:2co6,9 in the soil (Table 6-2). Although this research is consistent,
these studies have limitations because nearly all of this research has been conducted on
conifer species and most of the negative effects occur in studies using unrealistically high
6-28
-------
rates of N addition (>100 kg N/ha/yr). However, two studies of ectomycorrhizae in
Norway spruce forests in Europe have documented significant declines along gradients of
ambient N deposition (Bahr et al.. 2013; Kjoller et al.. 2012). In particular, (Ki oiler et al..
2012) observed an 80 and 90% decrease in ectomycorrhizal root tip abundance and
mycelial production, respectively, across a canopy throughfall N deposition gradient of
27 to 43 kg N/ha/yr in Denmark. Morrison et al. (2016) also observed a decline in the
relative abundance of ectomycorrhizal fungal DNA within the soil of the N addition (50
or 150 kg N/ha/yr for 25 years) plots at the oak (<9z/ercz«)-dominated Harvard Forest in
Massachusetts.
Given these decreases in ectomycorrhizal growth and productivity in response to added
N, it is not surprising that the amount of C allocated belowground by plants to
mycorrhizae can be sensitive to N availability. Hogberg (2012) twice added N
(100 kg N/ha) over a period of two growing seasons to boreal Scots pine forests in
Sweden as part of a short (~2 hour) 13C canopy labeling experiment. In the 5 weeks
following the labeling, 48% less of the 13C label was found in an ectomycorrhizal fungal
lipid biomarker in the soil of the N amended plots. Hasselquist et al. (2012) found that in
Scots pine forests in northern Sweden, 6 years of N additions at a rate of 100 kg N/ha/yr
decreased respiration from ectomycorrhizae by 40%, but N additions at a rate of
20 kg N/ha/yr increased this respiration by 120%. However, the low N addition did not
change the fractional contribution of ectomycorrhizae to total soil respiration. The low N
addition treatment also had no effect on ectomycorrhizal sporocarp production, but the
high N addition treatment nearly eliminated sporocarp production [99% decrease;
Hasselquist et al. (2012)1. The availability of N also appears to influence the transfer ofN
from mycorrhizal fungi to plants. In a boreal Scots pine forest in Sweden, Nasholm et al.
(2013) observed that a single N dose (100 kg N/ha) shifted the dominant sink for a 15N
tracer added 2 weeks later from the cytoplasm of ectomycorrhizal fungi and other soil
microorganisms to the pine foliage.
Although ectomycorrhizae are important in many high latitude and temperate
ecosystems, particularly forests, most species of terrestrial plants form arbuscular
mycorrhizae (Rillig. 2004). The impact of N deposition on arbuscular mycorrhizae has
received less research attention, perhaps partly because these fungi are best known for
their role in plant P acquisition (Rillig. 2004). Arbuscular mycorrhizae community
composition and production can be sensitive to added N rEgerton-Warburton and Allen
(2000); van Diepen et al. (2010); Table 6-31. but these effects may not be consistent, van
Diepen et al. (2010) reviewed eight previous studies of how intraradical (within root) and
extraradical arbuscular mycorrhizal biomass responded to N additions, predominantly in
forests, and found inconsistent effects. Intraradical biomass significantly declined in
response to N in three studies and increased in two studies, including in the work of
6-29
-------
Garcia et al. (2008). In comparison, extraradical biomass was either unresponsive or
declined (van Diepen et al.. 2010). Aside from the work of Garcia et al. (2008). other
researchers since 2008 have found either neutral or negative effects on metrics of
arbuscular mycorrhizal biomass and abundance (Table 6-3). Observations of decreased
arbuscular mycorrhizal biomass span from a subtropical broadleaf evergreen forest in
China that experienced a single year of N additions [25 or 50 kg N/ha/yr; Shi et al.
(2016a)l to a subalpine Englemann spruce (Piceci engelmctnnii) forest in Rocky Mountain
National Park that had received more than 15 years of N additions [25 kg N/ha/yr; Boot
et al. (2016)1.
Notably, the arbuscular mycorrhizal species studied by Garcia et al. (2008) were all
understory or subdominant canopy species growing beneath ectomycorrhizal loblolly
pines. The increased growth of these larger trees in response to added N may have
imposed other resource limitations on the arbuscular mycorrhizal plant species. In
general, the response of arbuscular mycorrhizae to N additions may depend upon the
relative availability of P or other nutrients. Johnson et al. (2003) found N additions
decreased arbuscular mycorrhizae when the soil N:P ratio was low (i.e., P-rich soils),
while increasing arbuscular mycorrhizae under high soil N:P (i.e., P-poor soils). Where P
is more limiting, plants may allocate more carbon to arbuscular mycorrhizae in order to
acquire it (Egerton-Warburton et al.. 2007; Johnson et al.. 2003). Thus, the varying
availability of soil P may explain in part the seemingly inconsistent response of
arbuscular mycorrhizae to N additions.
6-30
-------
Table 6-3 Growth, productivity, and carbon cycle responses of arbuscular
mycorrhizal fungi to nitrogen added via atmospheric deposition or
experimental treatments.
Reference
Study
Location
Vegetation
Ambient
Deposition
or
Addition
Nitrogen
Addition
Rate
(kg N/ha/yr)
Duration
(yr)
Endpoint
Effect of
Additional
Nitrogen
van Dieoen et al.
(2010)
Michigan
(four
sites)
Sugar maple
(Acer
saccharum)
Addition
30
13
Extraradical
biomass
production
Decrease
van Diepen et al.
(2010)
Michigan
(four
sites)
Sugar maple
(Acer
saccharum)
Addition
30
13
Extraradical
biomass
(16:1 oj5c
abundance)
Decrease
Chen et al. (2014)
China
Steppe
grassland
Addition
100
6
Hyphal length
Decrease
van Dieoen et al.
(2010)
Michigan
(four
sites)
Sugar maple
(Acer
saccharum)
Addition
30
13
Intraradical
biomass
(16:1 oj5c
abundance)
Decrease
Garcia et al. (2008)
North
Carolina
Loblolly pine
(Pinus
taeda)
Addition
100
1-2
Root
colonization
(%)
Increase
Mandvam and
Jumpponen (2008)
Kansas
C3 and C4
grasses
Addition
100
3
Root
colonization
(%)
Not
significant
Van der Heiiden et
al. (2008)
Holland
Dune
grasses
Addition
100
1
Root
colonization
(%)
Decrease
Camenzind et al.
(2014)
Ecuador
Evergreen
tropical
forest
Addition
50
3
Root
colonization
(%)
Decrease
Chen et al. (2014)
China
Steppe
grassland
Addition
100
6
Root
colonization
(%)
Not
significant
Garcia et al. (2008)
North
Carolina
Loblolly pine
(Pinus
taeda)
Addition
100
1-2
Soil glomalin
content
Not
significant
ha = hectare; kg = kilogram; N = nitrogen; yr = year.
Notes: single studies are reported more than once if multiple endpoints were measured. References ordered by endpoint. Only
statistically significant effects are listed as increases or decreases.
6-31
-------
Aside from mycorrhizal respiration, the other dominant component of autotrophic soil
respiration is root respiration. Although Janssens et al. (2010) observed a decrease in
autotrophic soil respiration in the presence of N, this decrease was not apparently related
to changes in root respiration. In a long-term N addition experiment (30 kg N/ha/yr for
+15 years) in four northern hardwoods forests, Burton et al. (2012) found that although
root N concentration was a strong predictor of the root respiration rate and N additions
had increased root N concentrations, N additions had not increased root respiration
because the added N changed the relationship between root N and respiration. Drake et
al. (2008) made a similar observation in loblolly pine forests in North Carolina: N
additions (100 kg N/ha/yr for 2 years) increased fine root N concentrations but did not
impact fine root respiration rates, resulting in an altered tissue N respiration relationship.
Frev et al. (2014) also reported that root respiration was unresponsive to long-term N
additions (50 or 150 kg N/ha/yr) in oak forests and pine forests in Massachusetts.
However, Hasselquist et al. (2012) found that a relatively modest N addition rate
(20 kg N/ha/yr for 6 years) increased both root and mycorrhizal respiration in a boreal
forest in Sweden that received relatively low rates of ambient N deposition
(<5 kg N/ha/yr). Thus, it appears that C allocation to mycorrhizae is more sensitive to N
availability than C allocation to root respiration.
In a meta-analysis of the effects of N additions on microbial biomass, Treseder (2008)
found that changes in soil respiration were significantly and positively correlated with the
response of microbial biomass. Mycorrhizal fungi are major components of forest soil
microbial communities, with Hogberg et al. (2010) estimating that ectomycorrhizal
mycelium made up 39% of total soil microbial biomass in a Swedish boreal forest. At
Harvard Forest, Morrison et al. (2016) observed that 59-72% of all fungal operational
taxonomic units (OTUs) belonged to ectomycorrhizal fungi. Given the widespread
negative effects of added N on mycorrhizal fungi, the results of Treseder" s meta-analysis
and the broadly negative or neutral effects of N additions on microbial biomass in the
studies published since 2008 are unsurprising (Table 6-4).
6-32
-------
Table 6-4 Abundance and carbon cycle responses of forest soil
microorganisms and soil invertebrates to nitrogen added in
experimental treatments.
Reference
Study Location
Ambient Nitrogen
Deposition Addition Rate Duration
Vegetation or Addition (kg N/ha/yr) (yr)
Effect of
Additional
Endpoint Nitrogen
Treseder (2008) Meta-analysis
Mostly Addition
boreal and
temperate
forests
1-600 0.5-57 Bacterial Not
biomass significant
Zhao et al.
(2014a)
China (Tibetan
plateau)
Spruce-fir Addition
(Picea
asperata,
Abies
faxoniana)
250
Bacterial
biomass
Decrease
Hesse et al.
(2015)
Michigan (Ml
gradient)
Northern Addition
hardwood
forests
(Acer
saccharum)
30
16
Bacterial
biomass
Not
significant
Treseder (2008) Meta-analysis
Mostly Addition
boreal and
temperate
forests
1-600
0.5-57
Fungal
biomass
Not
significant
Zhao et al.
(2014a)
China (Tibetan
plateau)
Spruce-fir Addition
(Picea
asperata,
Abies
faxoniana)
250
Fungal
biomass
Decrease
Hesse et al.
(2015)
Michigan (Ml
gradient)
Northern Addition
hardwood
forests
(Acer
saccharum)
30
16
Fungal Not
biomass significant
Enowashu et al. Germany
(2009)
Norway
spruce
(Picea
abies)
Subtraction 9.7 (-21)
16
(recovery)
Fungal
biomass
(ergosterol)
Increase
Bebber et al.
(2011)
U.K.
Broadleaf Addition
temperate
forest
(Fraxinus-
Acer,
Fagus)
2.8
Fungal Not
mycelium significant
growth
6-33
-------
Table 6-4 (Continued): Abundance and carbon cycle responses of forest soil
microorganisms and invertebrates to nitrogen added in
experimental treatments.
Ambient
Deposition
Reference Study Location Vegetation or Addition
Nitrogen Effect of
Addition Rate Duration Additional
(kg N/ha/yr) (yr) Endpoint Nitrogen
van Diepen et al.
(2017)
Massachusetts Temperate, Addition 50, 150
28
(Harvard
Forest)
mixed
hardwood
forest
(Quercus
velutina,
Quercus
rubra)
Fungal
mycelium
growth
Not
significant
overall;
varied by
fungal
isolate
Allison et al. Alaska Boreal Addition 140 5 Fungal Decrease
(2008) forest sporocarp
(Picea biomass
mariana)
Lin et al. (2017) China
Subtropical Addition
deciduous
and
coniferous
forests
47 10 mo
Invertebrate Increase
density (pre- on
dominantly coniferous
Collembola litter; not
and Acari) on significant
decomposing on
litter deciduous
litter
Allison et al. Alaska Boreal Addition 140 5 Microbial Not
(2008) forest biomass significant
(Picea
mariana)
Treseder (2008) Meta-analysis
Mostly Addition 1-600
boreal and
temperate
forests
0.5-57 Microbial Decrease
biomass
Keeler et al. Minnesota Temperate Addition 100 5 Microbial Not
(2009) (Cedar Creek) forests biomass significant
(Quercus
ellipsoidalis,
Pinus
strobus)
and
grassland
van Diepen et al. Michigan (Ml Northern Addition 30 12 Microbial Decrease
(2010) gradient) hardwood biomass
forests
(Acer
saccharum)
Hobbie et al. Minnesota Oak and Addition 100 5 Microbial Not
(2012) (Cedar Creek) pine forests biomass significant
(Quercus
ellipsoidalis,
Pinus
strobus)
6-34
-------
Table 6-4 (Continued): Abundance and carbon cycle responses of forest soil
microorganisms and invertebrates to nitrogen added in
experimental treatments.
Ambient
Nitrogen
Effect of
Deposition
Addition Rate Duration
Additional
Reference
Study Location
Vegetation
or Addition
(kg N/ha/yr) (yr)
Endpoint
Nitrogen
Zhao et al.
China (Tibetan
Spruce-fir
Addition
250 4
Microbial
Decrease
(2014a)
plateau)
(Picea
biomass
asperata,
Abies
faxoniana)
Allison et al.
Alaska
Boreal
Addition
114 7
Microbial
Decrease
(2010)
forest
biomass C
(Picea
mariana)
Wana et al.
China
Subtropical
Addition
50, 100 8
Microbial
Low dose:
(2015b)
(southern)
pine forest
biomass C
increase
(Pinus
High
masson-
dose: not
iana)
significant
Wana et al.
China
Subtropical
Addition
50, 100 8
Microbial
Not
(2015b)
(southern)
pine-
biomass C
significant
broadleaf
forest
(Pinus
masson-
iana)
Wana et al.
China
Subtropical
Addition
cn
o
100, 150 8
Microbial
Low, mid
(2015b)
(southern)
broadleaf
biomass C
dose: not
forests
significant
High
dose:
decrease
Pena et al.
China
Mid-
Addition
cn
o
150 2+
Microbial
Decrease
(2017)
(Sichuan
subtropical,
biomass C
Province)
evergreen,
broadleaf
forest
Lin et al. (2017)
China
Subtropical
Addition
47
10 mo
Microbial
Not
deciduous
biomass on
significant
and
decomposing
coniferous
litter
forests
Kana et al.
New
Northern
Addition
30
2,3
Microbial
Not
(2016)
Hampshire
hardwood
respiration
significant
forests
(Acer
saccharum,
Fagus
grandifolia)
6-35
-------
Table 6-4 (Continued): Abundance and carbon cycle responses of forest soil
microorganisms and invertebrates to nitrogen added in
experimental treatments.
Reference
Study Location
Vegetation
Ambient
Deposition
or Addition
Nitrogen
Addition Rate
(kg N/ha/yr)
Duration
(yr)
Endpoint
Effect of
Additional
Nitrogen
Gillet et al.
Switzerland
Norway
Addition
150
12
Saprobic
Increase
(2010)
spruce
fungal
and
(Picea
sporocarp
decrease
abies)
abundance
(N x yr)
van Dieoen et al.
(2010)
Michigan (Ml
gradient)
Northern
hardwood
forests
(Acer
saccharum)
Addition
30
12
Saprotrophic
fungal
biomass
Not
significant
C = carbon; ha = hectare; kg = kilogram; N = nitrogen; yr = year.
Notes: single studies are reported more than once if multiple endpoints were measured. References ordered by endpoint. Only
statistically significant effects are listed as increases or decreases.
However, forest soil microbial communities are taxonomically and functionally diverse
(see Appendix 6.3.3.3) and have exhibited varying responses to added N (Zhang et al..
2015c; Treseder. 2008). For instance, Wang et al. (2015b) added N (50, 100, or
150 kg N/ha/yr for 9 years) to pine, broadleaf, and mixed pine-broadleaf forests in
southern China; the 50 kg N treatment increased microbial biomass C in the pine forests
and the 150 kg N treatment decreased microbial biomass C in the broadleaf forests, but
there were no other significant treatment effects. Elsewhere in China, a single year of N
additions (25 or 50 kg N/ha/yr) in a subtropical broadleaf evergreen and temperate
broadleaf deciduous forest had no effect on total microbial biomass in either forest, but
decreased the abundance of Gram-negative bacteria, actinomycetes, and saprotrophic
fungi in the subtropical evergreen forest (Shi et al.. 2016a). The addition of
100 kg N/ha/yr to a regenerating Alaskan boreal forest decreased the microbial biomass
and the C:N ratio of microbial biomass (Allison et al.. 2010; Allison et al.. 2008). while
N additions (50 kg N/ha/yr) in Indiana mixed hardwood forests decreased microbial
biomass without affecting the microbial C:N ratio (Midglev and Phillips. 2016). In
Sweden, the addition of 100 kg N/ha/yr to boreal forests for 6 years decreased bacterial
biomass, but did not significantly affect fungal biomass (Wardle et al.. 2016). The
opposite result was observed in pine and hardwood stands at Harvard Forest and in
subalpine spruce-fir forests in Rocky Mountain National Park (RMNP). There, N
additions decreased fungal biomass, but bacterial biomass was unaffected [50 and 150 kg
N/ha/yr for 14 years at Harvard Forest; 25 kg N/ha/yr for 17 years in RMNP; Boot et al.
(2016)1. In Michigan, long-term (+15 years) N additions (30 kg N/ha/yr) to northern
hardwood forests decreased the abundance of actinobacteria in the surface mineral soil,
but not in the soil organic horizon (Eisenlord and Zak. 2010). Relative to mycorrhizal
6-36
-------
fungi, there is less information available regarding the growth and biomass responses of
individual domains and other taxonomic groupings within microbial communities
Table 6-4). Of this finer scale information, most observations are for fungi (sometimes
including mycorrhizal fungi), with positive, neutral, and negative effects of added N.
Atmospheric N deposition may affect forest microbial communities and litter
decomposition by altering litter chemistry through shifts in forest community
composition, increasing litter N concentrations, or changing the secondary chemistry of
litter. The two dominant forms of forest plant litter, leaves and fine roots, can each be
altered by N additions (Xia et al.. 2015). As noted in the aboveground processes section
of this appendix, there is abundant evidence that N deposition can increase foliar N
concentrations in trees and there is also a large body of research that the increase in foliar
N concentrations will result in increased leaf litter N concentrations (Acrts. 1996). Since
the 2008 ISA, there have been new observations of increased leaf litter N concentrations
both in long-term N addition experiments (van Diepen et al.. 2015; Xia et al.. 2015; Zak
et al.. 2008) and along an N deposition gradient [6.8-11.8 kg N/ha/yr; Talhelm et al.
(2012)1. Increases in leaf litter N concentrations with greater N deposition are not
universal [e.g., Watmough and Meadows (2014)1. but in a meta-analysis, van Diepen et
al. (2015) observed a >20% increase in tree leaf litter N concentrations in response to N
additions. Other changes in forest litter chemistry likely involve changes in more
complex physiological and biogeochemical mechanisms (Du and Fang. 2014). In an old
growth boreal forest in China, N additions (20, 50, or 100 kg N/ha/yr for 3 years)
decreased leaf litter P concentrations, an effect that could have been caused by changes in
either physiological or biogeochemical processes (Du and Fang. 2014). In a
meta-analysis, van Diepen et al. (2015) found that simulated N deposition significantly
decreased leaf litter concentrations of P, calcium, manganese, aluminum, and zinc, but
did not find significant changes in concentrations of potassium (K), magnesium, boron,
iron, or copper. It is not entirely understood how all of these changes in elemental
composition impact the composition and function of soil microbial communities, but
manganese and calcium are needed for the production of extracellular enzymes that
degrade lignin (van Diepen et al.. 2015).
There is a large volume of research about how major biochemical constituents of plant
litter such as lignin, cellulose, condensed tannins, and phenolics impact leaf litter
decomposition, but less information about how N additions change the abundance of
these compounds. Xia et al. (2015) researched the effects of long-term N additions
(30 kg N/ha/yr for +15 years) on sugar maple leaf litter and fine root chemistry in four
mature northern hardwood forests in Michigan. The N additions increased concentrations
of condensed tannins in leaf litter and in fine roots. At three of the four sites, N additions
increased the fraction of nonstructural cell wall material and decreased both cellulose and
6-37
-------
lignin in leaf litter. On an ecosystem basis, N additions decreased total annual litter
(leaf + fine root) fluxes of condensed tannins and soluble protein and increased the fluxes
of nonstructural carbohydrates (sugars + starch) and N (Xia et al.. 2015). Based on broad
decomposition chemistry patterns observed elsewhere, these changes in litter chemistry
were expected to increase initial rates of litter decomposition, which contrasts with the
decrease in litter turnover rates observed at these sites (Xia et al.. 2015). However,
changes in tissue chemistry are not ubiquitous; for example, Gricpcntrog et al. (2015)
observed that the abundance and composition of fatty acids in leaf and root tissues in
spruce and beech trees in Switzerland were not influenced by added N (70 kg N/ha/yr).
Further, while N additions can change plant tissue chemistry, links between litter
chemistry and microbial abundance and microbial function are complex [e.g., Baumann
et al. (2009)1.
6.2.3.3 Forest Lichens
Lichens are widely used as indicators of N deposition impacts on ecosystems, particularly
in forests. However, lichens also are important for ecosystem function. For insects, birds,
and mammals, lichens represent camouflage, building materials for nests, and a source of
food (Brodo et al.. 2001). Lichens absorb N, sulfur (S), and other elements from
atmospheric deposition and throughfall, and lichens that host cyanobacteria can add
significantly to ecosystem N inputs, providing N to other plants (Kobvlinski and Fredeen.
2015). In addition, lichens also have a role in hydrologic cycling, have many traditional
human uses, and have high potential for pharmaceutical use (McCune and Geiser. 1997).
Lichens are symbioses comprised of fungi (mycobiont) and a green alga and/or
cyanobacterium [photobiont; Palmqvist (2000); Sundberg et al. (2001)1. Much of the
lichen biomass is comprised of fungal (mycobiont) tissue, but the photobiont synthesizes
organic compounds, supplying the energy and structural C needed for growth of the
lichen. Both the photobiont and the mycobiont require N for growth (Palmqvist. 2000).
but the supply of C and N must be coordinated for the development of lichen thalli
(Sundberg et al.. 2001). Lichens with a cyanobacterial photobiont are N fixing, but those
with a green algal photobiont depend on atmospheric deposition for N.
Lichens can be classified based on their response to N pollution. Lichens occurring in
areas that receive high N deposition are considered nitrophytic or eutrophic; lichens
common in areas receiving low N input are designated acidophytic or oligotrophic (Gaio-
Oliveira et al.. 2005; van Herk. 2001). Atmospheric N deposition can impact lichens
through changes in physiological function caused by an increased supply of N or by
altering the pH of tree bark hosting the lichen (Jovan. 2008). Although lichens with a
green algal photobiont depend on atmospheric deposition as a source of N, these lichens
6-38
-------
can also be negatively impacted by N deposition via the accumulation of toxic
concentrations of NH44" within the thallus. Cyanobacteria can grow on either NO;, or
NH4+ sources when administered at nontoxic concentrations, but more rapid growth has
been observed with NIL+ than NO;, (Von Riickert and Giani. 2004). Ammonium is more
easily assimilated by lichens; both NO; and nitrite must first be reduced to NFL+ before
assimilation can occur (Von Riickert and Giani. 2004). The 2008 ISA noted that lichens
with a cyanobacterial photobiont appear to be more sensitive to adverse effects from
atmospheric N deposition than most other lichens (Dahlman et al.. 2002; Hallingback and
Kellner. 1992; Hallingback. 1991). In part, the sensitivity of lichens to increasing N
deposition is a function of the mechanisms with which that lichen can respond to high N
supply, such as decreasing N uptake or assimilating N into nontoxic forms such as
arginine (Gaio-Oliveira et al.. 2005; Dahlman et al.. 2002).
Since the 2008 ISA, new research on the impact of N deposition on the growth and
physiology of forest lichens has both confirmed that lichen abundance is sensitive to N
deposition and provided further insight on the growth and physiological changes that
occur when lichens are exposed to exogenous N (Table 6-5). For example, previous
research had suggested that lichens were most sensitive to N as ammonia (Sheppard et
al.. 2011; Jovan. 2008). However, a study of lichen communities on California black oak
(Onerous kelloggii) forests at 22 sites in the Los Angeles Basin in California, Jovan et al.
(2012) found that the abundance of eutrophic lichen species was only weakly related to
gaseous NH3 concentrations. Instead, the strongest N pollution-related predictor of
eutrophic lichen abundance was total N deposition (as canopy throughfall). Further, at the
relatively neutral bark pH levels in the Los Angeles Basin, there was no influence of pH
on the abundance of eutrophic lichen species. The abundance of eutrophic lichens was
also best correlated with total throughfall N deposition in southeastern Alaska
(Schirokauer et al.. 2014a). This research provides strong evidence that total N
deposition, not the deposition of a particular form of N, is the primary driver of changes
in the growth, physiology, and composition of epiphytic lichens.
Increases in lichen thalli N concentrations in response to added N have been widely
observed in the U.S. and Europe (Table 6-5). even at relatively low rates of atmospheric
N deposition. For instance, McMurrav et al. (2013) measured throughfall N deposition
and sampled lichen thalli N concentrations at sites near the Bridger-Teton National Forest
Wilderness that were at increasing distances downwind of a major oil and natural gas
production field. Although the observed rates of N deposition were only 0.8 to
4.1 kg N/ha/yr along this gradient, thalli N concentrations in Usnect lapponicci
approximately doubled from -1.2 to 2.4%. In southeastern Alaska, Schirokauer et al.
(2014a) found increases in thalli N concentrations along an even smaller range of N
6-39
-------
deposition (0.05 to 1.05 kg N/ha/yr), a pollution gradient attributed to cruise ship
emissions.
In southern California, Riddell et al. (2008) transplanted thalli from the oligotrophic
lichen Ramalina menziesii from relatively unpolluted sites into fumigation chambers
under moderate and high HNO3 concentrations (19.9-25 (ig/m3 and 26.4-35.3 (ig/m3,
respectively). The HNO3 fumigation caused significant declines in chlorophyll content
and C exchange capacity compared to thalli in control chambers. This research was later
expanded to six species known to vary in sensitivity to N pollution (Riddell et al.. 2012).
Fumigation with HNO3 (daily peaks near 50 ppb) decreased chlorophyll content,
chlorophyll fluorescence, gross photosynthesis, and dark respiration in three of the five N
sensitive species; while only photosynthesis declined in the other two N sensitive species.
Four of the N sensitive species were tested for fumigation effects on cell membrane ion
leakage; overall ion leakage, and specifically K+ ion leakage, were increased by HNO3
fumigation in all species.
Johansson and colleagues conducted a series of experiments in Sweden intended to
understand the physiological responses of epiphytic lichens to added N (Johansson et al..
2012; Johansson et al.. 2011; Johansson et al.. 2010). In a whole-tree N addition
experiment in a spruce forest in boreal Sweden, low rates of N addition (6,
12.5 kg N/ha/yr for 3 years) increased total lichen abundance, but higher N addition rates
(25, 50 kg N/ha/yr) decreased total lichen abundance (Johansson et al.. 2012). However,
there was considerable variation among lichen species in the N addition treatment rate
that resulted in optimal growth. For two of the lichen species in that experiment,
Alectoria sarmentosa and Platismatia glauca, Johansson et al. (2010) quantified
physiological responses to N additions. Thalli N concentrations increased in both species
at the two highest N addition rates (25, 50 kg N/ha/yr). There was a significant positive
relationship between cumulative N dose and chlorophyll content, but the N additions did
not affect thalli P concentrations. In a separate experiment with a single much higher N
addition rate (300 kg N/ha/yr for one season), Johansson et al. (2011) examined the
physiological and growth responses of three lichen species. All three lichen species
exhibited increased tissue N concentrations, increased chlorophyll concentrations,
increased photosynthesis, and increased growth of the photobiont. However, mycobiont
growth decreased in two species and was unchanged in the third. Total biomass changes
among the three species were positive, neutral, and negative.
6-40
-------
Table 6-5 Growth and physiology responses of forest epiphytic lichens to
nitrogen added via atmospheric deposition or experimental
treatments.
Reference
Study
Location
Vegetation
Ambient
Deposition
or
Addition
Nitrogen
Addition
Rate (kg
N/ha/yr)
Duration
(yr)
Endpoint
Effect of
Additional
Nitrogen
Nvbakken et
al. (2009)
Sweden and
Norway
Boreal
forests
Addition
50
1
C-based
secondary
compounds
Three
species: not
significant
One species:
decrease
Riddell et al.
(2008)
California
(Los Angeles
basin)
Oak forests
(Quercus
douglasii)
HNO3 gas
fumigation
0, 15,
30 mg/m3
0.08
Chlorophyll
content
Decrease
Nvbakken et
al. (2009)
Sweden and
Norway
Boreal forest
Addition
50
1
Chlorophyll
content
Three
species:
increase
One species:
not significant
Johansson et
al. (2010)
Sweden
Norway
spruce
(Picea
abies)
Addition
6, 12.5,
25, 50
3
Chlorophyll
content
Increase
Johansson et
al. (2011)
Sweden
Boreal forest
Addition
300
1
Chlorophyll
content
Increase
AsDlund et al.
(2010)
Sweden and
Norway
Boreal
forests
Addition
50
1
Gastropod
feeding
preference
Three of four
lichen
species:
decrease
One species:
increase
Strenqbom and
Nordin (2008)
Sweden
Boreal forest
Addition
150
(twice)
Additions
22 and
30 yr
prior to
surveys
Lichen
abundance
Decrease
Johansson et
al. (2012)
Sweden
Norway
spruce
(Picea
abies)
Addition
6, 12.5,
25, 50
4
Lichen
abundance
Six and 12.5
doses:
increase
25 and 50
doses:
decrease
Will-Wolf et al.
(2015)
Northeastern
U.S.
Forests
Ambient
Not stated
n/a
Lichen
abundance
Decrease
6-41
-------
Table 6-5 (Continued): Growth and physiology responses of forest epiphytic
lichens to nitrogen added via atmospheric deposition or
experimental treatments.
Reference
Study
Location
Vegetation
Ambient
Deposition
or
Addition
Nitrogen
Addition
Rate (kg
N/ha/yr)
Duration
(yr)
Endpoint
Effect of
Additional
Nitrogen
Riddell et al.
(2008)
California
(Los Angeles
Basin)
Oak forests
(Quercus
douglasii)
HNOsgas
fumigation
0, 15,
30 mg/m3
0.08
Membrane ion
leakage
Increase
Johansson et
al. (2011)
Sweden
Boreal forest
Addition
300
1
Mycobiont
growth
One species:
decrease
Two species:
not significant
Johansson et
al. (2011)
Sweden
Boreal forest
Addition
300
1
Photobiont
growth
Increase
Riddell et al.
(2008)
California
(Los Angeles
Basin)
Oak forests
(Quercus
douglasii)
HNO3 gas
fumigation
0, 15,
30 mg/m3
0.08
Photosynthesis
Decrease
Johansson et
al. (2011)
Sweden
Boreal forest
Addition
300
1
Photosynthesis
Increase
Riddell et al.
(2008)
California
(Los Angeles
Basin)
Oak forests
(Quercus
douglasii)
HNO3 gas
fumigation
0, 15,
30 mg/m3
0.08
Respiration
Decrease
Johansson et
al. (2010)
Sweden
Norway
spruce
(Picea
abies)
Addition
6, 12.5,
25, 50
3
Thalli N %
Six and
12.5 doses:
not significant
25 and
50 doses:
increase
Johansson et
al. (2011)
Sweden
Boreal forest
Addition
300
1
Thalli N %
Increase
McMurrav et
al. (2013)
Wyoming
Conifer
forests
Ambient
0.8-4.1
n/a
Thalli N %
Increase
Root et al.
(2013)
Western
North
America
Forests
Ambient
0.1-39.3
n/a
Thalli N %
Increase
Boltersdorf et
al. (2014)
Germany
Forests
Ambient
2.2-9.5
n/a
Thalli N %
Increase
Schirokauer et
al. (2014a)
Alaska
(southeast)
Conifer
forests
Ambient
0.05-1.05
n/a
Thalli N %
Increase
McMurrav et
al. (2015)
Idaho,
Wyoming,
Montana
Conifer
forests
Ambient
0.5-4.3
n/a
Thalli N %
Increase
6-42
-------
Table 6-5 (Continued): Growth and physiology responses of forest epiphytic
lichens to nitrogen added via atmospheric deposition or
experimental treatments.
Reference
Study
Location
Vegetation
Ambient
Deposition
or
Addition
Nitrogen
Addition
Rate (kg
N/ha/yr)
Duration
(yr)
Endpoint
Effect of
Additional
Nitrogen
Johansson et
al. (2010)
Sweden
Norway
spruce
(Picea
abies)
Addition
6, 12.5,
25, 50
3
Thalli P %
Not significant
Johansson et
al. (2011)
Sweden
Boreal forest
Addition
300
1
Thalli P %
Two species:
increase
One species:
not significant
Nvbakken et
al. (2009)
Sweden and
Norway
Boreal forest
Addition
50
1
Thallus growth
Three
species:
increase
One species:
not significant
Johansson et
al. (2011)
Sweden
Boreal forest
Addition
300
1
Thallus growth
One species:
increase
One species:
not significant
One species:
decrease
C = carbon; ha = hectare; HN03 = nitric acid; kg = kilogram; m = meter; mg = milligram; N = nitrogen; n/a = not applicable;
P = phosphorus; yr = year.
Notes: single studies are reported more than once if multiple endpoints were measured. References are ordered by endpoint. Only
statistically significant effects are listed as increases or decreases.
In another experiment in Scandinavia, lichens collected from Sweden and Norway were
exposed to added N (50 kg N/ha/yr for one season) to understand whether this addition
altered the concentrations of the C based secondary compounds (CBSCs) thought to
protect lichens from herbivores and whether it affected feeding preferences of gastropod
herbivores (Asplund et al.. 2010; Nvbakken et al.. 2009). The N additions decreased the
concentrations of CBSCs in one species, but had no effect on the other three lichen
species (Nvbakken et al.. 2009). The gastropod herbivores preferred to feed on lichens
from the control treatment for three of the four species, while preferring the thalli from
the N addition treatment for the fourth lichen species (Asplund et al.. 2010). Notably, the
species exhibiting the decrease in CBSCs was not the species that was preferable to
herbivores in the N addition treatment. Together, these results suggest that N deposition
may alter lichen community composition by shifting herbivore feeding preferences.
6-43
-------
6.2.3.4 Net Ecosystem Production and Carbon Sequestration
Response
The 2008 ISA concluded that N deposited onto terrestrial ecosystems increased net
primary productivity (NPP) and ecosystem C storage (kg C/ha). However, the extent to
which this additional N stimulates forest C sequestration was a topic of debate
[e.g., Nadelhoffer et al. (1999b); Magnani et al. (2007); De Schriiver et al. (2008); de
Vries et al. (2008); Sutton et al. (2008)1. Magnani et al. (2007) correlated net ecosystem
production (NEP; kg C/ha/yr) with estimated rates of N deposition for 20 forested sites in
North America and Europe, which resulted in an estimate of 725 kg C sequestered per kg
of added N (i.e., kg C/kg N). However, this estimate was widely criticized and a
reanalysis of these data produced an estimate of C sequestration more than an order of
magnitude lower [68 kg C/kg N; Sutton et al. (2008)1.
Since the 2008 ISA, several new syntheses and a number of field experiments and
modeling studies have provided further evidence that N deposition increases NPP, NEP,
and ecosystem C content, and that have more tightly constrained estimates of the
response of plant, soil, and ecosystem C content to N deposition. Nave et al. (2009b)
estimated that up to 15% of the N needed to support NPP in a northern Michigan aspen
forest was supplied by atmospheric N deposition. In China, N additions of 20, 50, or
100 kg N/ha/yr to an old growth boreal forest stimulated aboveground tree productivity
by 5, 7, and 23% during the second and third years of the experiment (Du and Fang.
2014). Chen et al. (2011) found that a Douglas fir stand in the Pacific Northwest
increased NEP by 2,500 kg/ha (+83%) in the first year after receiving 200 kg N/ha of
urea, both from decreased respiration (930 kg/ha) and increased gross primary production
(157 kg/ha). Increases in ecosystem C content have been noted in long-term N addition
experiments in Massachusetts (Frev et al.. 2014). Michigan (Pregitzer et al.. 2008). and
New York (Lovett et al.. 2013). Using a process model that neglected changes in stand
ontogeny, disturbances, and shifts in forest management, de Vries and Posch (2011)
concluded that N deposition was a dominant determinant of European forest productivity
throughout the 20th century.
Since the 2008 ISA, a variety of techniques have been used to quantify the C
sequestration response of forests to N deposition, particularly in Europe (Frev et al..
2014). Using a 15N-labeling technique, Gundale et al. (2014) observed a linear
relationship between C sequestration and N additions of up to 50 kg N/ha/yr that had a
slope of 16 kg C/kg N in a Swedish boreal forest. Hvvonen et al. (2008) synthesized soil
and plant C sequestration data from 15 long-term (14-30 year) N addition experiments in
boreal (Piceci abies, Pinns sylvestris) forests in Sweden and Finland and estimated that C
sequestration averaged 23 kg C/kg N for Piceci and 30 kg C/kg N for Finns, with an
6-44
-------
additional 11 kg C/kg N within the soil. Also in Sweden, Eliasson and Agren (2011)
applied an ecosystem model to Scots pine (Finns sylvestris) forests and estimated that
ecosystem C stocks increased by 24,123 kg/ha in response to a cumulative 224 kg N/ha
added over a century (108 kg C/kg N). Based on the stimulation of photosynthesis and
assumptions about the fraction of photosynthate that is ultimately transformed into tree
biomass, Fleischer et al. (2013) estimated that N deposition stimulates C sequestration by
25 kg C/kg N.
Forest inventory studies typically make use of one or more decades of tree growth data
from national or continental monitoring networks and relate the variation in this growth
to differences in N deposition and other environmental factors. Across western and
northern Europe, Solberg et al. (2009) found that variation in tree volume increment was
positively related to N deposition and summer temperature, particularly for pine {Finns)
and spruce (Picea), with similar, but weaker, relationships for beech (Fagns) and oak
(Onerous). The overall model estimated C sequestration was 19 kg C/kg N. Eastaugh et
al. (2011) analyzed Norway spruce (Picea abies) growth data from the Austrian National
Forest Inventory and estimated that N deposition sequestered 21.6 kg C/kg N in
aboveground tree biomass during the latter half of the 21st century. Etzold et al. (2014)
and Ferretti et al. (2014) took similar approaches with 18 inventory plots in Switzerland
and 25 inventory plots in Italy, respectively. Both studies found a positive relationship
between N deposition and NPP, but the authors were unable to isolate NPP effects from
other environmental factors.
Within the U.S., the analysis of forest inventory data in the northeastern U.S. by Thomas
et al. (2010) estimated that N deposition increased aboveground tree biomass at
61 kg C/kg N; a rough estimate including belowground C increased this to 73 kg C/kg N.
Pinder et al. (2012) modified the estimates of Thomas et al. (2010). using alternate values
for N deposition, more complex belowground biomass estimates, and changes in soil C
pools, to produce an enhancement of 65 kg C/kg N. One criticism of the inventory and
modeling studies is that they match variation in growth only to current levels of N
deposition and neglect the potential effects of previous N deposition at that site. Thus, the
influence of current N deposition may be exaggerated (Hogberg. 2012). In addition,
estimates of tree growth and biomass production typically rely on the application of
equations that predict tree biomass based on measurements of tree stem diameter, either
alone or in combination with tree height measurements. Ibanez et al. (2016) observed that
at the sites in Michigan where N additions had increased tree growth (Pregitzer et al..
2008). the N additions had also altered the allometric patterns of sugar maple tree growth
so that at a given tree stem diameter at breast height (1.4 m), the trees receiving N
additions were taller than those in the control plots. This meant that stemwood biomass
equations underestimated tree size for trees receiving N additions (Ibanez et al.. 2016). If
6-45
-------
these results apply in other environments and with other tree species, this implies that
allometric biomass equations may be underestimating tree biomass in regions
experiencing high rates of atmospheric N deposition and that estimates of C sequestration
due to tree growth increases [e.g., Thomas et al. (2010)1 may be too low.
In the long-term N addition experiment at Harvard Forest (Frcv et al.. 2014). the C
response to N deposition ranged from -2 to 30 kg C/kg N among the four treatment types
(50 or 150 kg N/ha/yr; Finns or Quercus forests) and the amount of C sequestered in the
soil was equal to or greater than the amount of C sequestered in trees. Similar results
showing more C sequestration in the soil than in trees were found in N addition
experiments at four Michigan forests [23 kg C/kg N soil vs. 17 kg C/kg N tree; Pregitzer
et al. (2008)1 and in the Catskills (Lovett et al.. 2013). Several syntheses of forest C
sequestration changes in response to N deposition have been conducted (Tian et al..
2016a; Frev et al.. 2014; Pinder et al.. 2013; Pinder et al.. 2012; Butterbach-Bahl et al..
2011; Janssens et al.. 2010; Liu and Greaver. 2009). producing estimates of 12-41 kg
C/kg N. Often these synthesis studies are meta-analyses; however, de Vries et al. (2014a)
took a novel approach of combining estimates of ecosystem N retention with N allocation
and C:N ratios to develop stoichiometric estimates ofN deposition for tropical,
temperate, and boreal ecosystems. Although this approach involves a number of
important assumptions about biogeochemistry and tree physiology, estimates for tree and
soil C sequestration response rates were similar to values from other syntheses
(Figure 6-3). In their study of aboveground NPP responses per unit of added N, Tian et
al. (2016a) found that the average forest response was a 3.75% increase in NPP per g of
N.
Variation in these estimates is likely influenced by the research approach. For instance,
the model used by Eliasson and Agren (2011) predicted almost no N losses from the
forests due to leaching or denitrification, which are both widely reported at high rates of
N deposition. However, environmental and ontological factors also appear to influence
the sensitivity of forest C sequestration to N deposition. Using the G'DAY model, Dezi et
al. (2010) found recent forest harvests, shifts in plant C allocation away from roots, and
canopy N uptake all increased the C sequestration response to N deposition. Among the
15 experiments synthesized by Hvvonen et al. (2008). the amount of C sequestered
ranged from -0.8 to 61 kg C/kg N, with greater increases in kg C/kg N at lower N doses.
The amount of additional tree C sequestered was small in plots that had low soil
O-horizon C:N ratios (near 25) and the amount of C sequestered by added N increased
with O-horizon C:N until C:N reached 35. Among Picea forests, young stands were more
responsive than old stands, but no such effect was apparent for Finns. In addition, C
sequestration increased more in plots simultaneously supplied with K and P (Hvvonen et
al.. 2008). Janssens et al. (2010) acknowledged the role of stand age in altering the
6-46
-------
dynamic of C cycling, explicitly removing young, rapidly expanding forests from some
portions of the meta-analysis.
A. Aboveground Forest Biomass
Modeling
oO o
-O-
Wamelink ct al. 2009b
Fleischer et al. 2013
Inventory
o
de Vries at al. 2006
O°o
COD
E*»t*ugh wt al. 2011
Solberg et al. 2009
Laubhann ct al. 2009
Experimental
c
oo (
oO
o
) o o
o
o
oaO
Oo
°o
CD O
O
Nadelhoffer et al. 1999
HOgberget al. 2006
Hyvonen et al. 2008
Pregitieret al. 2008
Lovett et al. 2013
Frey et al. 2014
Gundale et al. 2014
Fowler etal. 201S
Synthesis
o
•o
Butterback-Bahl et al. 2011
de Vries et al. 2014
dc Vries ct al. 2014
de Vries et al. 2014
-30 0 30 SO 90
Response Ratio (kg C kg"1 N)
B. Ecosystem Carbon
Modeling
CD
o
O
O O
—O
o
o
o
O
Levy et al. 2004
Sutton et al. 2DOB
Sutton et al. 2008 (one site]
wamelink etal. 20090
Wamdink etal. 2009a
Eliasson & Agren 2011
Inventory
O
o
de Vries etal. 2006
Sutton ct al. 2008 (Macnani)
Experimental
0
o
O
> oO o
O OO O1 °
o o O
O O 0 0
O O o O
o
o
o
0 0
Nadelhoffer et al. 1999
Hftgberg et al. 2006
Hyvonen et al. 2008
Pregltzeretal. 2008
Lovett etal. 2019
Freyet al. 2014
Gundale et al. 2014
Fowler et al- 2015
Synthesis
—o-
>
Liu & Greaver 2009
Butterbach-Bahl et al. 2011
de Vries etal. 2014
de Vries et al. 2014
dit Vriui |