Appendix A. Detailed description of methods for estimating empirical critical loads of nitrogen.
We summarized current knowledge relating nitrogen (N) inputs to ecological effects, from published literature, reports, and some recent unpublished studies for ecosystems across the United States. We included N gradient, long-term and fertilizer studies; for fertilizer studies, we focused on those with low N additions (i.e., close to ambient deposition). We included both publications that report a response and those that report the lack of response.
Empirical critical loads are determined by recording the deposition level at which detrimental ecological effects have occurred in the field. The lowest deposition level at which a detrimental response occurs is considered the critical load. In cases where there is a variation in the level of N input that causes a given response, a range for the critical load is often reported (e.g., Bobbink et al. 2003).
Empirical critical loads rely on observed cause-and-effect or correlative relationships rather than on models or other forms of indirect inference. Because the observed responses to a given deposition level are based on past deposition scenarios, they may not be indicative of the response that might occur in the future. For critical loads derived from gradient studies, the assumption typically is that the N deposition rate is the major factor contributing to changes in a particular response (e.g., foliar N concentrations, species richness) along the gradient, even though soil, climate, biogeochemical factors, and other pollutants, may also influence the response. (For lichens, these other factors were addressed in the gradient studies) For critical loads derived from fertilization studies, the effects of added N are isolated through controlled experimentation; however, the scale of these experiments is often fairly small and their numbers are limited, making extrapolation to the ecoregion scale difficult. Another pitfall of the empirical approach is that the observed response may be unique to the site at which it was measured (because of location-specific disturbance history or site characteristics such as soil thickness) and may not be truly representative of the ecosystem type. Observations used to determine or estimate empirical critical loads are generally made over a relatively short term. In contrast, the residence time of N within major ecosystem compartments (e.g., soil, sediment, forest biomass) is generally on the order of decades to centuries. As a result, empirical critical loads provide limited understanding of true ecosystem response to sustained elevated N inputs or their response to changes in these inputs. Critical loads modeling addresses some of the disadvantages of the empirical critical loads approach. However, current modeling capabilities do not enable large scale mechanistic modeling of critical loads for ecoregions across the United States. Thus we utilize multiple sources of current empirical information (multiple gradient and fertilization studies) to determine critical loads. To address these uncertainties, we present the ranges of responses observed. Where we have estimated an empirical critical load for N, we have made some assessment of the reliability of that estimate. We used the same categories for reliability as those used in European assessments (Bobbink et al. 2003, UBA 2004):
In all cases, accurate estimates of N deposition are essential for developing useful empirical critical loads of N.
In presenting the ecosystem responses and critical loads, we identified the response parameter and the threshold. Within an ecosystem, there could be very different types of response parameters that vary in sensitivity, and hence in their critical load. This poses a challenge in setting a critical load for the ecosystem. In some cases, it may be appropriate simply to choose the lowest critical load value, while in other cases, there may be some responses more characteristic or of greater concern which should be considered the primary indicator for that ecosystem. Furthermore, comparing critical loads that were set based on different ecosystem responses presents a challenge.
Several criteria were determined before assigning empirical critical loads to an ecosystem type. These included: what response variables are relevant, what types of studies are acceptable, how to choose the value for the critical load, and if and when we can extrapolate conclusions beyond the study area.
Approaches for setting empirical critical loads for N for specific ecosystem components are discussed below. Approaches for setting critical loads for aquatic ecosystem components follow.
A.2 Mycorrhizal Fungi
A.2.1 Response Variables
The impacts of N deposition on arbuscular mycorrhizal fungi (AMF) and ectomycorrhizal fungi (EMF) have been determined primarily by examining the changes in percent root colonization, soil spore density (AMF) or sporocarp production (EMF), spore or sporocarp species composition, DNA or immunologic assays of species composition and structure on root tips or of hyphae in soils, and functional analysis of fungal communities. Data from naturally occurring, anthropogenic N deposition gradients have been particularly useful in determining critical loads because of the ability to fit response curves. In deciding if N inputs are affecting mycorrhizal fungi, we focused mostly on community response (species richness and community structure changes) and functional responses. However, where these are lacking, we considered biomass responses, giving them less weight than the other factors.
A.2.2 Acceptable Studies
For information sources, we put more weight on results of long-term (>10 yr) fertilizations and gradient studies rather than short-term fertilizations, as the latter are unlikely to represent a steady-state response to N. Gradients have the advantage over typical two-level experiments in that response curves and threshold responses can be derived. However, even a two-level study with only a control and treatment can put an upper bound on critical loads if there is a response and lower bound if there is no response. Therefore we have included some data from experimental sources, recognizing their limitations.
A.2.3 How the Critical Load Value was Chosen
For our purposes, any detectable deviation in mycorrhizal community structure or composition from the baseline state is considered to have crossed a critical load threshold. For example, if the best-fit response curve in an N deposition gradient indicates a linear response starting from the baseline, then any addition above the baseline would exceed the critical load. If, in contrast, the data suggest a nonlinear response with a threshold below which no directional community change can be detected, then the critical load is set at this threshold. The criterion of any detectable directional change is a very strict criterion that could be open to debate, but it has the merit of being relatively easy to define.
One justification for using the community change criterion is that there appear to be functional consequences for host plants arising from N-driven community change. This has been determined by growing plants using inoculum that has been subject to elevated N. Nitrogen-affected inoculum not only has lower spore density, but also has lower diversity and shifts in species composition (Egerton-Warburton and Allen 2000, Egerton-Warburton et al. 2007, Sigüenza et al. 2006). Growth response studies showed that native late successional plants had reduced biomass with N-affected inoculum, while early successional plants were unaffected or sometimes had increased biomass. This indicates that long-term trends in losses of late seral (successional) vegetation and replacement by early seral or invasive species were in part promoted by shifts in mycorrhizal fungal species composition due to elevated N deposition (Corkidi et al. 2002, Sigüenza et al. 2006).
For ecological regions for which there are no data, we have cautiously extrapolated our critical load values to regions that have similar biota, climate, and soils. For example, we have suggested that ectomycorrhizal community response to N in an Engelmann spruce forest in the Rockies (Northwestern Forested Mountains ecoregion), for which there are no data of which we are aware, would be similar to that of communities in Alaskan white spruce (Marine West Coast Forest ecoregion) and northeastern red spruce- balsam fir (Northern Forest ecoregion) forests, for which we do have data. Thus we have used, with qualification, the critical loads for the latter two as the basis for a provisional estimate for the critical load for the former.
A.3 Lichens and Bryophytes
Lichens, especially forest epiphytes, have been systematically surveyed for air quality monitoring purposes by the U.S. Forest Service Forest Inventory Analysis (FIA) program and other U.S. Forest Service and National Park Service efforts, and many of these data are available electronically: the NPLichen database (Bennett and Wetmore 2005a, 2005b; U.S. Geologic Survey 2010), the NPElement database (U.S. Geologic Survey 2009), the U.S. Forest Service Forestry Inventory and Analysis Program (U.S. Forest Service 2005, 2007), and the U.S. Forest Service National Lichens and Air Quality Database (U.S. Forest Service n.d.). No equivalent efforts exist for bryophytes in the United States. Therefore, apart from the Tundra and Taiga ecoregions (for which we extrapolate some critical loads from European and Canadian studies) and the North America Desert ecoregion (in which bryophytes are integral components of cryptogamic crusts), we primarily reference lichen studies.
Lichens are adapted to different levels of N availability and most species can be sorted into one of three N indicator groups: oligotrophs are more or less restricted to nutrient-poor environments; mesotrophs have an intermediate nutrient requirement; and eutrophs thrive in nutrient-enhanced environments. A few lichens have broad ecological amplitudes and tolerate a large range of N availability. As anthropogenic N is added to background N levels, the excess N favors the small, cosmopolitan eutrophs at the expense of native oligotrophic and mesotrophic forage and N2-fixing lichens (McCune and Geiser 2009).
To avoid possible confusion from variable terminology in the literature, we used the broad terms ‘oligotroph’, ‘mesotroph’ and ‘eutroph’ in lieu of the more classic lichenological terms, ‘acidophyte’, ‘neutrophyte’, and ‘nitrophyte or nitrophile’, respectively.
A.3.2 Response Variables
To identify critical loads for the Marine West Coast Forests, Northwest Forested Mountains, and Mediterranean California ecoregions, we used two epiphytic lichen responses that are closely linked to ecological impact: N concentration in lichen thalli, and composition of lichen communities. Changes in thallus N concentrations and physiology usually precede measurable changes in lichen community composition. Local N deposition is correlated with N concentrations of lichen thalli, which passively accumulate N in excess of nutritional requirements (Bruteig 1993, Søchting 1995). This response is typically linear at low N deposition levels (Glavich and Geiser 2008). The upper limit of thallus N concentrations among clean sites is typically used as the response threshold. Increasing N deposition, likewise, causes changes in lichen community composition—such as a decrease in the number of oligotrophic species and an increase in the number of eutrophic species.
For the Taiga, Tundra, and North American Desert ecoregions, we included physiological and morphological measures of lichen status, such as changes in photosynthetic capacity (e.g., chlorophyll pigment ratios, carbon dioxide (CO2) uptake), membrane integrity, amounts of characteristic protective chemicals (e.g., usnic acid), apportionment of resources to algal vs. fungal partners, ultrastructural integrity, or growth rates.
A.3.3 Acceptable Studies
As with the mycorrhizal fungi, we primarily relied on gradient studies as information sources rather than short-term fertilizations. In addition, studies of epiphytes are considered superior to studies of ground-dwelling species due to epiphytes’ greater independence from soil nutrients. We obtained community data from regional, systematic surveys of epiphytic lichens, primarily conducted by the FIA program for the Rocky Mountain (11 sites), Marine West Coast Forests (1411 sites), and Mediterranean California ecoregions (118 sites). Most of the physiological studies were N-addition studies. Since they were short-term additions or added N to already elevated background deposition levels, we feel they do not adequately predict long-term effects. However, they do prescribe upper limits for critical loads.
A.3.4 How the Critical Load Value was Chosen
We determined the critical load by identifying the level of N deposition which led to specific ecological impacts (changes in community composition and increased thallus N concentration) using the datasets described above. The ecological impact at any N deposition level was quantified by doing a regression on N deposition against proportions of oligotrophic, mesotrophic, and eutrophic species, or an index separating climate-related influences from N deposition influences on community composition called the “air score” (Geiser and Neitlich 2007, Jovan 2008).The value of N deposition at the critical response threshold is the critical load. The critical thresholds were community shifts toward a higher proportion of eutrophic species and elevated thallus N concentrations. The exact value for the critical threshold thallus N concentration was typically set based on the 97.5% quantile for the range of thallus N concentrations observed at clean sites. The 97.5% quantile is the thallus N concentration below which 97.5% of clean site thallus N concentrations fall. The exact threshold for species composition varied because the baseline (species composition at an unpolluted site) varied in different ecosystems (see Table A.1). Because all the lichen response measures were linear with respect to deposition, we chose community-based critical response thresholds high enough to allow for natural variability across clean sites, but low enough to protect communities from a shift in dominance of indicator groups. We used the 95% confidence interval as the range for the critical load.
Forest Inventory and Analysis (FIA) Program lichen-community based “air scores” are a measure of air quality generated from lichen community data and region-specific interpretive models separating air pollution effects on community composition from climate, elevation and other site effects (Jovan 2008). Models are typically developed from a calibration subset of regional FIA lichen community data using non-metric multidimensional scaling and Sorenson’s distance measure (McCune and Mefford 1999) to ordinate sites in N-dimensional space. With the aid of overlays, regression, scatter plots and other tools, at least two strong gradients can usually be extracted from the ordination: one relating to air quality and a second relating to macroclimate. Air quality and climate are then scored as the distance of the community (site) along the respective gradient. Repeat or additional surveys can be fitted and scored at any time to assess spatial and temporal trends. Finished models are available for the southeastern United States (McCune et al. 1997), California’s greater Sacramento Valley (Jovan and McCune 2005) and Sierra Nevada (Jovan and McCune 2006), and western Oregon and Washington (Geiser and Neitlich 2007). In the western United States, nutrient N deposition is considered the dominant air pollutant affecting lichen communities (Jovan 2008, Geiser et al. 2010); in the southeastern United States, N- and sulfur (S)-containing acidifying and fertilizing compounds are considered most influential (McCune et al. 1997).
Lichen community and thallus N concentration appear to correlate better with N concentration (mg L-1) in deposition than with N load (kg ha-1 y-1) (Geiser et al. 2010). Because lichen thalli accumulate and leach N in dynamic equilibrium with concentrations of pollutants in precipitation, precipitation can have a diluting effect on the concentrations of pollutants to which lichens are exposed. Thus, the same load of N deposited in a higher volume of precipitation will have a smaller impact. Therefore, we accounted for precipitation volume when developing lichen critical loads based on measurements of bulk (NADP-style) or modeled N deposition.
In the Marine West Coast Forests, Northwest Forested Mountains, and Mediterranean California ecoregions, lichen-based critical loads were derived by correlating lichen responses (described above) to Interagency Monitoring of Protected Visual Environments (IMPROVE; fine particulate ammonium nitrates (NH4NO3) and sulfates ((NH4)2SO4) in air) and NADP monitoring networks (wet deposition of inorganic N), 10 year total annual average deposition of 16 N-species modeled by CMAQ*, or in-situ deposition measurements of ammonium (NH4+) and nitrate (NO3-) in bulk and canopy throughfall.
IMPROVE (Interagency Monitoring of Protected Visual Environments) monitors include an aerosol sampler that measures total fine particulate (< 2.5 µm diameter) NO3- and sulfate (SO4-) concentrations over 24 hours every third day. These anions are presumed balanced by NH4+ and best correlations between fine particulate N and lichen-based air scores were observed when mean annual averages included N from both sources (Improve 2010; University of California Davis 1995).
NADP (National Atmospheric Deposition Program); weekly samples of precipitation captured in buckets designed to open during rain and snow events are analyzed for concentrations (mg L-1) and total wet deposition (kg ha-1 y-1) of NH4+ and NO3- and total inorganic N (from NH4+ and NO3-). Data that meets quality control criteria can be downloaded from the website (National Atmospheric Deposition Program 2009).
* CMAQ (Community Multiscale Air Quality model). The CMAQ model forecasts wet and dry deposition of 16 primary and secondary N-containing air pollutants: HNO3, NH3, NO2, PAN, NO, RNO3, PAN2, N2O5, HONO, ANH4I, MA-PAN, ANO3I, PBZN, ANO3J, and ANH4J. Model output accounts for complex interactions among many variables (e.g., landscape, vegetation density, atmospheric chemistry, and climate). Mean annual deposition, based on 1990–1999 emissions data, was calculated on a 36 km grid as described in Porter (2007). For model details, see the on-line CMAQ science documentation (U.S. EPA 2009). Lichen response was related to the total sum of wet and dry deposition of all 16 air pollutants.
To make initial estimates of lichen-based critical loads for N in forested ecological regions where existing analyses were lacking, (i.e., Taiga, Northern Forests, Eastern Forests, and the Temperate Sierras), and to encompass more of the Northwest Forested Mountains ecoregion, we applied a model that was developed in western Oregon and Washington for the Marine West Coast Forests ecoregion (Geiser et al. 2010). This model has been shown to replicate previously published critical loads for wet oak woods of Scotland (Mitchell et al. 2005), mesic conifer forests of the California Sierras, and dry mixed hardwood-conifer forests of California’s Greater Sacramento Valley (Geiser et al. 2010). Input variables required are minimum and maximum mean annual precipitation and a lichen community response threshold (e.g., maximum acceptable percent eutrophs). Mean annual minimum and maximum precipitation within each ecoregion was estimated from Parameter-elevation Regressions on Independent Slopes Model (PRISM) national map of 1971–2000 normals (Daly et al. 2002, PRISM 2010). Expert judgment was used to choose reasonable community response thresholds (Table 4.1). The thresholds are a minimum allowable oligotroph or maximum allowable eutroph percentage of the community composition. Forests with hardwood components were assigned a higher response threshold than western Oregon and Washington coniferous forests because the more alkaline bark and richer canopy exudates of hardwoods favor eutrophs (Van Herk 2001, Goward and Arsenault 2000 ). Colder forests were assigned lower response thresholds than warmer forests because native epiphytes are typically oligotrophs adapted to cold- and low-nutrient environments—eutrophs are better adapted to nutrient-concentrated, heat and drought-stressed environments typical of warmer climates (Jovan and McCune 2004, Will-wolf et al. 2006). The most problematic area was the Eastern Forest ecoregion, for which systematically sampled lichen community data exist but are also influenced by current or historic levels of N and S deposition and atmospheric concentrations of sulfur dioxide. Because the cumulative effects of multiple pollutants on epiphytic lichens are not well understood, our critical load estimate for eastern forests, based on a western model that accounts for nutrient N deposition alone, are clearly provisional.
A.4 Herbaceous species and shrubs
A.4.1 Response Variables
Responses of herbaceous species to N deposition included changes in species abundances and composition, particularly increases in nitrophilic and invasive species, increases in productivity, and changes in foliar chemistry. In some herbaceous dominated communities changes in N cycling and inorganic N leaching below the rooting zone were also used. Typically the critical loads for these variables were ca. 10 kg ha-1 y-1 above the value estimated from changes in plant composition.
In more impacted ecoregions such as the forested areas of the eastern U.S. using herbaceous species for critical loads may be problematic, because, as previously stated, N deposition (in combination with S deposition) over the last half century has probably already significantly altered community composition. The critical loads obtained using herbaceous species in these more impacted regions will therefore potentially overestimate the level associated with the initial changes in plant diversity.
A.4.2 Acceptable Studies
We included studies across N deposition gradients and low level N fertilization experiments to evaluate herbaceous species and shrub responses to N deposition We excluded fertilization studies when single input rates of more than double the current ambient level were used. As noted above, an implicit assumption was made that the primary factor influencing variation in herbaceous species composition in gradient studies was N deposition. We considered studies that included a range of N experimental inputs to be more reliable (##) than those which used single fertilization values or gradient studies.
A.4.3 How the Critical Load Value was Chosen
The critical load value was estimated as the range between the input level at which no response was observed and the lowest level at which an alteration in species abundance or changes in N cycling occurred. In some cases, the input level where a response occurred could be quantitatively assessed by fitting the response data to simple mathematical functions (Bowman et al. 2006). Quantitative changes in species composition for both gradient and N fertilization studies are best determined using point-intercept or frequency estimates. Subtle changes may be missed or misinterpreted using more subjective estimates (e.g., visual estimates of cover). An increase in nitrophilic species cover was used to determine critical loads in tall-grass prairies, Mediterranean shrublands and deserts, and alpine tundra. Changes in community structure, using ordination analysis, was also used in alpine vegetation. In fire-prone western ecosystems, the increased productivity of invasive grasses has promoted increased fires, and the productivity threshold under N deposition that promotes fire has been set as the critical load (Rao et al. 2010). Decreases in the diversity of understory herbaceous species were used in eastern forests, again indicating that the initial increase in nitrophilic species has already occurred in these more impacted ecosystems.
Gradient studies were used for only a few ecoregions (e.g. coastal sage scrub). Critical loads were estimated as a range below the deposition rate where significant increases in nitrophilic species occurred. In general, this approach has not been widely used for herbaceous species composition, due, in large part, to the multitude of factors other than N deposition that influence the composition of herbaceous species composition and chemistry.
Critical loads based on changes in N cycling in an alpine ecosystem were estimated as below the N inputs where significant changes in net N mineralization, nitrification, and inorganic N leaching below the rooting zone were observed.
A.5.1 Response variables
The variables considered for forests include changes in percent N in trees and soils, alterations in soil N cycling, elevated NO3- losses, increased susceptibility to secondary stressors (freezing, drought, pests), declines in tree health, declines in tree growth, and increased mortality. There is considerable variation in N concentration in tissue by species and site and over time. Thus, setting a threshold value for tissue N concentration, as was done for lichens, is difficult.
Changes in species composition, such as those reported for communities of shorter-lived organisms, are infrequently observed for trees. We included some studies that reported decreased tree health and growth and increased mortality, which are observed only rarely.
A.5.2 Acceptable studies
We included N gradient, fertilization, and long-term studies. Results from long-term, low-level fertilization and gradient studies are considered the most reliable. The additions of N in most forest fertilization studies are much greater than ambient deposition. Gradient studies have the advantage of demonstrating changes in real time. When a single gradient study was used, the results could not be considered more than fairly reliable , #. When many gradient studies were included (e.g., Aber et al. 2003), the results could be considered reliable, ##.
A.5.3 How the critical load value was chosen
The method varied among ecoregions. In two ecoregions with extensive datasets (Eastern Temperate and Northern Forest ecoregions), we were able to identify a break point on the response curve, the point at which the response begins, and thereby identify the critical load with high certainty. We used this approach for NO3- losses from forest ecosystems which have been shown to increase above a threshold N deposition (Aber et al. 2003, Wright et al. 2001) This analysis included 354 sites in the Northern and Eastern Forest ecoregions. In regions with sparser NO3- loss data (e.g., Northwestern Forested Mountains ecoregion), we were able to identify only the range within which the critical load is most likely to lie. For gradient studies without a clear inflection point, for example, the tree growth and mortality studies in eastern and northern forests, we set the critical load to be greater than the low end of the range of deposition over which the response was observed. For fertilization studies, we identified the critical load as lying in the range between the highest N input for which we observed no response and the lowest N input for which we observed a response.
For the Tropical and Subtropical Humid Forests ecoregion for which there are no data, we have cautiously extrapolated our critical load values from regions that have similar biota, climate, and soils. Thus we have used, with qualification, the critical loads for the Southeastern Coastal Plain forests as the basis for a provisional estimate for the critical load for the Tropical and Subtropical Humid Forests.
A.6.1 Response Variables
Responses of wetland species to N deposition include increases in peat accumulation, N concentrations in plant tissues or soil, N2O emissions, productivity, and changes in bulk density, plant anatomy, species abundances and composition.
Coastal wetland species present some challenges for estimating critical loads, because some receive nitrogen loading from multiple sources that include atmospheric deposition, runoff carrying fertilizer and waste water, which combine to create N loads that may reach thousands of kg ha-1 year-1. Thus, the critical loads presented are the value of total N load.
A.6.2 Acceptable Studies
We included studies designed to evaluate N deposition gradients, experimental N inputs, and modeling to evaluate wetland responses to N deposition. Many gradient and fertilization studies are not useful because they do not include inputs on the low end of the gradient. Studies that reported responses across a gradient and those that included fertilizer application up to about twice ambient deposition are used in the evaluation of freshwater wetlands. As previously stated, N loading to coastal ecosystems can be quite high, therefore studies evaluating N loads up to about 10253 kg N ha-1 yr-1 are included in the evaluation (Wigand et al. 2003), however critical loads are set for those endpoints evaluated with treatments of about twice ambient deposition or less.
A.6.3 How the Critical Load Value was Chosen
The critical load value was estimated as the input level at which an alteration in a given endpoint is observed. When multiple studies evaluated the same endpoint, the range of input levels for the two lowest levels at which an alteration in a given endpoint is observed are reported.
A.7 Inland Surface Waters
A.7.1 Response variables
While the goal of establishing critical loads for inland freshwaters is to determine the effects of atmospheric N deposition on aquatic organisms, the response variable used was surface water NO3-, because elevated NO3- concentrations are evidence of leaching from the surrounding watershed. Both mechanistic (Aber and Driscoll 1997) and conceptual (Stoddard 1994) models of N saturation predict that elevated rates of N deposition will eventually produce elevated runoff concentrations of NO3- in otherwise relatively undisturbed catchments. Empirical cross-site watershed studies show the relationship between lake and stream N concentrations or N export as a function of atmospheric N deposition (Sickman et al. 2002, Aber et al. 2003, Fenn et al. 2003, Bergström and Janssen 2006). Such cross-site analyses were originally conducted for Europe (Wright et al. 2001), and typically show a “dog-leg” pattern—low surface water NO3- concentrations or leaching with low levels of N deposition, and elevated leaching losses and concentrations above a threshold of deposition of 10 kg N ha-1 yr-1.
A.7.2 Acceptable studies
We relied on papers and studies that linked aquatic biological, chemical, and ecological response to atmospheric deposition to connect the empirically-determined NO3- concentration critical load with biological freshwater lake and stream impacts. A search of the literature found 100 published papers that reported N deposition or N amendment results for aquatic chemical concentrations or organisms, or experimental dose-response studies of N amendments to algae, zooplankton, or fish. All of these papers were summarized to document the mechanisms and consequences of N enrichment to surface waters. The body of evidence amassed from these studies shows a very strong response of freshwater algae, especially diatoms, from nutrient-poor waters, to increased NO3-concentrations. Our selection of N critical loads used lake chemistry data from three regions, the Sierra Nevada, the Rocky Mountains, and the eastern United States, and is based on the following logical sequence: (1) In otherwise undisturbed headwater lakes with low N deposition, late summer lake NO3- is also low in concentration; (2) Increasing amounts of N deposition to otherwise undisturbed headwater lakes are positively correlated with increased late summer lake NO3- concentrations; (3) Increased lake NO3- concentrations, or nutrient enrichment, triggers biological responses, including increased algal productivity and loss of diversity; (4) Increased lake NO3- concentrations can also represent loss of watershed acid neutralizing capacity (ANC), and this triggers lake acidification and loss of diversity.
A.7.3 How the critical load value was chosen
Lake NO3- concentrations were plotted as a function of wet atmospheric N deposition using stream and lake data for the eastern United States from Aber et al. (2003) and from the western United States using data from Sickman et al. (2002). The “dog-legs,” or thresholds, as described above, where lake NO3- concentrations increased in response to N deposition, were determined by observing the deposition at which concentrations increased. Keeping in mind that the threshold will differ between lakes of strongly versus weakly vegetated catchments, the inflection point where NO3- concentrations increase in response to increasing N deposition defines the empirical N critical load.
TABLE A1. The input variables, assumptions, and rationale behind the calculation of lichen critical loads for U.S. ecoregions, derived from the Marine West Coast Forest model: Critical Load kg N ha-1 yr-1 = (0.0918 + 0.0024 [mean annual precipitation] + threshold air score) / 0.1493. Precipitation is estimated from PRISM 1961–1990 annual average precipitation maps; lichen response thresholds are based on expert opinion.
|Rationale / Assumptions|
Acidic coniferous substrates, cold temps favor oligotrophs.
Therefore, community response threshold set with comparatively
high oligotroph and low eutroph composition.
Climate and forest type similar to Marine West Coast Forest.
Therefore, use same threshold. Synergistic or confounding
effects of acid deposition of S and N, which is much greater
here than Marine West Coast Forest, not considered in this analysis.
|44||451||0.21||30–41||27–34||Model was developed for this region by Geiser et al. (2010)|
Lower thresholds needed for Alaska because
of cooler temperatures (see Taiga rationale).
Lower and upper thresholds calculated by Fenn et al. 2008.
Low humidity and high temperatures favorable to eutrophs
and unfavorable to oligotrophs except in high elevation fog zones.
Humid climate favors oligotrophs but hardwood dominance favors
mesotrophs and eutrophs, so response threshold allows
comparatively more eutrophs. Synergistic or confounding effects
of acid deposition of S and N are not considered in this analysis.
Humid climate and acidic conifer substrates favor oligotrophs.
Warm summer temperatures limiting to oligotrophs? Synergistic
or confounding effects of acid deposition of S and N are not
considered in this analysis.
mixed Conifer Forest
Climate and forest type closest to Northwestern Forested Mountains,
used Fenn et al. 2008 upper threshold to account for comparatively
warmer, drier conditions compared to the Sierras.
Threshold (50% eutrophs) selected from overlay of CMAQ data
on Jovan and McCune 2006 analysis of this ecoregion. Hot
dry summers, hardwood dominance favors mesotrophs and eutrophs.
Natural contribution of eutrophs to lichen community composition
presumed relatively high due to dry, hot climate.
Aber, J. D.; Driscoll, C. T. 1997. Effects of land use, climate variation and N deposition on N cycling and C storage in northern hardwood forests. Global Biogeochemical Cycles. 11:639–648.
Aber, J. D.; Goodale, C. L.; Ollinger, S. V.; Smith, M. L.; Magill, A. H.; Martin, M. E.; Hallett, R. A.; Stoddard, J. L. 2003. Is nitrogen deposition altering the nitrogen status of northeastern forests? BioScience. 53:375–389.
Bennett, J. P.; Wetmore, C. M. 2005a. Lichens of the U.S. National Parks. The Bryologist. 108(4):544–553.
Bennett, J. P.; Wetmore, C. M. 2005b. NPLichen: A database of lichens in the US National Parks. 2005 Final Report. http://www.nbii.gov/files/nplichen/npl_report.pdf
Bergström A., Jansson M. 2006. Atmospheric nitrogen deposition has caused nitrogen enrichment and eutrophication of lakes in the northern hemisphere. Global Change Biology 12:635–643.
Bobbink, R.; Ashmore, M.; Braun, S.; Flückinger, W.; van den Wyngaert, I. J. J. 2003. Empirical nitrogen critical loads for natural and semi-natural ecosystems: 2002 update. In: Achermann, B.; Bobbick, R., eds. Empirical critical loads for nitrogen. Environmental Documentation No. 164. Berne, Switzerland: Swiss Agency for the Environment, Forests, and Landscape: 43–170.
Bowman, W. D.; Larson, J. R.; Holland, K.; Wiedermann, M.; Nieves, J. 2006. Nitrogen critical loads for alpine vegetations and ecosystem response- are we there yet? Ecological Applications. 16(3):1183–1193.
Bruteig, I.E. 1993. The epiphytic lichen Hypogymnia physodes as a biomonitor of atmospheric nitrogen and sulphur deposition in Norway. Environmental Monitoring and Assessment. 26:27–47.
Corkidi, L.; Rowland, D. L.; Johnson, N. C.; Allen, E. B.. 2002. Nitrogen fertilization alters the functioning of arbuscular mycorrhizas at two semiarid grasslands. Plant and Soil. 240:299–310.
Daly, C.; Gibson, W. P.; Taylor, G. H.; Johnson, G. L.; Pasteris, P. 2002. A knowledge-based approach to the statistical mapping of climate. Climate Research. 22:99–113.
Egerton-Warburton, L. M.; Allen, E. B. 2000. Shifts in arbuscular mycorrhizal communities along an anthropogenic nitrogen deposition gradient. Ecological Applications. 10:484–496.
Egerton-Warburton, L. M.; Johnson, N. C.; Allen, E. B. 2007. Mycorrhizal community dynamics following nitrogen fertilization: a cross-site test in five grasslands. Ecological Monographs. 77:527–544.
Fenn, M. E.; Baron, J. S.; Allen, E. B.; Rueth, H. M.; Nydick, K. R.; Geiser, L.; Bowman, W. D.; Sickman, J. O.; Meixner, T.; Johnson, D. W.; Neitlich, P. 2003. Ecological effects of nitrogen deposition in the western United States. BioScience. 53:404–420.
Fenn, M. E., Geiser, L. H. In press. Temperate Sierra. In: Pardo, L. H., Robin-Abbott, M. J., Driscoll, C. T., eds. Assessment of effects of N deposition and empirical critical loads for nitrogen for ecoregions of the United States. General Technical Report NRS-80 Newtown Square, PA: U.S. Department of Agriculture, Forest Service, Northern Research Station, 291 p.
Fenn, M. E.; Jovan, S.; Yuan, F.; Geiser, L.; Meixner, T.; Gimeno, B. S. 2008. Empirical and simulated critical loads for nitrogen deposition in California mixed conifer forests. Environmental Pollution. 155:492–511.
Glavich, D. A.; Geiser, L. H. 2008. Potential approaches to developing lichen-based critical loads and levels for nitrogen, sulfur and metal-containing atmospheric pollutants in North America. The Bryologist. 111:638–649.
Geiser, L. H.; Jovan, S. E.; Glavich, D. A.; Porter, M. 2010. Lichen-based critical loads for atmospheric nitrogen deposition in western Oregon and Washington Forests, USA. Environmental Pollution 158:2412–2421.
Geiser, L. H.; Neitlich, P.N. 2007. Air pollution and climate gradients in western Oregon and Washington indicated by epiphytic macrolichens. Environmental Pollution. 145:203–218.
Goward, T. ; Arsenault, A. 2000. Cyanolichen distribution in young unmanaged forests: a dripzone effect? The Bryologist 103:28–37.
Jovan, S. 2008. Lichen bioindication of biodiversity, air quality, and climate: baseline results from monitoring in Washington, Oregon, and California. Gen. Tech. Rep. PNW -737. Portland, OR: U.S. Department of Agriculture, Forest Service, Pacific Northwest Research Station. 115 p.
Jovan, S.; McCune, B. 2004. Regional variation in epiphytic macrolichen communities in northern and central California forests. The Bryologist 107:328–339.
Jovan, S.; McCune, B. 2005. Air-quality bioindication in the greater Central Valley of California, with epiphytic macrolichen communities. Ecological Applications 15:1712–1726.
Jovan, S.; McCune, B. 2006. Using epiphytic macrolichen communities for biomonitoring ammonia in forests of the greater Sierra Nevada, California. Water, Air and Soil Pollution. 170:69–93.
IMPROVE (Interagency Monitoring of Protected Visual Environments). n.d. Data resources. Available at http://vista.cira.colostate.edu/Improve/Data/IMPROVE/improve_data.htm. (May 5 2010).
McCune, B.; Dey, J.; Peck, J.; Heiman, K.; Will-Wolf, S. 1997. Regional gradients in lichen communities of the southeast United States. The Bryologist. 100:145–158.
McCune, B .M.; Geiser, L. H. 2009. Macrolichens of the Pacific Northwest. 2nd Edition. Corvallis, OR: Oregon State University Press. 448 p.
McCune, B; Mefford, M.J. 1999. PC-ORD. Multivariate analysis of ecological communities, Version 4.24. MJM Software: Gleneden Beach, OR.
Mitchell, R. J., Truscot, A. M., Leith, I. D., Cape, J N., van Dijk, N., Tang, Y. S., Fowler, D., Sutton, M. A. 2005. A study of epiphytic communities of Atlantic oak woods along an atmospheric nitrogen deposition gradient. Journal of Ecology. 93:482–492.
NADP (National Atmospheric Deposition Program). 2009. NADP maps and data. Available at http://nadp.sws.uiuc.edu/data/. (May 5 2010).
Porter, M.K. 2007. Regional modeling of nitrogen, sulfur, and mercury atmospheric deposition in the Pacific Northwest. Pullman, WA: Washington State University. M.S. thesis. 127 pp.
PRISM Climate Group. 2010. PRISM climate group. Available at http://www.prism.oregonstate.edu/. (December 3 2010)
Rao, L. E.; Allen, E. B.; Meixner, T. 2010. Risk-based determination of critical nitrogen deposition loads for fire spread in southern California deserts. Ecological Applications. 20:1320–1335
Sickman J. O., Melack J. M., Stoddard J. L.. 2002. Regional analysis of inorganic nitrogen yield and retention in high-elevation ecosystems of the Sierra Nevada and Rocky Mountains. Biogeochemistry 57–58:341–374.
Sigüenza, C.; Crowley, D. E.; Allen, E. B. 2006. Soil microorganisms and fatty acid methyl ester profiles along a N deposition gradient. Applied Soil Ecology. 32:13–26.
Söchting, U. 1995. Lichens as monitors of nitrogen deposition. Cryptogamic Botany. 5(3):264–269.
Stoddard J.L. 1994. Long-term changes in watershed retention of nitrogen. Its causes and aquatic consequences. Pages 223–284 in LA B, ed. Environmental Chemistry of Lakes and reservoirs. Washington DC: American Chemical Society.
UBA (UmweltBundesAmt). 2004. Manual on methodologies and criteria for mapping critical levels/loads and geographical areas where they are exceeded. Berlin: Federal Environmental Agency (UmweltBundesAmt). Available at http://www.icpmapping.org.
University of California Davis. 1995. IMROVE data guide. Available at http://vista.cira.colostate.edu/Improve/Publications/OtherDocs/IMPROVEDataGuide/IMPROVEDataGuide.htm. (May 5 2010).
U.S. Environmental Protection Agency. 2009. CMAQ science documentation. Available at http://www.epa.gov/asmdnerl/CMAQ/CMAQscienceDoc.html. (May 5 2010).
U.S. Forest Service. N.d. National lichens and air quality database and clearinghouse. Available at http://gis.nacse.org/lichenair/. (May 5 2010).
U.S. Forest Service. 2005. Forest Inventory and Analysis National Program. Lichen communities. Monitoring air quality and biodiversity. Available at http://fia.fs.fed.us/lichen. (May 5 2010).
U.S. Forest Service. 2007. Forest Inventory and Analysis Program. Phase 3 Field Guide – Section 10. Lichen Communities, Version 4.0. Available at http://fia.fs.fed.us/library/field-guides-methods-proc/docs/2007/p3_4-0_sec10_10_2007.pdf. (May 5 2010).
U.S. Geologic Survey. 2009. NPElement, a database of lichen elemental concentrations in the U. S. National Parks. Version 1.0. Available at http://www.nwhc.usgs.gov/our_research/np_element.jsp. (May 5 2010).
U.S. Geologic Survey. 2010. NPLichen, a database of lichens in the U. S. National Parks. 2010. Version 4.5. Available at http://www.nbii.gov/nplichen. (May 5 2010).
Van Herk, C. M. 2001. Bark pH and susceptibility to toxic air pollutants as independent causes of changes in epiphytic lichen composition in space and time. Lichenologist 33:419–441.
Wigand, C., McKinney, R. A., Charpentier, M. A., Chintala, M. M., Thursby, G. B. 2003. Relationships of nitrogen loadings, residential development, and physical characteristics with plant structure in New England salt marshes. Estuaries 26:1494–1504.
Will-Wolf, S.; Geiser, L. H.; Neitlich, P.; Reis, A.H. 2006. Forest lichen communities and environment – how consistent are relationships across scales? Journal of Vegetation Science 17:171–184.
Wright, R. F.; Alewell, C.; Cullen, J. M.; Evans, C. D.; Marchetto, A.; Moldan, F.; Prechtel, A.; Rogora, M. 2001. Trends in nitrogen deposition and leaching in acid-sensitive streams in Europe. Hydrology and Earth System Sciences. 5:299–310.