Diatoms to human uses: linking nitrogen ... - Wiley Online Library

SPECIAL FEATURE: AIR QUALITY AND ECOSYSTEM SERVICES

Diatoms to human uses: linking nitrogen deposition, aquatic eutrophication, and ecosystem services CHARLES RHODES,1, ANDREW BINGHAM,2 ANDREA M. HEARD,3 JULIE HEWITT,4 JASON LYNCH,5 RANDALL WAITE,6 AND MICHAEL D. BELL2 1

Oak Ridge Institute for Science and Education, Office of Water, and Office of Research and Development, U.S. Environmental Protection Agency, Washington, D.C. 20460 USA 2 Air Resources Division, National Park Service, Denver, Colorado 80225 USA 3 Sierra Nevada Network, National Park Service, Three Rivers, California 93271 USA 4 Office of Water, U.S. Environmental Protection Agency, Washington, D.C. 20460 USA 5 Office of Air and Radiation, U.S. Environmental Protection Agency, Washington, D.C. 20460 USA 6 Office of Air and Radiation, U.S. Environmental Protection Agency, Durham, North Carolina 27711 USA Citation: Rhodes, C., A. Bingham, A. M. Heard, J. Hewitt, J. Lynch, R. Waite, and M. D. Bell. 2017. Diatoms to human uses: linking nitrogen deposition, aquatic eutrophication, and ecosystem services. Ecosphere 8(7):e01858. 10.1002/ ecs2.1858

Abstract. Nitrogen (N) loading to aquatic ecosystems can lead to eutrophication, changing the ecosystem within a waterbody, including primary productivity, water clarity, and food web dynamics. Nutrient loading often first affects the primary productivity of aquatic systems through shifts in phytoplankton communities. However, ecologically important changes in phytoplankton are often not relatable to the general public—whose behavior would need to change to alter patterns of nutrient loading. Therefore, we use the STressor–Ecological Production function–final ecosystem Services Framework to develop 154 chains that link changes in biological indicators of aquatic eutrophication (a shift in phytoplankton community) to final ecosystem services that people use or appreciate. We identify 13 ecological production functions (EPF) within three different ecosystems (alpine lakes, lakes, and estuaries) that connect changes in phytoplankton and algae to ecological endpoints that the general public and policy makers can appreciate. Using the Final Ecosystem Goods and Services Classification System, we identify 18 classes of human beneficiaries that potentially will be impacted by a change in one of these endpoints. We further assign strength-of-science scores to each link within the EPFs for the 154 chains according to how well each link is supported by current peer-reviewed literature. By identifying many pathways through which excess N loading in U.S. surface waters can affect ecosystems and ultimately the beneficiaries of ecosystem services, this work intends to draw attention to gaps in empirical ecological literature that constrain understanding of the magnitude of effects that excess N loading can have on human well-being. Results highlight the importance of intersections between the natural and social sciences when managers and policy makers evaluate impacts from ecological stressors. A balance between knowledgeable specialists proved key to applying this approach and will continue to remain important. Key words: aquatic eutrophication; critical loads; ecological endpoints; ecological production function; ecosystem services; final ecosystem goods and services (FEGS); nitrogen deposition; Special Feature: Air Quality and Ecosystem Services. Received 18 October 2016; revised 21 February 2017; accepted 27 February 2017; final version received 3 May 2017. Corresponding Editor: Debra P. C. Peters. Copyright: © 2017 Rhodes et al. This is an open access article under the terms of the Creative Commons Attribution License, which permits use, distribution and reproduction in any medium, provided the original work is properly cited. E-mail: [email protected]

❖ www.esajournals.org

1

July 2017

❖ Volume 8(7) ❖ Article e01858


INTRODUCTION

In order to protect aquatic ecosystems from eutrophication, humans must make choices that reduce N inputs into these systems. Even when knowing that common biological indicators, like phytoplankton (including diatoms), are an early link in a complex food chain, society often will not respond accordingly unless the change is made relevant to individual interests. Because of the number and complexity of interactions, tracing the effects of excess N from a primary chemical effect to a product of the ecosystem that people care about is rarely attempted. Studies have examined both trophic interactions, such as the effect that altered diatom communities have on small fish and macroinvertebrate communities, and the effect that changes in small fish and macroinvertebrate communities have on piscivorous wildlife (Scheffer et al. 1993, Jeppesen et al. 2000), but rarely look at the broader impacts of the change. To help non-ecologists realize the human impacts from aquatic eutrophication, we link changes to the ecosystem due to aquatic eutrophication to final ecosystem services, the “components of nature, directly enjoyed, consumed, or used to yield human well-being” (Boyd and Banzhaf 2007). Conceptually mapping ecological processes that occur far from the general public’s view to attributes and products of the environment that people know they care about can help decision makers and the beneficiaries of ecosystem goods and services to more easily understand and appreciate how elements of their own welfare depend on deep and complex ecological processes. For example, locally hypoxic waters (from eutrophication) may lower the fish component of raptor diets, harming raptor populations. Hikers may or may not care about diatoms, but they are likely to enjoy seeing piscivorous wildlife such as peregrines or eagles, or to notice that raptors are no longer denizens of the valley. While these may be important effects, they are not always presented to the public when discussing impacts of pollution. We use the STressor–Ecological Production function–final ecosystem Services (STEPS) Framework to connect the response of a biological indicator to the stressor of aquatic eutrophication, through ecological production functions (EPF), to an ecosystem component within affected waters that people consume, interact with, or enjoy (Bell et al. 2017, see

Excess nitrogen (N) in an ecosystem can have wide-reaching effects. The eutrophication of aquatic ecosystems occurs because many primary producers are N-limited and alleviating this limitation can favor some species over others, leading to changes in community structure and species abundance (U.S. EPA 2008). Excess N loading to surface waters—from wastewater discharges, overland runoff, groundwater seepage, and the marginal contributor of our focus, atmospheric deposition—is contributing to increased algal productivity and phytoplankton population shifts (including diatoms) throughout the United States (Saros et al. 2003, U.S. EPA 2008). The level of N inputs estimated to change a sensitive ecosystem component varies by environment, source of N, and even by size of the area of study (Gao et al. 2014). A critical load 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). Critical loads for eutrophication from atmospheric N deposition have been identified for a wide variety of individual ecosystems and biological indicators (U.S. EPA 2008, Pardo et al. 2011), at the species level as well as at the ecosystem level. While these critical loads are useful for indicating deposition levels at which harm to an ecosystem may begin to manifest, for example, for €m alpine lakes (Clow and Sueker 2000, Bergstro and Jansson 2006), they are relevant only to the biological indicator directly studied. Whether airborne N load entering a particular aquatic system as a pollutant proves to be enough to push the total N load to a point that it likely harms sensitive resources has yet to be specified for the full range of specific aquatic conditions. Eutrophic rivers, lakes, or estuaries may in turn affect aquatic vegetation, fish, and bird species and through these changes may directly impact humans who use and value aquatic ecosystems. In this paper, we explore the potential ramifications of changes in algal populations due to excess N loading in aquatic ecosystems, evaluating environmental impacts that a range of human beneficiaries care about. ❖ www.esajournals.org

RHODES ET AL.

2

July 2017



RHODES ET AL.

humans that depend on them, using the Final Ecosystem Goods and Services Classification System (FEGS-CS, Landers and Nahlik 2013). The FEGS-CS was developed as a tool to identify ecosystem classes and sub-classes, and the beneficiary class and sub-classes that interact with an ecosystem component, where beneficiaries are described as “the interests of an individual” (i.e., person, group, and/or firm) that drive active or passive consumption and/or appreciation of ecological endpoints, resulting in an impact (positive or negative) on their welfare. Thus, any one “FEGS” bridges a biophysical quality or feature

Fig. 1). These connections are documented in chains, each of which links a biological indicator that has associated critical loads for aquatic eutrophication to an ecological endpoint (Boyd and Krupnick 2009). These ecological endpoints are not “endpoints” in an ecological sense, but in the sense that humans interact with or react to them in some way, bringing them physically or conceptually into the realm of humans and of “value.” The ecological process by which a biological indicator is linked to an ecological endpoint is an EPF. These same chains then link final ecosystem goods and services (FEGS) through ecological endpoints to the

Fig. 1. A conceptual model of the STressor–Ecological Production function–final ecosystem Services Framework. The Stressor Module (red squares) consists of the chemical, environmental, and/or biological responses that are influenced by a stressor and lead to a change in the biological indicator. The SOSS score represents the scientific integrity of the relationships within the module (blue line). The Ecological Production Function (EPF) Module (purple squares) is the cascade of ecosystem effects due to the change in the biological indicator. The EPF can have zero-to-n additional steps, represented by the dotted lines connecting each component to the ecological endpoint. The yellow diamonds are the SOSE score for the relationship between two components. The orange circle represents the combination of all SOSE scores in the SOSEPF equation. The ecological endpoint feeds into the Final Ecosystem Goods and Services Classification System Module, which identifies a beneficiary class for each endpoint. The SOSC score represents the scientific integrity of the chain linking a change in biological indicator to a final ecosystem service, and is represented by the red line.


3

July 2017



to human well-being without need of further translation. We identify pathways of effect that begin with basic biochemistry of excess N loading, continue through internal ecosystem processes to ecological endpoints, and end in human beneficiaries who value ecosystem services. We expect this work contributes to the literature by highlighting how the ecosystem services perspective can move those conducting ecosystem services assessments further toward choosing specific relevant metrics for analysis, while also engaging narratives that can translate proven ecological effects to a context relevant to a non-scientific audience. Our approach further intends to draw attention to gaps in empirical ecological literature, gaps that constrain understanding of the magnitude of effects that excess N loading in U.S. surface waters can have on human well-being.

(Critical loads and total maximum daily loads [TMDLs] are similar in that they both determine a quantitative pollution loading below which harmful effects are expected not to occur for an aquatic system. However, critical loads serve as an assessment tool, while a TMDL is legally required for all waterbodies that fail to meet water quality criteria under Section 303(d) of the Clean Water Act.) Participant knowledge of the relatively small number of published critical loads for freshwater aquatic eutrophication facilitated this step. Once the critical loads and biological indicators were established, we brainstormed all of the environmental and trophic interactions that could result from exceeding the threshold for N as a biological indicator. Brainstorming at this stage was meant to consider interactions firmly established in the literature as well as to identify any hypothesized interactions based on expert opinion and the transfer of known relationships among ecosystems. The exchange between experts from different disciplines was crucial during this stage and often when an interaction was posited, another in the group would be able to point to evidence in the published literature for the speculated interaction. The next step was to construct EPF for ecological endpoints causally impacted at higher trophic levels by a critical change in our biological indicator, N (Fig. 1, orange circle). Input from the economists in the group was crucial at this stage in helping to determine which ecological processes and products were directly used or valued in some way by human beneficiaries and were thus true ecological endpoints (FEGS, within the FEGS-CS lexicon), from those that were merely intermediate ecological processes or products. Once an ecological endpoint had been identified (Fig. 1, green boxes), the group then proceeded to develop a list of beneficiaries of the same ecological endpoint before starting on the next chain (Fig. 1, blue boxes). The FEGS-CS Framework defines 52 separate beneficiaries in ten separate categories and includes agricultural and commercial extractors (such as livestock grazers and industrial processors) as well as subsistence, recreational, and nonuse consumers of FEGS (such as people that value the existence of nature). By considering that the identification of FEGS is an analytical step independent from the challenges of actually conducting environmental or valuation measurements,

METHODS Air Quality and Ecosystem Services Workshop Our author group is one of four cross-disciplinary subject groups that met at the Air Quality and Ecosystem Services (AQES) Workshop (February 2015) to finalize and put into practice the STEPS Framework, which links the ecological effects from N deposition in different types of environments to end users (Blett et al. 2016). The four subject groups each explored one mode of response to atmospheric N deposition, including aquatic acidification (O’Dea et al. 2017), aquatic eutrophication, terrestrial acidification (Irvine et al. 2017), and terrestrial eutrophication (Clark et al. 2017). As with other related papers in this volume, we developed links from primary effects of critical load exceedances to human beneficiaries, and we identified all of the possible chains linking biological indicators that have associated critical loads, through an ecosystem, to the goods and services upon which human beneficiaries depend. An overview of the STEPS Framework developed for the AQES workshop, and a summary of the effort, can be found in Bell et al. (2017).

Chain development Having selected atmospheric N deposition as the stressor (Fig. 1, red boxes), the group’s first step in developing chains was to identify relevant deposition critical loads for aquatic eutrophication. ❖ www.esajournals.org

RHODES ET AL.

4

July 2017



analysts are freed to consider a full spectrum of plausible FEGS combinations. The FEGS-CS is designed to facilitate nearly exhaustive (and nonduplicative) identification of potential FEGS for any legitimately feasible chain for ecological production of ecological endpoints with which humans concern themselves. This process continued until all relevant EPFs were associated with potential beneficiaries of the ecological endpoint.

populations. While the assignment of any SOS score was subject to the scrutiny of the entire group, often it was determined by one participant with specialized knowledge of the scientific literature pertaining to a specific interaction. Grounding thus in the SOSE and in the number of components in an EPF chain (EPF Length) that connect the biological indicator to this chain’s ecological endpoint, Eq. 1 is as follows: P 1 SOSE SOSEPF ¼ 1 . M EPF Length EPF Length (1)

Strength-of-science

An important contribution of the “chain” approach is the assignment of a “strength-ofscience” (SOS) score for each chain that relates how well the scientific literature supports the existence of a relationship from a change in a stressor to a change in a final ecosystem service, following the STEPS Framework. If an ecological endpoint is utilized by multiple beneficiaries, each has a different chain, so that the more chains associated with a biological indicator, the wider the impacts of the affected resource. Calculating the SOS for a chain score (SOSC; Fig. 1, red circle) is a multistep process that first involves calculating an SOS score for the environmental production function associated with the ecological endpoint that helps to define a chain (SOSEPF; Fig. 1, orange circle). Considering that longer EPFs involve more components (Fig. 1, purple boxes) that may attenuate the causality that can be attributed to a single line of relationships, a calculation should lower the SOSEPF as the string of EPF components grows. The SOS Effect (SOSE) characterizes the breadth and certainty of scientific evidence supporting one link in the EPF between components of an EPF (Fig. 1, yellow diamonds). SOSE is a qualitative ranking with high (=1) reflecting multiple strong lines of published scientific evidence supporting the relationship between two EPF components, with medium (=0.67) reflecting limited, inconsistent, or conflicting scientific evidence, and with low (=0.33) reflecting observations by experts or unpublished data that support the effect of one EPF component on the next in a hypothesized chain, but without a body of published scientific evidence. Assigning an SOSE score for each link of the EPF allowed our aquatic eutrophication group to assess the weight of scientific evidence for a given effect—for example, that a change in diatom community composition or abundance would affect macroinvertebrate ❖ www.esajournals.org

RHODES ET AL.

Here, M is a constant representing a chain length that is long enough to separate the biological indicator from the ecological endpoint beyond the ability to reasonably infer a direct line of causality. In our calculations, M is set to 8, as our longest chain was six components. With SOSE scores averaged and weighted by chainlength factors, the SOSEPF is thus by design and by inputs bound between 0 and 1 and tends to distribute as familiar fractions. The SOS Stressor (SOSS) identifies the quality of the science linking the change in a biological indicator to the exceedance of its critical load (Fig. 1, blue circle). The SOSS is represented by a high (1, scientific literature confirms the relationship), medium (0.67, scientific literature is reliable for a critical load derived through modeling), or low (0.33, based on expert judgment measured in a similar habitat or region to the one being evaluated) score, as the SOSE is. See Bell et al. (2017) for detail on any SOS score definition or calculation referred to here. The SOS of a full chain (SOSC), from a change in a stressor to effects on beneficiaries of specific final ecosystem services, represents the confidence in scientific data across the total chain (Eq. 2), with numeric indicators similar in form to those of SOSEPF scores. SOSC ¼

SOSS þ ðSOSEPF EPF LengthÞ . EPF Length þ 1

(2)

The full weight of the SOSS is averaged here with the value of each SOSEPF and diminished by the chain length, so SOSC scores still derive first from SOSE scores. The SOSC score is intended to be the main metric evaluating the relationships depicted through the STEPS Framework. 5

July 2017



in biological indicator “Altered abundance and species richness of phytoplankton,” and branched toward 13 ecological endpoints, four in alpine and subalpine lake systems, six in freshwater non-alpine lake systems, and three in estuarine systems (Table 1). The ecological endpoints are linked to 18 of the 19 beneficiary classes initially identified (Waste Water Treatment Plant Operators was removed). The SOSEPF scores for these 13 chains range from 0.45 to 0.86, with an average score of 0.61. The SOSC scores range from 0.58 to 0.93, with an average score of 0.78 (within a potential range from 0.33 to 1). We give examples in this section that explore the potential effects on ecosystem services for these three aquatic systems, ranging from least to most complex. Complexity varies in terms of the links in the chain and the ecological processes that need to be understood. These chains also vary on a spectrum defined by the level of ecological production and the economic inputs often necessary for beneficiaries’ enjoyment, including the degree to which impaired ecological production has an effect on economic production, how well that effect is understood, and ultimately how these changes affect human welfare. We begin with the least complex type, which has few nitrogen contributors other than airborne deposition: alpine and subalpine lakes. We proceed then to a more complex type that experiences other N sources, likely higher in magnitude than airborne deposition: freshwater lakes at lower altitudes. Estuarine systems round out the analysis—challenging due to the range of species and of variables that affect these aquatic living webs, where the great bulk of N often comes from agricultural and urban runoff.

The weakest link of the chain (SOSWL) is calculated by taking the lowest score from the SOSS and SOSE scores in a chain. This is based on the concept that a chain is only as strong as its weakest link (Bell et al. 2017). Compared together, the various SOS scores thus defined may be used to contrast chains where many strong links indicate a robust effect on ecosystem services with chains where a weak link or two may indicate the need for additional study, perhaps eliciting research that explores weaker links.

RESULTS During the AQES workshop, we first identified four different aquatic environments offered by the environmental sub-classes of the FEGS-CS (lakes and ponds, streams and rivers, estuaries and nearshore waters, and wetlands) that had published data on the impacts of N enrichment on a biological indicator. We then established a series of general chains based on aquatic trophic groups and duplicated the lists across the different aquatic environments, removing irrelevant chains for particular aquatic environments. In this way, we avoided a great narrowing of scientific focus on the strength of each link for each species in each environment. As most of the hypothesized chains were relevant for each environment, we totaled 589 individual chains potentially impacting 19 unique beneficiary types (Experiencers and Viewers, Resource-Dependent Businesses, Water Subsisters, Waders Swimmers and Divers, Spiritual and Ceremonial, Artists, Anglers, Boaters, Researchers, Educators, People Who Care (Option), People Who Care (Existence/Bequest), Residential Property Owners, Hunters, Waste Water Treatment Plant Operators, Drinking Water Treatment Plant Operators, Food Subsisters, Food Extractors, Aquaculturists). Thus, we built the case that even for proven eutrophic effects that may seem distant to many human interests, a wide array of human users may be affected when deposition critical loads are exceeded. With further analysis, we identified 13 EPFs within the alpine and subalpine lake systems, the freshwater non-alpine lake systems, and the estuarine systems that led to high-value ecological endpoints and had strong scientific backing. These represent 154 of our original 589 chains (Data S1). The chains all started with the change ❖ www.esajournals.org

RHODES ET AL.

Alpine and subalpine lakes Alpine and subalpine lakes in the western United States are especially sensitive to eutrophication from N deposition because they are highly oligotrophic, occur in basins with little soil and vegetation, and are frequently N-limited (Clow €m and Jansson 2006). and Sueker 2000, Bergstro There has been extensive research to understand the effects of increased N deposition on these sensitive ecosystems as they are habitat for aquatic species, popular recreational destinations, headwaters for primary water supplies throughout the western United States (U.S.), and predominantly 6

July 2017



RHODES ET AL.

Table 1. Numeric summary of SOS scores for chains based on ecological endpoints. Ecological endpoints Alpine and subalpine lakes Change in desired diatom species abundance and composition Changes in desired macroinvertebrate abundance and composition Decrease in water clarity Decreased probability and abundance of terrestrial consumers (e.g., birds) Freshwater non-alpine lakes Decrease in water clarity Decreased probability of occurrence and abundance of algae and bacteria Decreased probability of occurrence and abundance of phytoplankton and zooplankton Increased probability of occurrence and abundance of small fish Decreased probability of occurrence and abundance of game fish Decreased probability of occurrence and abundance of SAV Estuaries Decrease in SAV abundance Decrease in SAV-dependent fish and waterfowl Decrease in water clarity

Beneficiary sub-classes

Chain length

SOSEPF

SOSC

SOSWL

9 9 14 10

1 3 3 4

0 0.56 0.83 0.45

1 0.71 0.89 0.58

1 0.67 1 0.33

14 8

3 1

0.83 0

0.89 1

1 1

8

4

0.80

0.68

0.33

12 12 12

5 5 5

0.69 0.58 0.69

0.62 0.54 0.62

0.33 0.33 0.33

16 14 16

3 5 2

0.83 1 0.86

0.89 0.75 0.93

1 0.33 1

Note: SOS, strength-of-science; SOSC, SOS for a chain score; SOSEPF, SOS score for the environmental production function; SOSWL, SOS score for the weakest link of the chain; SAV, submerged aquatic vegetation.

environments (Saros et al. 2005). Algal growth thresholds are crossed at lake N concentrations ranging from 0.18 to 3.1 lmol/L in the Sierra Nevada and Rocky Mountains (Arnett et al. 2012, Nanus et al. 2012, Heard 2013), which are linked to critical loads for diatoms in these ecosystems ranging from 1.0 to 2.5 kg Nha1yr1 (wet deposition; Baron et al. 2011, Pardo et al. 2011, Saros et al. 2011). As algal abundance increases, the lake water clarity, typically monitored as Secchi depth, decreases (Carlson 1977, Goldman 1988). In Lake Tahoe, transparency has decreased over the last 50 yr and is partially attributed to atmospheric N deposition influencing an increase in phytoplankton growth (Goldman 1988, Jassby et al. 1999, Swift et al. 2006). It is likely that similar trends are occurring in N-affected mountain lakes throughout the west; however, long-term water clarity data are lacking at larger spatial scales. We identified 15 beneficiaries for the water clarity chain, and all are applicable to alpine and subalpine lake ecosystems (Fig. 2). The broad scope of beneficiaries is not surprising given that water clarity is one of the primary, and possibly only, aspects of water quality that beneficiaries will have the opportunity to directly observe and use

located on protected federal lands such as national parks and wilderness. Since the primary source of anthropogenic N to these ecosystems is atmospheric deposition, these aquatic systems provide a unique opportunity to study air pollution effects and critical loads without the complications of multiple nutrient sources. We highlight two chains important to alpine and subalpine lakes. Both follow shifts in diatom communities, with the first focusing on changes in lake water clarity, and the second on higher trophic interactions (Fig. 2). Nitrogen is deposited in remote mountain watersheds via wet and dry deposition, where it is leached into receiving waters primarily as nitrate (Baron et al. 2000, Mast et al. 2014). There is extensive research demonstrating that increases in atmospheric deposition and subsequent increases in lake water N concentrations are causing shifts in diatom species composition and increased algal growth in mountain lakes throughout the western United States (Saros et al. 2003, 2011, Wolfe et al. 2003, Nydick et al. 2004, Heard 2013). Sediment records show a notable shift post-1950 from oligotrophic species to more mesotrophic species, predominately Asterionella formosa and Fragilaria crotensis, which thrive in higher N ❖ www.esajournals.org

7

July 2017



RHODES ET AL.

Fig. 2. Flow chart summarizing example chains of effects with demonstrations from literature for alpine and subalpine lakes. Residential Property Owners in the second Types of Beneficiaries (blue) box applies only to the terrestrial consumers’ ecological endpoint.

2001, Lafrancois et al. 2004). Less grazing pressure from zooplankton can cause increased phytoplankton growth, further exacerbating eutrophication issues. Eutrophication effects from N may also be further enhanced by acidification (Lafrancois et al. 2004). The limited research on the effects of N deposition on macroinvertebrates in alpine and subalpine lakes is inconclusive (Lafrancois et al. 2003), but shifts in pelagic phytoplankton abundance and species composition have the potential to affect macroinvertebrate communities in several ways. Increased algal growth in the water column can shade benthic primary producers and thus shift primary production from the benthic to pelagic (Vander Zanden et al. 2003). This shift has the potential to affect benthic macroinvertebrates that depend on algal production in

to draw conclusions about the quality of a waterbody and their willingness to use it for recreation (Quick and Johansson 1992, Smith et al. 1995). The second chain focuses on cascading changes in the food web—from diatoms to bird species—linking aquatic and terrestrial ecosystems (see Fig. 2). As N concentrations increase, the lake shifts from N limitation to phosphorus (P) limitation, favoring phytoplankton taxa that are more adapted to P-limited environments. Shifts from smaller chrysophytes and diatoms to larger chlorophytes are observed (Lafrancois et al. 2004, Nydick et al. 2004), and seston C:P and N:P ratios increase (Elser et al. 2001, 2009a). The N-altered phytoplankton population is generally less palatable to grazers, and decreases in zooplankton, especially Daphnia spp., have been observed in nutrient experiments (Elser et al. ❖ www.esajournals.org

8

July 2017



Freshwater non-alpine lakes

lake bottoms and littoral zones for food (e.g., grazers). In mountain lakes, the food source for predatory macroinvertebrates is zooplankton and other benthic macroinvertebrates (Finlay and Vredenburg 2007), both of which have the potential to be affected by cascading changes in the food web. Another effect of eutrophication is lowered dissolved oxygen levels, which can also affect macroinvertebrates. Ephemeroptera (i.e., mayflies) are especially sensitive and are often used as indictors of early eutrophication (Bauernfeind and Moog 2000). Aquatic macroinvertebrates are important food sources for higher-level terrestrial consumers including amphibians, bats, spiders, and birds (Nakano and Murakami 2001, Finlay and Vredenburg 2007). These species establish an important link between the aquatic and terrestrial ecosystems, a link that is especially important in nutrient-poor terrestrial alpine ecosystems that rely more heavily on the flow of energy and nutrients from aquatic habitats (Finlay and Vredenburg 2007, Epanchin et al. 2010). One example of the aquatic–terrestrial connection that has the potential to be disrupted is the connection between mayflies and Gray-Crowned Rosy-Finch (Leucosticte tephrocotis dawsoni). Mayflies are an important food source (up to 38%) for the Gray-Crowned Rosy-Finch during the critical breeding season (Epanchin 2009). Epanchin et al. (2010) observed significant declines in Gray-Crowned RosyFinches at lakes where mayfly populations were greatly reduced (by non-native fish). We identified 10 beneficiary types that would be affected by cascading food web effects leading to a change in terrestrial consumers, such as birds (Fig. 2). These include groups who frequent the lakes based on the presence of wildlife and homeowners whose property values increase with high-functioning aquatic ecosystems nearby. One complication when considering alpine and subalpine lakes is that many mountain lakes now contain non-native fish due to stocking practices in the 20th century (Knapp et al. 2001). We have opted to focus on the food web for fishless lakes since this is the natural state in many high-elevation systems. However, it should be recognized that non-native fish have strong topdown effects on the aquatic food web and N effects in a fish stocked lake may differ (Knapp et al. 2001, Finlay and Vredenburg 2007). ❖ www.esajournals.org

RHODES ET AL.

Decreased water clarity of non-alpine lakes.— Oligotrophic to eutrophic lakes (non-alpine) are also susceptible to eutrophication through a combination of N and P, coming from joint sources of runoff, direct discharges, and atmospheric deposition. Because these lakes are typically co-limited by N and P, enhanced N can lead to increases in productivity, but only until a threshold where P limitation restricts growth, at which €m point N will accumulate in the lake (Bergstro €m and Jansson 2006, Elser et al. 2005, Bergstro et al. 2007, 2009b). For these lake systems, we are highlighting the critical load changing the production of phytoplankton communities and linking it to a change in lake water clarity and a change in fish populations. As with alpine lakes, it is known that increased N inputs can lead to an increase in the productivity of phytoplankton (i.e., algal and bacterial abundance). This is a particular concern in nutrient-poor lakes where the primary productivity is often N-limited and N from atmospheric deposition and runoff can change the N-limited status (Elser et al. 1990, Axler et al. 1994, Jassby et al. 1995). In the absence of N loading, submerged aquatic vegetation (SAV) and benthic algae dominate the productivity, but as nutrient levels increase, macrophytes and phytoplankton body size and abundance also increase. Submerged aquatic vegetation becomes taller and their biomass increases along with phytoplankton, causing benthic production to decline and a shift to primary production in the pelagic zone of the lake (Moss et al. 1994, Vadeboncoeur et al. 2003). This shift in the location of phytoplankton productivity to the upper section of the lakes causes a decrease in light attenuation in the upper water column, lowing water clarity. Chlorophyll-a concentration is used as an indicator of water clarity via phytoplankton biomass (Paerl and Piehler 2008). However, the amount of nutrient loading from atmospheric deposition or runoff that can cause a shift in water clarity is highly variable and depends on properties that can vary lake to lake, making transfer of ecological responses among lakes difficult. Often shallow small mesotrophic to eutrophic lakes with slow recharge are most strongly affected by N loading (Moss et al. 1994), and humic lakes respond differently compared to clear-water lakes (Faithfull et al. 2015). We 9

July 2017



RHODES ET AL.

Fig. 3. Flow chart summarizing example chains of effects with demonstrations from literature for non-alpine oligotrophic and eutrophic lakes.

benthic to pelagic (Vadeboncoeur et al. 2001), along with an increase in algal production as phytoplankton community composition shifts toward one of blue-green algae and away from a diatom-rich community (Jeppesen et al. 2000). Nutrient enrichment can impact the number, biomass, and size of fish in the lake as their grazing patterns shift from piscivorous toward planktivorous/benthivorous fish species abundance. This shift can often lead to fish becoming smaller (Jeppesen et al. 2000, Søndergaard et al. 2005) as a consequence of increased predation on zooplankton and grazer macroinvertebrates (Jones and Sayer 2003, Jeppesen et al. 2005). This reduced grazing pressure leads to further reduction in grazing pressure on epiphytes and phytoplankton, resulting in less light attenuation into the lake impacting submerged plants and increasing biomass of benthic-feeding fish (Lammens 1999). The fish community is further impacted by the reduction or loss of SAV,

linked decreased water clarity to 14 unique beneficiary classes (see Fig. 3). As non-alpine lakes are usually closer to population centers relative to alpine lakes, the potential amount of use by each of these groups is higher. Trophic changes from primary production to fish populations.—The second chain for non-alpine lakes, ranging from oligotrophic to eutrophic lakes, focuses on cascading trophic changes from primary production to fish populations. Because these lakes are almost always co-limited by N and P, enhanced N loading only leads to increases in productivity where P limitation does €m et al. 2005, not restrict productivity (Bergstro €m and Jansson 2006, Elser et al. 2007, Bergstro 2009b). N-driven increase in the productivity (bottom-up control) can cause increased predation (top-down control) that alters the trophic cascade, leading to shifts in zooplankton to fish populations. Eutrophication in lakes often results in changes in the location of production from ❖ www.esajournals.org

10

July 2017



particularly, in shallow lakes or the lateral zone in deeper lakes, as phytoplankton dominate in response to nutrient loading. Submerged aquatic vegetation provides shelter and forage for piscivorous fish taxa and in its absence can lead to their decline, as planktivorous/benthivorous fish then dominate. This further increases predation on zooplankton and reinforces the dominance of phytoplankton (Scheffer et al. 1993). While the set of changes just described has been well studied, there is marked variability in the biological response to increased nutrients across lakes (Van de Bund et al. 2004). For example, a nutrient addition of 1 mg/L P and 10 mg/L N resulted in increased total phytoplankton, algal biomass, and SAV, but these responses varied among studied lakes. The literature does not lay out precise levels of nutrient loading that drive ecological changes for each of a variety of aquatic conditions. Nutrient loading that causes changes described here depends on many lake characteristics, including lake depth (Moss et al. 1994), lake productivity (Chase 2003), and most importantly, P concentration as it relates to N limitation in the lake. We link changes in fish populations and SAV populations to separate lists of 12 potential beneficiaries. Several papers have investigated the relationship between fish populations and recreational fishing demand (Massey et al. 2006, Zhang 2011). As with many FEGS, the strength of the relationship between fish population and angler use is likely to be highly species- and place-specific, and thus could vary significantly, which means that relevant chains would need to be downscaled to a species level for use at a local scale or even a specific site.

non-alpine lakes receive N in complex ways, the mix of sources is often more complex for estuarine systems, as all major N sources (atmospheric and not) likely contribute to the total load. To understand the impact of atmospheric deposition on an estuary, one must first determine the total load to an estuary from all sources including direct deposition to the surface of the estuary as well as deposition to and consequential runoff from the watershed. Atmospheric deposition can then be understood as a percentage of the total load and the total N load can be used to examine impacts to the estuary. The Chesapeake Bay is a well-studied example of the impacts of N loading to an estuary. In 2008, total estimated N loading from the watershed to the bay was 284 million pounds, about a third of which originated from atmospheric deposition (U.S. EPA 2009). Nitrogen inputs increase water column N that in turn increases primary productivity and affects water quality. Primary productivity (phytoplankton) in estuarine systems is generally limited by N availability (although, phosphorus, and silicon can also be important depending on location within the estuary), hydrology, and seasonality (Elser et al. 2007, U.S. EPA 2008). While the link between excess nitrogen in estuaries and phytoplankton growth is well studied, the exact levels of N in the water column and the corresponding growth are unique to each estuary and each location within an estuary, limiting the transferability of data among systems. Estuarine productivity is often closely related to the health and vitality of SAV, which provides habitat for a variety of aquatic organisms, serves as nursery grounds for invertebrates and fish, absorbs excess nutrients, and traps sediments (Handley et al. 2007, U.S. EPA 2008). Clear water with sufficient light penetration is essential for SAV growth (Dennison et al. 1993, Kemp et al. 2004). Some opportunistic macroalgae grow very rapidly during times of N availability (Abreu et al. 2011), which can lead to increases in phytoplankton biomass and macroalgae abundance and type that can block sunlight and smother or outcompete SAV (Kennison et al. 2011). Submerged aquatic vegetation loss can cause loss of habitat and food sources affecting ecosystem structure and function, with changes in aquatic fauna and waterfowl dependent on SAV. Fish and blue crabs utilize the SAV for shelter.

Estuarine systems Estuaries are ecologically important areas where freshwater inflows meet saltwater, creating a salinity gradient. Total N loading to an estuary presents a complex problem because of contributions from multiple sources, including atmospheric deposition, leachate from septic tanks, discharges from sewage treatment plants and industries, and runoff from farms, urban, and natural areas. The atmospheric deposition portion may contain emissions from on- and off-road vehicles, ships and watercraft, industries, power plants, animal feeding operations, and fertilized fields (U.S. EPA 2009). Although many freshwater ❖ www.esajournals.org

RHODES ET AL.

11

July 2017



RHODES ET AL.

provide water clarity and an aesthetic beauty to an area provide a sense of serenity and life beneath the water. When they are lost, mud flats take the place of green strands of grass waving to onlookers, affecting the enjoyment of boaters, artists, photographers, and shoreline property owners. Changes in water quality, aquatic fauna, and waterfowl represent FEGS that affect 16 of the 18 beneficiary groups we identify across all chains (see Fig. 4). The fact that estuaries are in the transition zone of aquatic and marine systems leads to high plant and animal diversity and the potential for wide-ranging species-level effects if trophic levels are downscaled to a local site.

While migratory waterfowl depend on SAV and associated aquatic fauna for food, SAV also stabilizes sediments, helping to provide clear water, and to absorb wave energy, which can reduce coastal erosion (http://www.chesapeakebay.net/ issues). The links between phytoplankton growth, macroalgae growth, and decreases in SAV abundance can vary depending on location and conditions. Submerged aquatic vegetation loss and population decreases in SAV-dependent fauna strongly correlate, but specific relationships are dependent on many local factors. As water quality deteriorates, SAV declines and the habitat changes, ecosystem production and function is altered, creating stresses on aquatic fauna that depend on the SAV. These changes create a ripple effect through the ecosystem, catalyzing changes that impact human uses of the ecological endpoints. Hunters of waterfowl see diminished flocks. Fishers see desired fish and shellfish relocate or disappear. Seagrasses that

DISCUSSION The presence of both ecologists and economists in our group presented welcome and productive challenges, as a great deal of discussion

Fig. 4. Flow chart summarizing example chains of effects with demonstrations from literature for estuarine systems.


12

July 2017



and interaction was required for each person to understand how the other group was using a similar term differently, while both ecologists and economists struggled to accommodate how the different constraints of biological and economic systems might be rectified within a single model. As vocabulary and basic knowledge common to each field were shared across the divide, the group learning curve bridged understanding, allowing group members to proceed confidently in the task of tracing ecological effects to potential impacts on human welfare. The proliferation of potential FEGS in our chains is driven in part by the natural dynamics of N deposition in aquatic systems, which varies with catchment shape, waterbody size, and water flow variables. While a certain response to a change in a biological indicator is hard to confidently attribute for any particular body of water, laying out the full range of potential FEGS from aquatic eutrophication opens the available ecological science to a wider understanding of the potential effects on humans.

assessments based on a very narrow set of ecosystem services values are useful, but limited in their ability to describe the entire suite of impacts that a stressor such as air-pollutioninduced aquatic eutrophication may have. Completing these chains and assessing the SOS are important steps, but the ultimate goal of this inquiry is to provide decision makers with the best possible information describing how actions to limit atmospheric N deposition loads will ultimately affect the environment and human welfare. While not all valuation measures need to be monetized, a common practical application of such knowledge is benefit–cost analysis, which does require estimating monetary values for benefits provided by ecosystem services. (A number of valuation studies include water clarity, e.g., as a direct driver of values. See Gibbs et al. 2002 for an example of a hedonic property value study.) The U.S. EPA has conducted a stated-preference survey specifically for valuing improvements from reduced nutrient loadings to the Chesapeake Bay (Moore et al. 2015). Although studies such as Moore et al. may greatly aid decision makers in assessing the value of these waters to people, conducting high-quality, original stated-preference research is both timeconsuming and costly. Using a beneficiary approach to the identification of ecosystem services preceding the quantification and valuation components of a full ecosystem services assessment may reduce study time and cost by tightening definitions that in turn may focus the choice of metrics. One method that the U.S. EPA has found useful for estimating monetary values is benefits transfer—transferring benefits of a policy studied at one location, perhaps with adjustments, to a similar policy applied to another location. A recent benefit-transfer study of national regulations to improve surface water quality conducted a meta-analysis of the environmental economics literature, including 51 studies with 140 distinct valuation estimates (see Chapter 4 and Appendix H of U.S. EPA, 2015), where seven of these studies specifically value lakes (New York, Iowa, Minnesota, South Dakota, and Wisconsin), and an additional six value lakes and other waterbodies (Colorado, Ohio, Wisconsin, Florida, and Montana). However, because none of the studies included in the meta-analysis specifically address

Strength-of-science scores Strength-of-science chain scores are useful for comparing chains to assess confidence in the described effect and as a means to highlight where further research may effectively contribute. While the SOSE scores of many of the ecological changes reported here are high, this effort focused on areas where these scores likely would be high, as a means of demonstrating the STEPS Framework in an aquatic eutrophication context. These chains therefore should be looked at as a potential response to nutrient loading, but specific site characteristics may limit the transferability of SOS scores beyond sites where ecological responses were measured. Relatively lower chain scores here draw attention to chains with steps that require broader scientific support to be accepted as established when working at a regional or national scale.

Considerations for applying this research approach to valuation work The beneficiary approach to ecosystem services analysis used here characterizes ecosystem services in a way that they can be comprehensively accounted for in subsequent valuation processes. Studies that conduct valuation ❖ www.esajournals.org

RHODES ET AL.

13

July 2017



is most impacted by impaired ecological production, and to focus on developing models and measurements that support detailed analysis. The balance between natural and social science specialists in our approach was necessary, at least until the ecosystem services literature thickens with more clearly defined biophysical relationship chains, and more metrics and indicators relating ecological endpoints to human use and appreciation. It remains important moving forward that clear definitions of overlapping economic and biological terminology be established.

western alpine and subalpine lakes, there are no estimates or parameters for significant tracts in the west. The transfer of benefits from a more general class of waters to a specific class may add uncertainty to a valuation exercise when there are no parametric bounds derived from meta-analysis, even when using completed ecological chains such as those from our work here.

CONCLUSIONS We used the STEPS Framework to identify 154 chains, each linking an exceedance of a critical load of atmospheric N deposition within aquatic systems to FEGS and their beneficiaries. We determined that all chains go through changes in phytoplankton communities as an indicator of exceedance of the critical load as they affect aquatic ecosystems in complex ways. Whereas our original work at the AQES workshop developed 589 chains with the intention of being comprehensive, we narrowed our focus here to 154 chains depicting ecosystem relationships supported by peer-reviewed literature, within three FEGS-CS Environment sub-classes, in order to demonstrate the likely variety in strengths of science that these well-supported ecosystem relationships or categories of chains would exhibit. This approach highlights for non-ecologists and for policy makers the potential effects on human welfare that exceedance of nitrogen critical loads may have. By demonstrating that chains from these three aquatic environments show a range of strengths of science, we hope to stimulate new data collection and research where the SOS is weakest, and where these effects are likely to be important but have not already been heavily studied, as the Chesapeake Bay has been. We also note that a key aspect of using the FEGS-CS to classify ecological endpoints is that different beneficiary categories may each value characteristics of ecological endpoints differently, highlighting the variety of impacts to human well-being. A greater understanding of the relationship between ecological and economic production for these beneficiaries will also be important for understanding the magnitude of changes in human well-being. Understanding this relationship is necessary to identify those beneficiaries whose economic production or well-being ❖ www.esajournals.org

RHODES ET AL.

ACKNOWLEDGMENTS This work resulted from a workshop supported by a National Science Foundation, Directorate for Biological Sciences Award: NSF-DEB-1547041. We thank Irina Irvine and the staff at the Santa Monica Mountains National Recreation Area in Thousand Oaks, California, for their hospitality in supporting our workshop. The views presented here are those of the authors and do not represent official views or policy of the U.S. Environmental Protection Agency (EPA) or any other U.S. federal agency. As an ORISE post-doctoral fellow, the contributions by Charles Rhodes are supported by an interagency agreement between U.S. EPA and U.S. DOE.

LITERATURE CITED Abreu, M. H., R. Pereira, A. H. Buschmann, I. SousaPinto, and C. Yarish. 2011. N uptake responses of Gracilaria vermiculophylla (Ohmi) Papenfuss under combined and single addition of nitrate and ammonium. Journal of Experimental Marine Biology and Ecology 407:190–199. Arnett, H., J. Saros, and M. Alisa Mast. 2012. A caveat regarding diatom-inferred nitrogen concentrations in oligotrophic lakes. Journal of Paleolimnology 47:277–291. Axler, R. P., C. Rose, and C. A. Tikkanen. 1994. Phytoplankton nutrient deficiency as related to atmospheric nitrogen deposition in northern Minnesota acid-sensitive lakes. Canadian Journal of Fisheries and Aquatic Sciences 51:1281–1296. Baron, J. S., C. T. Driscoll, J. L. Stoddard, and E. E. Richer. 2011. Empirical critical loads of atmospheric nitrogen deposition for nutrient enrichment and acidification of sensitive US lakes. BioScience 61:602–613. Baron, J. S., H. M. Rueth, A. M. Wolfe, K. R. Nydick, E. J. Allstott, J. T. Minear, and B. Moraska. 2000.

14

July 2017



as barometers of Chesapeake Bay health. BioScience 43:86–94. €m, Elser, J. J., T. Andersen, J. Baron, A. Bergstro M. Jansson, M. Kyle, K. Nydick, L. Steger, and D. Hessen. 2009a. Shifts in lake N: P stoichiometry and nutrient limitation driven by atmospheric nitrogen deposition. Science 326:835–837. Elser, J. J., M. E. S. Bracken, E. E. Cleland, D. S. Gruner, W. S. Harpole, I. I. H. Hillebrand, J. T. Ngai, E. W. Seabloom, J. B. Shurin, and J. E. Smith. 2007. Global analysis of nitrogen and phosphorus limitation of primary producers in freshwater, marine, and terrestrial ecosystems. Ecology Letters 10:1135–1142. Elser, J. J., K. Hayakawa, and J. Urabe. 2001. Nutrient limitation reduces food quality for zooplankton: Daphnia response to seston phosphorus enrichment. Ecology 82:898–903. Elser, J. J., M. Kyle, L. Steger, K. R. Nydick, and J. S. Baron. 2009b. Nutrient availability and phytoplankton nutrient limitation across a gradient of atmospheric nitrogen deposition. Ecology 90:3062– 3073. Elser, J. J., E. R. Marzole, and C. R. Goldman. 1990. Phosphorus and nitrogen limitation of phytoplankton growth in freshwaters of North America: a review and critique of experimental enrichments. Canadian Journal of Fisheries and Aquatic Sciences 47:1468–1477. Epanchin, P. N. 2009. Indirect effects of nonnative trout on an alpine-nesting passerine bird via depletion of an aquatic insect subsidy. Dissertation. University of California, Davis, California, USA. Epanchin, P. N., A. K. Roland, and P. L. Sharon. 2010. Nonnative trout impact an alpine-nesting bird by altering aquatic-insect subsidies. Ecology 91:2406– 2415. Faithfull, C. L., P. Mathisen, A. Wenzel, A. K. €m, and T. Vrede. 2015. Food web efficiency Bergstro differs between humic and clear water lake communities in response to nutrients and light. Oecologia 177:823–835. Finlay, J. C., and V. T. Vredenburg. 2007. Introduced trout sever trophic connections in watersheds: consequences for a declining amphibian. Ecology 88:2187–2198. Gao, W., R. W. Howarth, B. Hong, D. P. Swaney, and H. C. Guo. 2014. Estimating net anthropogenic nitrogen inputs (NANI) in the Lake Dianchi basin of China. Biogeosciences 11:4577–4586. Gibbs, J. P., J. M. Halstead, K. J. Boyle, and J. C. Huang. 2002. An hedonic analysis of the effects of lake water clarity on New Hampshire lakefront properties. Agricultural and Resource Economics Review 31:39–46.

Ecosystem responses to nitrogen deposition in the Colorado Front Range. Ecosystems 3:352–368. Bauernfeind, E., and O. Moog. 2000. Mayflies (Insecta: Ephemeroptera) and the assessment of ecological integrity: a methodological approach. Hydrobiologia 422:71–83. Bell, M. D., J. Phelan, T. F. Blett, D. H. Landers, A. M. Nahlik, G. Van Houtven, C. Davis, C. M. Clark, and J. Hewitt. 2017. A framework to quantify the strength of ecological links between an environmental stressor and final ecosystem services. Ecosphere 8:e01806. https://doi.org/10.1002/ecs2.1806 €m, A. K., P. Blomqvist, and M. Jansson. 2005. Bergstro Effects of atmospheric nitrogen deposition on nutrient limitation and phytoplankton biomass in unproductive Swedish lakes. Limnology and Oceanography 50:987–994. €m, A. K., and M. Jansson. 2006. Atmospheric Bergstro nitrogen deposition has caused nitrogen enrichment and eutrophication of lakes in the northern hemisphere. Global Change Biology 12:635–643. Blett, T. F., M. D. Bell, C. M. Clark, D. Bingham, J. Phelan, A. Nahlik, D. Landers, C. Davis, I. Irvine, and A. Heard. 2016. Air Quality and Ecosystem Services Workshop Report: Santa Monica Mountains National Recreation Area, Thousand Oaks, CA— February 24–26, 2015. National Park Service, Fort Collins, Colorado, USA. Boyd, J., and S. Banzhaf. 2007. What are ecosystem services? The need for standardized environmental accounting units. Ecological Economics 63:616–626. Boyd, J., and A. Krupnick. 2009. The definition and choice of environmental commodities for nonmarket valuation. RFF DB 09-35. Resources for the Future, Washington, D.C., USA. Carlson, R. E. 1977. A trophic state index for lakes. Limnology and Oceanography 22:361–369. Chase, J. 2003. Strong and weak trophic cascades along a productivity gradient. Oikos 101:187–195. Clark, C. M., M. D. Bell, J. Boyd, J. Compton, E. Davidson, C. Davis, M. E. Fenn, L. Geiser, L. Jones, and T. F. Blett. 2017. Nitrogen-induced terrestrial eutrophication: cascading effects and impacts on ecosystem services. Ecosphere, https://doi.org/ 10.1002/ecs2.1877, In press. Clow, D. W., and J. K. Sueker. 2000. Relations between basin characteristics and stream water chemistry in alpine/subalpine basins in Rocky Mountain National Park, Colorado. Water Resources Research 36:49–61. Dennison, W. C., R. J. Orth, K. A. Moore, J. C. Steven€m, and R. son, V. Carter, S. Kollar, P. Bergstro Batiuk. 1993. Assessing water quality with submersed aquatic vegetation. Habitat requirements


RHODES ET AL.

15

July 2017



and unintended consequences. Ecosystems 4: 275–278. Lafrancois, B. M., D. M. Carlisle, K. R. Nydick, B. M. Johnson, and J. S. Baron. 2003. Environmental characteristics and benthic invertebrate assemblages in Colorado mountain lakes. Western North American Naturalist 63:137–154. Lafrancois, B. M., K. R. Nydick, B. M. Johnson, and J. S. Baron. 2004. Cumulative effects of nutrients and pH on the plankton of two mountain lakes. Canadian Journal of Fisheries and Aquatic Sciences 61:1153–1165. Lammens, E. 1999. The central role of fish in lake restoration and management. Hydrobiologia 395/396:191–198. Landers, D. H., and A. M. Nahlik. 2013. Final Ecosystem Goods and Services Classification System (FEGS-CS). EPA/600/R-13/ORD-004914. Office of Research and Development, U.S. Environmental Protection Agency, Washington, D.C., USA. Massey, D. M., S. C. Newbold, and B. Gentner. 2006. Valuing water quality changes using a bioeconomic model of a coastal recreational fishery. Journal of Environmental Economics and Management 52: 482–500. Mast, M. A., D. W. Clow, J. S. Baron, and G. A. Wetherbee. 2014. Links between N deposition and nitrate export from a high-elevation watershed in the Colorado Front Range. Environmental Science & Technology 48:14258–14265. Moore, C., D. Guignet, K. Maguire, C. Dockins, and N. Simon. 2015. A stated preference study of the Chesapeake Bay and Watershed Lakes. NCEE Working Paper Series 15-06. U.S. Environmental Protection Agency, Washington, D.C., USA. Moss, B., S. McGowan, and L. Carvalho. 1994. Determination of phytoplankton crops by top-down and bottom-up mechanisms in a group of English lakes, the West Midland meres. Limnology Oceanography 39:1020–1029. Nakano, S., and M. Murakami. 2001. Reciprocal subsidies: dynamic interdependence between terrestrial and aquatic food webs. Proceedings of the National Academy of Sciences USA 98:166–170. Nanus, L., D. W. Clow, J. E. Saros, V. C. Stephens, and D. H. Campbell. 2012. Mapping critical loads of nitrogen deposition for aquatic ecosystems in the Rocky Mountains, USA. Environmental Pollution 166:125–135. Nilsson, J., and P. Grennfelt. 1988. Critical loads for sulphur and nitrogen. Workshop Report, UNECE/ Nordic Council, Skokloster, Sweden. Nydick, K. R., B. M. Lafrancois, J. S. Baron, and B. M. Johnson. 2004. Nitrogen regulation of algal biomass, productivity, and composition in shallow

Goldman, C. R. 1988. Primary productivity, nutrients, and transparency during the early onset of eutrophication in ultra-oligotrophic Lake Tahoe, California-Nevada. Limnology and Oceanography 33:1321–1333. Handley, L., D. Altsman, and R. DeMay, editors. 2007. Seagrass Status and Trends in the Northern Gulf of Mexico: 1940–2002. U.S. Geological Survey Scientific Investigations Report 2006–5287 and U.S. Environmental Protection Agency 855-R-04-003. USGS, Reston, Virginia. Heard, A. M. 2013. Global change and mountain lakes: establishing nutrient criteria and critical loads for Sierra Nevada lakes. Dissertation. University of California, Riverside, California, USA. Irvine, I. C., T. Greaver, J. Phelan, R. D. Sabo, and G. Van Houtven. 2017. Terrestrial acidification and ecosystem services: effects of acid rain on bunnies, baseball and Christmas trees. Ecosphere 8:e01857. https://doi.org/10.1002/ecs2.1857 Jassby, A. D., C. R. Goldman, and J. E. Reuter. 1995. Long-term change in Lake Tahoe (CaliforniaNevada, USA.) and its relation to atmospheric deposition of algal nutrients. Archives of Hydrobiology 135:1–21. Jassby, A. D., C. R. Goldman, J. E. Reuter, and R. C. Richards. 1999. Origins and scale dependence of temporal variability in the transparency of Lake Tahoe, California-Nevada. Limnology and Oceanography 44:282–294. Jeppesen, E., J. Jensen, M. Søndergaard, and T. Lauridsen. 2005. Response of fish and plankton to nutrient loading reduction in eight shallow Danish lakes with special emphasis on seasonal dynamics. Freshwater Biology 50:1616–1627. Jeppesen, E., J. Peder Jensen, M. Søndergaard, T. Lauridsen, and F. Landkildehus. 2000. Trophic structure, species richness and biodiversity in Danish lakes: changes along a phosphorus gradient. Freshwater Biology 45:201–218. Jones, J., and C. D. Sayer. 2003. Does the fish-invertebrate-periphyton cascade precipitate plant loss in shallow lakes? Ecology 84:2155–2167. Kemp, W. M., et al. 2004. Habitat requirements for submerged aquatic vegetation in Chesapeake Bay: water quality, light regime, and physical-chemical factors. Estuaries 27:263–377. Kennison, R. L., K. Kamer, and P. Fong. 2011. Rapid nitrate uptake rates and large short-term storage capacities may explain why opportunistic green macroalgae dominate shallow eutrophic estuaries. Journal of Phycology 47:483–494. Knapp, R. A., P. S. Corn, and D. E. Schindler. 2001. The introduction of nonnative fish into wilderness lakes: good intentions, conflicting mandates,


RHODES ET AL.

16

July 2017



RHODES ET AL.

directive: ecological classification of Danish lakes. Journal of Applied Ecology 42:616–629. Swift, T., J. Perez-Losada, S. G. Schladow, J. Reuter, A. Jassby, and C. Goldman. 2006. Water clarity modeling in Lake Tahoe: linking suspended matter characteristics to Secchi depth. Aquatic Sciences 68: 1–15. U.S. EPA. 2008. 2008 Final Report: Integrated Science Assessment (ISA) for Oxides of Nitrogen and Sulfur Ecological Criteria. EPA/600/R-08/082F. U.S. Environmental Protection Agency, Washington, D.C., USA. U.S. EPA. 2009. The next generation of tools and actions to restore water quality in the Chesapeake Bay. U.S. Environmental Protection Agency, Washington, D.C., USA. U.S. EPA. 2015. Benefit and cost analysis for the effluent limitations guidelines and standards for the steam electric power generating point source category. EPA-812-R-15-0005. U.S. Environmental Protection Agency, Washington, D.C., USA. Vadeboncoeur, Y., E. Jeppesen, M. J. Vander Zanden, H. Schierup, K. Christoffersen, and D. Lodge. 2003. From Greenland to green lakes: cultural eutrophication and the loss of benthic pathways in lakes. Limnology Oceanography 48:1408–1418. Vadeboncoeur, Y., D. Lodge, and S. Carpenter. 2001. Whole-lake fertilization effects on distribution of primary production between benthic and pelagic habitats. Ecology 82:1065–1077. Van de Bund, W. J., et al. 2004. Responses of phytoplankton to fish predation and nutrient loading in shallow lakes: a pan-European mesocosm experiment. Freshwater Biology 49:1608–1618. Vander Zanden, M. J., S. Chandra, B. C. Allen, J. E. Reuter, and C. R. Goldman. 2003. Historical food web structure and restoration of native aquatic communities in the Lake Tahoe (CaliforniaNevada) Basin. Ecosystems 6:274–288. Wolfe, A. P., A. C. Van Gorp, and J. S. Baron. 2003. Recent ecological and biogeochemical changes in alpine lakes of Rocky Mountain National Park (Colorado, USA): a response to anthropogenic nitrogen deposition. Geobiology 1:153–168. Zhang, J. 2011. Behavioral response to stock abundance in exploiting common-pool resources. B.E. Journal of Economic Analysis and Policy 11: 1–27.

mountain lakes, Snowy Range, Wyoming, USA. Canadian Journal of Fisheries and Aquatic Sciences 61:1256–1268. O’Dea, C. B., S. Anderson, T. Sullivan, D. Landers, and C. F. Casey. 2017. Impacts to ecosystem services from aquatic acidification: using FEGS-CS to understand the impacts of air pollution. Ecosphere 8:e01807. https://doi.org/10.1002/ecs2.1807 Paerl, H. W., and M. F. Piehler. 2008. Nitrogen and marine eutrophication. In D. G. Capone, M. Mulholland, and E. Carpenter, editors. Nitrogen in the marine environment. Volume 2. Academic Press, Orlando, Florida, USA. Pardo, L. H., M. J. Robin-Abbott, and C. T. Driscoll. 2011. Assessment of nitrogen deposition effects and empirical critical loads of nitrogen for ecoregions of the United States. USDA Forest Service, Newton Square, Pennsylvania, USA. Quick, A., and A. Johansson. 1992. User assessment survey of a shallow freshwater lake, Zeekoevlei, Cape Town, with particular emphasis on water quality. Water SA-Pretoria 18:247. Saros, J., D. Clow, T. Blett, and A. Wolfe. 2011. Critical nitrogen deposition loads in high-elevation lakes of the western US inferred from paleolimnological records. Water, Air, & Soil Pollution 216: 193–202. Saros, J. E., S. J. Interlandi, A. P. Wolfe, and D. R. Engstrom. 2003. Recent changes in the diatom community structure of lakes in the Beartooth Mountain Range, USA. Arctic Antarctic and Alpine Research 35:18–23. Saros, J. E., T. J. Michel, S. J. Interlandi, and A. P. Wolfe. 2005. Resource requirements of Asterionella formosa and Fragilaria crotonensis in oligotrophic alpine lakes: implications for recent phytoplankton community reorganizations. Canadian Journal of Fisheries and Aquatic Sciences 62:1681–1689. Scheffer, M., S. Hosper, M. Meijer, B. Moss, and E. Jeppesen. 1993. Alternative equilibria in shallow lakes. Trends in Ecology and Evolution 8:275–279. Smith, D. G., G. F. Croker, and K. McFarlane. 1995. Human perception of water appearance: 1. clarity and colour for bathing and aesthetics. New Zealand Journal of Marine and Freshwater Research 29:29–43. Søndergaard, M., E. Jeppesen, J. Peder Jensen, and S. Lildal Amsinck. 2005. Water framework

SUPPORTING INFORMATION Additional Supporting Information may be found online at: http://onlinelibrary.wiley.com/doi/10.1002/ecs2. 1858/full


17

July 2017
