Grantee Research Project Results
Final Report: Testing Indicators of Coastal Ecosystem Integrity Using Fish and Macroinvertebrates
EPA Grant Number: R828675C003Subproject: this is subproject number 003 , established and managed by the Center Director under grant R828675
(EPA does not fund or establish subprojects; EPA awards and manages the overall grant for this center).
Center: EAGLES - Great Lakes Environmental Indicators Project
Center Director: Niemi, Gerald J.
Title: Testing Indicators of Coastal Ecosystem Integrity Using Fish and Macroinvertebrates
Investigators: Johnson, Lucinda , Richards, Carl , Schuldt, Jeffrey A. , Ciborowski, Jan , Morrice, John , Yurista, Peder , Hrabik, Thomas , Breneman, Dan , Kelly, John R. , Scharold, Jill , Sierszen, Michael , Tanner, Dan , Trebitz, Anett , Brady, Valerie J
Institution: University of Minnesota , U. S. Environmental Protection Agency , University of Wisconsin - Green Bay , University of Windsor , Minnesota Sea Grant College Program
EPA Project Officer: Packard, Benjamin H
Project Period: January 10, 2001 through January 9, 2005 (Extended to January 9, 2006)
RFA: Environmental Indicators in the Estuarine Environment Research Program (2000) RFA Text | Recipients Lists
Research Category: Ecological Indicators/Assessment/Restoration , Water , Aquatic Ecosystems
Objective:
Despite the incomplete historical data record on Great Lakes habitats, there is consensus that the lakes are rapidly changing (Environment Canada and U.S. Environmental Protection Agency [EPA], 1999). No single indicator can capture the diverse information necessary to evaluate ecosystem condition. A major challenge is to select a subset of many proposed indicators that will effectively and efficiently measure the major components of ecosystem health and can diagnose causes of impaired community function. A second challenge is to link those indicators to assessment endpoints. We have combined the strengths of two common approaches (multimetric and multivariate) to generate derivative, ecologically relevant indicators that have the greatest possible discriminatory power to distinguish degraded systems from least-impaired systems.
We employed a multitiered sampling and modeling strategy, integrating data collected at regional scales via satellite imagery, local scales, and site scales via field sampling. The goals of our project were to: (1) evaluate the applicability of relevant State of the Lake Ecosystem Conference (SOLEC)-derived and complementary indicators in the context of the ecosystem types making up the Great Lakes coastal region; (2) rigorously evaluate a suite of indicators across the range of Great Lakes coastal habitats; and (3) recommend indicators of specific ecological conditions keyed to assessment endpoints and stressors in the Great Lakes coastal region.
Specific objectives of this research project were to: (1) characterize fish and macroinvertebrate communities; (2) develop a fish index of biotic integrity and test Scirpus and Typha-dominated wetland indices of fish biotic integrity; (3) assess macroinvertebrate responses to stress; (4) assess fish community responses to local and landscape stressors; (5) assess habitat changes that stem from environmental disturbances; (6) detect scales of responses of macroinvertebrate and fish indicators to landscape stressors; and (7) develop multivariate methods to identify coastal ecosystem conditions using fish communities and macrobenthos communities.
Summary/Accomplishments (Outputs/Outcomes):
Methods
Between 2001 and 2003, we sampled 116 sites at 101 unique locations across the greater than 7,000 km of U.S. Great Lakes coastal margin. Fifteen sites were revisited to quantify temporal variation. In addition, 53 sites were sampled as part of a parallel study (EPA Science To Achieve Results Grant R828777) to define reference condition at Great Lakes coastal margins. Benthic samples were collected using sweep nets, sediment coring tubes, and petite ponar grab samplers from which at least 226 genera (excluding Chironomidae and Oligochaeta) were identified. Approximately 1,100 overnight fyke net sets were fished, resulting in capture, identification, and release of more than 100,000 fish representing 110 species. Habitat attributes and characteristics of sampling location and the surrounding landscape were recorded at more than 1,500 benthos sampling points, 800 net locations, and 3,000 additional randomly selected points. Water quality was measured at approximately 2,000 locations. Fish community composition (numbers of individuals of each species) was noted at each fyke net; the catch was standardized by net size (small vs. large) and catch per unit effort. Fish and macroinvertebrates were summarized with respect to relative abundance of each taxon per site, as well as using a variety of taxon metrics describing trophic habits, life history features, behavioral characteristics, and community composition.
Synoptic habitat structure measurements of water depth, substrate, vegetation, hydrologic connections to the lake and landscape, and disturbances were made. Physicochemical variables were measured at each biotic sample location. More than 100 individual habitat-associated attributes of each site were summarized by principal components analysis (PCA) to yield four sets of eigenvectors representing land use/land cover, physical structure, vegetation cover, and anthropogenic disturbance.
Community Structure: Zoobenthos
The extensive collection records of zoobenthos (more than 4,000 samples from 143 locations) have provided important biogeographic and taxonomic information over and above the primary use of the data to develop and test indicators of anthropogenic stress. We discovered one invertebrate species new to the Great Lakes, mapped the range expansion of a second invader, and used the distributions and associations among several other nonindigenous invading species (NIS) together with our stressor data to test hypotheses about the mechanisms that promote establishment of new species in the Great Lakes.
Quantifying Aquatic Habitat Alteration Stemming From Anthropogenic Disturbance. A redundancy analysis was conducted to explain how landscape-scale features (ecoregion, hydogeomorphic type, anthropogenic disturbance) influenced site-specific variation in local synoptic aquatic habitat attributes. In all, 33 percent of local aquatic habitat variation could be accounted for by the landscape and stressor features. Approximately two-thirds of this explained variation (19% of the overall variation) could be accounted for specifically by the effects of anthropogenic stress—15 percent was uniquely attributable to stress, and 4 percent could also be explained by covariation with spatial pattern, geomorphic type, or complex size (Figure 1A). Overall, anthropogenic disturbance exerted small but meaningful changes in habitat attributes that themselves influence fish and macroinvertebrate community structure.
Community Structure. Fishes
Two Fish Indicator Indices Assess Great Lakes Coastal Wetland Condition. Scientists with the Great Lakes Coastal Wetlands Consortium had previously proposed that because submergent plant communities adapt quickly to changing water levels, perhaps the fish communities associated with plant types could be used as indices of wetland condition. They proposed a fish index of biotic integrity (IBI) for wetlands dominated by cattails (Typha) and another IBI for those in which bulrushes (Scirpus) were most common.
The fish IBI scores calculated for these wetlands varied predictably but only according to specific classes of human-related stress. Fish communities in cattail-dominated wetlands became degraded as a disturbance variable that combined population density, road density, and urban development in the watershed surrounding the wetland increased. In contrast, fish communities of bulrush-dominated wetlands reflected nutrient and chemical inputs associated with the degree of agriculture in the surrounding landscape. Great Lakes water levels varied by up to 100 cm over the study years, confirming the indices’ effectiveness under changing water conditions. The fish IBI scores in bulrush wetlands were much lower once a threshold level of agricultural-input stress had been exceeded. In contrast, the IBI scores in cattail wetlands gradually declined with increasing population disturbance (Figure 2). The Uzarski, et al. (2006) IBI scores reflect specific classes of anthropogenic stress in coastal wetlands dominated by cattails and bulrushes, most notably agriculture or population density effects rather than generalized disturbance. Diagnosing causes of water quality impairment is an important component of the Great Lakes water quality agreement of the governments of Canada and the United States and the U.S. Federal Clean Water Act. These results address one of the weaknesses of the IBI approach, in that a single value representing ecological condition does not address the cause of impairment.
Describing Patterns of Variation of Fish Communities and Responses to Stress. Effective environmental indicators exhibit consistent patterns of variation as a function of particular stressors and have a clearly identified scale of response (Jackson, et al., 2000). Matching the appropriate scale of responses to the stressors of interest is a critical step in the development of an effective indicator. We have addressed this question using several approaches to study variation pertaining to fish community responses to stress in Great Lakes coastal areas: (1) What are the relative influences of wetland geomorphic setting, geography, and landscape-scale stressors on the absolute and relative abundances of selected fish species and species traits? (2) What landscape-scale stressors best explain the distribution of fish species in Great Lakes coastal wetlands? and (3) What is the relative influence of local habitat versus landscape stressors and spatial position on fish species across Great Lakes coastal areas?
Figure 1. (A) Percent Variation Explained in Great Lakes Environmental Indicators (GLEI) Project Habitat Principal Components by Geomorphic Type/Complex Size, Landscape/Stress, and Spatial Variables. Note that one-half of the variation in habitat features is associated with landscape/stress factors. (B) Percent variation explained in GLEI fish presence/absence data (with rare species removed) by local, landscape, and spatial variables.
* Negative values caused by a not strictly linear relationship between habitat principal components and environmental variables, geomorphic type, and complex type when spatial variables are held as covariates (complex relationships between predictor variables).
Figure 2. Typha IBI for Fishes Decreases With Increasing Population Disturbance
Pronounced gradients in climate and landforms across the Great Lakes cause strong biogeographic differences across the basin. Danz, et al., 2006 (see the report for Grant No. R828675C001) examined the distribution of landscape stressors and selected environmental indicatorsand demonstrated strong geographic patterns in five types of anthropogenic stressors across the Great Lakes basin. Specific to fish communities, 30 percent of the total variation in fyke net fish catch per unit effort was explained by the combination of stress, geographic location, and geomorphic setting. They showed that five NIS, along with turbidity-tolerant species, were associated with increasing levels of those anthropogenic stressors. Danz, et al., 2006 (see the report for Grant No. R828675C001) identified a replacement series of taxa whose relative abundance changed as the adjacent landscape ranged from natural to disturbed land covers. These analyses demonstrated that the patterns of association were confounded by species’ geographic position within the basin; however, because of the coarse nature of the stressor data (i.e., derived from regional data sources and summarized at segment shed scales) the spatial scales of the stressors explaining the observed patterns could not be distinguished.
Using fish indicators that summarized species composition, native species richness, and proportional abundance values reflecting fish size, life history, behavior, feeding guild, and tolerance to turbidity found that the values of six fyke-fish indicators were best predicted by the Great Lake being sampled, whereas wetland type and amount of anthropogenic stress explained the highest amount of variance for two indicators each. Interaction effects between lake and wetland type were prevalent. Among these selected fyke net fish indicators bluegill, carp/goldfish, and rock bass relative abundances had the largest amount of variance accounted for by a composite measure of stress assessed at the segment shed level. This analysis revealed that both individual species and species-based composite indicators exhibited a broad range of responses to spatially-relevant features (i.e., wetland geomorphic type, Great Lake, and ecoregion) as well as to an overall measure of stress.
Analyses left open the question – what is the specific spatial scale of responses to individual stressors and geographic factors? This question was addressed using classification and regression trees to predict responses of fish indicators to geographic (i.e., ecoregion, lake) variation and to stressors characterized at differing spatial extents around coastal wetlands, ranging from 100 m to full watersheds draining into coastal wetlands. An indicator of turbidity intolerance had the largest amount of variation explained of all the fyke net fish variables. The model predicted 54 percent of the total variance in fish relative abundance. Watershed area, percent of development at the 500 m buffer extent, and percent of row crop within the watershed were the best explanatory variables for this index. Percent of nest-guarding spawners was the next best indicator, with 46 percent of the variance explained exclusively by physical variables, including wetland area and type, watershed area, and lake.
Two individual fish species blue sunfish (Lepomis machrochirus) and northern rock bass (Ambloplites rupestris) and a species group including the European carp and goldfish proved to be responsive as potential indicators. Bluegill occurrence was best predicted by the ecosection, percent of development and percent of rowcrop agriculture. Interestingly, development was either positively or negatively correlated with this species, depending on the spatial extent. At small spatial extents (100 m and 500 m buffers), development was positively correlated with bluegill abundance. At the 5,000 m extent and the watershed, the bluegill was negatively correlated with both development and rowcrop agriculture. This suggests a potential stimulation or fertilization effect of development at the local scale.
The northern rock bass exhibited a more spatially consistent response to disturbance (with 44% variance explained). Occurrence was negatively associated with development at the watershed scale. This model also included lake and wetland area. In a separate analysis examining the spatial scale of responses to land use, the northern rock bass was found to correlate negatively with increasing development at very large scales, with the strongest negative correlation found at a 50 km2 extent.
The goldfish/European carp species group was positively correlated with rowcrop agriculture at the 500 m extent, but negatively associated with rowcrop at the 5,000 m scale (35% of variation explained). We found that the frequency of occurrence of both species increased with amount of rowcrop agriculture at the scale of entire watersheds.
The analyses effectively partitioned variation caused by the spatial extent of stressors and geographic position; however, this study was limited to coastal wetlands and did not address the potential impact of local habitat features on the fish community. We studied variation in fish community composition to address the question—what are the dominant environmental and geographic factors structuring the fish community in Great Lakes coastal areas, including high energy zones, embayments, and wetlands? Landscape and stressor data were summarized for watersheds delineated for each sample site. Fish community composition was analyzed for 143 sites, including 41 high energy, 19 embayments, and 83 coastal wetlands. Canonical correspondence analysis was used to examine species-environment relationships, implementing variance apportioning procedures (Cushman and McGarigal, 2002) to separate unique and shared variation among environmental predictors. Eigenvalues for four distinct classes of habitat variables were used as local-scale predictors along with geomorphic type and size.
Variance Decomposition. Very strong (and identical) patterns in fish presence/absence were observed with respect to local, landscape/stress and spatial variables whether analyses were conducted on fish species presence/absence or relative abundance. Forty-six percent of the total variation in fish species presence/absence was explained by the three sets of variables. Of that variation, more than half (27%) was attributable to local variables (geomorphic type and size, surrounding land use, physical habitat structure and water quality, and vegetation), and a little less than half (19%) was attributed to landscape/stress. Landscape/stress variables and spatial location shared 10 percent of their variation in common. Thus, 50 percent of the variation in spatial position also was explained by the landscape/stressor variables (Figure 1B).
Of the local-scale variation, geomorphic type accounted for 35 percent of the variance, land use accounted for 43 percent, aquatic habitat accounted for 32 percent, physical characteristics and unit size accounted for 28 percent, and anthropogenic activities (local disturbance) accounted for 13 percent. There was considerable overlap in the explained variation in local scale variables.
Fish Community Responses. Three primary fish associations were identified to have stable responses to local and landscape/stress variables independent of spatial location. These groups have characteristic thermal and habitat preferences, as well as consistent relative tolerance to disturbance. Many species exhibit a “wedge” distribution relative to intensity of stress—variation increases (or decreases) along the stress gradient. This indicates that unmeasured variables are limiting the distribution (Cade and Noon, 2003). Further analyses will be conducted to characterize this response envelope.
Potential Indicators. Several species exhibit consistent response patterns to stress across geomorphic types and the basin. These include the nonnative carp/goldfish and white perch; yellow perch and native species emerald shiner, northern pike, burbot, and bluegill. By quantifying the relative influence of each of these three factors, we have been able to identify species responses to the local habitat as well as to landscape stressors, independent of the influence of geography. This has enabled us to identify species that have consistent responses to stress independently of their spatial location in the basin (Figure 3).
Developing Multivariate Fish and Macroinvertebrate Indicators Using A Priori Classification of Reference Condition and Degraded Condition Sites
Effective indicators of environmental condition should be scored against accepted standards. The standard used to assess the integrity of a community is whether or not the numbers and kinds of species are similar to the community found in an environmentally similar area. Such sites are said to represent the “reference condition.”
Multivariate indicators of environmental condition are typically developed by defining the bounds of composition of the biological community expected within a suite of sampling sites selected to represent the reference condition. A test site is classed as “nonreference” if its community falls outside that range; this is expressed as a probability. However, because such measures are unbounded, one cannot tell how degraded a nonreference community is. We identified classes of sites representing the two extremes of anthropogenic stress in Great Lakes coastal wetlands. Stress at a site was the relative maximum (relmax) intensity of road density, residential/commercial land use, agricultural land use, human population density, and distance to point sources in the contributing watershed. The 20 percent of wetlands with the lowest relmax scores were classified as reference condition sites.
To identify fish community indicators of reference and degraded conditions, we used the relative abundances of each fish species at 133 locations. Cluster analysis of fish community composition at 46 reference condition sites only identified five unique assemblages. Discriminant function analysis was then used to find the suites of environmental variables that best characterized the sites supporting each of the assemblages. Seven environmental variables, summarizing primarily ecoregion attributes and wetland type were able to correctly classify all but one of the reference condition sites. The derived discriminant functions were then used to assign the 87 other locations to 1 of the 5 unique assemblages. The assignation represented the fish assemblage expected at an “other” location if that location was equivalent to reference.
Example Indicators | ||
Figure 3. Examples of Fish Species Responses to Stressors Summarized for Watersheds Contributing to Coastal High Energy, Embayment or Wetland Sites. The most common response is a “wedge,” indicating that unmeasured factors are limiting at either end of the stress gradient (Cade and Noon, 2003). The response envelope will be characterized for individual species and species traits at a later date.
We developed PCA-derived composite estimates of urban disturbance and of agricultural disturbance for each of the 133 sampling locations using the data of Danz, et al. (2005; R828675C001). To identify the indicator species characteristic of reference conditions and of sites degraded by urban development and agricultural activity, respectively, we performed Bray-Curtis ordinations on the fish community data on an assemblage-by-assemblage basis. We first designated the sites with the lowest agricultural disturbance score and the highest agricultural disturbance score as the two end points of a “fish condition index of agricultural stress.” The Bray-Curtis analysis then ordinated the sites within the assemblage along this index. The fish species whose relative abundances correlated most highly with changes in the index score were designated “indicator species.” The process was then repeated using endpoints for urban disturbance scores for the assemblage. In all, 10 ordinations were performed (two per fish assemblage).
There was considerable variation in our ability to statistically define indicator species representative of the reference and degraded ends of the disturbance gradients. There were marked differences within assemblages of three clusters (correlations ranging from 0.6 to 0.7, p < 0.05). Furthermore, taxa indicative of reference condition sites in one assemblage (e.g., spottail shiners in cluster 1) were sometimes representative of disturbed conditions in another assemblage (spottail shiners in cluster 3, Table 1). This reflects the broad correlation between latitude and productivity characteristic of areas minimally affected by humans across the Great Lakes.
The multivariate approach to identifying biological indicators of environmental condition is unique in that faunal data are used to group together reference sites that have similar species composition, thus providing an objective strategy for assessing habitat quality based on species assemblages. We were able to develop models of fish species relative abundances that can be used to evaluate the quality of sites in response to specific anthropogenic stressors.
Identifying Macroinvertebrate Indicators. We have concentrated on developing indicators for Great Lakes wetlands because wetlands exhibit less variation in depth than other Great Lakes margin habitats and hence less compositional variability. Indicator development to date has focused on D-frame dip net samples, collected at depths of 0.5–1 m along transects set perpendicular to shore throughout each wetland. Most taxa were identified to genus. Chironomidae and Oligochaeta were identified to family.
Cluster number (# sites) | Reference species | Disturbed (urban) | Disturbed (agriculture) |
1 (n = 12) | Northern rockbass | Alewife | European carp Pumpkinseed sunfish |
2 (n = 9) | Sand shiner | European carp | Bluegill sunfish |
3 (n = 5) | White sucker | Northern rockbass Largemouth bass | Alewife |
4 (n = 5) | Slimy sculpin | Northern rock bass | None |
5 (n = 11) | Brown bullhead | Emerald shiner | Spotfin shiner |
Nonmetric multidimensional scaling analysis indicated that macroinvertebrate assemblages in northern wetlands (Laurentian Mixed Forest; n = 42) (Keys, et al., 1995) were highly significantly different from those in the southern Great Lakes (Eastern Broadleaf Forest; n = 41, p < 0.0001). Furthermore, the assemblages of Lake Erie wetlands differed from those of Lake Ontario (p < 0.0001), and assemblages varied significantly and consistently among wetland types (riverine, protected, or open lacustrine) in the northern wetlands (p < 0.0001). Potential macroinvertebrate indicator metrics were less variable but still differed markedly between the northern and southern wetlands (p < 0.0001). Thus, wetland macroinvertebrate assemblages were diverse and distinct from one another when assessed solely on where wetlands occurred and on their hydrogeomorphic type. Similar trends were observed in eight zoobenthic indicator taxa by noting highly significant Great Lake by wetland-type interactions.
Several of the 24 potential macroinvertebrate indicator metrics assessed thus far show promise. In particular, relative abundance of the common mayfly genus Caenis may be a useful indicator for Lake Erie wetlands. The proportion of Caenis mayflies in dip net samples was negatively correlated with the amount of unvegetated area in the 14 Lake Erie wetlands sampled (r = 0.47), the amount of riprapped shoreline (r = 0.45), larger watershed sizes for wetlands (r = 0.65), amount of row crop agriculture in the watershed (0.74), our watershed urbanization PC-1 (0.67), and positively correlated with the amount of emergent wetlands and deciduous forest in the watershed (r = 0.79 and 0.65, respectively) (Figure 4). The negative correlation between the proportion of Caenis mayflies and wetland watershed size appears to result from the increasing likelihood of larger watersheds being more developed and hence subject to greater disturbance, particularly in the southern Great Lakes. Overall, the proportion of a sample comprised of mayflies (all genera combined) was negatively correlated with watershed size in the southern ecological province, 222 (r = 0.52).
Other promising metrics include:
- For Lake Erie wetlands, the proportion of a sample comprised of all Ephemeroptera, Trichoptera, Odonata and Sphaeriidae (proportion ETSO).
- For wetlands in the northern Great Lakes ecological region (Keys, et al., 1995):
- Proportion of Caenis mayflies.
- Proportion Ephemeroptera.
- For the southern Lake Superior uplands ecological region (Keys, et al., 1995):
- Proportion Caenis mayflies.
- Proportion Coenagrion and Enallagma damselflies.
- Proportion Trichoptera.
- Taxonomic richness at the family or genus level.
Several metrics with potential have not borne out. In particular, measures of community evenness have proven to be less useful than Shannon-Weaver diversity and richness measures.
Macroinvertebrates also were collected in wetlands using cores and petite ponars, and indicators developed based on these methods will be compared to those based on the dip net collection method. We also collected macroinvertebrates from wave-swept beaches and shoreline areas along embayments.
Zoobenthic community composition is strongly regulated by hydrological features such as exposure to waves and currents. Water depth (which ranged from 0.5 to 10 m in our sampling protocols) is expected to exert as strong an influence on community composition as many landscape and biogeographic features. Consequently, we anticipate that the application of our multivariate analytical approach described in detail for fish indicator development (Bhagat, et al., in review) will be especially useful in accounting for this covariation.
Figure 4. Variation in Caenis Mayflies Relative to Habitat and Watershed Characteristics for Lake Erie Wetlands (n = 14). (A) The proportion of Caenis mayflies was higher in wetlands with less open water (r = 0.47). (B) The proportion of Caenis mayflies was higher in wetlands with higher amounts of emergent marsh in their watersheds (r = 0.79). (C) Caenis mayfly proportions were higher in wetlands with more deciduous tree cover in the watershed (r = 0.65). (D) Caenis proportions were lower in wetlands with more watershed urbanization (r=0.67).
Figure 5. Structured Equation Model Relating Direct and Indirect Effects (Humans to Plants)
Direct and Indirect Impacts of Human Activities on Larval Odonata–A Structured Equation Modelling Approach. Honors undergraduate thesis student Carolyn Foley used structural equation modelling to determine quantitatively whether anthropogenic activities along the edges of wetlands in the Great Lakes affected organisms directly or indirectly through alteration of habitat. Structural equation models were developed to test the effect of human activities at the shoreline on nymphal Odonata and on the structural habitat provided for them by aquatic macrophytes at the 0-0.5 m and 0.5-1 m depths in coastal, riverine, and protected wetlands. Each model contained four latent variables: “human activities,” “abiotic habitat,” “plants,” and “odonates” (Figure 5). We used 244 sampling points in the 0-0.5 m model, and 233 sampling points in the 0.5-1 m model. Structural equation models were compiled and analyzed using the Structural Equation Modeling and Path Analysis (SEPATH) module in Statistica™ version 7.1 (Statsoft Inc., 2005). The maximum likelihood estimation was used to estimate path coefficients based on the covariance matrix of the variables, with the root-mean-square error of approximation used as a goodness of fit index. The adjusted population gamma index was also examined for goodness of fit.
Overall, the models indicate that human activity in the Great Lakes wetlands sampled tended to increase the density of macrophytes (likely through eutrophication effects) that reduces the condition of the larval odonate community. The indirect effects are likely more important than direct (e.g., toxic discharge) effects of human activity on odonate communities. These patterns are broadly consistent with findings of Brady et al. (in preparation, see above), who reported that the proportion of damselflies in samples tended to decrease as a function of increasing generalized environmental disturbance.
Additional Applications
GLEI and the Lake Erie Lakewide Management Plan (LaMP). Active interactions between GLEI principal investigators Jan Ciborowski and Lucinda Johnson and the Lake Erie LaMP’s technical workgroup have resulted in a strong integration of GLEI concepts and approaches into various aspects of the Lake Erie implementation and indicators strategies. The Lake Erie LaMP has recognized the strong link between land-based stress and both coastal and basinwide ecosystem processes. Accordingly, they are adopting the stressor measures developed by Danz (see the Final Report for Grant No. R828675) as integral indicators of the condition of watersheds tributary to Lake Erie. The Lake Erie LaMP has allocated funding to Drs. Johnson and Ciborowski to compile and crosswalk Canadian-based GIS stressor information and develop conversion factors that will permit the entire Lake Erie basin to be assessed according to five classes of anthropogenic stress. Negotiations also are underway to incorporate information on Canadian portions of the other Great Lakes as opportunities arise. Once compilation is complete we will have contributed greatly to the mandate of SOLEC to provide indicators that can report on basinwide measures of biological integrity. Because our fish and invertebrate indicators are calibrated against the Danz measures of anthropogenic stress, our derivative biological measure of ecosystem condition will also be applicable to the entire Great Lakes basin.
GLEI and Tiered Assessment of Aquatic Life Uses (EPA Office of Water). The EPA Office of Water has been engaged in a 4-year study to develop biologically based criteria (condition of algae, macroinvertebrates, fishes, etc.) to assess the quality of wadeable streams throughout the continental United States (EPA, 2005). Its goal is to develop a classification strategy by which regional Agency scientists can assign the biota characteristics of their region of jurisdiction into tiers representing five classes of biological integrity reflecting environmental condition ranging from equivalent-to-reference to strongly degraded. Drs. Johnson and Ciborowski have participated in adapting the conceptual approaches and applications achieved by the GLEI project and companion Reference Condition project for definition of a human disturbance gradient against which the biological tiers can be calibrated. GLEI provided a method for the quantification and summary of multiple stressors into a smaller number of independent stress classes that can be related to biological endpoints, and a methodology for summarizing the classes to define the bounds of a reference condition. These ideas have been incorporated into a draft document (EPA, 2005) and are being presented by the Office of Water as guidelines at numerous regional development workshops. The assessment criteria that ultimately are derived from these workshops are expected to increasingly guide water use designation in waterways across the United States over the next decade.
References:
Cushman SA, McGarigal K. Hierarchical, multi-scale decomposition of species-environment relationships. Landscape Ecology 2002;17:637-646.
Environment Canada and U.S. EPA. State of the Great Lakes 1999. U.S. Environmental Protection Agency EPA Report, 905-R-99-008, 1999.
Jackson LE, Kurtz JC, Fisher WS. Evaluation guidelines for ecological indicators. EPA/620/R-99/005. U.S. Environmental Protection Agency, Office of Research and Development, Research Triangle Park, NC, 2000, 107 pp.
Keys Jr. J, Carpenter C, Hooks S, Koenig F, McNab WH, Russell W, Smith ML. Ecological units of the eastern United States - first approximation, Atlanta, GA: U.S. Department of Agriculture, Forest Service, 1995.
U.S. EPA. Use of biological information to better define designated aquatic life uses in state and tribal water quality standards: tiered aquatic life uses. U.S. EPA Draft Document – August 10, 2005, 207 pp.
Journal Articles on this Report : 8 Displayed | Download in RIS Format
Other subproject views: | All 36 publications | 9 publications in selected types | All 8 journal articles |
---|---|---|---|
Other center views: | All 279 publications | 67 publications in selected types | All 58 journal articles |
Type | Citation | ||
---|---|---|---|
|
Bhagat Y, Ciborowski JJ, Johnson LB, Uzarski DG, Burton TM, Timmermans ST, Cooper MJ. Testing a fish index of biotic integrity for responses to different stressors in Great Lakes coastal wetlands. Journal of Great Lakes Research 2007;33(Suppl 3):224-235. |
R828675C003 (Final) |
Exit Exit Exit |
|
Brady VJ, Ciborowski JJH, Johnson LB, Danz NP, Holland JD, Breneman DH, Gathman JP. Optimizing fishing time: one vs. two-night fyke net sets in Great Lakes coastal systems. Journal of Great Lakes Research 2007;33(Suppl 3):236-244. |
R828675 (Final) R828675C003 (Final) |
Exit Exit Exit |
|
Cade BS, Noon BR. A gentle introduction to quantile regression for ecologists. Frontiers in Ecology and the Environment 2003;1(8):412-420. |
R828675C003 (Final) |
Exit |
|
Grigorovich IA, Mills EL, Richards CB, Breneman D, Ciborowski JJH. European valve snail Valvata piscinalis (Müller) in the Laurentian Great Lakes basin. Journal of Great Lakes Research 2005;31(2):135-143. |
R828675C003 (2004) R828675C003 (Final) R828777 (2003) |
Exit Exit Exit |
|
Grigorovich IA, Kang M, Ciborowski JJH. Colonization of the Laurentian Great Lakes by the amphipod Gammarus tigrinus, a native of the North American Atlantic coast. Journal of Great Lakes Research 2005;31(3):333-342. |
R828675C003 (2004) R828675C003 (Final) |
Exit Exit |
|
Host GE, Schuldt J, Ciborowski JJH, Johnson LB, Hollenhorst T, Richards C. Use of GIS and remotely sensed data for a priori identification of reference areas for Great Lakes coastal ecosystems. International Journal of Remote Sensing 2005;26(23):5325-5342. |
R828675C003 (Final) R828777 (2003) |
Exit Exit |
|
Kang M, Ciborowski JJH, Johnson LB. The influence of anthropogenic disturbance and environmental suitability on the distribution of the nonindigenous amphipod, Echinogammarus ischnus, at Laurentian Great Lakes coastal margins. Journal of Great Lakes Research 2007;33(Suppl 3):198-210. |
R828675C003 (Final) |
Exit Exit Exit |
|
Uzarski DG, Burton TM, Cooper MJ, Ingram JW, Timmermans STA. Fish habitat use within and across wetland classes in coastal wetlands of the five Great Lakes: development of a fish-based index of biotic integrity. Journal of Great Lakes Research 2005;31(Suppl 1):171-187. |
R828675C003 (Final) |
Exit Exit Exit |
Supplemental Keywords:
Great Lakes, monitoring, indicators, risk assessment, stressor, ecological effects, animal, plant, diatoms, toxics, aquatic ecosystem, aquatic ecosystems, atmospheric pollutant loads, climate variability, coastal ecosystem, coastal environments, diatoms, ecological assessment, ecological condition, ecological response, ecosystem assessment, ecosystem impacts, ecosystem indicators, ecosystem stress, environmental consequences, environmental stressor, environmental stressors, estuarine ecosystems, hierarchically structured indicators, human activities, hydrologic models, hydrological, hydrological stability, nutrient stress, nutrient supply, nutrient transport, toxic environmental contaminants,, RFA, Scientific Discipline, ENVIRONMENTAL MANAGEMENT, Geographic Area, Water, ECOSYSTEMS, Ecosystem Protection/Environmental Exposure & Risk, Nutrients, exploratory research environmental biology, Ecosystem/Assessment/Indicators, Ecosystem Protection, Ecological Effects - Environmental Exposure & Risk, Environmental Monitoring, Ecological Monitoring, Ecological Risk Assessment, Ecology and Ecosystems, Great Lakes, Ecological Indicators, Risk Assessment, coastal ecosystem, diatoms, ecological condition, aquatic ecosystem, hydrological stability, nutrient supply, nutrient transport, fish, ecosystem assessment, hierarchically structured indicators, wetland vegetation, environmental stressor, hydrological, macroinvertebrates, coastal environments, environmental consequences, ecological assessment, ecosystem indicators, estuarine ecosystems, nutrient stress, aquatic ecosystems, toxic environmental contaminants, water quality, ecosystem stressProgress and Final Reports:
Original AbstractMain Center Abstract and Reports:
R828675 EAGLES - Great Lakes Environmental Indicators Project Subprojects under this Center: (EPA does not fund or establish subprojects; EPA awards and manages the overall grant for this center).
R828675C001 Great Lakes Diatom and Water Quality Indicators
R828675C002 Vegetative Indicators of Condition, Integrity, and Sustainability of Great Lakes Coastal Wetlands
R828675C003 Testing Indicators of Coastal Ecosystem Integrity Using Fish and Macroinvertebrates
R828675C004 Development and Assessment of Environmental Indicators Based on Birds and Amphibians in the Great Lakes Basin
R828675C005 Development and Evaluation of Chemical Indicators for Monitoring Ecological Risk
The perspectives, information and conclusions conveyed in research project abstracts, progress reports, final reports, journal abstracts and journal publications convey the viewpoints of the principal investigator and may not represent the views and policies of ORD and EPA. Conclusions drawn by the principal investigators have not been reviewed by the Agency.
Project Research Results
8 journal articles for this subproject
Main Center: R828675
279 publications for this center
58 journal articles for this center