IP029 (STE-1117-759456) Structuring casual analysis for success in practical application-a case study, Bogue Homo, Jones County, Mississippi.
Start time: 1:50 PM
Hicks, M.¹, Whittington, K.², Thomas, J.¹, Carroll-Perkins, A.², Cook, D.³, Folmar, H.², Stephens, W.², Kurtz, J.⁴, ¹ RMA, Inc., Brandon, MS² Mississippi Department of Environmental Quality, Jackson, MS³ IMS, Inc., Vicksburg, MS⁴ US EPA, Gulf Breeze, FL
Total Maximum Daily Load (TMDLs) are required by 303(d) of the Clean Water Act for those pollutants identified in the listing process. To compute TMDLs for biological impairment listings, the stressor(s) must be identified. The stressor identification (SI) process was developed by the United States Environmental Protection Agency (US EPA) to determine the probable cause(s) once biological impairment has been established. Bogue Homo was initially listed based upon evaluated (unmonitored) information. A conceptual model was developed to present the most probable sources of stress, causal pathways, proximal stressors, and realized effects measured as biological responses of aquatic macroinvertebrates. Data utilized in the SI included benthic macroinvertebrate community metrics, causes and observed effects measured as biological responses of aquatic macroinvertebrates, qualitative habitat assessment scores, sediment particle size counts, various water quality measurements, land use and land cover percentages, and other miscellaneous watershed information. Georeferenced site classes (bioregions) have been previously determined for Mississippi and described in the Development and Application of the Mississippi Benthic Index of Stream Quality (M-BISQ). Impairments were indicated in a segment of Bogue Homo utilizing comparisons of M-BISQ least-disturbed (LD) reference conditions and site specific comparators (SSC). Comparability tables were utilized to identify impairment and stress with the observed parameter levels of the 25th or 75th percentile of the least-disturbed condition values dependent upon the parameter of concern. This case study describes the impairments, identifies candidate causes, evaluates the relationships with biological characteristics, and identifies the most likely stressors through a logical approach of elimination and strength of evidence. Upon completion of the analysis, organic enrichment/low dissolved oxygen, nutrient enrichment, and sediment were identified as the most critical stressors requiring TMDLs for this segment of Bogue Homo. Subsequent to this SI, TMDLs will be developed for the Bogue Homo for those applicable pollutants identified in the process.

IP030 (COR-1117-819948) Ecoepidemiological screening to design a sampling plan for TMDLs: Groundhouse River, MN.
Start time: 1:50 PM
Cormier, S¹, Lane, C¹, Neimela, S², Chirhart, J², ¹ U.S. Environmental Protection Agency, Office of Research and Development, Cincinnati, OH, U.S.A.² Minnesota Pollution Control Agency, Minnesota, U.S.A.
The Groundhouse River in central Minnesota did not meet minimum criteria for aquatic life and beneficial uses and was listed as impaired under Clean Water Act (CWA) Section 303(d). A screening-level causal assessment was undertaken to organize and assess the available data, to eliminate impossible causes, and to identify additional data that would increase confidence in the strength-of evidence analysis. Four candidate causes were considered evaluated. (1) Chronic or episodic toxic exposures were not likely to have occurred since there were no known point sources and ammonia levels were below concentrations reported to cause effects. (2) Hypoxia associated with excessive nutrients was unlikely because excessive algal growth and/or decaying material were not observed, and dissolved oxygen levels were above levels reported to cause most effects (> 5.0 mg/L). (3) Altered food resources were unlikely because excessive algal growth was not observed, and nutrient levels were similar to others sites where the impairment was not observed. (4) The loss of suitable habitat from unstable or unsuitable substrates was most probable because fine-grained sediment (>2mm in diameter) were three times greater than at the upstream location and 50% greater than published effect levels for some species of fish. The median particle size (D50) decreased from 22 mm to 1mm. However, since the analysis was based on unreplicated observations obtained in different years, uncertainty was high. Additional sampling was recommended with special attention to determining the boundaries of the biological impairment, to characterizing natural sediment regimes, and to developing exposure-response associations using state monitoring data. Additional biological, physical and chemical sampling will be used to confirm or refute excess sediment as the probable cause of impairment in Groundhouse River.

IP031 (COR-1117-811114) The role of toxicity tests in ecoepidemiological investigations: Naugatuck River, Connecticut.
Start time: 1:50 PM
Bellucci, C¹, Dunbar, L¹, Cormier, S², ¹ Connecticut Department of Environmental Protection, U.S.A.² U.S. Department of Environmental Protection, National Center for Environmental Assessment, Cincinnati, OH, U.S.A.
The Naugatuck River in Thomaston, Connecticut was biologically impaired for five miles and the cause was unknown. There were three industrial points sources, one publicly-owned treatment works (POTW), and three known sites of contaminated groundwater. Biological condition progressively deteriorated downstream with successive outfalls. Using the U.S. EPA Stressor Identification Guidance, four candidate causes were assessed: asphyxiation from low dissolved oxygen due to impoundment or algal growth, and toxicity from ammonia, metals, or a complex mixture. Biological impairments were near outfall locations, effluents were acutely toxic in laboratory tests, and Monte Carlo analysis identified whole effluent as the pollutant with highest probability of causing toxic impacts. Dissolved oxygen was above levels known to cause most biological effects, and nutrient levels were similar to background levels. Most ammonia concentrations were at background levels, and higher levels were not concurrent with measured toxicity. No individual metal concentrations were correlated with measured effluent toxicity and Monte Carlo analysis indicated a low probability of individual metal parameters causing toxic in-stream conditions. Furthermore, metal concentrations did not exceed criteria. As a result of the causal analysis, a Total Maximum Daily Load (TMDL) was developed for whole effluent toxicity (WET), and the National Pollution Discharge Elimination System (NPDES) limits for WET were revised for three permitees based on wasteload allocations in the TMDL. New NPDES limits for metals were also set for the three permitees and the POTW. The TMDL implementation plan provided guidance for clean-up levels at three groundwater impacted sites. WET tests provided critical evidence to support the identification of causal relationships and to complete the TMDL allocation for the Naugatuck River.

IP032 (ZIE-1117-750483) Causal analysis of biological impairment in Long Creek, a sandy-bottomed stream in coastal southern Maine.
Start time: 1:50 PM
Varricchione, J¹, Ziegler, C², Norton, S², Meidell, S¹, Schofield, K², ¹ Maine Department of Environmental Protection, Portland, ME, USA² EPA Office of Research and Development, Washington DC, DC, USA
Determining the cause of biological impairment can be a daunting task in complex cases involving multiple possible causes, differing biological effects at different sites, irregularly collected data, natural variability (e.g., seasonality, weather, and land formation) and lack of basic scientific information needed to diagnose environmental problems. This poster presents results from such a complex case: the Long Creek watershed in Portland, Maine. Long Creek is a low-gradient, sandy-bottomed stream that flows through an urbanizing commercial area. Different patterns of biological effects were observed at different sites in the stream, indicating that a site-by-site causal analysis would be required. The case study focused on three locations that did not meet state standards for macroinvertebrate community condition. Seven candidate causes were evaluated: altered basal food source, altered flow regime, increased temperature, increased fine sediments, decreased woody debris, decreased dissolved oxygen, and increased toxic substances. Available data relevant to each candidate cause were evaluated to determine whether candidate causes co-occurred with biological effects, and whether effects increased with increasing exposure. Where possible, literature on stressor-response relationships and organismal life history attributes were added to the stressor identification evidence base. The stressor identification process identified strong evidence for three of the seven candidate causes: altered flow regime, decreased dissolved oxygen, and increased toxic substances. The stressor identification guidance provided a useful structure for organizing the evidence available in this complex case. A lesson learned from this case study is that more evidence would have been available if data collection had proceeded in two stages: biological monitoring first to define the location of the biological effects, and a second stage linking the hypothesized causes with the effects. The stressor identification process would provide a sound basis for designing such a two-stage analysis.

IP033 (GER-1117-730131) Multiple exposure-response associations for inferring probable causes. Coal River, West Virginia.
Start time: 1:50 PM
Gerritsen, J¹, Bailey, J², Boschen, C³, Burton, J³, Lowman, B², Ludwig, J⁴, Wilkes, S⁴, Wirts, J², Zheng, L¹, ¹ Tetra Tech, Inc., Owings Mills, MD, USA² West Virginia Department of Environmental Protection, Charleston, WV, USA³ Tetra Tech, Inc., Fairfax, VA, USA⁴ Tetra Tech, Inc., Charleston, WV, USA
To infer causes of biological impairment in West Virginia streams, we examined macroinvertebrate community response to single and multiple stressors, using a statewide dataset. We developed exposure-response (dose-response) relationships between potential stressors and invertebrate assemblages - collected from a range of locations representing ecoregional reference sites to extremely impaired sites. The exposure-response models were used to define 2 thresholds: 1) below which few sites were biologically impaired; and 2) above which most sites were biologically impaired. We also used a similarity analysis approach to compare species composition at an assessment site to the composition of both minimally stressed and "dirty" reference site groups. The relative similarities were used to infer a ranking of stressors acting at a site. We integrated the empirical models of biological impairment with onsite field observations of biota, habitat, water quality, and watershed observations; using a strength of evidence approach to infer probable causes of impairment. Strongest inferences were obtained where the independent predictive models agreed with field observations of stressor measures. Final stressor determinations for each watershed will be used for the development and implementation of TMDLs.

IP034 (DIA-1117-729777) Excess sediment and elevated temperature are probable causes of biological impairment in theTouchet River, Washington.
Start time: 1:50 PM
Wiseman, C¹, LeMoine, M², Diamond, J³, ¹ Washington Dept of Ecology, Lacey, WA, USA² WWU Huxley College, Bellingham, WA, USA³ Tetra Tech, Inc., Owings Mills, MD, USA
The Washington Department of Ecology (Ecology) currently practices single entry Total Maximum Daily Load (TMDL) studies, where multiple water quality impairments in a watershed are addressed in a single sampling season. In 2002, a single entry TMDL occurred in the Touchet River, an agricultural (mainly wheat production) watershed in eastern Washington State, USA. In order to determine cumulative impacts and prioritize pollutants for implementation projects, benthic macroinvertebrate assessment information were examined along with chemical and physical habitat data from several sites on the Touchet River. A stressor identification (SI) approach was used to decipher likely cause and effect relationships between potential stressors and observed biological responses. Many expressions of biological condition declined from upstream to downstream sites, coinciding with several different and potentially correlated gradients including elevation, temperature, and surrounding agricultural land use stressors: nutrients, sediment, and pesticides. After reconciling chemical, physical, and geomorphologic results with the biological responses, some potential stressors, including toxics and algal-driven eutrophication, could be eliminated. Increased river temperature and sediment-driven habitat degradation were the most probable stressors causing deteriorating biological conditions based on several lines of evidence in the SI process including: (1) better biological condition where natural riparian vegetation was maintained and sediment loading was less; (2) temperature-tolerant species were correlated with temperature and not elevation and (3) other, more naturally vegetated rivers in the region at the same elevation, had much cooler water temperatures on average and correspondingly more cool water fish and macroinvertebrate species. These results indicate that managers should focus resources primarily on restoring shading and lowering sediment delivery to the river to restore beneficial aquatic life uses. Future biological surveys in this watershed will determine the effectiveness of TMDL implementation efforts.

IP035 (COR-1117-812575) Stream removed from 303d list of impaired waters: Willimantic River, Connecticut.
Start time: 1:50 PM
Bellucci, C¹, Cormier, S², Menzie, C³, ¹ Connecticut Department of Environmental Protection² U.S. Department of Environmental Protection, National Center for Environmental Assessment³ Menzie Curra and Associates
Self-monitoring reports submitted to the Connecticut Department of Environmental Protection by the Stafford sewage treatment plant indicated that the plant was probably exceeding Connecticut's Water Quality Criteria for copper, lead and zinc in the Willimantic River downstream of the discharge. Based on the self-monitoring reports, the Willimantic River was listed among the 1998 Connecticut Waterbodies Not Meeting Water Quality Standards. Subsequent biological monitoring confirmed the impairment but also revealed additional biological impairments upstream from the treatment plant. A screening level stressor identification analysis implicated an episodic exposure as a likely cause. Additional biological sampling isolated the location where the impairment began and a previously unknown point source was uncovered. The discharge from the point source was rerouted to the waste treatment plant. Subsequent monitoring over the next three years have shown progressively improving condition based on benthic macroinvertebrate sampling. This case is a good example of using an adaptive management approach within the stressor identification framework to solve environmental problems.