You are here:
Differentiating Impacts of Watershed Development from Superfund Sites on Stream Macroinvertebrates
Detenbeck, N., C. Rosiu, L. Hayes, AND J. Legros. Differentiating Impacts of Watershed Development from Superfund Sites on Stream Macroinvertebrates. U.S. EPA Office of Research and Development, Washington, DC, EPA/600/R-13/352, 2014.
This report presents the results of an EPA Region 1 RARE project (The use of EMAP and NCA data to refine an index of urban intensity for New England, and an application of the index to improve measures of aquatic community impairment). The article describes the analysis of historical macroinvertebrate community data compiled from New England sources along urbanization gradients. These analyses helped to establish the most appropriate predictor and response indicators for detecting urbanization effects on macroinvertebrate communities. Results demonstrated that the analytical approach could be used to distinguish the effects of diffuse nonpoint-source-pollution from upstream development from those of site-specific contamination (e.g., Superfund sites), thus allowing the success of remediation activities at those sites to be tracked.
Urbanization effect models were developed and verified at whole watershed scales to predict and differentiate between effects on aquatic life from diffuse, non-point source (NPS) urbanization in the watershed and effects of known local, site-specific origin point sources, contaminated sediment sites. Using U.S. Environmental Protection Agency (US EPA) Environmental Monitoring and Assessment Program (EMAP) data from the New England Wadeable Stream Survey and datasets from States of Maine (ME) and Connecticut (CT), we derived macroinvertebrate community response curves for watersheds with different levels of urban development (n = 731 watersheds). Taxonomy, resolution, and indicator values were standardized across datasets to facilitate watershed comparisons. We classified study sites by Omernik Ecoregion, USFS Ecological Unit, Nature Conservancy Aquatic Habitat Class, and hydrologic regime class based on predicted peak and low flow statistics to compare differences in sensitivity of response. Within the hydrologic regime classes, six peak-flow classes and three low-flow classes could be distinguished based on watershed attributes. We applied Non-Metric Dimensional Scaling ordination and Indicator Analysis of macroinvertebrate community metrics to narrow down macroinvertebrate community endpoints to a subset explaining most of the variation within each dataset. We applied boosted regression trees (BRT) to develop response models, allowing us to simultaneously differentiate interactions among variables and successfully identify biological effect thresholds. The best predictors of watershed development impacts were percent Impervious Area (%IA) at the watershed- (NEWS, CT) or local- (ME) scale and percent High Density Residential Area in the stream buffer. When these indicators were seen operating at both watershed and local scales, they tended to have synergistic (more than additive) effects. Loss of dispersal corridors is expected to have an even greater impact in streams within high %IA watersheds and frequent hydrological disturbance, leading to an interaction between riparian buffer effectiveness and %IA impacts. For the first time, we were able to demonstrate the effects of road density and road-stream crossings independent of impervious area effects. We also demonstrated declines even in aggregate community metrics at very low levels of urbanization (<1 – 2% IA), once effects of moderating variables had been factored out. Percent forested riparian buffer was a significant moderating influence on impacts, with sensitivity modified by watershed area, slope class, ecounit (NEWS), and low flow class. BRTs were powerful enough to discriminate local impacts (Superfund contaminated sediment sites) from upstream development with 95% confidence, once stressor-specific indicators were incorporated and upstream Superfund site effects accounted for. Our study is the first published case demonstrating the cumulative effects of upstream Superfund sites on stream macroinvertebrate community composition at the whole watershed scale, and distinguishing these effects on aquatic life from those of urban development in the watershed. Application of these urbanization effects models with watershed bio-monitoring data in New England offers potential as diagnostic tools for assessment of in-stream biological condition differentially sensitive to point source and NPS pollution. The purpose of the assessment may be to identify streams where biological impacts are greater than predicted for the level of watershed development (identify potential site point sources) or as measures of the effectiveness of site remediation or restoration, or watershed best management practices (BMPs).