Anthropogenic landscape and climatic disturbances translate into in-stream stresses that can be described by changes in the chemical composition of river water or sediment, in-stream or riparian habitat, hydrologic and hydraulic characteristics, and food (energy) sources. The purpose of this research was to develop a regional, layered, hierarchical model system linking anthropogenic stresses with biotic integrity measures based on probabilistic risk propagation. Research focused on describing and quantifying effects of selected in-stream stresses on macroinvertebrate communities using an intermediate layer of risks associated with water quality, sediment quality, and habitat. A relational database, STARED, developed in the earlier phase of the research to facilitate analyses of relationships between physical, chemical, and habitat parameters of a water body and its biological integrity, was enhanced with built-in queries identifying selected monitoring stations with data from different categories (e.g., water chemistry or habitat), as well as extracting the necessary data to calculate biological indexes. An automated procedure was developed to match stations to characterize stream reaches using Geographical Information Systems (GIS). Available habitat data, as well as water and sediment concentrations for selected constituents (copper, lead, and zinc), represent measurable responses of anthropogenic disturbances and stresses on watersheds and streams. Ecological risks to aquatic biota associated with these responses were quantified for individual stressors using a probabilistic approach, calculating a joint probability of environmental conditions and exposure effects associated with these conditions. Effects of environmental variables on biotic indexes were then quantified using multiple regression analysis with backward selection. Two biotic indexes using information on macroinvertebrate communities were used: the Macroinvertebrate Biotic Index (MBI) representing tolerance indexes and the Invertebrate Community Index (ICI) representing multi-metric indexes. Both direct effect through environmental variables and indirect effect through risk variables were investigated. In northern Illinois, the direct effect of environmental variables resulted in more significant multiple regression equations, explaining higher percentages of variability in data than when using risk variables. In all cases, less than 55% of the variability in biotic indexes or metrics was explained, but all relationships were statistically significant. The individual regression equations investigating direct effects had three common variables: concentration of copper in sediment, stream width, and percent substrate as medium gravel. Sediment toxicity dominated by the risk associated with copper concentration also appeared in the two regression equations investigating indirect effect on biotic indexes and in seven out of 10 regression equations investigating indirect effect on ICI metrics. Risk to filter-feeding macroinvertebrates due to percent substrate as clay was also common to both regression equations investigating indirect effect on biotic indexes. Risk to scrapers due to aquatic vegetation was found important in three ICI metrics despite its very low variability. The highest correlations among the ICI metrics were achieved for those involving mayflies and caddisflies. The scope of this study included only variables identified as significant in previous study, whether related to habitat, flow regime, or chemical composition of water or sediment. Inclusion of additional habitat characteristics, such as percentages of canopy cover, brush debris jams, terrestrial vegetation, and rock ledge, is recommended to further increase the explanatory power of multiple regression equations as these characteristics appear significant in northern Illinois. Original variables were selected based on data from Wisconsin.
|Name||ISWS Contract Report|