Estimating ecological risk to aquatic biota from physical and chemical impairment due to urbanization

Alena Bartosova, Marquette University

Abstract

Urbanization has a negative impact on ecological integrity of streams. Increased human activities affect chemical as well as physical characteristics of urban streams. The research has focused on risk-based methods of evaluating ecological integrity. The ecological risk is expressed in terms of the loss of species. The methodology of estimating risks from chemical impairment has been refined to incorporate the effect of hardness on toxicity of heavy metals. Various theoretical distributions have been fitted to the probability of organisms being impacted. The methodology has also been modified to accommodate effects of sediment contamination assuming the exposure to benthic species through interstitial pore water. The main contribution of this dissertation lies in developing a methodology for estimating risks imposed on aquatic biota through alteration of physical habitat. The proposed methodology is based in distributional ecology using the Maximum Species Richness (MSR) concept. Measures of species richness (the number of mayfly taxa, the number of caddisfly taxa, and the total number of taxa) are significantly affected by selected habitat characteristics (% clay in substrate, rubble or smaller substrate, % aquatic vegetation, average unit area flow). The MSR line determines the maximal species richness expected for given habitat conditions. The percentage of species living under specific habitat conditions then determines the community probability of survival. Data collected by the Department of Natural Resources have been used to develop and test the methodology. Statistically significant effects of urbanization have been documented on water temperature, aquatic vegetation, substrate composition, flow regime, and the multi-metric Michigan Index of Biotic Integrity (MIBI). The individual risk components have been calculated for the Oak Creek and the Menomonee River watersheds. Various combined regression models describe 55-98% variability in the MIBI. The proposed methodology represents the first attempt to numerically describe the risk from physical habitat alteration to aquatic biota. Further research is necessary to refine the method and to include additional habitat variables, uncertainties and variability in these variables, and to test the method on additional sites and watersheds. The method can be eventually used in watershed management, translating habitat indices into numbers comparable with other engineering evaluation methods.

This paper has been withdrawn.