A test of watershed classification systems for ecological risk assessment

To facilitate extrapolation among watersheds, ecological risk assessments s hould be based on a model of underlying Factors influencing watershed respo nse, particularly vulnerability. We propose a conceptual model of landscape vulnerability to serve as a basis for watershed classification systems to predict resistance and resilience of aquatic ecosystems to hydrology-relate d stressors. Watershed area, storage capacity, channel slope, and soil perm eability determine sensitivity of lotic systems to stressors associated wit h land-use activities that impact hydrologic regimes. Natural hydrologic di sturbance regimes also influence the resilience of aquatic systems by selec ting for life history strategies associated with rapid recolonization follo wing disturbance. Variability in some of these physiographic driving factor s can be partitioned by landscape classification schemes such as the U.S. F orest Service Ecological Unit Classification System, while others (watershe d storage) may explain remaining variability within landscape units. We are conducting a comparative watershed study to examine simple and interactive effects of physiographic units, watershed storage (lakes + wetlands), and land-clearing activities in watersheds surrounding the western arm of Lake Superior. Initial results for second-order watersheds indicate significant watershed class effects on baseflow water quality, percent motile biraphid diatom species in periphyton communities, habitat quality, and fish communi ty integrity. Future studies have been designed to examine cumulative effec ts downstream.