Fish Habitat in Midwestern Streams: How Forecasting Tools Can Help Managers Prepare for a Changing Climate

We modeled fish response to climate and land use changes resulting from changes in stream temperature and flow.  The methodology behind our project has been used successfully over smaller regions elsewhere though our data coverage is much richer than previously available and the scale at which we forecast results is unprecedented.  We provide consistently-developed output for river reach in Michigan, Wisconsin, and Minnesota (~165,000 reaches).  Our main goals were to (1) develop neural network models to predict daily summer stream temperature; (2) develop state-scale models to predict exceedence flows; (3) develop fish distribution that link fish, water temperature regime, and flow regime; and (4) develop fish habitat scores that allow us to compare the fish assemblages, water temperature, and flow regimes under current and future climate and land-use conditions.  Our results include (1) predicted temperature and flow regime maps; (2) predicted distribution maps for important fish groupings; and (3) potential vulnerability ranking and fish species shifts.  The endpoint is a tool that will provide resource managers the ability to prioritize restoration and/or conservation efforts for valuable fish habitat based on the potential vulnerability of said habitats.