Development of a Framework for Assessing Fish Habitat in the Midwest Glacial Lakes Partnership

Characterizing lake resources and assessing their condition is critical for developing effective management and conservation strategies.  Towards that end, the Midwest Glacial Lakes Partnership has developed a framework to better understand lake resources and to prioritize fish habitat efforts.  First, we characterized the landscape context of over 40,000 lakes across multiple spatial scales.  Next, we developed a model to predict lake surface temperatures, which allowed us to characterize a key habitat variable structuring lakes.  We then used boosted regression trees to predict presence-absence and abundance of selected lake fishes based upon landscape variables.  Model accuracies ranged from 60 to 80% and included variables that characterize lake position, size, connectivity, and thermal regime.  We used these models to estimate statewide patterns of species distribution and abundance.  We also used a classification scheme developed for Michigan to classify all lakes in the partnership into one of six fish assemblage types.  Finally, we developed a ranking system to prioritize habitat management in lakes based on lake type, species abundance, landuse in the catchment, proximity to public lands, and geographic location.