Landscape based fisheries classification for Michigan lakes

We present a landscape based approach to classify lake types for fisheries management.  Using multivariate regression tree modeling and data from a subset of 200 lakes, we identified lake types based on fish species assemblages and determined the landscape features that structure these types.  Using these fish assemblage/landscape relationships, we then classified all 10,000 lakes in Michigan.  Multivariate regression tree modeling identified seven fish assemblage types that could be classified based on length of the growing season, mean water temperature, and lake surface area.  Lakes having a longer growing season were dominated by warmwater species whereas lakes having a shorter growing season were dominated by coolwater species.  Smaller lakes were dominated by sunfish whereas larger-bodied species such as walleye and white suckers were more prominent in larger lakes.  Species composition among lakes having similar growing season length and surface area differed based on mean summer temperature.  Our approach can be used to efficiently identify the fisheries potential for many lakes across broad regions.  This information is critical for developing statewide sampling programs, quantifying the amount and location of different resource types, and for determining appropriate management action on individual waterbodies.