Factors influencing fish distributions in shallow lakes in prairie and prairie-parkland regions of Minnesota, USA

Abstract

Fish exert strong influences on shallow lakes, but managers lack empirical models useful for predicting fish distributions at landscape scales. We used classification and regression tree analysis (CART), and regression to predict fish presence/absence (P/A), richness, and community composition in 82 shallow lakes distributed among two regions (prairie and prairie-parkland) along the eastern margin of the Prairie Pothole Region in western Minnesota, U.S.A. A CART model for fish P/A using downstream connections to fish sources and maximum depth correctly classified ≥92% of our study sites, indicating the rare fishless sites observed in our study were either isolated or shallow. Fish richness was positively related to both lake and watershed size. Given that many fish species have strong negative influences on shallow lake ecological characteristics, we conclude that future conservation efforts should focus on protecting shallow, isolated basins, or reducing surface connectivity among basins as these factors were decisive in promoting fish populations. Such management strategies should help to maintain current levels of fish richness and enhance richness of aquatic birds, amphibians, plants, and invertebrates. © 2010 Society of Wetland Scientists.