Spatial habitat dynamics affecting bluegill abundance in Minnesota bass-panfish lakes

A holistic approach to fisheries management requires an understanding of factors related to fish abundance over several spatial scales. We used geographic information systems to extract data describing habitat influences across three different spatial scales for a selected ecological class of Minnesota bass-panfish lakes (n = 113). These data were then analyzed by regression tree analysis to describe relationships between habitat and trap-net catch per effort (CPE) of bluegills Lepomis macrochirus. At the landscape scale of analysis, bluegill CPE increased among lakes with decreases in hydrologic connectivity (landscape position) and increases in geographic northing and easting gradients that corresponded to regional differences in geomorphology and edaphic characteristics. At the watershed-lake scale of analysis, a regression tree model with variables describing watershed area, cultivated land cover, forested land cover, and lake area explained 55% of the variation in bluegill CPE among lakes. At the site scale, a regression tree model with variables describing submerged plant cover and detritus substrates explained 57% of the variation in bluegill CPE among 72 sites spread among six lakes. However, much of the site-scale habitat influences on bluegill abundance was explained by broader landscape- and watershed-scale factors that influence the plant and bottom substrates in lakes. This study reinforces the importance of identifying habitat limitations and the influences Of human activities at the landscape and watershed scales in addition to more commonly addressed site-scale habitat deficiencies.