Catch-per-unit-effort and size spectra of lake fish assemblages reflect underlying patterns in ecological conditions and anthropogenic activities across regional and local scales

Across broad geographic scales, ecological indicators for fish assemblages should represent causal ecological processes, be sensitive enough to show patterns across the landscape, and reflect underlying biotic or abiotic conditions that influence those patterns. We assessed the responses of commonly applied ecological indicators for lake fish assemblages (mean body size, catch-per-unit-effort (CPUE), and normalized length size spectrum (NLSS) slope) to regional (climate, water chemistry, and watershed stress due to human activities) and local (lake morphometry, water quality, and angling pressure) ecological and anthropogenic variables. The indicators were estimated using fish assemblage catch data acquired via a standardized gillnetting protocol implemented within 693 lakes in Ontario, Canada. To our knowledge, our study is the first size-based or catch-based indicator evaluation to include detailed observations of angling pressure on hundreds of inland lakes. Boosted regression tree models showed that CPUE of large-bodied organisms and NLSS slope best described underlying patterns in the regional and local variables. Models developed with a mix of regional and local variables performed better than models developed with regional or local variables alone. The relative influences of the variables and responses varied among indicators, but in general, ecological variables had greater influence on the indicators than anthropogenic variables. These results emphasize the complex and multiscaled nature of factors and ecological processes affecting body size, habitat-community production, and trophic dynamics in lake fish assemblages.