Concordance among fish and macroinvertebrate assemblages in streams of Indiana, USA

The use of multiple taxa rather than a single taxon of stream organisms in ecological studies appears to be necessary to interpret independent environmental influences and interactions. We tested if macroinvertebrate assemblages in Indiana, USA, streams were better predicted from co-occurring fish assemblages or environmental variables. We used multivariate analyses to identify significant environmental predictor variables for macroinvertebrate and fish assemblages. Macroinvertebrate distribution and relative abundance were best predicted by in-stream cover and turbidity, and fish distribution and relative abundance were best predicted by fine sediments, bedrock, water temperature, and pool habitat. Patterns in fish assemblages were not significant predictors of macroinvertebrate distribution and relative abundance. Mantel tests for covariation among fish assemblage composition and macroinvertebrate assemblage composition resulted in significant, but low correlations. Our results suggest that macroinvertebrates respond to local environmental variation, and less to local presence of fishes indicating the surrogate taxa approach has little use at the Eastern Cornbelt Plain ecoregion of Indiana. Stream surveys of multiple taxa and environmental variables are appropriate assessment methods for ecosystem integrity.