Temporal variability in fish assemblages from disturbed and undistrubed streams

Fish assemblages were sampled by electrofishing over a two- to ten-year period in undisturbed and anthropogenically disturbed South Carolina coastal plain streams. Jaccard similarity, Bray-Curtis similarity, and Spearman rank correlations among samples collected from the same sites over time were significantly greater at undisturbed sites than at disturbed sites, suggesting greater fish assemblage persistence and stability at the undisturbed sites. The Index of Biotic Integrity (IBI) also exhibited significantly less variation over time at undisturbed sites than at disturbed sites. Physical habitat structure changed more over time at disturbed sites than at undisturbed sites, and this variability was directly related to temporal variability in fish assemblage structure. Comparisons between multiple and single pass electrofishing samples suggested that only a small proportion of the temporal variability observed at the study sites was caused by inefficient sampling. Assessment of temporal variation in fish assemblage structure can serve as an indicator of environmental disturbance and facilitate the distinction of substantive ecological change from normal backgrourrd variation.