The response of fish and macroinvertebrate assemblages to multiple stressors: A comparative analysis of aquatic communities in a perturbed watershed (Onondaga Lake, NY)

• Fish and macroinvertebrate assessments of freshwater stream quality were compared.
• Differences in assemblage-specific metric scores were analyzed.
• Fish and macroinvertebrates responded differently to environmental stressors.
• Study scale is important for accurate ecological assessments.
• Assessments of water quality should include more than one assemblage.

Stream bioassessment studies usually use a single taxonomic group, most commonly fish or macroinvertebrates. However, differences in the life histories, habitat preferences, and physiological and behavioral traits of each assemblage may cause differences in their responses to environmental degradation. These factors can lead to discrepancies in overall assessments of ecological integrity. The Onondaga Lake Watershed (NY) has had a diverse history of pollution from multiple impacts, and the lake and several subsites are listed as USEPA Superfund sites. The response of fish and macroinvertebrate assemblages to environmental variables in the Onondaga Lake Watershed was assessed at 17 stream reaches in 2007, 2010, and 2011 using assemblage-specific metrics and indices of biotic integrity. Land use and substrate composition variables were significantly correlated (p < 0.05, r > |0.3|) with the first two macroinvertebrate PCA axes and were related to a downstream gradient of increased habitat degradation for the major tributaries. Water chemistry variables, canopy cover, and stream width were significantly correlated (p < 0.05, r > |0.3|) with the first two fish PCA axes. PCA axes for fish and macroinvertebrate assemblages were not significantly correlated, suggesting that the major environmental variables affecting fish and macroinvertebrate assemblages differed. Pearson correlations between the macroinvertebrate Biological Assessment Profile (BAP) and the fish Index of Biotic Integrity (IBI), and the macroinvertebrate Percent Model Affinity (PMA) and IBI were not significantly correlated. The macroinvertebrate metrics Percent Model Affinity (PMA) and BAP were positively and significantly correlated with Visual Habitat Assessment (VHA) scores, and the fish IBI was not significantly correlated with the VHA, further illustrating differences of fish and macroinvertebrates in their response to environmental variables. Results suggest that the use of multiple assemblages may be warranted for future assessments of stream quality and the development of effective monitoring programs in the Onondaga Lake Watershed.