Response of taxonomic groups in streams to gradients in resource and habitat characteristics

Compared to other ecosystem types, a wide variety of biomonitoring methods are available for rivers. Few studies have, however, compared the response of bioindicators (i.e. different river inhabiting taxonomic groups) to different types of environmental stress. We regressed diversity, taxonomic distinctness and assemblage composition of fish, invertebrates, macrophytes and benthic diatoms to multivariate gradients in nutrient enrichment and habitat degradation using data from 66 lowland and 77 mountain streams. In lowland streams, the composition of benthic diatom assemblages, measured by Detrended Correspondence Analysis, showed the strongest response to elevated nutrient concentrations (R(2) = 0.830), followed by macrophytes (0.711), fish (0.443) and invertebrates (0.391). Macrophyte richness was slightly better correlated with the habitat gradient (0.253) than diatom richness (0.235) and both were better predictors than either fish (0.147) or invertebrate (0.140) assemblage composition. For mountain streams, invertebrate assemblage composition was the best predictor of changes in nutrient concentrations (R(2) = 0.749), followed by macrophyte (0.396) and benthic diatom (0.325) assemblages and fish diversity (0.099). Invertebrate assemblage composition was also slightly better correlated with the habitat gradient (0.391) compared to macrophyte richness (0.323) and both were better than either benthic diatom assemblage composition (0.206) or fish abundance (0.161). Macrophyte and benthic diatom assemblages in lowland streams and fish and invertebrate assemblages in mountain streams were correlated with the other taxonomic groups, implying that these taxonomic groups may be used as surrogates for indicating wholesale change in diversity. Synthesis and applications. Our findings show that response trajectories differ between taxonomic groups and stressor, and even with stream type. For this reason, streams and other ecosystem type's response signatures and first and second principle relations should be considered in the selection of robust, complementary and cost-effective measures for biomonitoring. For instance, our findings of asymmetric response to stress show how different taxonomic groups can be used to strengthen inference of change.