Rivers and their accompanying riparian ecosystems are hydrologically linked to the surrounding landscape at a variety of scales.  For example, upstream and upland land cover and use have documented impacts on aspects of water chemistry that can affect fish and benthic macroinvertebrates (BMIs). Recent studies indicate the importance of cover type at a watershed scale, while hinting at potential trade-offs at the riparian level.  Determining which environmental conditions are most significant for fish and components of the BMI community, as well as the relationship between fish and BMIs under different conditions, can contribute to developing effective strategies for conservation of these diverse and vulnerable systems. 

Fish and benthic macroinvertebrates were sampled over the course of 1-4 years from a total of ten riffle sites on four rivers in Connecticut, along with basic water chemistry and flow.  Frequency was 2x-5x/year.  Fish were collected through electrofishing and identified to species; BMIs were collected in rock bags and Surber samples and identified to family.  Sites were characterized by a variety of habitat and landscape variables including substrate composition, bank vegetation, canopy, and stream width.  GIS was used to determine impervious surface and land cover at a variety of scales within the watershed.
Results/Conclusions

Environmental conditions relevant to fish and BMI communities are different, but overlapping, and results imply a bottom-up effect of BMIs on fish abundance. 

Sites were significantly different from one another with respect to the suite of environmental conditions measured and BMI and fish taxa that characterized them.  The BMI community was restricted to hydropneustic taxa; a total of 83 families were identified for the study period.  Most significant correlations with environmental conditions were for segments of the community identified as sensitive to water quality.  Confirming this designation, EPT taxa (insects orders Ephemeroptera, Plecoptera and Trichoptera) and a subset (“most sensitive”), were significantly correlated with percent watershed forested and developed, for example.  Path analysis revealed differences between manifest variables affecting these components of the BMI community, though the percent developed was consistently negatively correlated.

Fish abundance and richness explains surprisingly little of the variation in BMI abundance or richness. Variability in fish abundance and richness can be explained by a different, though overlapping, suite of environmental conditions, including the degree to which the watershed is forested and developed, and the abundance of EPT taxa.