Testing the Severity of Ill Effects Model for Predicting Fish Abundance and Condition

Abstract

We tested the severity of ill effects (SEV) model as a predictor of the potential impacts of suspended sediment in small streams in an agricultural region of New Brunswick. Proposed by Newcombe and Jensen, the SEV uses the concentration of suspended sediment and the duration of exposure to predict the impacts on fish. In 2007, eight streams with brook trout Salvelinus fontinalis and slimy sculpin Cottus cognatus were used to test the SEV model. Total suspended solids measured from May to August were used to calculate the SEV index using two different methods. Our analyses indicated poor correlations between the SEV index and measures of fish densities and condition factor. A survey of an additional eight streams in the region over 2 years indicated that local habitat and landscape features and temperature were better predictors of fish density and species occurrence. These results demonstrate the complexity of interpreting stressor impacts and the challenges inherent in developing indices of stress for fish in natural streams.

Received January 20, 2010; accepted January 28, 2011

ACKNOWLEDGMENTS

We are grateful to Mark Gautreau and many assistants for help in the field collections. We would like to thank J. Culp, G. Benoy, A. Sutherland, E. Luiker, R. Allaby, and D. Hryn, of Environment Canada and Agri-Food Canada for our environmental data. We thank B. Brua, W. Monk, A. Alexander, K. Heard, and D. Drolet for help with data analyses, and W. Siddall for assistance in the map creation. Reviews by K. Munkittrick and the journal's editors were very beneficial.