Microhabitat-scale numerical and behavioral responses by juvenile salmonids to instream habitat restoration

Background/Question/Methods

In-stream habitat restoration is a commonly employed technique locally, regionally, and globally for the recovery of stream fish populations; however relatively few of these restoration projects is followed by monitoring to determine the project’s effectiveness.  In the Entiat River (Washington, USA), in-stream habitat restoration structures were installed to enhance juvenile rearing habitat, particularly for Chinook salmon (Oncorhynchus tschawytscha) and steelhead trout (O. mykiss).  To complement large-scale monitoring in this basin, we examined small-scale fish density, growth, and movement for both species.  We studied patterns of fish density in a reach treated with structures and in similarly sized microhabitats that lacked structures during three study seasons (2009-2011).  We compared the effect of treatments on fish density through 1) spatially random censuses of abundance within a treated reach compared with abundance at structures, 2) comparisons of abundance at structures with random censuses taken in 3-4 other reference reaches, and 3) comparisons between a fixed set of untreated pools in a reach adjacent to the treated reach. We used linear models to examine physical habitat correlates of fish density and estimated growth and habitat affinity for both treated and untreated microhabitats using mark-recapture techniques.

Results/Conclusions

Several patterns emerged and were consistent across multiple comparisons of fish density. In each study year, Chinook salmon were substantially more abundant at microhabitats with structures than in those that lacked structures; however this difference in density declined during the course of the rearing season due to factors likely unrelated to habitat.  Steelhead density lacked consistent patterns and was frequently higher in microhabitats lacking structures.  Chinook salmon responded positively to the increase in depth created by the structures, as indicated by multiple regression models, whereas steelhead showed inconsistent relationships with physical variables both within and between seasons.  There were species and habitat differences in both growth and habitat affinity.  Both species were recaptured more frequently in microhabitats with structures, even when steelhead density was highest at untreated microhabitats. Steelhead also showed density dependence in habitat affinity whereas Chinook salmon density was too low in untreated microhabitats to have a significant effect.  Low recaptures of Chinook salmon in untreated microhabitats made a comparison of growth intractable, but parameterization of the von Bertalanffy growth model showed subtle differences between microhabitats in parr-to-smolt stage growth.  Steelhead growth was higher in treated microhabitats, despite higher density in those lacking treatments, suggesting that impacts to fish populations beyond numerical responses are detectable at a small scale.