Macroinvertebrate drift patterns and fish habitat use were examined in the River Raisin watershed, southeastern Michigan during the summer of 1992. Three study streams of similar drainage area, slope, and location were chosen for their contrast in instre am habitat features. Macroinvertebrate drift samples were collected during the day and night at sites in the Raisin River, Iron Creek, and Evans Creek. Stream fish assemblages at these sites were sampled by means of alternating current (AC) electrofishi ng grids prepositioned in the streambed, in order to record habitat preferences (depth, current, habitat type) of given fish species. In this study, diel patterns of macroinvertebrate drift were consistent with the risk of predation hypothesis (Allan 1978). The genera Tricorythodes, Caenis, and Gammarus demonstrated patterns of individuals drifting at night being larger and more abunda nt than individuals drifting during the day. A greater number of taxa were collected in the night drift than in the day drift in all three study streams. The stream fish faunas were comprised of visually foraging invertivores which could be expected to exert some predation pressure on macroinvertebrates. The use of drift sampling was shown to have advantages over gravel basket collections at the same streams, in that a greater abundance and variety of surface and benthic taxa were collected.

Significant differences were seen in the fish assemblages and habitat use at the three study sites. Raisin Rt. 50 had the greatest number of species, percent of intolerant species, and trophic diversity. The fish community of the channelized Evans Creek was dominated by more sediment-tolerant species, and smaller individuals of species with more streamlined body morphologies. Overall biomass was highest in Evans Creek, but the average weight per fish was approximately one third of that found in Raisin Rt. 50 and Iron Creek. Whereas fish in Raisin Rt. 50 and Iron Creek were fairly evenly distributed between available habitat types (riffles, runs, undercut banks, pools with and without cover), 92% of fish in Evans Creek were found in pools. Habitat sim plification at Evans Creek reduced the ranges of depths, current velocities, and habitat types available for fish use.