Abstract
The robust redhorse Moxostoma robustum, originally described by Edward Cope in 1870 from specimens collected from the Yadkin River, North Carolina, apparently went unnoticed until 1991 when it was rediscovered in the lower Oconee River, Georgia. Despite extensive surveys and an ongoing, decade-long restoration program, wild-spawned juveniles 30–410 mm in total length have not been collected. This project experimentally evaluated one hypothesis (flow class use) that seeks to explain the absence of juvenile robust redhorses from the catch. Two experimental mesocosms were used to determine whether juvenile robust redhorses use flow classes in proportion to their availability. Pond-reared juveniles were exposed to four flow-based habitat types (eddies = −0.12 to −0.01 m/s, slow flow = 0.00–0.15 m/s, moderate flow = 0.16–0.32 m/s, and backwaters) in four 10-d trials, in which 16 pond-reared test fish were used per trial, with replacement. Location data were recorded hourly during daytime hours for each fish in all trials and evaluated with a log-linear, chi-square model. In winter, test fish showed a preference for eddies and backwaters and avoided slow to moderate flows. In early spring, test fish showed a preference for eddies and avoided the moderate flows. Current field sampling for juvenile robust redhorses has not targeted the flow classes used by fish in this experiment; however, collection of wild-caught juveniles may be improved by sampling in eddies and their associated transitional areas.