Vagrancy of Hatchery-Stocked Steelhead Trout in Lake Erie

Pacific salmonids are noted for their long-distance ocean migration and then high fidelity to natal spawning locations. Olfactory use of chemical cues has been clearly determined in these stocks. In Lake Erie, most steelhead trout (Oncorhynchus mykiss) smolts are produced in hatcheries (>2 million annually) using water very different (watershed) from the stocking locations.  Additionally, stocked tributaries vary considerably in size, and fish are released into tributaries at varying distance from confluence with the Lake (i.e., <2 to >40 km), so residence time in stocked tributaries is potentially variable.  We hypothesized that fidelity (and the converse, vagrancy) could be lower (higher) in small tributaries where steelhead trout smolts were released close to Lake Erie.  We first determined the stock identity of steelhead trout smolts using unique otolith chemical signatures (Sr, Ba, and Pb; laser-ablation-inductively-coupled-plasma-mass-spectrometry; LA-ICPMS ) of the hatchery and naturally reproduced smolts to correctly identify fish source with high accuracy (90-100%; discriminant analysis, overall MANOVA: Wilk’s λ F_27,427 = 78.91; P < 0.0001). From these results, we then determined the within-stream stock distribution of spawning steelhead trout (~500) in up to six Lake Erie tributaries (West to East; Huron MI, Vermilion OH, Conneaut OH and PA, 16-mile PA, Chautauqua NY, and Cattaraugus NY) from both fall- and spring-run adults.  Preliminary fall-run data suggests strong evidence of homing of returning adult steelhead trout in some systems (8% vagrants, N = 49), but much higher proportions (>60%, N = 79) of vagrants in other tributaries.  Most of these vagrants come from Pennsylvania releases where most fish are stocked into small tributaries (and released relatively close to Lake Erie) with short expected residence time after stocking.  Although not conclusive, our data suggest that non-return to release sites (vagrancy) is somehow coupled with residence time in the release tributary.  Given the costs and benefits associated with this fishery in Lake Erie (and elsewhere), further understanding the sensory systems used by salmonids to exhibit site fidelity and timing of information acquisition is warranted. Stock discriminating tools like otolith chemistry signatures (generating hatchery-specific signatures) should be utilized along with other techniques to determine optimal salmonid release strategies.