Influence of Juvenile and Adult Experiences on Tributary Overshoot and Fallback by Steelhead in the Columbia River Basin

Abstract

Tributary overshoot occurs when adult fishes homing to natal sites continue upstream past the mouth of their natal stream. Using multistate release-recapture models, I examined the prevalence of overshooting and fallback to home by 37,806 PIT-tagged steelhead from 14 tributaries of the Columbia River basin in the years 2005—2015. For stocks that overshot at rates > 5%, I used generalized linear models and conditional inference trees to examine the influence of juvenile and adult experiences on overshooting and fallback to home. More than 40% of adult John Day, Umatilla, Walla Walla, Wenatchee, and Tucannon river steelhead overshot upstream dams. Average annual fallback rates to home ranged from 17.8% (SE 1.9%) for Walla Walla hatchery steelhead to 75.0% (SE 2.6%) for Umatilla wild steelhead. Overshooting was associated with factors related to the reservoir environment. Incorrect shoreline orientation within 24 rkm of the natal stream resulted in increased overshooting. Steelhead were also more likely to overshoot when water temperatures were higher (significant in 5 of 7 tributaries, P < 0.05). In contrast with adult experiences, juvenile experiences linked to imprinting disruption did not consistently increase overshooting. Hatchery steelhead were more likely to overshoot and less likely to fall back home than their wild counterparts. However, overshooting was only elevated in hatchery stocks reared upstream of release sites. Attraction to upstream areas was decreased with endemic broodstock and long-term acclimation. Juvenile barging was found to decrease overshooting relative to in-river out-migration. Longer ocean residency was associated with increased overshooting, but effects were biologically small. Finally, there was some evidence that spill during March increased fallback to home. A significant positive effect of spill during March was found for hatchery (χ28 = 15.82, P = 0.032) but not for wild steelhead (χ210 = 12.87, P = 0.231).