Biological, environmental, and geomorphological factors influencing reach-specific survival of spring Chinook Salmon smolts upstream of the Columbia–Snake River hydrosystem

Abstract Objective Pacific salmon Oncorhynchus spp. are exhibiting catastrophic declines throughout much of the Pacific Northwest, and juvenile mortality contributions are disproportionately high. Many Pacific salmon smolt populations are exhibiting high natal stream mortality; however, detailed natal stream reach-specific survival knowledge is lacking. Our objectives were to estimate natal stream reach-specific survival and identify influential covariates for spring Chinook Salmon O. tshawytscha smolts. An additional objective was to evaluate (i.e., estimate postrestoration cumulative survival) a novel and strategic large-scale river restoration approach (i.e., stream confluence restoration) in two Pacific Northwest spring Chinook Salmon natal streams. Methods Using radiotelemetry techniques during March–June (2011–2017), we estimated natal stream reach-specific survival for spring Chinook Salmon smolts in Catherine Creek and the Grande Ronde River, northeast Oregon. We examined the interrelated influences of temporal, biological, environmental, and geomorphological covariates on the behavior, travel time, and reach-specific survival of spring Chinook Salmon smolts that were emigrating through two hydrologically altered natal streams in the interior Columbia River basin. For each natal stream smolt tag-group, Cormack–Jolly–Seber reach-specific survival estimates that were adjusted for premature tag failure (i.e., bias-corrected) were generated. Subsequently, smolt effects were estimated using reach-specific travel times and survival estimates coupled with individual time-varying covariates from radio-tagged smolts and occupied reaches using an existing predator–prey model (i.e., mean free-path length [XT] model). Complementary population-specific principal coordinate analyses (PCoAs) were conducted to facilitate visualization of multicovariate resemblances in ordination space. Last, we employed our best-fitting XT model to model cumulative changes in smolt survival that were associated with a large-scale stream channel restoration scenario (i.e., restoration of the historical Catherine Creek and Grande Ronde River confluence). Results In aggregate, Catherine Creek smolts exhibited high mortality throughout Catherine Creek but near-100% survival upon entering the Grande Ronde River. In contrast, Grande Ronde River smolts generally exhibited low mortality upstream from the Grande Ronde Valley and high mortality throughout the Grande Ronde Valley. Our best-fitting XT model indicated that smolt survival was positively correlated with discharge, body size, and current velocity but negatively correlated with avian predation. Natural-origin smolts from Catherine Creek displayed higher λ-values (i.e., distance between predator–prey encounters) than their significantly larger and faster emigrating hatchery conspecifics, indicating that hatchery smolts may be more susceptible to predation. Our population-specific PCoAs further revealed that novel water (i.e., unique water chemistry) and emigration rate were the strongest predictors of reach occupancy in multivariate ordination space. Under a stream restoration scenario of restoring the historical hydrological template (i.e., restored channel configuration), our best-fitting XT model predicted changes in population-specific cumulative survivals to the lower Grande Ronde Valley (Imbler, Oregon) ranging from −38.4% to 69.0% for Grande Ronde River and Catherine Creek spring Chinook Salmon smolts, respectively. Conclusions Our best-fitting XT model and PCoAs identified the relative importance of biological, environmental, and geomorphological factors to both natural- and hatchery-origin Chinook Salmon smolt survival in two neighboring interior natal streams in the Columbia River basin. Our research indicates that concurrent habitat restoration and fish management strategies, including focused attention on smolt-rearing nursery habitat restoration, piscivorous avian control plans, adaptive management strategies for hatchery smolt releases, and discharge and novel water regime restoration projects, may yield survival benefits to “in-basin” Chinook Salmon smolts. Our research can inform origin-type-specific Chinook Salmon smolt management, habitat restoration, and future research decisions throughout the upper Grande Ronde River subbasin and potentially the Pacific Northwest.