Search

SummaryAssessing the degree to which at-risk species are regulated by density dependent versus density independent factors is often complicated by incomplete or biased information. If not addressed in an appropriate manner, errors in the data can affect estimates of population demographics, which may obfuscate the anticipated response of the population to a specific action.We developed a Bayesian integrated population model that accounts explicitly for interannual variability in the number of reproducing adults and their age structure, harvest, and environmental conditions. We apply the model to 41 years of data for a population of threatened steelhead troutOncorhynchus mykissusing freshwater flows, ocean indices, and releases of hatchery-born conspecifics as covariates.We found compelling evidence that the population is under strong density dependence, despite being well below its historical population size. In the freshwater portion of the lifecycle, we found a negative relationship between productivity (offspring per parent) and peak winter flows, and a positive relationship with summer flows. We also found a negative relationship between productivity and releases of hatchery conspecifics. In the marine portion of the lifecycle, we found a positive correlation between productivity and the North Pacific Gyre Oscillation. Furthermore, harvest rates on wild fish have been sufficiently low to ensure very little risk of overfishing.Synthesis and applications.The evidence for density dependent population regulation, combined with the substantial loss of juvenile rearing habitat in this river basin, suggests that habitat restoration could benefit this population of at-risk steelhead. Our results also imply that hatchery programs for steelhead need to be considered carefully with respect to habitat availability and recovery goals for wild steelhead. If releases of hatchery steelhead have indeed limited the production potential of wild steelhead, there are likely significant tradeoffs between providing harvest opportunities via hatchery steelhead production, and achieving wild steelhead recovery goals.

To identify potential actions for conserving Chinook Salmon Oncorhynchus tshawytscha in the Skagit River, Washington, we used a 16-year time series of streamflow data, adult escapement, and out-migrant abundance to understand how out-migrant abundance and life history diversity were related to spawner abundance and incubation flows. Three freshwater rearing strategies were distinguished based on body size at out-migration: fry (≤45 mm FL), subyearling parr (46–100 mm FL), and yearling smolts (>100 mm FL). Density-independent and density-dependent processes were hypothesized to influence survival in sequence, with density-independent mechanisms operating during incubation and density-dependent mechanisms operating between emergence and out-migration. A model selection process compared spawner–recruit models with and without different incubation flow metrics. Density-independent models that included measures of flow duration and magnitude were strongly supported (Akaike's information criterion [AIC] difference ≤ 3). Sustained flow events of moderate magnitude (1-year recurrence interval) were an equivalent if not better predictor of freshwater survival than short-duration flow events of high magnitude (peak flows). A second model selection process evaluated density dependence of each life history type. The composition of out-migrants (fry, subyearling parr, and yearling smolts) was a density-dependent function of spawner abundance. Fry out-migrant abundance was density independent, and subyearling parr out-migrant abundance was density dependent. Neither model was supported for yearling smolts. At least one out-migrant life history, subyearling parr, should benefit from continued restoration of freshwater habitats in the Skagit River system. Factors contributing to the yearling smolt life history will benefit from additional study.Received October 12, 2014; accepted February 3, 2015