Captive breeding is a widely used strategy for buffering rapidly declining populations against short-term extinction risk. In the case of Pacific salmon and other exploited species, captive breeding takes the form of supplementation hatchery programs, which are intended to increase population size while minimizing artificial selection in breeding and rearing practices. Supplementation is an increasingly common tool for Pacific salmon conservation, but substantial uncertainty remains about its effectiveness and potential impacts on wild populations. Directly measuring these impacts is challenging, in part because of the difficulty of estimating the relative reproductive contributions of wild- and hatchery-reared individuals that breed naturally. To address this, we analyzed time series (13 - 26 yr) of adult density from 23 populations of spring/summer Chinook salmon (Oncorhynchus tshawytscha) in the Snake River basin, USA, which have experienced a range of supplementation levels (including no supplementation). We fit models that predict total naturally derived recruitment as the sum of offspring produced by wild- and hatchery-reared adults (these are distinguished in the abundance data). We compared alternative hypotheses about density-dependence or -independence and the equivalence of wild- and hatchery-reared fish with respect to two key parameters: productivity at low density and the per-capita strength of density dependence.

Results/Conclusions

Using a hierarchical Bayesian framework, we found the strongest support for a model in which wild- and hatchery-reared fish differ in both intrinsic productivity and density-dependent effects. At low density, hatchery fish are estimated to produce only 33% as many surviving mature offspring as wild fish, and the estimated carrying capacity of a population composed entirely of hatchery fish is 17% lower than that of an all-wild population. This asymmetry in population-dynamic attributes of the two types of fish suggests underlying differences in their ecology, behavior, or life history, despite the efforts by the supplementation program to minimize such divergence. Our results indicate that the introduction of hatchery-reared fish into wild populations may reduce productivity, and thus supplementation programs may face a trade-off between short-term increases in abundance and erosion of the long-term potential for population rebuilding.