|Document Type:||Journal Article|
|Title:||Exploring latent mortality of juvenile salmonids related to migration through the Columbia River hydropower system|
|Author:||Benjamin P. Sandford, Richard W. Zabel, Lyle G. Gilbreath, Steven G. Smith|
|Journal:||Transactions of the American Fisheries Society|
|Keywords:||latent mortality, salmon, survival analysis, modeling, weibull, Kaplan-Meier, Columbia River, stress, barge transportation, hypass system|
The ability to manage anthropogenic actions that affect the dynamics of animal populations requires the identification and understanding of life–stage–specific mortality. This understanding can be confounded when the expression of mortality is removed, in time or space, from its cause. For years, researchers studying endangered Snake River springsummer Chinook salmon Oncorhynchus tshawytscha have debated the magnitude of mortality that is related to—but expressed after—passage through the Snake and Columbia River hydropower system (latent mortality). We conducted experiments with Chinook salmon to assess the magnitude of latent mortality from two sources: passage through juvenile bypass structures at dams, and transportation with larger juvenile steelhead O. mykiss present in the barge holds. Nearly 129,000 juvenile Chinook salmon (passive integrated transponder tagged) were exposed to different treatment conditions during downstream migration. Study fish were then held in seawater tanks for up to 223 d, and time to mortality was noted for each individual that died. We analyzed survival patterns by using statistical procedures for time–to–event data (i.e., survival analysis). Differential survival between treatment groups was taken to indicate latent mortality caused by the specific treatment. We used a nonparametric Kaplan–Meier analysis to visualize survival patterns and a parametric logistic regression analysis to model the effects of multiple factors. Chinook salmon that were transported with steelhead had significantly lower survival than those that were transported alone. However, there was little evidence for differential latent mortality between Chinook salmon that were not detected at the bypass systems of five dams along the migration route and those that were detected at two to five bypass systems. Our application of survival analyses to individuals subjected to various treatments and held for extended periods produced an effective combination that can be used to test for latent mortality; these results may serve as an initial assessment for further conservation investigations and as a guide to more targeted research.