Using Stochastic Life Cycle Models to Evaluate the Potential Response of Chinook Salmon Populations to Recovery Actions That Include Hatchery Supplementation

Stochastic life cycle models can be valuable tools for evaluating potential population performance under alternative future management and environmental scenarios.  We have developed stochastic life cycle models for a set of Columbia Basin Chinook populations to assess the projected effects of changes in habitat on population abundance and risk, either in isolation or in combination with actions aimed at other life stages.  All of the models are calibrated to recent spawner-to-adult return data.  The Grande Ronde spring Chinook population models also incorporate derived empirical relationships for three sequential tributary life history stages.  Each of the models includes parameter uncertainty and annual environmental variation.  Recovery strategies for three of the populations include hatchery supplementation using locally derived broodstocks.   It is generally recognized that properly designed hatchery supplementation programs can contribute to maintaining populations while actions to change key limiting factors take effect, however not without risks.  A key uncertainty is at what point in a recovery program, do the long term diversity risks of supplementation outweigh increased total spawner demographic benefits.  We use simple indices representative of demographic and diversity risks for summarizing sensitivity analyses of parameter uncertainty and evaluation of projected performance under particular management strategies.