A Life History Simulator to Assess the Robustness of Salmon Stock Assessment

Stock assessments are widely applied throughout fisheries management and are vital for evaluating the status of stocks and projecting the future catch of a stock. Regardless of the model utilized for the assessment, assumptions about the life history parameters, structure and population dynamics of the stock are necessary since many of the biological processes are not directly observable (e.g., natural mortality rates). Disparity between the assumptions made in the stock assessment tool and the true dynamics of the stock may give rise to inaccurate assessment results, which influence harvest policy and determine the future status of the stock. I use a life history simulator for Pacific salmon (Oncorhynchus spp.), to test the robustness of stock assessment tools to these assumptions. Results using a cohort reconstruction for the assessment indicate that assumptions made about size dependence of the natural mortality rate influence the accuracy of the assessment results. Additionally, the quantity and type of data available for the assessment can greatly influence the results. The use of this simulator and the data it generates has a wide potential to applications beyond assessments. For instance, using this simulator in conjunction with other modeling techniques, such as state dependent life history theory, allows us to investigate optimal life history trajectories and growth strategies for Pacific salmon under a variety of conditions and to predict how various strategies within a cohort will influence population dynamics.