Simulation Testing of Two Methods for Representing Spatial Catch-at-Age in a Stock Synthesis Assessment

If there is no or limited movement of post-settlement fish, population age-structure and catch age-compositions will vary spatially when there is spatial variation in fishing or recruitment.  Selectivity, which measures relative age-specific vulnerability to fishing, can be used to model catch-at-age either at the level of the population (population selectivity) or at regional spatial scales (areas-as-fleets selectivity).  In general, temporal variation in regional fishing or recruitment will cause temporal variation in both forms of selectivity.  This study uses a three-region operating model to simulate data and evaluate how assessment results are influenced in Stock Synthesis models based on these two alternative approaches for modeling selectivity.  One Synthesis model accommodates temporal variations in population-selectivity using random-walk processes to represent the selection-at-age coefficients.  The experiment uses a factorial design to consider the relative effect on assessment estimates of two life-history types (short-lived and more productive vs long-lived and less productive); two trends in spawning biomass (increasing versus decreasing); and two levels of variability in selection.  Despite the large numbers of estimated parameters, Stock Synthesis is able to produce reasonably accurate estimates of final year spawning biomass and depletion, but accuracy is greatly influenced by the life-history traits and trends in spawning biomass.