Simulation testing the robustness of a multiregion, tag-integrated assessment model that exhibits natal homing and estimates natural mortality and reporting rate

Abstract

The influence of model complexity on integrated tagging and catch-at-age analysis (ITCAAN) parameter estimation is poorly understood for populations exhibiting natal homing. We investigated ITCAAN performance under varying levels of movement, degree of similarity in population productivities, data quality, spatial complexity in parameterization, and whether natural mortality and (or) reporting rates were fixed at actual values, estimated, or misspecified. Dynamics of four populations with natal homing that intermixed during periods of harvest were simulated based on Lake Erie walleye (Sander vitreus). Our results suggest, when high-quality tagging data are available, ITCAAN models are able to simultaneously estimate movement rates, natural mortality, and tag reporting rates, though accuracy and precision of model estimates will decrease with greater model complexity and fewer tags released. Additionally, ITCAAN models may have difficulty estimating individual population abundances under certain movement rates when population productivities are vastly different. ITCAAN models that estimate natural mortality and reporting rates may perform best with similar sized populations and when data are available to assist the estimation of reporting rates.