Testing the Performance of a Spatially Explicit Tag-Integrated Model of Yellowtail Flounder through Simulation Analysis

In any stock assessment application, the implicit assumptions about spatial population structure must be carefully evaluated.  Tag-integrated models offer a promising approach for incorporating spatial structure and movement patterns in stock assessments, but the complexity of the framework makes implementation challenging and the appraisal of performance difficult.  Additionally, many important sources of bias in tag-integrated models remain unexplored.  A flounder-like fishery was simulated to emulate the metapopulation dynamics of the three yellowtail flounder (Limanda ferruginea) stocks off of New England, and the robustness of spatially explicit tag-integrated models was compared with closed population assessments. Different movement parametrizations and data uncertainty scenarios were simulated, while the ability of the tag-integrated model to estimate reporting rate and time-varying movement were also evaluated.  Results indicated that the tag-integrated framework was robust for the simulated fishery, and that tag reporting rates were accurately estimated.  Closed population models also demonstrated limited bias.  Therefore, spatially explicit approaches may not be warranted, even when regional connectivity is occurring.  Tag-integrated models may be better suited as comparative tools for validating the assumptions and results of closed population models rather than as the sole basis for fishery management decisions.