Evaluating the Performance for a Yellowtail flounder (Limanda ferruginea) Movement-Mortality Model Based on Simulations

Larry A. Alade, Steven X. Cadrin, Thomas J. Miller, Eric B. May, and Christopher M. Legault

A movement-mortality model was developed for a yellowtail flounder tagging study to evaluate movement and survival rates in the northeast, United States. As the impact of stock mixing on population dynamics parameters is unknown, a simulation study was used to gauge the effects of measurement and process errors on datasets with similar characteristics to the tagged yellowtail flounder population. Stochastic simulations were developed in which tagged “virtual” fish were released and subsequently recaptured in the model. Estimates from the statistical model were then compared to the “true” reference values to evaluate statistical measure of bias and precision. Results from the simulations show that when moderate movement rates (10-25% per year) is simulated with moderate precision of input data (10% to 25% CV), estimates of fishing mortality and movement rates were well estimated (<10% bias). However, for scenarios that simulate no movement, low movement (5%) or high movement (45%) rates, the model estimates tend to be confounded, indicating that movement rates and fishing mortality could not be effectively differentiated in the model. The implication these results indicate that there is a substantial penalty for imprecise input data, and parameter correlation may be a problem for some applications.