A Revised Operating Model for Sablefish in British Columbia, Canada in 2016

Abstract

Fisheries and Oceans Canada (DFO) and the British Columbia (BC) Sablefish (Anoplopoma fimbria) fishing industry have collaborated on a management strategy evaluation (MSE) process since 2009. This process is used to develop and implement a transparent and sustainable harvest strategy for the multi-gear Sablefish fishery. Variations of age-structured models have been used in simulation testing candidate management procedures and therefore represent the cornerstone of the MSE process. In this paper, we revise the Sablefish operating model to account for potential structural model misspecification and lack-of-fit to key observations recognized in previous models. Specific modifications include: (i) changing from an age-/growth group operating model to a two-sex/age-structured model to account for differences in growth, mortality, and maturation of male and female Sablefish, (ii) adjusting model age-proportions via an ageing error matrix, (iii) testing time-varying selectivity models, and (iv) revising the multivariate-logistic age composition likelihood to reduce model sensitivity to small age proportions. These structural revisions to the operating model improved fits to age-composition and at-sea release data that were not well-fit by the previous operating model. Accounting for ageing errors improved the time-series estimates of age-1 Sablefish recruitment by reducing the unrealistic auto-correlation present in the previous model results. The resulting estimates clearly indicate strong year classes of Sablefish that are similar in timing and magnitude to estimates for the Gulf of Alaska. Two unanticipated results were obtained. First, time-varying selectivity parameters were not estimable (or necessarily helpful) despite informative prior information from tagging. Second, improved recruitment estimates helped to explain the scale and temporal pattern of at-sea release in the trawl fishery. The latter finding represents a major improvement in our ability to assess regulations and incentives aimed at reducing at-sea releases in all fisheries. Estimates of Sablefish stock status, productivity, and trends over the past several years are consistent with previous harvest strategy simulations. Estimated exploitation rates for years 2011-2015 varied across seven data scenarios (~8-10%) but are consistent with exploitation rates projected for the current U60-40+Floor management procedure under the former operating model.