Initial tests of the robustness of the provisional harvest control rule in Canada’s Sustainable Fisheries Policy to process and measurement errors using simulated depleted fish populations

Abstract

Canada’s Department of Fisheries and Oceans (DFO) Sustainable Fisheries Framework and the associated Decision Making Framework Incorporating the Precautionary Approach policies (DMF), implemented in 2009, provide a context with potential to improve fisheries management. A Provisional Harvest Control Rule (PHCR) is proposed in the DMF to allow adjustments of the annual total allowable catch based on a scientific assessment of the state of the stock. The DMF defines three spawning stock biomass Zones (Critical, Cautious and Healthy). The PHCR adjusts fishing mortality dependent on the Zone within which the spawning stock biomass is estimated to fall. Elements of the PHCR have been incorporated in the scientific advice and management approaches for a number of Canadian fish stocks. In this study, initial evaluation of the PHCR was carried out on three simulated depleted fish populations with different life histories under a variety of combinations of process error on recruitment and measurement error on spawning stock biomass. The simulations represent “bestcase” scenarios because reference points were assumed to be known exactly and the magnitude of the errors was moderate. The simulation results suggested that fish stocks in the Critical Zone should rebuild to the Healthy Zone under the PHCR with high probability (>0.78) irrespective of life history differences and the combinations of process and observations errors. However, the time to rebuild was up to twice as long as it took in the absence of fishing and the PHCR was not effective in ensuring the DMF requirement of a low probability (<0.1) of the population returning to the Cautious Zone. The PHCR was also not effective in keeping fishing mortality below the level that generates maximum sustainable yield when the stock was in the Cautious Zone and subject to measurement error. Variation in the annual catch generated by the PHCR in the simulations increased with increasing process and observation errors to a maximum CV of 0.6, which may be inconsistent with the fishing industry’s desire for low variation in annual catch.