An ASPIC Based Assessment of Redfish (S. mentella and S. fasciatus) in NAFO Divisions 3LN (assuming that the highest apparently sustained historical average level of catch is a sound proxy to MSY)

There are two species of redfish in Divisions 3L and 3N, the deep-sea redfish (Sebastes mentella) and the Acadian redfish (Sebastes fasciatus) that have been commercially fished and reported collectively as redfish in fishery statistics. Redfish in Div. 3LN is regarded as a management unit composed of two Grand Bank populations from those two very similar redfish species. The present ASPIC assessment is based on the logistic form of a non-equilibrium surplus production model (Schaeffer, 1954; Prager, 1994), adjusted to a standardized catch rate series (Power, 1997) and to most of the stratified-random bottom trawl surveys conducted in various years and seasons in Div. 3L and Div. 3N from 1978 onwards. Both CPUE and surveys were used with all observations of each series.\n\t\t\t\t This assessment is not a follow up of the previous ones (Ávila de Melo et al., 2012 and 2010). The logistic Schaefer production model (1954) incorporated in ASPIC operating model (Prager, 1994) can not cope anymore with the most recent biomass increases observed in both spring and (mainly) autumn Canadian 3LN surveys, unless it is allowed to provide unrealistic assessment results. And continuing to strip off the highs of each one of these series, in order to get a picture in line to what is the perception of the stock history from commercial and survey data trends, is no longer a valid option, as reflected on the last STACFIS research recommendation on this matter (NAFO, 2012).\n\t\t\t\t Being so, input has been reframed opening room to a new combination of Canadian autumn 3L and 3N surveys. The inclusion of the Spanish spring survey on Div. 3N and the removal of the historical CPUE series have also been considered. Two selected frameworks options have finally run with MSY kept constant at an initial starting guess, instead of being estimated by the model. Before entering the latest (2013) ASPIC Suite flow, the input selected from exploratory analysis was submitted to a sensitivity test in order to evaluate the robustness of the new framework against variability on random number seed, start user guesses for key model parameters and last year survey biomass.\n\t\t\t\t The consistency of the new ASPIC assessment with their predecessors was checked by comparison of biomass and fishing mortality fit trajectories against previous ones from the 2012 and 2010 assessments. A 2014-2012 retrospective analysis was also performed with good results (small retro bias on relative biomass and fishing\n\t\t\t\t 2\n\t\t\t\t mortality in response to the general increase of the still standing survey series), and the assessment pursued successfully to bootstrap mode (again good consistency with previous results) and projections.\n\t\t\t\t A medium term management plan is finally proposed, based on bi-annual increases of the catch from the present TAC level of 6 500 t up to target catch/TAC of 18 100 t, the 2014 equilibrium yield from the present assessment, that should be in place by 2019-2020. This management plan allows, with a very high probability, that biomass is kept above Bmsy and fishing mortality below Fmsy.