SCAA/ASPM assessments of white hake incorporating data to 2007

Abstract

The preliminary SCAA/ASPM assessment of white hake presented to the previous GARM meeting is refined and updated, being advantaged by the greater availability of catch-at age information for the surveys and now also for commercial catches. Six assessment variants are presented, which reflect whether or not to take survey catchat-length information into account in fitting the assessment models, whether to use a Ricker or a Beverton-Holt form for the stock-recruitment relationship, and whether or not to constrain the multiplicative bias factor for the autumn NEFSC survey sweptarea estimates of biomass to preclude the possibility of herding. Imposing this last constraint leads to an appreciable deterioration in the overall likelihood of the model; the likelihood also indicates a slight preference for the Ricker over the Beverton-Holt form. Inclusion of the survey catch-at-length data in the likelihood leads to a considerable improvement in estimation precision, but also shows a marked overestimate by the model of the proportion of white hake of 20 cm and less in the autumn NEFSC surveys. Both variants are put forward as candidates to provide the basis for scientific management advice. The variant which includes the catch-at-length information reflects lesser abundance in absolute terms, a greater retrospective pattern, but also current abundance at a greater proportion of the MSY level, when compared to its counterpart. The estimation of current resource status in relation to the MSY level is critically dependent on the determination of the parameters which specify the starting (1963) numbers-at-age vector for the assessment.1