| Summary: | The aim of this WD was to explore surplus production models to assess the western component of the anchovy 27.9a stock. Models were fitted to catch per quarter or semester (1991 – 2021) and to one biomass index, the spring acoustic survey (1999 – 2021), or two biomass indices, the acoustic survey and the autumn groundfish survey (1991 – 2018) using SPiCT. Various assumptions regarding the shape of the production curve, the initial biomass depletion and the intrinsic growth rate of the population were combined such that models varied from nearly unconstrained (more complex) to increasingly constrained (less complex). Bi-annual catch data and two survey indices lead to a higher number of convergent models. Several models passed all ICES criteria to accept a SPiCT assessment, except for a higher level of uncertainty in F/FMSY than the agreed one for long-lived stocks. A model assuming a Schaefer production curve, a prior on r from a meta-analysis and, an initial depletion rate of 80%, showed better retrospective analysis, survey hindcast cross-validation and convergence performance than other candidate models. The results indicated that F/FMSY was below 1 across most of the period, B/BMSY fluctuated well below 1 until 2010 and above 1 since 2016. The present results may be considered for further work in a benchmark workshop.
|
|---|