Abstract
Several data-poor stock assessment methods have recently been proposed and applied to data-poor fisheries around the world. The Mauritanian pink spiny lobster fishery has a long history of boom and bust dynamics, with large landings, stock collapse, and years-long fishery closures, all happening several times. In this study, we have used catch, fishing efforts, and length-frequency data (LFD) obtained from the fishery in its most recent period of activity, 2015–2019, and historical annual catch records starting in 2006 to fit three data-poor stock assessment methods. These were the length-based Bayesian (LBB) method, which uses LFD exclusively, the Catch-only MSY (CMSY) method, using annual catch data and assumptions about stock resilience, and generalised depletion models in the R package CatDyn combined with Pella-Tomlinson biomass dynamics in a hierarchical inference framework. All three methods presented the stock as overfished. The LBB method produced results that were very pessimistic about stock status but whose reliability was affected by non-constant recruitment. The CMSY method and the hierarchical combination of depletion and Pella-Tomlinson biomass dynamics produced more comparable results, such as similar sustainable harvest rates, but both were affected by large statistical uncertainty. Pella-Tomlinson dynamics in particular demonstrated stock experiencing wide fluctuations in abundance. In spite of uncertain estimates, a clear understanding of the status of the stock as overfished and in need of a biomass rebuilding program emerged as management-useful guidance to steer exploitation of this economically significant resource into sustainability.