An uncertainty-based decision support tool to evaluate the southern king crab (Lithodes santolla) fishery in a scarce information context

摘要

Decision-making in fishery management depends on the reliable quantification of uncertainty, which emerges from a variety of sources. For small-scale fisheries (SSF) in developing countries, the main sources of uncertainty are the available data, which are often untrustworthy, and poor institutional controls, which allow illegal fishing activities. Despite the global spread of SSF, the management of these socio-ecological resources is usually based on insufficient assessments estimated using scarce data. In this study, we identify and quantify the uncertainty associated with the harvested biomass of southern king crab (SKC) (Lithodes santolla), as declared by artisanal fishers of the Magallanes Region, Chile, in order to assess the harvestable biomass under a precautionary (conservative) approach to its exploitation in a scarce information context. This analysis includes the application of (i) fuzzy-Bayesian machine learning to quantify the uncertainty related to the harvested biomass per vessel per fishing cruise (C), and (ii) nonlinear polynomial regression models to project the spatial distribution of C according to the estimated uncertainty limits (minimum and maximum). The uncertainty analysis shows a 2000 kg threshold for C; beyond that, the declared harvest does not follow the estimated relationship between biomass and habitat characteristics (temperature, turbidity). The maps created from this analysis convey crucial information that would allow a more rational exploitation of the SKC stock and could be used to estimate a sustainable fishing quota, i.e., allocating effort according to the harvestable biomass. Our results suggest that the proposed tool can reliably inform management faced with uncertainty in stock exploitation by SSF, particularly when information is scarce.