RISK AND UNCERTAINTY IN CAPTURE-MORTALITY PRODUCTION MODELS

Abstract

The paper explores the feasibility of different methods for evaluating uncertainty in some common Biological Reference Points (BRPs) derived from yield-mortality surplus production models. The bootstrap method was applied by resampling (a) the original data set and (b) the residuals from the model fit, in order to obtain confidence bounds. Both bootstrap approaches were very useful in estimating confidence intervals, but the former appears to be more conservad ve. Criteria for management related to the total productionof thestock,suchas the"Yield at Máximum Biological Production" and derived BRPs, resulted more conservativo that those related to the Máximum Sustainable Yield level, and therefore are suggested as useful benchmarks in a context of precautionary fishery management. A simple approach towards the formulation of riskaverse management strategies was explored using Decisión Theory. For this purpose, we introduced the concepts of Maximax, Maximin and Máximum Regret Criterion as alternativo criteria for choice under uncertainty. We concluded that these criteria seem to be well adapted to formulation oí risk-averse management strategies and precautionary fisheries management.