Bayesian surplus production model with serial autocorrelation

Presentation is made of a simple surplus production model called Surplus Production Model with Serial Autocorrelation (MPECAS in Spanish) since it considers as a unique assumption that the surplus production shows a serial correlation and has no explicit functional relation with biomass. Its application requires only an abundance index proportional to a given power of the actual mean abundance of the resource and the corresponding annual catches series. The estimate of the model parameters is presented within a Bayesian context using the SIR (Sampling Importance Resampling) algorithm. Simple risk criteria are proposed to estimate the Maximum Biologically Acceptable Catch (MBAC) and the risks associated to each hypothetical catch level considered. A simulation exercise was performed to assess the statistical capability of MPECAS to reproduce the information provided by a Schaefer operational surplus production model considered as an actual one. Finally, an application example with the white croaker (Micropogonias furnieri) is presented and the MBAC for 5 and 10% risk of biomass decline the year following the assessment year calculated with the Schaefer and MPECAS models are shown.