Gaining information from commercial catch for a Bayesian two-stage biomass dynamic model: application to Bay of Biscay anchovy

Abstract

A two-stage biomass dynamic model for Bay of Biscay anchovy is presented. Compared with the model currently applied by ICES for the assessment of that stock, the new model separates the growth and natural mortality processes and allows them to differ by age class. Stochastic equations involving the observed weights by age class in surveys are incorporated to provide information on growth rates. The fishing process is modelled separating fishing mortality into year and age-class effects in each semester, and observation equations are introduced for total catch and catch proportion by age class (in biomass) by semester. The model is first tested on simulated data, then applied to real data for the years 1987-2008. Although the results are affected by survey catchability and natural mortality assumptions, estimates of population trends, when expressed in relation to the value in a given year, are robust. The new model has significantly more parameters, requiring longer computational time for its fitting, which is done in a Bayesian context. However, it does allow the testing of different assumptions on natural mortality, which is of special interest after the recent fishery closure, and estimating new parameters, which could provide further insight on stock and fleet dynamics.