A Bayesian estimation and decision analysis for an age-structured model using biomass survey data

Abstract

The use of Bayesian estimation and decision analysis is described and applied to estimate the probability distribution of virgin stock size, and expected consequences of future quotas for an age- and sex-structured model fit to biomass survey data. It is shown that Bayesian methods can be easily used with complex models when uncertainty in a single parameter is considered. An example from the fishery for hoki ( Macruronus novaezelandiae) in New Zealand is presented, in which the weight, vulnerability schedule, fecundity schedules, natural mortality rate and the sensitivity of recruitment to spawning stock are assumed known. Results with a uniform prior probability on stock size are compared with those where an informed prior distribution is available. Three alternative harvest policies are compared and the expected values of several indicators of performance are computed for different quotas. The problems in extending the model to calculate Bayesian posterior distributions with multiple indicators of abundance, many more free parameters, and a stochastic model are considered.