BUGS in Bayesian stock assessments

This paper illustrates the ease with which Bayesian nonlinear state-space m odels can now be used for practical fisheries stock assessment. Sampling fr om the joint posterior density is accomplished using Gibbs sampling via BUG S, a freely available software package. By taking advantage of the model re presentation as a directed acyclic graph, BUGS automates the hitherto tedio us calculation of the full conditional posterior distributions. Moreover, t he output from BUGS can be read directly into the software CODA for converg ence diagnostics and statistical summary. We illustrate the BUGS implementa tion of a nonlinear nonnormal state-space model using a Schaefer surplus pr oduction model as a basic example. This approach extends to other assessmen t methodologies, including delay difference and age-structured models.