This paper develops a Bayesian mark-recapture model for estimating both fishing and natural mortality, and integrates catch-at-age data into the estimation model, to aid in distinguishing between natural and fishing mortality. The model essentially follows tagged cohorts, with the estimated parameters being the fishing and natural mortality vectors for that particular cohort.
The model was tested against simulated data, and performed as expected. Mark-recapture data from toothfish at South Georgia was analysed with the model. The ages covered by the analysed cohorts ranged from 5 to 13. In the model runs, while fishing mortality-at-age was estimated, only one value of the natural mortality was permitted, as allowing both to vary can produce either non-sensical or nonconvergent results. For all the cohorts the model was applied to, a consistent picture emerged. Both natural and fishing mortality were estimated to be lower than those estimated in previous stock assessments using both mark-recapture and catch data. There was additionally some suggestion of agestructure in the natural mortality-at-age.