Model complexity and catch-age analysis

Catch-age analysis has become a standard tool of fisheries stock assessment. Modern catch-age models can contain dozens or even hundreds of parameters to be estimated. In this paper, we compare the performance of a typically complex catch-age model with a simpler model in which the population age structure is represented only by the annual mean age. We find that the population trajectory estimated from the simple model is in general agreement with the trajectory from the catch-age analysis. We suggest that the simple model can serve as both an intuitive and quantitative aid to interpreting results from the complex model. For example, the contrast between models highlights the role of technically complex assumptions. To some extent, errors in the age structure data cancel in calculating the mean age. Thus, our analysis suggests options available for investigating stocks where only summary age data are reliable.