Performance of age-only state-space assessment models under diverse somatic growth scenarios

Abstract

Recent developments have allowed state-space assessment models (SSAMs) to incorporate processes such as growth, size-based selectivity and maturity; however, many assessments continue to approximate them as age-based ("age-only" SSAMs). In this study, we use a simulation experiment to evaluate how different factors related to the sampling scheme and the type of growth variability affect the performance of age-only SSAMs. We followed two simulation approaches: "traditional", which assumes all processes in the simulation are age-based, and "stepwise", which aims to approximate the age-length dynamics and sampling process. We found that the traditional approach may produce overly optimistic performance by ignoring the age-length dynamics. Also, a length-stratified sampling scheme for ageing improves recruitment estimates, while a random sampling scheme may be preferable for estimating population mean weight-at-age. Modelling time-varying selectivity when variability in somatic growth is present is critical to improving recruitment variability and SSB estimates. Our results offer practical guidance when implementing SSAMs with age-specific data and highlight the importance of accounting for growth dynamics and sampling design in the assessment process.