Random effects on numbers-at-age transitions implicitly account for movement dynamics and improve stock assessment and management

Implementing operational assessment models that account for spatial structure and movement dynamics is challenging, especially with limited tagging data. Random effects on numbers-at-age (NAA) transitions in state–space models offer a potential solution to circumvent direct movement estimation by attributing movement variation to NAA random effects. However, whether this approach reliably achieves desirable management outcomes remains unclear. In this study, we conducted a management strategy evaluation that emulated a generic medium-lived fish that exhibit natal homing dynamics, using assessment models with varying levels of spatial complexity. We compared the performance of each spatial implementation with and without NAA random effects to evaluate their effectiveness in achieving management outcomes. Our results showed that models with NAA random effects consistently outperformed those without, although the benefits of NAA random effects degraded at high rates of movement. Therefore, NAA random effects could serve as a practical intermediate solution when explicit movement modeling is not feasible due to insufficient movement information. Our findings suggest that incorporating NAA random effects should be a default starting point in state–space stock assessments.