Misspecifying Sex‐Structured Dynamics in Stock Assessment Models

Abstract

Sex‐specific variation in population demography and life‐history traits (e.g., growth, natural mortality) is common in many fish populations. Differences in these processes by sex can dictate population dynamics and influence how harvesters interact with the resource. Across various fisheries management systems, stock assessment models (SAMs), which mathematically represent population age and/or size structure, are widely utilised to estimate fish population status and provide sustainable harvest recommendations. However, few studies have examined the implications of alternative modelling assumptions when incorporating sex‐specific dynamics in SAMs. For instance, the impacts of simultaneously ignoring sex‐specific variations in growth, selectivity, and natural mortality on SAM performance have not been explored. In this study, a simulation‐estimation framework was developed for a sexually dimorphic fish population to: (1) assess the consequences of ignoring sexual dimorphism (i.e., growth, natural mortality, and selectivity) and the benefits of using sex‐specific catch data to inform the estimation of these processes, (2) evaluate the implications of incorrect modelling assumptions regarding sex ratio at birth, and (3) develop advice for parameterising observation likelihoods to describe sex‐specific composition data. Correctly parameterising sex‐specific variation in life‐history traits led to more robust population estimates and catch advice. Conversely, SAMs ignoring these variations yielded biased estimates of biomass and harvest recommendations. Collectively, our results underscore that oversimplified assumptions about sex‐specific variations in SAMs can lead to poor management advice. Moreover, results emphasise the need for routine collection of sex‐specific data to support the development of biologically realistic models.

Bibliographic Information