Management implications of modelling fisheries recruitment

The representation and parameterisation of the stock-recruitment relationship is highly influential in fisheries stock assessments. We overview important management implications arising from choices and assumptions made when modelling recruitment and propose pragmatic solutions. An age-structured population model is used to highlight some additional considerations including (1) incorrect assumptions about the source of recruits and (2) ways to account for environmental correlates in stock assessment models, either directly or indirectly. Existing stationary approaches ignore the sometimes considerable influence of the environment on population processes such as recruitment, growth and mortality, all of which are expected to shift under changing climate. We show that where environmental relationships are explicitly included in models, they need to be rigorously assessed and model parameterization carefully evaluated. Variability in annual recruitment estimates (specified by the standard deviation sigma(R)) must be adjusted to control the partitioning of total variation into environmental and unexplained variation, as reference point estimates are particularly sensitive to the choice of sigma(R). Finally, we evaluate two recommended solutions to these issues: Harvest Control Rules (HCRs) that are robust to recruitment uncertainty and variability, and empirical HCRs that use pre-season survey inputs as a way to more directly provide forecasts without needing to quantify the complex underlying details of the environmental-recruitment relationships.