Spatiotemporal patterns of movement are not just a biological curiosity, they have profound implications upon how population models are structured and interpreted, as well as model performance. In some fisheries, assuming a single, aggregated selectivity pattern may be sufficient. But when selectivity varies by sex, fishing sector, latitude, or season, as is the case in stocks exhibiting seasonal spawning migrations, such simplifications can bias estimates of fishing mortality and biomass. Spanish mackerel on Australia’s east coast exemplify this complexity.
To assess such a stock, the movement patterns need to be understood and the structure of the model tailored to reflect those dynamics. In the 2025 assessment, a 19-year dataset of conditional age-at-length compositions linked to latitude and date of capture was used to reconstruct and illustrate the seasonal migration of Spanish mackerel along Queensland’s east coast. This enabled the creation of a spatiotemporal heatmap that visualised age structure over latitude and month, revealing a clear and consistent migratory pattern. While the movement of Spanish mackerel is widely known by fishers in Queensland, this explicit visualisation of the pattern is a novel resource.
These insights directly informed the design of a spatially structured population model using a ‘fleets-as-areas’ approach. By enabling region and season specific selectivity at a spatiotemporal resolution directly informed by biological data, the model was able to capture these complex variations in selectivity. This approach also minimised the risk of bias arising from estimating these patterns on aggregate, without due appraisal of these dynamics. This analysis was a significant advance in assessing this migratory stock, making use of a rich age-at-length dataset, where broad-scale tagging data were not available to inform us on movement.