A comprehensive method to integrate unbiased fisheries data in spatially-explicit population dynamics models

Stock assessment models and other quantitative models rely heavily on fishery-dependent data, particularly in regions where fishery-independent data (e.g., scientific surveys) are unavailable. However, the relationship between catch-per-unit-effort (CPUE) and fish abundance is impacted by variations in catchability and selectivity across different fishing operations. We present a comprehensive methodology for preparing unbiased fisheries data for use in spatially-explicit population dynamics models such as SEAPODYM, a spatiotemporal model of population dynamics with age structure. Our approach addresses two key challenges: first, by systematically grouping fishing data into distinct fisheries with consistent catchability and selectivity patterns, and second, by leveraging high-resolution spatial data to maintain linear relationships between catch and biomass density at the grid cell level. We demonstrate this methodology using operational longline data from Pacific Island countries and distant-water fishing nations targeting yellowfin tuna in the Pacific Ocean. The approach incorporates covariates such as hooks between floats and target species to account for fishing-driven changes in catchability, while assuming remaining variability is driven by environmental factors and the heterogeneity in population density that SEAPODYM explicitly accounts for. By combining these operational data with coarse resolution aggregated data from all gears, we ensure comprehensive coverage of fishing mortality while integrating fine-scale spatial resolution data needed for parameter estimation. This approach improves how fisheries data are structured for spatially-explicit stock assessment models, improving the accuracy of population dynamics estimates.