A framework to map fishing activity from incomplete data

Knowledge of the timing and location of fishing activity underpins management strategies that address both ecological sustainability and socio-economic objectives. However, fishing records often lack accurate geographic information, and the scale of the reporting area is generally too broad for use in statistical models and spatial planning. Consequently, researchers often rely on qualitative data and proxies to map fishing activity. This study presents a simple and flexible framework to integrate information from diverse datasets to create a heterogenous layer of spatial sampling weights over which fishing activity is allocated (Spatial Weight Informed Mapping, SWIM). SWIM was applied to four case studies that ranged a spectrum of data completeness, from having comprehensive geographic data to lacking any coordinates. Using information acquired from logbooks, landings, harvester interviews, and species surveys, resulting maps displayed weekly variability in fishing activity. Results also illustrated how fisheries evolve over the season, responding to changes in target species distributions and quota availability. A validation test produced results that more closely matched true fishing patterns compared to methods that distribute fishing activity evenly over fishing areas. These results demonstrate that SWIM is a practical and scalable solution for identifying spatiotemporal patterns of fishing activity in data-limited environments, thus providing an important tool for evidence-based management. SWIM may be particularly useful for situations in which realized fishing activity is unlikely to be evenly distributed within management areas, as in many coastal fisheries.