Small-scale fisheries contribute about half the global catch, yet their spatial footprints remain poorly monitored in many regions. As fishing activities extend into remote and poorly monitored waters, improving our ability to quantify fishing effort in data-limited regions is increasingly important. Here, we aimed to map fishing activity using open-access vessel detection derived from satellite imaging (Sentinel-2) and machine learning, providing greater insight into the spatial distribution of small-scale fisheries and their potential impacts. Using artisanal fisheries in the eastern Red Sea as a case study, we modelled vessel occurrence in relation to distance to ports, distance to reefs, and water depth, and mapped the spatial distribution of untracked vessels and their proximity to coral reefs. We also identified vessel detection hotspots and assessed their overlap with marine protected areas (MPAs). We found that vessel detections varied substantially among regions, with the southern region accounting for the largest share of fishing activity and 37% of identified hotspots. Approximately half of vessel detections occurred within 1 km of reef structures, while 60% of hotspots overlapped with proposed or designated MPAs. Vessel occurrence was negatively correlated with fish biomass, supporting the use of satellite-derived detections as a proxy for the spatial distribution of fishing pressure. Our findings demonstrate the potential of satellite imagery to provide a practical and cost-effective approach for monitoring small-scale fisheries across large marine regions, particularly in data-limited areas and waters of high conservation priority.
[CB1]edit to reduce words
[CB2]inefficient use of words