Monitoring fishing activity is crucial for fisheries management and governments to ensure regulatory compliance and sustainable marine ecosystems. Analysing vessel movements provides insights into fishing dynamics, aiding decision-making. Additionally, measuring unmonitored fishing activity (hidden fishing) helps counteract the underestimation of fishing pressure. Big data analysis can reveal fishing patterns and hidden activities from vessel position and speed data, such as those transmitted by fleets carrying Automatic Identification Systems (AIS). We used an Open Science-compliant (reproducible, repeatable, and reusable) cloud computing-based big data analysis to estimate the manifest, total, and hidden fishing distributions of AIS-carrying vessels in the Mediterranean Sea from 2017 to 2022, processing about 1.6 billion vessel speed and position data. We estimated the principal hotspots of hidden fishing over the years and the potentially involved stocks from these data. We also assessed whether the hotspots corresponded to illegal fishing or AIS communication issues and concluded that most hotspots potentially corresponded to illegal fishing. Our manifest fishing distribution agreed with another produced through machine learning by the Global Fishing Watch. We developed a fast and reusable approach that can produce new information to help management authorities understand the extent of hidden fishing.

 

Monitoring; Data mining; Fisheries; Machine learning; Biological system modeling; Data models; Big Data.