Characterizing pelagic pair trawl partnerships using coordination metrics and behavior classification in the Black Sea

Pelagic pair trawling involves two vessels towing a single midwater net in coordinated motion to target schooling pelagic species. Monitoring and distinguishing pair-trawl operations remain challenging for fisheries management, particularly when analyses rely on single-vessel data. Focusing on the Turkish Black Sea, this study presents an AIS-based framework to identify and characterize cooperative behavior in pelagic pair trawling fleets. We combine Gaussian Mixture Models for speed-profile classification with dyadic coordination metrics based on spatial proximity, heading alignment, and speed synchronization. The results indicate a network of 70 vessels and 51 dyads that reveal a structured pattern of recurrent partnerships and transient collaborations, with short-term coordination events occurring at separations under 1 km and speed differences below 1 knot. Weekly networks revealed a pronounced winter-spring activity pulse, a shutdown during the national summer closure, and a post-ban resurgence in autumn. Cohesion peaked within compact subfleets structured around a few central hubs, while spatial footprints and port linkages indicated regional organization along the south-western and south-central coasts. Leader-follower scores clustered near parity, consistent with paired-gear constraints, and cross-recurrence of speed series showed high determinism and laminarity, reflecting stable, repetitive operational routines. By integrating behavioral, spatial, and temporal indicators, the framework enables objective detection of cooperative fishing and provides a transferable, data-driven approach to support monitoring, spatial management, and sustainable governance of cooperative pelagic trawl fleets.