A GIS Based Fuzzy Membership Approach for Mapping Yellowfin Tuna (Thunnus Albacares) Potential Zones in Sri Lanka's Exclusive Economic Zone

Abstract

The Sri Lankan offshore fishing industry faces numerous challenges including dynamic weather conditions, financial constraints, and limited technological adoption for locating productive fishing grounds within the country's Exclusive Economic Zone (EEZ). This study aimed to develop a GIS based model integrating fuzzy membership analysis to identify the Yellowfin Tuna (Thunnus albacares) potential zones in EEZ of Sri Lanka. Six oceanographic parameters were used in the model: sea surface temperature (SST), chlorophyll-a concentration (CHLO-a), dissolved oxygen (OXY), salinity (SAL), density (DEN), and bathymetry (BATHY). Monthly data for 2019 were obtained from Copernicus Marine Service and General Bathymetric Chart of the Ocean (GEBCO) data catalogs. Parameter maps were reclassified using fuzzy membership functions based on species-specific optimal ranges derived from literature (SST: 22-30°C; CHL-a: 0.05-0.25 mg/m³; BATHY: <160 m; SAL: 32-37 PSU; DEN: 1020-1027 kg/m³; OXY: >5 mg/L). Fuzzy overlay analysis generated monthly habitat suitability maps, which were combined to produce an annual forecast map. Validation used 63,105 fishery dependent catch locations from the Department of Fisheries (2019) and 67,310 random points. Results showed that an 80.3% accuracy within forecasted suitable zones. Forecast maps revealed seasonal variability in potential fishing grounds, with sensitivity of 75.7% and specificity of 84.5%. The Kappa coefficient of 0.60 indicated substantial agreement beyond chance, while ROC analysis yielded an Area under the Curve (AUC) of 0.86, demonstrating excellent discriminatory ability. With persistent high suitability areas identified in eastern and southern EEZ regions. The model successfully generated potential fishing zone predictions at 4 km resolution. As the forecasted areas are mostly within close range of the shore, they will reduce travel and search time, resulting in beneficial fuel savings for fishermen. This GIS-based fuzzy approach provides a cost-effective, scientifically robust tool for identifying Yellowfin Tuna potential zones, with direct applicability for sustainable fisheries management in Sri Lanka and potential adaptability for other data-limited regions.