Environmental modeling of yellowfin tuna fishing grounds integrating oceanographic and biological parameters using generalized additive model

BACKGROUND AND OBJECTIVES: Marine environmental variability influences the distribution, productivity, and habitat suitability of yellowfin tuna (Thunnus albacares). However, fishing-ground assessment based primarily on environmental conditions or production may overlook the biological suitability of harvested fish. The study objectives were to develop an integrated generalized additive model framework for assessing yellowfin tuna fishing grounds. The framework combined oceanographic conditions and fishing operational factors with production potential and fish size through two complementary Generalized additive models.METHODS: The study was conducted in the waters of West Sumatra, eastern Indian Ocean, Indonesia. Fisheries and oceanographic data covered ۲۰۱۹–۲۰۲۴, while biological observations covered ۲۰۱۹, ۲۰۲۰, and ۲۰۲۴. Oceanographic predictors included sea surface temperature, chlorophyll-a concentration, sea surface salinity, and sea surface height. Fisheries predictors included fishing year, month, gear, spatial location, and vessel capacity. A total of ۸,۷۷۱ yellowfin tuna fork-length measurements were analyzed. A fork length of ۹۹ centimeters was used as the maturity-based threshold. Two complementary generalized additive models were developed. The Fishing Ground–Environment, Technology, and Production model used catch as the production response. The Fishing Ground–Environment, Technology, and Size model used fork length as the biological response.FINDINGS: Sea surface temperature ranged from ۲۹.۳۷–۳۰.۴۵ degrees Celsius. Chlorophyll-a concentration ranged from ۰.۲۱–۰.۴۳ milligrams per cubic meter. Sea surface salinity ranged from ۳۳.۲۹–۳۳.۸۷ practical salinity units. Sea surface height ranged from ۰.۴۶–۰.۵۹ meters. Overall, ۹۴ percent of fish were harvest-suitable. The production model had an Akaike information criterion of ۱۵۵,۴۸۲.۹ and cumulative deviance explained of ۱۱.۵ percent. The size model had an Akaike information criterion of ۷۷,۹۱۳.۷۳ and cumulative deviance explained of ۲۸.۷۴ percent. Medium to high fishing-ground potential occurred more frequently around the Mentawai Islands. Temporal integration identified highly potential and potential fishing-period statuses based on relative production potential and the landing-based biological size indicator.CONCLUSION: The integrated generalized additive model framework links environmental conditions, fishing operations, production, and biological size. This approach provides a more selective basis for adaptive and sustainable management of yellowfin tuna fisheries.

کلیدواژه ها:

Environmental modeling ، Fisheries management ، Generalized additive model (GAM) ، Yellowfin tuna ، Fishing grounds ، Oceanographic variability.