Spatial Modeling of Yellowfin Tuna in the Banda Sea Based on Oceanographic Factors Using MaxEnt

Abstract

This study models the spatial distribution of yellowfin tuna (YFT) in the Banda Sea using the MaxEnt approach, addressing critical questions about its predictive capability, the influence of environmental variables such as sea surface temperature (SST) and chlorophyll-a concentration, and temporal patterns. MaxEnt was chosen for its ability to predict potential distribution areas based on presence data and environmental factors. Data utilized include fish catch records obtained from the fishing logbook of the Ministry of Marine Affairs and Fisheries of the Republic of Indonesia, chlorophyll-a concentration, and SST data sourced from ocean color satellite observations. Model performance was evaluated using the Area Under the Curve (AUC) metric. Study results reveal that significant spatial and temporal variations in YFT distribution are influenced by oceanographic factors, with the model performing best in July (AUC 0.72) and lowest in April, September, and December (AUC ~0.60). SST was the dominant variable in November (82.35%), while chlorophyll-a had the highest contribution in April (83.02%). These findings highlight the dynamic link between tuna distribution and environmental conditions. The spatial maps offer insights for optimizing fishing practices, reducing pressure on overexploited stocks, and supporting sustainable fisheries management through data-driven approaches like MaxEnt. However, the MaxEnt model has limitations such as sensitivity to multicollinearity, overfitting, and low transferability. Future research could enhance accuracy and robustness by using advanced methods like Spatial Maxent, Monte Carlo Variable Selection, or ensemble modeling to support adaptive fisheries management.