Advancements in computing power and improved biophysical dispersal models, have enhanced our ability to realistically simulate distributions and behaviors of fish larvae. In this study, a 1 km high-resolution ocean model capable of capturing the ocean's mesoscale and sub-mesoscale motions is integrated with a biophysical dispersal model that considers a range of larval behaviors. Together they are used to investigate the dispersal and connectivity of Red Snapper (Lutjanus campechanus) larvae, a key species for both commercial and recreational fisheries in the northern Gulf of Mexico (GOM). We quantify how various larval behaviors influence the spatiotemporal dispersal, connectivity and settling of Red Snapper larvae focusing on egg buoyancy, larvae swimming capability, and ontogenetic vertical migration. Alongside habitat preferences, the ocean advection of Red Snapper larvae is crucial in shaping their dispersal patterns. Moreover, our simulations suggest different settling and connectivity characteristics between the eastern and western GOM. These results, indicate the need to divide these regions into distinct entities for stock management, rather than treating them as a unified stock as conventionally done.