<p>Invasive species pose a significant threat to biodiversity and ecosystem stability worldwide, challenging resource managers to develop strategies that maximize effective control measures with limited resources. While understanding connectivity is foundational to managing and conserving populations, the same network dynamics that sustain native communities can be exploited to manage invasive species. Here, we combined biophysical larval dispersal modeling, network analysis, and population modeling to optimize lionfish control within a network of 17 offshore banks in Flower Garden Banks National Marine Sanctuary, a marine protected area extending over 400 km² in the Gulf of Mexico. Dispersal simulations revealed seasonal shifts in circulation dynamics, with winter and spring favoring connectivity among banks within the sanctuary and summer and fall promoting long-distance dispersal outside of the sanctuary through enhanced loop-current eddy activity. By increasing lionfish removal efforts during months of naturally low larval connectivity within the sanctuary, we were able to shift the modeled lionfish population from growing to declining with minimal effort. Network analysis further characterized the structure of larval connectivity in the sanctuary, identifying banks where removals would have a greater impact on population persistence versus locations where removals would only be impactful locally. This study illustrates how linking oceanographic processes, network theory, and population modeling can guide efficient and scalable interventions for invasive species control.</p>