Small, widely-dispersing, pelagic larvae are difficult to monitor through direct observation, but biophysical models can provide greater understanding of dispersal patterns. Here, we used a technique of reversing time in a biophysical model to estimate larval release sites of the blue
crab, Callinectes sapidus, in the northern Gulf of Mexico by backtracking from known locations of postlarval settlement. We modeled surface ocean movement in reverse time from four settlement sites (Grand Isle, Louisiana; Ocean Springs, Mississippi; Dauphin Island, Alabama; and Pensacola,
Florida). In the Gulf of Mexico, blue crab fisheries are managed at the state level. We found on average 73% of the larvae were traced back to larval release sites in the same state as settlement. Barrier island release sites accounted for 11% of settlement at Grand Isle and 17% at Dauphin
Island. Less than 0.1% of larvae that settled in Grand Isle were tracked back to a known blue crab spawning area located to the south west, Ship Shoal. The distance traveled by larvae was highest in Pensacola and Grand Isle. The coefficient of variation of distance traveled was highest in
Dauphin Island. The average distance traveled to each settlement location ranged from 31 to 124 km. Distance traveled explained 24% of empirical settlement numbers (F1,89 = 28.82, P < 0.001;R2 = 0.24). Settlement success was higher at release sites
predicted by this study than at randomly generated sites suggesting female selection of larval release sites.
