Abstract
Many highly migratory predator stocks that occupy the Gulf of Mexico are at risk, and the collapse of stocks could harm fisheries and ecosystems. Two pelagic longline spatial closures within the pelagic waters of the Gulf of Mexico have been established to protect pelagic species. In 2000, a permanent closure was established around DeSoto Canyon, with the management objectives of reducing catch and rebuilding biomass of bycatch and incidental catch species while minimizing impact to catch of target species. In 2015, a seasonal closure was established off the Louisiana shelf (Spring Closure), with the management objectives of reducing catch and rebuilding biomass of bluefin tuna (Thunnus thynnus). Pelagic spatial closures are relatively untested management tools. Science-driven analysis, including the investigation of ecosystem impacts through mathematical modeling, is necessary to address their utility. This dissertation presents research used to parameterize an ecosystem model, Atlantis, for the Gulf of Mexico marine ecosystem, followed by a study that used the Gulf of Mexico Atlantis model to conduct a policy exploration of the utility of Gulf of Mexico pelagic longline spatial closures. Chapter 2 described the collection of Gulf of Mexico historical, species-specific landings data for the calibration of the Gulf of Mexico Atlantis model, and investigated areas of uncertainty and bias, focusing on outputs from the Gulf of Mexico Atlantis model and landings-based indicators, due to unidentified landings and lack of data. U.S. landings not identified to species did not appear to bias landings-based indicators, nor does the aggregation of landings into Gulf of Mexico Atlantis functional groups. Chapter 3 described Gulf-wide spatial distributions of pelagic predatory functional groups. Distributions were estimated with generalized additive models fitted with U.S. bottom longline survey catch data (coastal models), and U.S. pelagic longline commercial catch data (pelagic models). This work advanced our knowledge on the correlations between the spatial distribution of pelagic predators within the Gulf of Mexico and the environment, and improved upon the spatial distributions previously used for the Gulf of Mexico Atlantis model. Finally, Chapter 4 described a policy exploration assessing if current pelagic longline spatial closures within the Gulf of Mexico, DeSoto Canyon and Spring Closure, could meet management objectives and evaluated possible ecosystem impacts. DeSoto Canyon was more successful at achieving management objectives and had more influence to ecosystem performance metrics than Spring Closure. Closures reduced Gulf-wide catches of bycatch and incidental groups with little reduction to catches of target groups. Rebuilding biomass of particular stocks may require additional reductions in fishing mortality. The Atlantis framework allowed for the detailed, spatially-explicit representation of biota, fleets and spatial closures, and provided a means to explore broad-scale ecosystem impacts. This dissertation found that pelagic spatial closures could be viable means to achieve management objectives for protecting highly mobile pelagic predators from fishing pressure.