Species Distribution Model of the North Pacific Loggerhead in North American Waters and Implications for Fisheries Bycatch

Lavelle A (2024) Species Distribution Model of the North Pacific Loggerhead in North American Waters and Implications for Fisheries Bycatch. Masters Thesis, Duke University

The North Pacific loggerhead (Caretta caretta) is a vulnerable Regional Marine Unit (RMU) according to the IUCN Red List. The most pressing risk to North Pacific loggerhead populations is incidental capture, or bycatch, in fisheries throughout their range. An area of particular threat to this RMU is the Baja California Peninsula, where Peckham et al. (2007) estimated that more than a thousand loggerheads perish in artisanal fisheries in the Gulf of Ulloa annually. Thus, it is critical to more accurately understand and predict the density and timing of loggerhead occurrence in the California Current Large Marine Ecosystem (CCLME) to inform the most efficacious spatial extent and timing for protective measures, such as fisheries closures, to reduce sea turtle bycatch. Previous studies have found a tight parabolic relationship between loggerhead presence in the east Pacific and water temperature, suggesting that sea surface temperature (SST) could be used as an effective proxy to describe the habitat of loggerheads habitat. To create a robust predictive species distribution model, we integrated four datasets (2003–2023): aerial surveys, shipboard surveys, satellite-tagged individuals, and citizen sighting with environmental co-variates of loggerhead presence. Using both generalized additive models and MaxEnt models, we found that in addition to SST, loggerhead presence in November–March during El Niño years within the southern CCLME is positively associated with chlorophyll-α, net primary productivity of carbon, sea surface elevation, magnitude of seafloor depth gradient, eastwest gradient of seafloor depth, magnitude of wind velocity, west-east seawind and north-south Seawind components. Loggerhead presence within the southern CCLME is negatively associated with the magnitude of seafloor depth gradient, particulate inorganic carbon, and north-south gradient of seafloor depth. To our knowledge, this is the first species distribution model for loggerheads in the eastern Pacific. Our analysis revealed an important overlap between the predicted hotspots of loggerhead occurrence and two designated conservation areas: the Fishing Refuge Zone in the Gulf of Ulloa, Mexico, and the Pacific Loggerhead Conservation Area in Southern California. This concordance underscores the importance of these areas for the conservation of loggerhead sea turtles. However, our results also suggest that these measures could be expanded geographically and temporally more dynamically. Our study also demonstrates the power of citizen engagement in reporting species sightings, potentially increasing the predictive power of our models. We recommend incorporating species distribution modeling and additional environmental variables to inform dynamic fishery closures and protect the multiple uses of North American coastal waters.