Integrating Empirical Data and Ocean Drift Models to Better Understand Sea Turtle Strandings in Virginia
Master of Science (M.Sc.)
Virginia Institute of Marine Science
David M. Kaplan
Marjorie A.M. Friedrichs
Kevin C. Weng
Katherine L. Mansfield
Hundreds of stranded turtles wash up deceased on Virginia’s coastline each year, yet the causes of most stranding events are poorly understood. In this thesis, a carcass drift model was developed for the Chesapeake Bay, Virginia, to predict likely locations of mortality from coastal sea turtle stranding records. First, field studies were carried out to better parameterize the drift characteristics of buoyant sea turtle carcasses, namely, probable oceanic drift times and the impact of direct wind forcing on carcass drift. Based on the duration that tethered, free-floating turtle carcasses were buoyant, we determined that oceanic drift duration of turtle carcasses was highly dependent on water temperature and varied from 2-15 days during typical late spring to early fall bay water conditions. The importance of direct wind forcing for turtle carcass drift was assessed based on track divergence rates from multiple simultaneous deployments of three types of surface drifters: bucket drifters, artificial turtles and real turtle carcasses. Turtle drift along-wind leeway was found to vary from 1-4% of wind speed, representing an added drift velocity of approximately 0.03-0.1 m/s for typical bay wind conditions.
The information obtained from these field studies were used to parameterize the oceanographic carcass drift model, which was applied to reported strandings during 2009-2014. Predicted origin of stranding records with probable cause of death identified as vessel strike were compared to commercial boating data. Locations of potential hazardous turtle-vessel interactions were identified in high traffic areas of the southeastern Chesapeake Bay and James River. Commercial fishing activity of various gear types with known sea turtle interactions were compared in space to predicted mortality locations for stranded turtles classified with no apparent injuries, suggesting possible fisheries-induced mortality. Probable mortality locations for these strandings were found to vary between spring peak and summer off-peak stranding periods, but two distinct hotpots were identified in the southwest and southeast portions of the lower bay. Spatial overlap was noted between potential mortality locations and gillnet, seine, pot, and pound net fisheries. These predictions provide clear space-time locations for focusing future research and prioritizing conservation efforts. Nevertheless, the lack of fine temporal and spatial resolution fishing data limited our ability to quantitatively assess most likely causes for specific stranding events. This study both highlights the importance of addressing these data gaps and provides a meaningful conservation and management tool that can be applied to stranding data of sea turtles and other marine megafauna around the globe.
© The Author
Santos, Bianca Silva, "Integrating Empirical Data and Ocean Drift Models to Better Understand Sea Turtle Strandings in Virginia" (2017). Dissertations, Theses, and Masters Projects. William & Mary. Paper 1516639566.