Date of Award


Document Type


Degree Name

Master of Science in Coastal Marine and Wetland Studies


Coastal and Marine Systems Science


College of Science

First Advisor

Robert F. Young

Second Advisor

Matthew E. Kimball

Third Advisor

Sharon L. Gilman


Strand-feeding is a unique, cooperative foraging strategy used by some common bottlenose dolphins (Tursiops truncatus) at low tide in South Carolina and Georgia salt marshes, where a group of dolphins charges schools of fish, thrusting them onto the bank with their bow wave and then capturing prey items onshore. Strand-feeding dolphins are estuarine residents and frequent predictable stranding sites, suggesting an intimate knowledge of the area. Dolphin behavior prior to stranding suggests they do not herd the fish before charging, but instead prey on fish already shoaling near the bank. The use of active acoustics from strand-feeders in order to locate fish schools, and therefore strand-feeding sites, has never been elucidated. Boat-based acoustic fish surveys, mapping surveys, and focal follows of strand-feeding dolphins were conducted during 21 days in July and September 2021 in a side channel of Bull Creek, SC, to explore the parameters involved in strand-feeding site selection and describe strand-feeding dolphin behavior. Acoustic fish and mapping surveys were conducted using a Humminbird side imaging sonar and an Adaptive Resolution Imaging Sonar (ARIS). Nine strand-feeders were identified, with 4 of them participating in 94.6% of strand-foraging bouts. All strand-feeding events (n = 244) occurred within three hours of low tide, with 68% during flood tide. Multiple categories of pre-stranding behavior were identified, some with prior investigation of the stranding site (physically and/or with echolocation clicks), some with no prior investigation, and none involving herding of fish prior to stranding. ARIS recordings allowed for confirming fish distributions within the side channel of Bull Creek using side scan imaging. Prey fish were abundant along much of the 8.9 km of surveyed creek banks as well as in the central channels, though dolphins routinely ignored large schools in central channels. Strand-feeding events occurred throughout the creeks, but three primary sections (<300 m length in total) accounted for 29% of strand-feeding events. These sections, however, were not the areas with the most abundant or consistent fish presence. Instead, creek bathymetry played the largest role in site selection, with the slope of the creek bank and the proximity to deeper sections of the creek identified by a Principal Component Analysis (PCA) as the most significant indicators. Echolocation buzzes were detected in 100% of acoustically recorded events before and during the charge, indicating that in most cases, dolphins assess fish presence at strand-feeding sites in real time when deciding to strand.