Use of a geographic information system (GIS) to examine bottlenose dolphin community structure in southeastern North Carolina

From June 1987 to April 1988 an epizootic of bottlenose dolphins (Tursiops truncatus) occurred along the eastern coast of the United States from New Jersey to Florida. As a result of this mortality, the mid-Atlantic “coastal migratory stock” of bottlenose dolphins was identified and estimated to have been reduced by more than 50%, which resulted in its classification as “depleted” under the U.S. Marine Mammal Protection Act. Since that time, researchers have documented the existence of multiple coastal stocks of Atlantic bottlenose dolphins, which are believed to include a complex mix of residents, seasonal inhabitants, and transient animals. One of these putative stocks occurs approximately between Beaufort, NC and Myrtle Beach, SC; 40 of the animals belonging to this stock were the focus of this study. Sighting histories for these 40 bottlenose dolphins have been compiled over the past nine years as part of an ongoing photoidentification study near Wilmington, NC. Two hypotheses were tested for dolphins in the Wilmington, NC area: (1) a single community of dolphins exists and (2) dolphins exhibit no preference for specific locations within the study area. To account for survey effort, a weighted index was developed to standardize the data. To investigate dolphin community structure, a variety of area use methods were tested using a Geographic Information System (GIS). Most common in the literature are the adaptive kernel estimator (ADK) and the minimum convex polygon (MCP) methods, which have become standards in animal movement studies. Conversely, geographers and statisticians have developed point pattern and density estimation techniques. These approaches were compared, and the geographically-based interpolation methods were found to most accurately represent the dolphins’ distributions. Based upon the area use results and dolphin association values (CoAs)