Integration of Genetic and Behavioral Data for Bottlenose Dolphin Population Unit Definition
By Karen Martien, PhD, Anna Sellas, MSc, Patricia Rosel, PhD, and Randall Wells, PhD

Defining population units of bottlenose dolphins in such a way as to be biologically-meaningful as well as useful to wildlife managers has long been a problem. We used a new approach of parameter estimation to define genetic population units for bottlenose dolphins from the west coast of Florida. The study area centered on Sarasota Bay, Florida, an area from which extensive behavioral and photo-identification data are available. The genetic samples were first stratified into hypothesized units using Boundary Rank (BR), a geographically constrained clustering method. The resulting units were then evaluated by estimating dispersal rates between them using the software program Migrate. The BR analysis requires that the genetic samples first be grouped into small initial units, a step that can be difficult when samples are continuously distributed. In order to determine the impact that the definition of initial units had on our results, we ran two analyses using two different sets of initial units. For the 'na?e' analysis, initial units were defined strictly on the basis of the geographic proximity of samples. For the 'expert' analysis, extensive behavioral data, including the sighting histories of the animals sampled in the study, as well as information on the fine-scale habitat variation across the study area were integrated into the definition of initial units. Analyses using the two sets of initial units produced very similar results, suggesting that BR could be successfully applied for the definition of management units in areas from which extensive behavioral data are not available. Dispersal rate estimates between the population units suggested by BR were on the order of a few genetic dispersers per generation between Tampa Bay, Sarasota Bay and Charlotte Harbor. These low estimates of dispersal suggest that bottlenose dolphins in the Gulf of Mexico are structured on a very fine geographic scale, and should therefore be managed on a fine scale if they are to be adequately protected. Support for this project was provided by NOAA Fisheries and the Harbor Branch Oceanographic Institution’s Protect Wild Dolphins program