Abstract
Sexing cetaceans typically requires extended periods of observation or expensive molecular methods. An alternative approach using photo‐identification may provide a cost‐effective, noninvasive method for assigning a sex to free‐ranging individuals. We investigated two methods for predicting the sex of common bottlenose dolphins ( Tursiops truncatus ) from Namibia, using the rate of change in dorsal fin (1) trailing edge marks or (2) surface features (scarring and lesions) and binary logistic regression models. Our results suggest animals with frequent and extensive changes to dorsal edge marks, or a high number/proportion of the dorsal fin covered in surface features are likely male, while those with little to no changes in dorsal edge marks or low numbers/proportion of surface features are likely female. Both sex prediction models performed reasonably well (accuracy: edge marks: 100%, surface features: 83%), with a high degree of consensus between models in sex predictions (78.6%). Our models are straightforward to implement and robust for the study population, providing useful insights into individuals' sex and population demographics. This reliable, inexpensive, and minimally invasive framework for assigning sex could be widely applicable, if population‐specific information of known sex individuals is included, due to the variable social systems and behaviors exhibited by Tursiops spp.