Modelling the emergence and stability of a vertically transmitted cultural trait in bottlenose dolphins

An apparently vertically, socially transmitted foraging specialization (‘sponging’) in bottlenose dolphins (Tursiops sp.) is observed in two adjacent gulfs within Shark Bay, Western Australia, where sponging has possibly spread from independent innovations. We designed an individual-based model based on empirical data, to investigate the conditions (of learning fidelity and fitness benefits for spongers) under which sponging could be established and maintained. Simulations show that sponging is unlikely to be established from a single innovation event but the probability increases with independent innovation events. Once established, however, it can be maintained in the absence of fitness benefits for spongers, if learning fidelity of daughters is virtually 100%. Smaller learning fidelities can be compensated for with fitness benefits for spongers, but these benefits must be 5% and 10%, respectively, to compensate for learning fidelities of 96.25% and 92%. Furthermore, we estimated the time since the emergence of sponging by tracking the average pairwise relatedness among spongers over time and comparing it to empirical estimates. For the eastern gulf of Shark Bay, we show that sponging might have been in place for at least 120 years if it originated from a single innovation event. For comparison of vertical, social transmission to other trait acquisition methods, we ran simulations in which sponging was either a genetic trait or acquired solely by innovation. In these simulations, sponging could be maintained but the simulation results and empirical data did not match. Hence vertical social transmission is a more feasible mechanism to explain the spread of sponging.