Variation in Response to a Repeated Intruder in a Terrestrial Frog and Explaining Variation in Assessment Mode During Animal Contests

Stephen Heap

Department of Zoology, University of Melbourne

I will give two 10 minute talks. The first is a practice talk for an upcoming conference detailing a field experiment conducted on frog calling behaviour. The second is an outline of a current problem regarding assessment abilities during animal contests for which Mike Mesterton-Gibbons and I are collaborating on a game-theoretic model to help explain.

Experiment: Territorial animals must optimise their investment towards exploiting their territory and defending it from competitors. However, the manner in which territorial investment changes in response to repeated interactions with a rival remains unclear. We experimentally exposed nest-defending terrestrial toadlets (Pseudophryne bibronii) to playback that simulated a repeated intruder and measured the change in a resident's calling behaviour attributable to the speaker. Results suggested that residents independently adjust their exploitative and defensive behaviors in response to intruders as they gain information over successive interactions. This process will be important to consider when there is variation in the number of times a given intruder will be encountered.

Model: Animals must often engage in contests that end when one contestant decides to retreat. However, the information that an individual uses as the basis for this retreat decision appears to vary across contexts. Some animals decide to retreat once they reach a personal cost threshold and thus only use information on their own state. In contrast, others compare their own state to that of their opponent and then decide to retreat if they are outmatched. There is currently no explanation regarding the functional significance of these different assessment strategies and how or why they vary depending on environmental and social contexts. Mike Mesterton-Gibbons and I are currently collaborating on a game-theoretic model that aims to address this knowledge gap and provide testable predictions for future empirical work.