Bergman . J . C . Beehner . D . L . Cheney .

To date, research on testosterone and behavior has focused on individuals, even when studying social behaviors that necessarily involve multiple participants. Here, we explore male responses to other males of different dominance ranks and testosterone levels in a population of wild baboons. In chacma baboons (Papio hamadryas ursinus) of the Okavango Delta, a male’s testosterone is related to his rank trajectory and, therefore, the threat he poses to other males. To examine the effects of testosterone and rank on male–male interactions, we used playback experiments to measure how a target male responded to the simulated approach of another male, scoring responses by whether or not the subject moved away from the speaker in the first minute. High testosterone subjects did not move away from the speaker more often than low testosterone subjects, but high testosterone callers elicited a move more often than low testosterone callers. When the combined testosterone of the subject and caller was high, moves were most common. The rank relationship between subject and caller did not predict moves, but the effect of combined testosterone on moving was most pronounced in adjacently ranked males. Adjacently ranked, high testosterone males are the most likely to be competing for each others’ rank, and our experiments on these dyads elicited the most moves. Both behavioral and experimental observations indicate that testosterone may be more important than the rank relationship in predicting the outcome of male–male interactions. Furthermore, combined information on the testosterone of both males was the best predictor of results, highlighting the utility of dyadic analyses when relating testosterone to behavior.