How life in a tolerant society affects the attention to social information in baboons

Abstract

An animals' ability to classify relationships between conspecifics according to kinship, affiliation and dominance constitutes an important adaptation to life in social groups. Yet, variation in the degree of competition as well as in association patterns is assumed to put differential pressure on social cognition. Because little is known about social cognition in more tolerant societies, we investigated the social knowledge of Guinea baboons, Papio papio, ranging in the Niokolo Koba National Park, Senegal. Guinea baboons live in a fluid multilevel society with ‘units’, ‘parties’ and ‘gangs’. Units comprise a ‘primary’ male, a small number of females and young. Male–female associations may last between a few weeks to several years, and females freely transfer between males at all levels of the society. Male–male relationships are characterized by high spatial tolerance and low level of competition. We examined whether males tracked changes in male–female associations at the level of the unit. We played back grunt exchanges between a female and a primary male to 25 males and recorded male responses. The grunt sequences simulated an affiliative interaction between individuals from different units (potentially indicating a female transfer to a new male, hereafter ‘inconsistent’ condition) or from individuals of the same unit (‘consistent’ condition). Surprisingly, males looked longer when exposed to the consistent compared to the inconsistent condition. In the more competitive chacma baboons, Papio ursinus, in contrast, animals responded more strongly to experimentally simulated deviations from existing social patterns. Our study suggests that the value of social information indeed varies with the degree of competition of Guinea baboons: because of female choice, little is at stake for males when a female transfers to a new male. Furthermore, the high gregariousness of the animals may favour the classification of deviant interaction patterns, at least initially, as ‘social noise’.