Sociobiology of Rhesus Monkeys Iii: the Basic Communication Network

Abstract

AbstractThe observed distribution among age-sex classes of socially significant behavior in free-ranging rhesus monkeys on Cayo Santiago was tested against a series of progressively stronger null hypotheses. The first hypothesis was of purely random interactions, in which it was assumed that all individuals were equally likely to interact, all were as likely to be actor as to be object of any social signal, and all chose their social partners at random. This model proved untenable. In particular, adult males and females interacted with each other much more often than expected from the random model, and adult males interacted more often than expected with other males, both adult and juvenile. Juvenile females tended to interact relatively infrequently with other animals of any age-sex class. Adult males and juveniles of both sexes seldom interacted with infants. A second model, which took into account these age changes and sex differences in reactivity, but ignored the possibility of social preferences and assumed that interactions could be predicted from the overall reactivity or sociability of the monkeys involved, was also inadequate to account for the observed results. Apparently rhesus social behavior not only changes with age and differs with sex, but also is affected by "attractions" and "repulsions" within the group. The result was frequent interaction between adult males and females, between adult females and infants, and among juvenile males. Juvenile females, despite their general nonsociability, were apparently attracted to juvenile males, with whom they intracted more often than expected. Males, both adult and juvenile, tended to avoid infants, and vice versa. Adult males tended to avoid and to be avoided by juveniles of both sexes. Finally, the effects on paired interactions of these age changes, sex differences attractions and repulsions constitute a new model expressed in the form of an action constant and a receptivity constant for each age-sex class, and two interaction constants for each pair of age-sex classes. A method is given for extrapolating the results to populations with other compositions.