Associations among wild orang-utans: sociality, passive aggregations or chance_?

Encounters and associations between adult male, adult female and subadult male orang-utans, Pongo pygmaeus, were examined at the Kutai National Park and Gunung Palung Nature Reserve, Indonesia. Observed rates of encounters, durations of associations and proportions of time spent in associations of two members of each age-sex class from each population were compared with expected values generated from null models of associations. Encounter rates, durations of associations and proportions of time spent in associations were greater than expected by chance. Two factors appear to generate these non-random patterns: subjects at both sites met conspecifics at food sources more often than expected, and associations frequently involved social behaviour. Three differences in association patterns existed between populations. Orang-utans at Kutai encountered each other more frequently, engaged in longer associations, and spent a greater proportion of time together than conspecifics at Gunung Palung. Variations in association patterns between populations reflected age-sex class differences in behaviour. In contrast to adult and subadult males at Kutai, males at Gunung Palung avoided conspecifics. Similarly, adult females and subadult males at Gunung Palung spent less time with others than did their Kutai counterparts. An analysis of the nature of associations revealed two additional behavioural differences between populations. Mating did not occur at Gunung Palung, but was observed frequently at Kutai, and subadult males at Kutai received aggression, while those at Gunung Palung did not. These data support the hypothesis that variations in the reproductive states of adult females account for differences in the patterns of association between orang-utans. Sociality among orang-utans appears primarily the result of aggregation at common resources, either food or mates. The description and the analysis of social systems are two major areas of ethological research. Animals of many species form non-random associations, ranging from breeding pairs to large hunting parties, but individuals of several other species are found both alone and in aggregations (see reviews in Wilson 1975). For highly gregarious species, the description of social patterns is relatively straightforward. In contrast, progress in understanding the social systems of animals that form only temporary groups has been hindered by the lack of a standard against which to measure the degree of sociality. Recent models developed to examine mixed-species associations of primates, #To whom correspondence should be addressed. 0003-3472/91/070033 + 14 $03.00/0 however, provide such a standard, and an appropriate test with which to assess the cohesiveness among animals (Waser 1982, 1984, 1987; Whitesides 1989). These models generate null hypotheses of the expected encounter frequencies, durations of associations and proportions of time spent in associations based on the assumption that individuals move independently and randomly. In this paper we employ these models to investigate associations between orang-utans, Pongo pygmaeus, an apparent exception to the near universal pattern ofsociality among higher primates (see reviews in Smuts et al. 1987). George Schaller (1961) initiated modern field studies of orang-utans with brief observations in Borneo. Schaller was able to discern little about 9 1991 The Association for the Study of Animal Behaviour