Comparison of multi-group social network analyses of grooming behaviour in captive chimpanzees (Pan troglodytes) and bonobos (Pan paniscus)

Abstract

Despite their close genetic relatedness, chimpanzees (Pan troglodytes) and bonobos (Pan paniscus) differ markedly in their social organization, forms of cooperation, and dominance structures. Grooming, one of the major affiliative behaviours in primates, plays a central role in reinforcing social bonds and maintaining group cohesion. Inter-species differences in grooming patterns and network structure have already been documented (Rodrigues & Boeving, 2019), yet multi-group comparative analyses in captivity remain scarce.

Our first objective was to assess the effect of species on grooming network measures. We then examined the influence of additional individual factors (sex, age, rearing history) as well as one group-level variable (group size) on several centrality metrics (eigenvector centrality, in-/out-degree, in-/out-strength), using data from 18 social groups (10 chimpanzee groups and 8 bonobo groups).

Our results highlight four general patterns. First, interspecific differences emerged: bonobos occupied more central positions and received grooming from a greater number of partners than chimpanzees, which aligns with their more tolerant social structure. Sex-related effects remain generally limited. However, the only significant interaction between sex and species concerns the number of incoming grooming partners (in-degree): in chimpanzees, females received grooming from fewer partners than males, whereas no comparable sex difference was observed in bonobos. For the other metrics (out-degree, in-strength, out-strength, eigenvector centrality), no sex effects were detected in either species. Second, the analysis revealed a strong and recurrent quadratic effect of age on all grooming network measures considered here (degree, strength, and eigenvector centrality). This indicates that these metrics do not vary linearly with age: they generally increase up to a middle-aged peak and then decline in older individuals. Third, rearing history influenced network position, with mother-reared individuals showing higher grooming strength, degree, and eigenvector centrality than those reared in atypical conditions. Finally, centrality decreased with increasing group size. In other words, the larger the group, the more social relationships are “diluted,” and the less possible it is for an individual to maintain a highly central position within the network.

This large-scale comparative study highlights species-specific characteristics while also suggesting substantial variability among groups within the same species. This indicates that life-history trajectories and demographic context jointly contribute to shaping social integration. It also highlights the value of multi-group social network analysis for disentangling the interactions between individual-, group-, and species-level determinants of primate sociality. Moreover, studies conducted in captivity provide valuable opportunities to control ecological variation and thereby isolate more clearly the effects genuinely associated with interspecific differences.