Time-matched grooming in female primates? New analyses from two species

The parcelling model of reciprocity predicts that grooming partners will alternate between giving and receiving grooming within grooming bouts, and that each partner will perform approximately as much grooming as it receives within each bout ('time matching'). Models of allogrooming based on biological markets theory predict that individuals of lower dominance rank will exchange grooming for tolerance from high-rankers, and therefore an inverse relation will be found between grooming partners' dominance rank distance and how closely they match each other's grooming contributions within each bout. We used weighted logistic regression and weighted least-squares regression to test these predictions using data from female white-faced capuchins, Cebus capucinus, and bonnet macaques, Macaca radiata. Only 5-7% of macaque grooming bouts, and 12-27% of capuchin grooming bouts, were reciprocated. However, (1) the duration of grooming by the first groomer significantly predicted whether the groomee would reciprocate at all, and (2) when bouts were reciprocated, the duration of grooming by the first groomer significantly predicted the duration of grooming by the second groomer. Grooming was most balanced among females of similar dominance ranks. Both the time-matching and rank-related effects were weak, although significant. These results indicate that although some form of time matching may be a general characteristic of grooming in female-bonded primate species, time matching accounts for relatively little of the variation in the distribution of grooming within bouts. We also draw attention to weighted regression as a technique that avoids pseudoreplication while using all available data. (C) 2004 The Association for the Study of Animal Behaviour. Published by Elsevier Ltd. All rights reserved.