Biol. Lett. doi:10.1098/rsbl.2009.0204 (2009)

Game theory can explain the evolution of cooperation in large, randomly structured groups through the balance between costs and benefits. However, groups of primates that exhibit cooperative behaviour are often neither large nor randomly structured.

Bernhard Voelkl and Claudia Kasper at the Strasbourg arm of the CNRS, France's basic-research agency, examined data on 70 primate groups ranging in size from 4 to 35 individuals, taking in a total of 30 species. They modelled the groups to see whether, over time, individuals would be better off cooperating or not.

They found that 61 of the groups were more likely to cooperate than they would have been in a less structured group of the same size, suggesting that the structure assists in the fixation of cooperative behaviour.