Nonconvergence in the evolution of primate life history and socio-ecology

The goal of this study was to investigate the extent of convergence in four basic life history and socio-ecological traits among the primates of Africa, Asia, South America and Madagascar. The convergence hypothesis predicts that similar abiotic conditions should result in similar adaptations in independent taxa. Because primates offer a unique opportunity among mammals to examine adaptations of independent groups to tropical environments, we collected information on body mass, activity pattern, diet and group size from all genera for quantitative tests of this hypothesis. We revealed a number of qualitative and quantitative differences among the four primate groups, indicating a lack of convergence in these basic aspects of life history and socio-ecology. Our analyses demonstrated that New World primates are on average significantly smaller than primates in other regions and characterized by a lack of species larger than about 10 kg. Madagascar harbours significantly more nocturnal species than the other regions and is home to all but one of the primates with irregular bursts of activity. Asia is the only region with strictly faunivorous primates, but lacks primarily gummivorous ones. The Neotropics are characterized by the absence of primarily folivorous primates. Solitary species are not represented in the New World, whereas solitary and pair-living species make up the majority of Malagasy primates. Lemurs live in significantly smaller groups than other primates, even after controlling for differences in body size. The lack of convergence among the major primate groups is neither primarily due to phylogenetic constraints as a result of founder effects, nor can it be sufficiently explained as a passive consequence of body size differences. However, because the role of adaptive forces, such as interspecific competition, predation or phenology in shaping the observed differences is largely unexplored, we conclude that it is premature to discard the convergence hypothesis without further tests.