Infanticide and the Evolution of Pair Bonds in Nonhuman Primates

Explanations for the evolution of male-female bonds originally emphasized the apparent importance of reproductive context, or mating system. Durable ‘‘pair bonds’’ were regarded as typical, or even diagnostic of monogamous species3 such as Asian gibbons (Hylobates spp.)4 and neotropical titi monkeys (Callicebus spp.).5 On the other hand, polygynous monkeys such as guenons (Cercopithecus spp.)6 and squirrel monkeys (Saimiri spp.)7 exemplified relatively undifferentiated or weak heterosexual relationships. The theoretical rationale for this distinction derived from the fundamentally divergent reproductive interests of the sexes, as articulated by Trivers8: Males are more likely to increase reproductive success by acquiring multiple mates, whereas females enhance fitness by discriminative choice of individual mates. Consequently, to the extent that an enduring bond with an anestrous female limits a male’s sexual access to additional estrus females, this social arrangement should be relatively rare in polygynous systems. Conversely, when, for whatever reason, a male restricts his mating to a single female, strong heterosexual bonds are not only less costly to males, but may also offer fitness advantages to both participants. The proposed benefit to the female of this arrangement is the parental care her offspring receive from a male that is certain of paternity.9 Thus, cohesive male-female relationships were originally viewed as part of a co-evolved suite of behaviors including monogamy, biparental care, and, in gregarious animals, ‘‘nuclear families’’ of parents and offspring.3 Essentially the same argument has been offered in some models of hominid evolution.10–12 This argument has been challenged from at least two directions. First, avian studies reveal that mating exclusivity and certainty of paternity are not natural corollaries of ‘‘social monogamy’’13,14 and that males may,15 but more commonly do not, adjust their parental effort accordingly.16 Furthermore, recent modeling and computer simulations suggest that substantial parental care by monogamous males does not invariably result from a high certainty of paternity combined with a male’s ability to increase offspring survival through such care.17 Thus, in some birds,18 as well as in nonhuman primates,19 it is not always clear whether male care of infants constitutes parental effort or mating effort. Second, it is now clear that variation in male-female social relationships is not meaningfully differentiated by mating system.20,21 Not only may monogamy entail weak heterosexual attachment,22 but strong bonds between males and anestrous females occur in a number of polygynous primates, including ring-tailed lemurs (Lemur catta),23 rufous lemurs (Eulemur fulvus),24 capuchin monkeys (Cebus spp.),25–27 rhesus and Japanese macaques (Macaca spp.),28–31 mountain gorillas (Gorilla gorilla),32–34 and savanna baboons (Papio cynocephalus).35–38 Because paternal care in polygynous primates is typically, though not uniformly, less direct and extensive than it is in monogamous species, recognition of these patterns has directed analytical attention to other potential adaptive causes of malefemale bonds. One of these is infanticide by adult males. Infanticide, the killing of unweaned young by conspecifics, is widespread among animals.39–41 Hrdy42 hypothesized that infanticide creates breeding opportunities for males that, as a consequence of intrasexual competition, have limited sexual access to fertile females. With some possible exceptions, such as the chimpanzee, Social relationships between adult males and females vary widely among mammals. In general, interactions between the sexes, particularly those of an affiliative nature, are associated with and, indeed, often limited to the period of copulation or female estrus.1 Nevertheless, cohesive male-female bonds persist beyond estrus in some species, particularly nonhuman primates,2 for reasons that remain largely obscure. Protection from male infanticide has been offered as a potential benefit to females of bonds with males in a variety of primates, including mountain gorillas and gibbons. Here I evaluate this hypothesis within a comparative framework that considers alternative costs and benefits of social relationships between the sexes.