KroeberAnthropological Society Papers, Nos. 71-72, 1990 Colobine Socioecology and Female-bonded Models of Primate Social Structure Craig B. Stanford Ecological models ofprimate social systems have been used extensively to explain the variationsfound in social organization among living primates and to accountforprimate sociality itself. Recent attempts to characterize primate social systems as either 'female-bonded" or "non-female-bonded" establish a typology that does notfully consider the variation inpatterns ofsex-biased dispersal seen in the Primate order. Thispaper uses as its example the Old World monkey subfamily Colobinae to show that ecologi- cal models are basedprimarily onfrugivorous, territorialpinmates. The models are inadequate to explain patterns ofintergroup competition and should not be usedfor setting general rulesforprimate societies. Apreliminary alternative view ofsex-biased dispersal that is basedonfrequency dependence is offered. INTRODUCTION bonded according to each species' typical pattern Trivers (1972) and Emlen and Oring (1977) of sex-biased dispersal. Males are assumed to outlined the hypothesis that females and males have a greater lifetime reproductive potential than have been selected to invest their lifetime energies females, and in most species they invest less in differently: females in maintaining access to offspring than do females (Trivers 1972). Since maintenance and growth resources, in order to food intake for an individual female primate is invest most in their offspring; and males in a maximized by feeding singly, the evolution of reproductive strategy that maximizes access to fe- primate sociality suggests that there is some bene- males, investing relatively little in offspring. In fit accruing to females who forage as a group. short, females should compete for food, and Wrangham (1980) considered this benefit, which males should compete for females. Wrangham outweighs the costs of intra-group competition, extended the logic of ranking evolutionary pres- to be the ability of females in groups to displace sures on social systems in this way to nonhuman and defend high-quality food patches from other primate societies, first to explain ape social sys- groups. Because the joint defense of food pat- tems (1979) and later to explain primate societies ches requires females to act in unison for in general (1980, 1987). His female-bonded individual reward, kin selection theory was in- model of the evolution ofprimate social structure voked to account for the fitness advantages of observed that most primate species are found in female kin cooperation. Other corollaries of the groups ofrelated females accompanied by males, model are that in female-bonded species, females and hypothesized that this grouping pattern is the should take an active role in territorial defense product of the competing costs and benefits of and are expected to influence the timing and di- living and foraging socially rather than singly rection of group travel (Wrangham 1980). (Wrangham 1980, 1987). The model thus depends on certain ecolo- This paper examines the applicability of this gical conditions, such as discreetly located theory to the Old World monkey subfamily Co- patches of food which are large enough to accom- lobinae and argues that ecological models of modate a group of females and are defensible by female-bonded primate groups fail to explain the them. In order to explain the evolution of the range of social adaptation found in this large pri- social systems of primate species that cannot be mate taxon. While acknowledging that a male/ considered "female-bonded", one or more of female asymmetry in mating strategies should be these ecological conditions must not be true. expected, I will argue that the evidence suppor- Common chimpanzees (Pan troglodytes), moun- ting both intergroup feeding competition and sex- tain gorillas (Gorilla gorilla beringei), hamadryas biased dispersal is weak, and that using field data baboons (Papio hamadryas) and red colobus from colobine studies to test the model leads to monkeys (Procolobus badius) are four species conclusions quite different from those of the labelied "non-female-bonded". They are consid- female-bonded model. By implication, the model ered to live predominantly in male kin groups is held insufficient as a unitary explanation for the with female dispersal for the ecological reason evolution ofprimate societies. given above, unlike female-bonded, male-disper- The female-bonded model characterizes sal species. The rest of this paper considers piimate societies as either female or non-female whether this dichotomy is warranted. 22 NON-FEMALE-BONDED SPECIES local ecologies, tend to subsist on a diet that is low in mature leaves but switch to preferred fruit Of the four primate species identified as non- and leaf flush whenever possible. It also sug- female-bonded or female transfer, the hamadryas gests that the colobine digestive adaptation may baboon should probably be considered a special vary among species in form and function rather case in that female transfers occur almost exclu- than be a uniform suite of morphological and sively between higher level strata of the social physiological traits that aid in the digestion of organization ("clans") and thus, females may not low-quality foods. be permanently leaving the local breeding pool. Most colobines live in territorial one-male In mountain gorillas and common chimpanzees, groups (Table 1). Males disperse from these females do not have affiliative relationships, and groups at or before sexual maturity, or are driven females of both species may emigrate from their off by the resident male, though exceptions exist natal groups, although in common chimpanzees in which subadults remain in their natal groups to the transfer may be more of a "visit" with a later breed. Females transfer between groups in many return to the natal group than a permanent emi- species; the earlier view of social units that are gration for reproductive purposes. The fourth essentially closed except for the dispersal of species, the red colobus monkey, seems to be young males and invasions by outside males is highly variable in social structure, and the same almost certainly simplistic. Little data exist on population may show quite different interannual female dominance relationships and where they levels of female emigration (Marsh 1979; Smith have been documented dominance is most often pers. comm.). The incidence of intensive pre- simply labelled "strong" or "weak". Female dation on some red colobus populations by dominance relations, or the lack of them in most chimpanzees adds to the difficulty in establishing colobines, are important because they may be re- functional bases for social structure; the male- lated to the relatively high rates of female bonded kin groups at the core of some red emigration observed among colobines (Stanford colobus populations may be the evolutionary re- in preparation). McKenna (1979) viewed the sult of advantages of male cooperation against relationships between female colobines to be attacking chimps. The relative effectiveness of strongly influenced by the uniform distribution of predator mobbing by related males versus single- their leafy diets, which minimized intragroup male anti-predator attacks has not been tested. competition and allowed a weakening ofthe dom- inance hierarchy. The assumption that colobines use leaves as a year-round staple diet has been COLOBINE SOCIOECOLOGY questioned, however, and such a dietary basis for dominance is not clear (Stanford 1989). The anthropoid subfamily Colobinae con- As field data accumulate on more colobine sists of approximately 35 species of African and species, most species living in one-male groups Asian monkeys. The range of social and mating have been reported to have a regular if unknown systems found in this group varies from mono- rate offemale transfer (Stanford 1989; Moore and gamy (Presbytispotenziani) to multi-male groups Ali 1984; Davies 1987; Bennett 1983). It is with female dispersal (some red colobus popu- therefore probably incorrect to classify the social lations) to territorial one-male groups (most systems of this subfamily as either "female-bon- Presbytis spp.), with some species displaying ded" or "female-transfer". Although Wrangham great intraspecific variation (Presbytis entellus). (1980) allows for a "restricted" amouni of trans- The group is characterized by a specialized gut fer by females in female-bonded groups, there is anatomy that is able to break down plant cell wall a continuum of female transfer rates. For exam- materials and, using symbiotic gut flora, digest ple, some red colobus populations have 100% mature leaf material that is typically avoided or female dispersal (Kanyawara, Kibale, Uganda; indigestible by other primates (Bauchop 1978). Struhsaker 1975), while others are characterized This adaptation has led to an unwarranted gen- by a much lower and perhaps variable frequency eralization that the colobines are leaf-eaters. (Tana River, Kenya; Marsh 1979; Smith pers. Although clearly adapted to coping with a comm.). Despite numerous field studies and seasonally poor diet, a diversity of diets from thousands of observer hours female transfer has seed-eating (Presbytis rubicunda, Davies 1984; rarely been reported in Hanuman langurs. Thus, Colobus satanus, McKey and Waterman 1982) to there appears to be wide variation in the pattern unripe fruit (Procolobus badius, Struhsaker and frequency of female transfer