Recent Developments in the Study of Wild Chimpanzee Behavior

Evolutionary Anthropology 9

ARTICLES

JOHN C. MITANI, DAVID P. WATTS, AND MARTIN N. MULLER

Chimpanzees have always been of special interest to anthropologists. As our organization, genetics and behavior, closest living relatives,1–3 they provide the standard against which to assess hunting and meat-eating, inter-group human uniqueness and information regarding the changes that must have oc- relationships, and behavioral endocri- curred during the course of human evolution. Given these circumstances, it is not nology. Our treatment is selective, and surprising that chimpanzees have been studied intensively in the wild. Jane Good- we explicitly avoid comment on inter- all4,5 initiated the first long-term field study of chimpanzee behavior at the Gombe population variation in behavior as it National Park, Tanzania. Her observations of tool manufacture and use, hunting, relates to the question of chimpanzee and meat-eating forever changed the way we define humans. Field research on cultures. Excellent reviews of this chimpanzee behavior by Toshisada Nishida and colleagues6 at the nearby Mahale topic, of central concern to anthropol- Mountains National Park has had an equally significant impact. It was Nishida7,8 ogists, can be found elsewhere.12–14 who first provided a comprehensive picture of the chimpanzee social system, including group structure and dispersal. SOCIAL ORGANIZATION No single issue in the study of wild Two generations of researchers erything about the behavior of these chimpanzee behavior has seen more have followed Goodall and Nishida apes in nature. But this is not the case. debate than the nature of their social into the field. As a result, chimpanzees In fact, we are entering a new and system. We now know that chimpan- are now one of the best and most extremely exciting era in the study of zees live in a “fission-fusion” society. widely studied of nonhuman pri- wild chimpanzee behavior. The pur- Individuals form socially and geo- mates. Long-term field research has pose of this review is to highlight graphically circumscribed “unit- been conducted at six sites by several some intriguing findings that have groups” or “communities,” within researchers spanning 42 years (Fig. 1). emerged through recent study. We fo- which they associate in temporary Shorter field studies have also been cus specifically on results from our subgroups or “parties” that vary in carried out in some areas.9–11 With own field research conducted in the size, composition, and duration. this extensive body of research, one Kibale National Park, Uganda, giving Males are philopatric, whereas fe- might think that we have learned ev- special emphasis to five areas: social males typically disperse. This seem- ingly clear-cut picture of chimpanzee society did not emerge easily. Given John C. Mitani is in the Department of Anthropology at the University of Michigan. His research involves studies of primate behavioral ecology and vocal communication. He has the fluid nature of chimpanzee soci- conducted fieldwork on all five species of living apes, including gibbons and orangutans in ety, it took exceedingly long for field Indonesia, gorillas in Rwanda, bonobos in the Democratic Republic of Congo, and chimpan- observers to discern regularities in zees in Tanzania and Uganda. Along with David Watts, he currently co-directs the Ngogo chimpanzee project, Kibale National Park, Uganda. grouping, dispersal, associations, and range use. David Watts received his BA from the University of Pennsylvania and his MA and PhD in 15 Anthropology from the University of Chicago. He is currently a Professor of Anthropology at Kortlandt was the first to report Yale. He has done fieldwork on white-faced capuchin monkeys in Panama and, for many temporary associations among wild years, on the social behavior and ecology of mountain gorillas in Rwanda. He also served as chimpanzees based on early observa- Director of the Karisoke Research Centre in Rwanda for two years. Since 1993, he has been conducting research on the behavior and ecology of chimpanzees at Ngogo, in Kibale tions in the Belgian Congo. Here he National Park, Uganda, in collaboration with Dr. John Mitani and Dr. Jeremiah Lwanga. described groups of 1–30 individuals Martin N. Muller has studied chimpanzees in Tanzania and Uganda. He recently completed that either contained members of his dissertation, “Endocrine Aspects of Aggression and Dominance in Chimpanzees of the both sexes or consisted of “nursery” Kibale Forest,” for the Department of Anthropology, University of Southern California. He is currently a postdoctoral fellow at Harvard University, where he continues his research on the groups of females and their young. behavior of wild chimpanzees. Kortlandt was prescient in his de- scriptions of the temporary and fluid nature of chimpanzee aggregations Key words: chimpanzees, behavioral ecology, Pan troglodytes and laid the groundwork for further Evolutionary Anthropology 11:9–25 (2002) study. Subsequent reports by Good- 10 Evolutionary Anthropology ARTICLES

the choice of companions. Males are more gregarious than females and prefer each other’s company, except when females are in oestrus. Females are less sociable and spend most of their time with their own offspring— except when cycling, at which time they become very sociable.” Consideration of sex differences in reproductive strategies and the costs of feeding competition provide a theoretical rationale to explain observed sex differences in chimpanzee association patterns. Females forage alone because the potential reproductive costs of scramble and contest feeding competition are higher for them than they are for males.21,26,27 Males may be more willing to assume the costs of feeding competition to gain mating opportunities and to derive social benefits from associating with other males.21,26,27 Part of the power of this model lies in its ability to accommo- date intraspecific variation in associa- Figure 1. Chimpanzee study sites. tion patterns. The theoretical predic- tion is that such variation will occur between study sites because for fe- all16 and the Reynolds17 tended to blur ties are not closed. While males typi- males the costs of grouping are ex- distinctions between communities. cally spend their entire lives in their pected to vary with differences in local After five years of field observations at natal communities, females com- resource abundance and distribu- Gombe, Gooodall16 wrote: “Since monly transfer to neighboring ones tion.28,29 chimpanzee groups in the reserve during adolescence.8,18,19 In keeping with this model, consid- freely unite from time to time without While effectively laying to rest per- erable intraspecific variation exists in signs of aggression, they cannot be sistent questions regarding the exis- chimpanzee association patterns. As divided into separate communities. It tence of chimpanzee communities, far back as 1979, Sugiyama and seems likely that only a geographic Nishida7 simultaneously described Koman30 reported that females were barrier would constitute a limiting sex differences in association. He as sociable as males in a small, iso- factor on the size of a community, noted that males associated more fre- lated community at Bossou, Guinea. although individuals living at opposite quently with each other than females Ghiglieri31 found no evidence of sex ends of the range might never come did with other females. From this he bias in grouping tendencies in the un- into contact.” concluded that strong bonds form be- provisioned and largely unhabituated Nishida7 altered this picture by de- tween males and that males compose Ngogo community at Kibale. More re- fining the social group of wild chim- the core of chimpanzee society. Sub- cently, Boesch32 and Boesch-Acher- panzees at Mahale. In contrast to sequent field research at Gombe,20,21 mann,33 working in the Taı¨ National Goodall and the Reynolds, Nishida Mahale,22,23 and the Kanyawara study Park, Ivory Coast, observed that most emphasized the stable nature of chim- area of Kibale National Park, Ugan- parties contained adults of both sexes, panzee communities. Although he da,24 validated and expanded Nishi- while parties consisting of only males never recorded all community mem- da’s picture of chimpanzee society. or only females were rare. This led bers together at a single place and Male chimpanzees in these popula- them to conclude that members of time, his longitudinal observations of tions are more gregarious and distrib- this population are “bisexually bond- association between individuals re- ute their activities more widely and ed.” Given a lack of information revealed an unambiguous social net- evenly over their territories than do garding feeding behavior and food work and structure (Fig. 2). From females. Goodall25 aptly summarized availability across study sites, we are these observations, Nishida7 con- the standard picture of chimpanzee not presently capable of evaluating cluded that, “The chimpanzees live a society to emerge from these studies: whether sex differences in the costs of clear-cut social unit which consists of “The most deep-seated principles un- competition account for these re- adult males, adult females, and imma- derlying chimpanzee community ported intraspecific differences. Tem- ture animals.” Subsequent field work structure are those concerned with poral variation in association patterns indicated that chimpanzee communi- sex differences in sociability and in at Taı¨ appears to be consistent with ARTICLES Evolutionary Anthropology 11

Figure 2. Early observations at the Mahale Mountains established the ﬁssion-fusion nature of chimpanzee society. Here a temporary subgroup or “party,” consisting of individuals of all ages and sexes, is shown.

this hypothesis,34 but again firm con- adic association indices.40 Inferring in individual gregariousness and clusions are elusive due to an absence social affinities from these indices, group demography.39 For example, of measures of food availability. Using high indices between males may indi- such measures, our own observations cate genuine social preferences or conform to expectation by showing In keeping with past may simply result from either a ten- that anestrous and lactating females dency to aggregate or a male-biased at Ngogo become more sociable dur- studies, we have found community sex ratio. ing periods of food abundance.35 that males at Ngogo are To overcome these problems, we Problems in recording, measuring, have used recently developed associa- and analyzing associations among more gregarious than tion indices and randomization tech- chimpanzees, a species with a fluid are anestrous females. niques to analyze associations among social system, have plagued field re- After controlling for the chimpanzees living in an unusually searchers for years. Accordingly, large community at Ngogo, Kibale some of the reported intraspecific overall general National Park, Uganda. In keeping variation may be more illusory than gregariousness of males, with past studies, we have found that real. For example, some of the differ- males at Ngogo are more gregarious ences may simply reflect differences however, we found that than are anestrous females.39 After in how observers record party mem- males do not associate controlling for the overall general gre- bership.36 Other differences can be at- with other males more gariousness of males, however, we tributed to the different ways that found that males do not associate field researchers have defined age-sex often than would be with other males more often than categories.37 More serious problems expected by chance. would be expected by chance.39 Alter- ensue from attempts to assay and an- natively, anestrous females associate alyze association patterns to make in- with each other more frequently than ferences about social affinities and chance expectation after taking their bonds.38,39 Standard methods, pio- relatively low levels of sociability into neered by Nishida7 in his original re- however, is problematic because they account.39 We note that even these research, typically involve the use of dy- are subject to biases due to variations sults are vulnerable to methodological 12 Evolutionary Anthropology ARTICLES criticism. As in other studies, our ob- important effects on the genetic struc- also frequently mate in this context. servations of female behavior were ture of chimpanzee populations.43–45 Males additionally mate in three re- unlikely to compose a random sam- strictive situations. In one, high-rank- ple. Given their relatively asocial na- ing males attempt to control mating GENETICS AND BEHAVIOR ture, female chimpanzees are elusive access to females, while mate-guard- and difficult to observe. Our ran- Recent advances in extracting, am- ing them against others in a group domization procedures nonetheless plifying, and sequencing DNA from setting. This “possessive” pattern dif- provide the kind of unbiased analyti- hair and fecal samples collected non- fers from consortships, in which cal method necessary to determine invasively in the wild are beginning to male-female pairs move away from whether intraspecific variations in as- revolutionize our understanding of other community members and en- sociation can be attributed to method- the behavior of animals.46,47 The gage in a nearly exclusive mating re- ological differences and the degree of study of chimpanzees has contributed lationship over the course of a single concordance in patterns across study to this revolution.43 Two particularly estrous cycle. Tutin48 originally de- sites. informative areas of recent investiga- scribed both males and females as While considerable intraspecific tion have involved the integration of equally interested and involved in variations in associations appear to genetic and behavioral data. Recent maintaining consortships. While this exist, differences in dispersal patterns findings have now produced clearer clearly is true in some cases, we now have also emerged. Observations from pictures of the chimpanzee mating know that males often display consid- Gombe reveal that female dispersal is system and the role of maternal kin- erable aggression toward females be- the norm there, but that only 60% of fore “persuading” them to follow.25,51 females disperse from their natal Based on our observations of the communities, with 10% doing so only Tutin originally described Ngogo chimpanzees,50 we can add a after giving birth for the first time.19 both males and females third restrictive tactic to the two al- Although all females have been previ- ready described. When the Ngogo ously described as dispersing at Ma- as equally interested chimpanzees form extremely large hale,41 new observations indicate that and involved in parties, high-ranking duos and trios a few remain and give birth in their begin to mate-guard estrous females. natal communities (T. Nishida, per- maintaining These mating-guarding coalitions ap- sonal communication). Recently com- consortships. While this pear to form, both successfully and piled data from Taı¨ show that only a clearly is true in some unsuccessfully, when male party sizes single female failed to disperse from become so large that it becomes pro- the main study community during 16 cases, we now know hibitive for a single male to mate- years of observation.33 that males often display guard successfully (Fig. 3). Research at Bossou provides a con- Early on, Tutin48 speculated that al- trast to the results from Gombe, Ma- considerable aggression though the vast majority of all copu- hale, and Taı¨. Observations over 21 toward females before lations occur opportunistically in a years suggest one successful immigra- group context, most conceptions actu- tion by an adult male, along with vis- “persuading” them to ally take place during consortships. its by two extra-community males.42 follow. Using more recent behavioral obser- Moreover, 86% of all adolescent males vations from Gombe, Wallis52 has have disappeared from the Bossou suggested that most conceptions can community. Based on these observa- be attributed to matings that occur in tions, Sugiyama42 has argued that multi-male parties, a finding that con- male dispersal occurs regularly at ship in structuring aspects of male so- curs with behavioral reports from Ma- Bossou because the community has cial behavior. hale.49,53 In chimpanzees, as with no immediate neighbors and no need It took 20 years of field study for the other primates, multiple mating and for territorial defense. Consequently, mating habits of wild chimpanzees to internal fertilization make it difficult selection for male cooperation has come into focus. Working at Gombe, to determine paternity using behav- been relaxed, and young males dis- Caroline Tutin48 was the first to de- ioral observations alone. Recent stud- perse to reduce within-group compe- scribe the alternative mating tactics ies employing genetic markers as- tition for mates. Male transfer be- employed by male chimpanzees. At sayed through noninvasive sampling tween communities is rare or absent Gombe and elsewhere,33,49,50 the vast regimes have permitted more reliable at Gombe, Mahale, and Taı¨.Were- majority of observed copulations oc- paternity assignments, which, in turn, quire more studies across a broader cur in a group context. Here chimpan- allow more definitive statements range of habitats to evaluate whether zees gather in relatively large parties about the payoffs of different male variation in the costs and benefits of containing members of both sexes, mating tactics. territoriality leads to systematic dif- and estrous females mate repeatedly Perhaps the most startling claim ferences in the frequency of male dis- with multiple males. All adult males made in recent studies of wild chim- persal. Irrespective of resolving this present typically share in such oppor- panzees involves paternity determina- question, dispersal is likely to have tunistic matings; adolescent males tions among the Taı¨ chimpanzees.54,55 ARTICLES Evolutionary Anthropology 13

Gombe. Their analyses indicate that all 14 infants were sired by resident males. Vigilant and colleagues67 have reported similar results in a sample of 41 infants from three communities at Taı¨. Here paternity could be assigned with a probability greater than 99% for 34 of these infants. All 34 infants had probable fathers who resided within the infant’s community, although complete sampling of potential fathers was achieved in only 13 cases. Of the seven remaining infants for which fathers could not be as- cribed, extra-group paternity was implicated in one for whom all potential fathers were sampled and excluded. Taken together, these new data indicate that extra-group paternity in chimpanzees occurs only rarely in the two populations investigated thus far. Additional research at other sites will Figure 3. Mate-guarding form and success vary with male party size. Mean (Ϯ 1 SE) values are shown for single male and coalitionary mate-guarding episodes at Ngogo, Kibale be required to evaluate the generality National Park, Uganda. of this finding. The discrepant results between these two studies and the earlier Taı¨ analysis have been attributed to mul- Using genetic data derived from 11 might be the case,25,64,65 however, tiple sources of error.67 Some samples microsatellite markers, Gagneux and long-term observations in the wild used in the original Taı¨ paternity colleagues assigned the paternity of gave no reason to think that extra- study may have been handled improp- 13 offspring born into the Taı¨ chim- group matings and conceptions were erly, leading to contamination and panzee community between 1990 and so common. The Taı¨ result has also sample mix-up.67 Genotyping errors 1995. Their surprising results sug- been perplexing because females who due to amplification artifacts during gested that 7 of the 13 young chim- engage in such behavior would run polymerase chain reaction68 and al- panzees did not have fathers residing the risk of receiving severe aggression lelic dropout69,70 may have exacer- in the community. The clear implica- from males in their own communities, bated these problems. The acknowl- tion drawn from this analysis was that and if discovered, of having those edgment that imperfect laboratory females actively seek matings with males kill their infants. Because re- procedures and analyses led to an er- males who live outside their own so- productive opportunities are so rare, roneous conclusion regarding the im- 33,54,55 cial group. with interbirth intervals typically av- portance of extra-group paternity Given the apparently unassailable eraging five to six years in the among wild chimpanzees67 will un- 33,41,52 nature of the genetic data, this aston- wild, we expect males to be es- doubtedly result in more careful stud- ishing result has been widely reported pecially wary, to guard against extra- ies in the future. This should also pro- 56 in major scientific publications, group matings, and to punish females mote critical evaluation of additional monographs,33 and textbooks,57 as who engage in them. In principle, the claims that are not easily accommo- well as the popular58 and news59 me- benefits of mate choice might out- dated by theory or empirical observa- dia. As a result, belief that extra-group weigh these risks,33 but females could tion. paternity is a common feature of wild gain these same benefits by simply Besides helping to resolve some out- chimpanzee behavior is now wide- transferring to another group. Alter- standing questions regarding chim- spread both inside and outside the ac- natively, extra-group paternity may panzee paternity, Constable and col- ademic community. Reconciling be- provide insurance against between- leagues66 study is notable for havioral observations and life history community infanticide.33 Any benefits providing additional information re- information with the high frequency derived here, however, would be off- garding the effectiveness of alterna- of extra-group paternity reported at set by increased costs of aggression by tive mating tactics employed by male Taı¨ has nevertheless been difficult. We males living within communities. chimpanzees and the relationship be- would not be surprised to find that, as Two recent and independent stud- tween male rank and reproduc- in many species of birds and mam- ies promise to put these questions to tion. Data regarding the mating tactic mals,60–63 female chimpanzees occa- rest. Constable and colleagues66 deter- employed by fathers was documented sionally conceive offspring with non- mined the paternity of 14 infants born for 12 conceptions. Offspring were resident males. Despite hints that this into the Kasekela community at conceived during opportunistic mat- 14 Evolutionary Anthropology ARTICLES ings five times (42%), possessive mat- tion with males from other communi- points raise the possibility that males ings four times (33%) and consort- ties.25,44,72,73 close in dominance rank may also be ships three times (25%). High-ranking Male chimpanzees develop strong so- inclined to form strong affiliative and males produced half of the infants. cial bonds with other males in their cooperative relationships.78 The alpha male sired 36% of all off- own community (Fig. 4). These bonds We have recently examined the ef- spring (5/14), four times via posses- are manifest in several contexts, includ- fects of age and rank on aspects of sive mating and once through mating ing association, grooming, proximity, male social behavior using observa- the mother opportunistically. Another coalitions, meat-sharing, and territorial tions of male chimpanzees living in high-ranking male used opportunistic boundary patrols.7,21,33,51,73–81 Given the extremely large community at matings to father two offspring (14%). the well-known effects of kinship on Ngogo.86 We have found that mem- Interestingly, he conceived one of primate behavior, chimpanzee male bers of the same age cohort and indi- these infants with his mother. Middle- bonds have often been assumed to form viduals that are close in rank are more ranking males were responsible for an between close genetic relatives.25,82 likely to affiliate and cooperate than additional 36% of all offspring, using Two independent field studies have re- are males belonging to different age and rank classes. Additional analyses opportunistic matings twice and con- cently questioned this assumption. In replicate earlier findings and show sortships twice; the mating tactic em- observations of the Kanyawara com- that males who affiliated and cooper- ployed by one middle-ranking male to munity in Kibale, Goldberg and Wrang- 83 ated were not close maternal relatives produce another infant was unknown. ham showed that male chimpanzees as assayed by mtDNA haplotype shar- Finally, two low-ranking males fa- who maintain proximity and frequently ing. A role for kinship might still be thered two infants (14%), once during groom each other are not necessarily related through the maternal line as as- implicated if reproductive skew is a consortship. high and male chimpanzees selec- This study may additionally provide sayed by mtDNA haplotype sharing and genetic distances. Our own observa- tively cooperate with age-mates who an explanation for the higher fre- are paternal siblings.87 Evaluating quency of consortships that has been tions of males at Ngogo support this finding and extend it by revealing that this possibility will require additional reported at Gombe than at other sites. mtDNA genetic relatedness is not sig- information regarding nuclear and Y Constable and colleagues66 note that nificantly correlated with levels of coop- chromosome genetic markers. Until females tended to consort with low- eration as measured by participation in these data are available, we can only ranking males when they had high- coalitions, meat-sharing, and patrols.84 conclude that demographic and social ranking male relatives in the commu- These studies raise two important factors may constrain patterns of nity. For such females, consortships and related questions: Why don’t male male chimpanzee social behavior to a may be a tactic to avoid inbreeding chimpanzees selectively bias their be- greater extent than does maternal kin- because male chimpanzees some- havior toward kin? What factors ac- ship. times attempt to force copulations count for the observed patterns of af- with their unwilling mothers or ma- filiation and cooperation among wild ternal sisters.25 Two predictions fol- male chimpanzees? Although chim- HUNTING, MEAT-EATING, AND low. First, between communities, con- panzee demography and life history MEAT-SHARING sortships should be more common in have the potential to provide answers cases where females disperse less of- Hunting and meat-eating by wild to both of these questions, they have chimpanzees captures widespread an- ten. Second, within communities, na- received scant attention. Chimpan- thropological attention because of its tal females should show higher rates zees are an extremely long-lived and obvious relevance to the study of hu- of consortship than immigrant fe- slowly reproducing species. With an man origins and evolution.88–94 Re- males. Additional study will be needed equal sex ratio at birth and high mor- search on chimpanzee hunting and to evaluate these predictions. tality among infants and juve- meat-eating also contributes more While genetic data have furnished 25,41,52 niles, the probability is rela- broadly to our understanding of tradi- novel insights into the chimpanzee tively low that a female will give birth tional ecological and ethological prob- mating system, similar information successively to sons that reach adult- lems, such as predator-prey relation- has begun to clarify the effects of ma- hood together. Thus, males will only ships and the evolution of cooperation ternal kinship on male chimpanzee rarely live with maternal kin who can in animals.95–98 behavior. Kinship plays a large and effectively join them in behaviors that Since Goodall’s4 seminal observa- important role in the lives of humans have important fitness consequences. tions, field workers at sites scattered as well as other primates,71 and chim- If kin are not generally available, then across Africa have documented the panzees are frequently used to illus- males might solicit and use others op- regular occurrence of hunting and trate the effects of kinship on primate portunistically.74,85 Individuals be- meat-eating by chimpanzees.99–104 As social behavior. Among chimpanzees, longing to the same age cohort may be a result, we now possess extensive enduring and long-lasting bonds form particularly attractive social partners data on chimpanzee prey choice, as between mothers and their off- because they grow up together, are well as hunting frequency, success, spring,5,25 while genetically related generally familiar with each other, and participation.33,98,104,105 With re- males living within the same social and share similar social interests and spect to prey choice, we know that group cooperate together in competi- power throughout their lives. Similar chimpanzees selectively hunt red ARTICLES Evolutionary Anthropology 15

colobus monkeys (Procolobus badius) everywhere they live sympatrically (Fig. 5). Hunts of red colobus occur frequently, on an average of 4 to 10 times per month. Chimpanzees are ex- traordinarily successful in preying on red colobus; hunting success rates average over 50% across study sites. Inde- pendent ﬁeld studies have consistently shown that adult male chimpanzees are responsible for the vast majority of all kills. Despite this wealth of information about chimpanzee predatory behavior, two fundamental questions have remained unanswered. First, what factors affect decisions to hunt? This question arises because chimpanzees sometimes quickly pursue red colobus on encounter, yet forego hunting at-

Hunts are costly in terms of time and energy, and during predation attempts chimpanzees take considerable risks, given that red colobus males mob them. Why, then, do chimpanzees relinquish meat, a scarce and valuable resource, to others?

tempts at other times.106,107 Second, why do chimpanzees share meat with conspecifics? Hunts are costly in terms of time and energy, and during predation attempts chimpanzees take considerable risks, given that red colobus males mob them.108,109 Why, then, do chimpanzees relinquish meat, a scarce and valuable resource, to others? Five nonmutually exclusive hypotheses have been advanced to provide answers to these two questions. Geza Teleki110 first made the simplest and Figure 4. Male chimpanzees form strong social bonds with each other. Here one adult male most direct proposal. Teleki, who con- chimpanzee hugs another (a) to seek reassurance (b). ducted the first systematic investigation of chimpanzee hunting behavior at Gombe, hypothesized that chim- 16 Evolutionary Anthropology ARTICLES

both of the variables of interest could lead to the posited association. Using long-term observations from Gombe, Pusey and colleagues116 have shown that female chimpanzee dominance rank is related to lifetime reproductive success. Given that high-ranking females are likely to obtain more meat than low-ranking females do, dominance rank is a probable confounding variable that produces a spurious correlation between female reproduction and meat eating. If neither hunger nor sex provides

Using long-term observations from Gombe, Pusey and Figure 5. Chimpanzees prey selectively on red colobus monkeys. Here an adult male colleagues have shown chimpanzee feeds on a portion of an adult female red colobus. that female chimpanzee dominance rank is panzees hunt because they are hun- chimpanzees hunt in order to obtain related to lifetime gry. Chimpanzees are frugivores that meat that they can swap for matings. rely on sugar-rich fruits that are sea- The meat-for-sex hypothesis has reproductive success. sonally available.111 Teleki proposed generated intriguing claims and has Given that high-ranking that chimpanzees hunt to compensate received considerable publicity,113,114 for the nutritional shortfalls they ex- but neither the claims nor the hypoth- females are likely to perience, primarily during the sea- esis appear to stand up well under obtain more meat than sonal troughs of fruit availability. De- close scrutiny. Unlike the situation at low-ranking females do, spite the elegant simplicity and Gombe, the presence of estrous fe- intuitive appeal of this hypothesis, it males does not predict the tendency of dominance rank is a has remained untested for more than the Ngogo males to hunt.107 More- probable confounding 25 years. Our recent studies at Ngogo over, the predicted behaviors occur have provided the ﬁrst direct infrequently at Ngogo and elsewhere. variable that produces a test.107,112 Our results show that in- At Ngogo, estrous females do not reli- spurious correlation ably obtain meat from their begging stead of hunting during fruit-poor between female times, chimpanzees actually increase efforts nor do matings typically ensue 107 the frequency of their hunting at- following meat exchanges. Most reproduction and meat tempts when fruit is abundant. tellingly, however, our observations indicate that males at Ngogo do not eating. A second hypothesis advanced to gain any mating advantage by sharing explain why chimpanzees hunt and meat with estrous females.107 share meat was also developed by One of the more surprising sugges- Teleki.110 This hypothesis has recently tions stemming from the meat-for-sex been revived by Stanford,92,98,106 who the prime motivation to hunt, what hypothesis is that female chimpan- also studied the Gombe chimpanzees. does? We suggest that a male chim- zees have a keen interest in acquiring Stanford’s work there showed that the panzee’s decision to hunt is affected meat because eating meat improves by his assessment of the likelihood single best predictor of a male chim- their reproductive performance.106 of success.107 Field studies consis- panzee’s decision to hunt was the Data compiled by McGrew115 based tently have shown that party size presence of estrous females. Stanford on observations made from Gombe25 and the number of male hunters are has used this ﬁnding, along with the revealed a positive relationship be- good predictors of hunting suc- additional observation that male tween the amount of meat a female cess.101,104,106,117 Male chimpanzees chimpanzees occasionally exchange obtains and her reproduction (Fig. 6). appear to swamp red colobus prey de- meat for matings, to argue for a pro- Care needs to be taken when inter- fenses with strength in numbers, so vocative “meat-for-sex” hypothesis. preting this result, however, for an im- that hunting success increases when According to this hypothesis, male portant confounding factor related to chimpanzees form large parties with ARTICLES Evolutionary Anthropology 17

chimpanzee hunting has been hard to operationalize. We lack an objective way to distinguish whether males who pursue prey actually coordinate their efforts to capture the prey or simply pursue their own selﬁsh strategies while taking into account each other’s movements and the response of the prey to these movements. Observers at Gombe,98 Mahale,117 and Ngogo112 report that individuals often switch between standing by and pursuing prey during the course of the same

Still open is the question of why male chimpanzees share Figure 6. Female chimpanzee reproduction and meat acquisition. Data adapted from Goodall25 (Table 11.15, p. 310). meat readily with conspecifics. One many male hunters. Ecological con- red colobus groups that they encoun- hypothesis invokes an straints limit the formation of large ter in forest with a generally low and important role for parties, however. During periods of broken canopy than those they meet low food availability, when the ecolog- in mature forest where the canopy is cooperation. ical costs of feeding competition are tall and continuous.112 Chimpanzees at Taı¨ high, chimpanzees form relatively Still open is the question of why appear to hunt small parties.27,35,118 It is only during male chimpanzees share meat readily periods of relative food abundance with conspecifics. One hypothesis in- cooperatively on a that chimpanzees regularly gather in vokes an important role for coopera- regular basis, both in the large parties. The positive relationship tion. Chimpanzees at Taı¨ appear to between party size and fruit abun- hunt cooperatively on a regular basis, sense that individuals dance provides a transparent explana- both in the sense that individuals co- coordinate their tion for our finding that hunting fre- ordinate their behavior with each quency at Ngogo increases during other and that per capita energy gain behavior with each periods of food abundance. We con- from hunting increases with the num- other and that per clude from these considerations that ber of active hunters.97,101 These ob- at Ngogo male chimpanzees hunt servations led Boesch97 to propose capita energy gain from when they are likely to be successful, that male chimpanzees selectively hunting increases with which is when they are in large parties share meat with others who have co- with several male hunters. Males operated with them to make kills. Ac- the number of active forego most hunting attempts when cording to this hypothesis, selective hunters. they are in smaller parties because the meat sharing ensures that “bystand- odds of success are greatly reduced. ers,” individuals who are present but Another finding from Ngogo is also fail to participate in hunts, do not ob- consistent with the hypothesis that tain meat and thus cannot exploit the hunting decisions depend on the like- hunting efforts of cooperators. hunt. This makes it difficult to differ- lihood of success and provides quan- We recognize that distinguishing entiate cooperators and cheaters reli- titative support for the longstanding cooperative hunters from bystanders ably because to do so requires com- proposal that chimpanzees preferen- is theoretically important. We also ac- plete information on the activities of tially hunt red colobus in situations knowledge that the cooperative hunt- all individuals present. Prevailing ob- where the monkeys have few or no ing hypothesis may explain meat shar- servation conditions often preclude escape routes.119 Hunts that occur in ing among the Taı¨ chimpanzees. obtaining complete records; during areas where the tree canopy is broken Evaluating the generality of this hy- hunts, multiple chimpanzees fre- are generally more successful than pothesis, however, will be difficult due quently pursue red colobus prey high those where the canopy is continuous to conceptual and methodological in the tree canopy over areas that and closed.112 Consequently, chim- reasons. As Stanford98 has noted, the sometimes cover several hundred panzees are much more likely to hunt notion of behavioral cooperation in meters. 18 Evolutionary Anthropology ARTICLES

The use of outcome-based criteria the meat-for-sex hypothesis lies in mination of one community at the to investigate whether hunting is co- its claim that male chimpanzees use hands of another at Gombe76; a simi- operative in an ecological sense—that meat to achieve matings. The male lar process is inferred at Mahale.72 is, whether success increases as a social bonding hypothesis, however, The functional significance of territo- function of the number of hunters— provides a more compelling ratio- riality and lethal aggression is un- also presents problems. Researchers nale for how males might use meat clear, but several fitness benefits fall have used various measures, includ- in a mating strategy. The great deal under the imbalance-of-power hy- ing some that apply to hunting parties of male sexual coercion among pothesis, which holds that attacks on as a whole, such as the probability of chimpanzees25,51 probably renders members of neighboring communi- at least one kill, the total number of female mate choice relatively unim- ties result from a motivation to dom- 80 kills, and the total amount of meat portant in this species. In contrast, inate others. This could lead to im- obtained. Alternatively, others have levels of male-male competition are proved safety, improved access to applied measures that are based on the high.25,123,124 Other male chimpan- food, and incorporation of more fe- 19,25,80,129 behavior of individuals, such as the zees are the main obstacle facing in- males into a community. amount of meat obtained per individual Assessing the imbalance-of-power dividuals who attempt to increase or the net energy gained per individual. hypothesis requires analysis of male their mating success. Forging long- The results of some studies satisfy some patrolling behavior, which is an inte- term alliances with other males via of these criteria,92,97,112,120 but others gral part of chimpanzee territoriality. meat-sharing provides an indirect do not.92,96,97,112 As Boesch33,97 has Male chimpanzees occasionally form yet effective way to improve mating noted, net energy gain per individual parties that move to and along the constitutes a theoretically sound mea- periphery of their territory, where sure, and, employing this criterion, they search for signs of chimpanzees cooperation appears to occur at Taı¨ Behavior during patrols from other communities. Patrols but not Gombe.97 However, con- is striking and unusual. sometimes make deep incursions into straints on visibility during hunts and Males are silent, tense, neighboring territories. Behavior dur- the common occurrence of multiple ing patrols is striking and unusual. kills101,104,112,117,121 make it difficult to and wary. They move Males are silent, tense, and wary. monitor the activities and prey intake in a tight file, often They move in a tight file, often pause of all chimpanzee hunters. As a result, to look and listen, sometimes sniff the field observers have been unable to pause to look and ground, and show great interest in estimate per capita net energy gain listen, sometimes sniff chimpanzee nests, dung, and feeding consistently across study sites. Until remains. Goodall and colleagues76 these data become available, ques- the ground, and show were the first to describe patrolling, tions about the extent of cooperation great interest in but this behavior has received little will remain open. chimpanzee nests, subsequent attention and is still A final hypothesis proposed to ex- poorly understood. New studies have plain meat sharing implicates the use dung, and feeding started to rectify this situation. of meat as a political tool. Using ob- remains. Boesch and Boesch-Achermann33 servations of the alpha male of M have described patrolling by chim- group at Mahale, Nishida and col- panzees in the Taı¨ National Park. Pa- leagues122 suggested that male chim- trols occurred an average of about panzees share meat strategically with once a month during 45 months of opportunities. In chimpanzees, alli- others in order to curry their favor study, but rates of patrolling varied ances are important in the establish- and support. These researchers’ ob- over time. Much of this variation was ment and maintenance of rank,74,85 servations indicated that a particu- a result of changes in the composition and high rank appears to confer larly shrewd male shared meat non- of the study community. As the num- mating and reproductive advan- randomly and selectively with other ber of adult males declined, the re- tages.48–50,66,73,77 males he depended on for support in maining males became more cau- long-term alliances. Our recent obser- tious. When male numbers were vations at Ngogo are largely consis- INTERGROUP AGGRESSION reduced to four or fewer, they tent with this male social bonding hy- Chimpanzees are well known for switched from searching for neigh- pothesis.107 There as elsewhere,101,122 their territorial behavior. They are bors to avoiding encounters. Although the vast majority of all sharing events among the few animals that engage in patrols at Taı¨ consisted mostly of take place between males. Males share lethal between-group coalitionary ag- males (Fig. 7), females joined to an nonrandomly and selectively with gression.80 Encounters between com- unusual extent; female participation only certain others and sharing is re- munities are typically hostile and at other sites is generally rare.25,73 Pa- ciprocated at a group level. In addi- sometimes result in the death of both trolling chimpanzees typically made tion, male chimpanzees at Ngogo ex- infants and adults, particularly deep incursions, sometimes well over change meat for coalitionary support. males.76,80,99,123–128 Extreme inter- 1 km, into the territories of other Part of the interest generated by group aggression has led to the exter- groups, but contacted neighbors on ARTICLES Evolutionary Anthropology 19

ipation in patrols is positively related to joint participation in grooming bouts and coalitions. Patrolling effort is also positively correlated with frequency of participation in red colobus hunts and with success at capturing prey. Willingness to pursue prey, which involves risk, and hunting skill should give others some indication of a male’s willingness and ability to take risks in intergroup aggression.73 The ostensible goal of patrols is to seek information or contact with members of other communities. At Figure 7. Intercommunity variation in patrol size and composition. The mean number of Ngogo, males encounter members of adult male patrollers and average patrol size are shown for three different communities. other communities, either aurally or visually, on 30% to 40% of all patrols.73 They approach or attack mem- only about a quarter of all patrols. male mating skew is high, however, bers of the other group in roughly half Especially intriguing are the complex males would not share evenly any in- of all encounters, and either avoid tactics Taı¨ patrollers used to attack creases in mating opportunities them or flee in the other half. Part of neighbors. These included direct fron- gained by recruiting females. Males the variation in the tendency to attack tal attacks, selective “lateral” attacks would also obtain unequal shares of or flee can be attributed to the patrol- on the smallest of several parties increases conferred on the reproduc- lers’ assessment of their relative nearby, and more complicated as- tive success of resident females by im- strength. The likelihood of attack in- saults supported from the rear by fel- proved foraging efficiency. Given the creases when patrols are large, low patrollers. Use of these tactics var- potential for unequal benefits, males ied with the size and composition of might be inclined to adjust their par- whereas small patrol parties are more patrols, with large patrols containing ticipation in patrols to reflect their ex- likely to flee. These data are consistent many males being those most likely to pected current and future reproduc- with the hypothesis that intergroup make direct frontal attacks.33 tive gains.73 aggression among chimpanzees oc- We have also described patrolling Until recently, assessing how indi- curs whenever the potential costs are 130 by males at Ngogo in the Kibale Na- vidual male chimpanzees weigh these low. These observations provide tional Park.73 At Ngogo, as elsewhere, costs and benefits and make decisions only a weak test of this hypothesis, males are the primary participants in to patrol has been difficult. The high however, given our lack of data on the patrols (Fig. 7). Patrols at Ngogo are frequency of patrols at Ngogo has al- respective size of groups with which larger and contain more males than lowed us to document patrolling ef- the Ngogo chimpanzees have inter- do those at Taı¨ or Gombe. This is a fort by individual males to an extent acted. simple consequence of the fact that not previously possible.73 We found Stronger support for the low-cost the Ngogo community is much larger significant interindividual variation in hypothesis comes from experimental than the communities at Taı¨ and patrolling effort. Some males patrol play-back studies conducted with the Gombe and has many more males. quite frequently, while others do so Kanyawara community in Kibale.131 The large number of males probably less often. Some males are especially Here a male’s decision to participate helps to explain the relatively high likely to join patrols whenever the op- in simulated territorial encounters de- rate of patrolling at Ngogo, about portunity occurs, whereas others are pended primarily on whether a favor- three times per month. less inclined to participate. Our obser- able numerical asymmetry existed. Patrol participants minimally lose vations support the hypothesis that When the call of a single extra-group time, expend energy, and face oppor- this variation arises partly because male was played back, parties con- tunity costs. At worst, patrollers risk males do not derive equal benefits taining three or more males consis- injury or even death. These costs from patrolling. Patrol participation tently chorused and approached the could be offset by several benefits, in- is correlated with mating success. speaker. Parties with fewer adult cluding improved safety, increased ac- Males who mate frequently and may males usually remained silent and ap- cess to food, female recruitment, and have the most offspring in the group proached the speaker less often. In improved foraging efficiency for resi- to protect now and in the future ap- these simulated encounters, the posi- dent females.19,73,129 All patrollers and pear to be motivated to patrol often. tion that males occupied in parties other members of their communities In addition, males seem to minimize that approached the speaker varied would benefit from improved safety the costs of patrolling by doing so with significantly among individuals. This and increased access to food. If these partners with whom they have strong position was independent of current are the only benefits and males share social bonds and on whom they can dominance rank, although two former them equitably, then a collective ac- rely to take risks. Our observations alpha males had the highest mean ap- tion problem is likely to ensue.73 If from Ngogo indicate that joint partic- proach rates. Willingness to respond 20 Evolutionary Anthropology ARTICLES

Fatalities during intergroup aggression are a well-documented aspect of chimpanzee behavior. In some cases, attacks on females and their infants result in infanticide.25,73,99,126–128 In- fanticide has also been documented within communities.132 Chimpanzees also kill adult members from other groups.25,76,80,124 Intraspeciﬁc killing by chimpanzees is unusual among primates in that the perpetrators also cannibalize their victims.25,99,123,125–128 Our understanding of the causes of lethal attacks and cannibalism is in- complete. Infanticide among chimpanzees does not obviously satisfy the conditions of the sexual-selection hy- pothesis133 because female secondary transfer appears to be rare19 and males often kill infants in their own communities.132 Instead of increasing

Infanticide among chimpanzees does not obviously satisfy the conditions of the sexual- selection hypothesis because female secondary transfer appears to be rare and males often kill infants in their own communities.

Figure 8. A group of male chimpanzees on patrol. male reproductive opportunities, between-group infanticide may be a male tactic to push females from to intrusions did not differ among in- territory and all simulated the pres- neighboring communities away from dividuals, and all males approached ence of only one extra-group male. boundary areas, thereby allowing fe- when they overwhelmingly outnum- Field observations indicate that inter- males in their own communities safer bered the caller.131 group encounters are more likely in and better access to food in these ar- These results suggest that male border areas, and frequently involve eas.129 Success in between-group ag- chimpanzees obtain mutual benefits multiple males on both sides.25,33,73 gression could have this effect in gen- from cooperating in intergroup ag- Under such conditions the potential eral. In support of this argument, gression,131 but further observations costs of involvement to individual long-term data from the Kasakela of actual territorial encounters are males may be greater, and individual community at Gombe show that fe- needed to exclude the alternative hy- differences in participation thus more male reproductive performance was pothesis that males participate and likely to emerge. At Taı¨, Gombe, and higher when the community’s terri- benefit unequally because their pro- Ngogo, communities neighboring the tory was relatively large than it was spective reproductive gains differ. The main study groups have been or are when pressure from neighbors re- play-back experiments did not resolve being habituated, which will facilitate duced the size of the territory (A. Pu- this issue because most were per- recording the details of both sides of sey, personal communication). Irre- formed well within the community’s actual intergroup encounters. spective of the many questions that ARTICLES Evolutionary Anthropology 21 surround between-group infanticide, and found that the patterns observed difference? One possibility is that the within-group infanticides continue to were influenced by the stability of the fission-fusion nature of chimpanzee remain a major puzzle in the study of dominance hierarchy. During a pe- social organization introduces an ele- chimpanzee behavior.132,134 riod in which the dominance hierar- ment of unpredictability to male so- chy was stable, high-ranking males cial relationships that is not present in were less aggressive than low-ranking baboon society. Because baboons BEHAVIORAL males and exhibited lower circulating travel and forage in relatively cohesive ENDOCRINOLOGY levels of cortisol and testosterone. groups, it presumably is easy for a During a period of extreme domi- male baboon to make accurate assess- Recent developments in sampling nance instability, however, high-rank- ments of his position in the domi- and assaying steroid hormones have ing males were more aggressive and nance hierarchy. Furthermore, a provided new insights into the physi- had higher levels of cortisol and tes- high-ranking male can continually ological mechanisms of primate be- tosterone than did low-ranking males. monitor developing social relation- havior.135 Urinary and fecal assays In our studies at Kanyawara, we ships between other individuals that have proven particularly attractive to have observed that across the adult might affect the hierarchy in the fu- field workers. Urine and fecal sam- males rates of aggression are posi- ture. Chimpanzee males, on the other pling are generally easy and inexpen- tively and significantly correlated with hand, frequently break into small sive to conduct, allow for frequent re- rank, even when the dominance hier- groups that may not come together sampling of individuals, and integrate archy appears to be stable; that is, again for hours, days, or even weeks. the short-term fluctuations in hor- Thus, it is more difficult for a high- mone production that can confound ranking chimpanzee to monitor the serum measurements. More impor- Similar correlations political maneuverings among poten- tantly, such sampling is noninvasive tial challengers. A dominant male is and does not adversely affect the wel- between aggression constantly at risk from opportunistic fare or behavior of study subjects. and rank during periods coalitions formed by lower-ranking There is, consequently, a burgeoning individuals and must continually as- literature on the hormonal correlates of relative dominance sert his dominance through agonistic of behavior in wild primates,136–139 to stability have been which studies of chimpanzees are display. This idea is supported by the contributing.140 reported from Gombe, high frequency of aggression that We have recently completed a set of takes place in the context of re- Mahale, and Ta¨ı.At 123,124 studies on the behavioral endocrinol- unions. In sum, even a stable ogy of chimpanzees in the Kanyawara Kanyawara we have chimpanzee hierarchy may be unsta- community, Kibale National Park. also found that ble in comparison to a baboon hierarchy, despite having a low rate of ac- This research has focused on the ste- dominance rank roid hormones testosterone and corti- tual reversals. 146 sol. Testosterone influences numer- consistently correlates Of course, as with baboons, dom- ous aspects of male reproductive positively with measures inance rank probably is not the entire physiology. Cortisol has been widely story when it comes to chimpanzee employed as a marker of both physio- of urinary cortisol. stress physiology. The personalities of logical and psychological stress.141 different individuals may partially The stress response is of special inter- mediate the relationship between est. Although this response clearly is rank and stress. Ongoing studies at adaptive in many situations, its Kanyawara and Gombe will soon al- when there are no rank reversals and chronic activation can lead to in- low us to address this issue. In the creased susceptibility to pathology, the rate of reversals in decided agonis- meantime, it appears that increased 124 including gastric ulcers, atherosclero- tic bouts is low. Similar correla- stress imposes a general cost on social sis, and suppressed immune func- tions between aggression and rank dominance for male chimpanzees, tion.141 Thus, there are circumstances during periods of relative dominance which must be set against the repro- in which cortisol measurements can stability have been reported from ductive benefits described earlier. help to assess the costs of particular Gombe,144 Mahale,77 and Taı¨.33 At An additional cost of social domi- behavioral strategies. Kanyawara we have also found that nance suggests itself in the form of Sapolsky’s142,143 research on stress dominance rank consistently corre- chronically elevated testosterone lev- physiology in baboons provides the lates positively with measures of uri- els. At Kanyawara, urinary androgen most complete picture available of the nary cortisol.145 Taken together, these levels also correlate with dominance interaction between hormones and findings suggest that the trials and rank. Although the correlation is not behavior in a wild primate, and fur- tribulations of life at the top of the as robust as that with cortisol, there is nishes a context for our findings from hierarchy are persistently more stress- a clear alpha-male effect,147 with the the Kanyawara chimpanzees. Sapol- ful for male chimpanzees than for highest-ranking male persistently ex- sky measured basal cortisol levels in male baboons. hibiting the highest levels of testoster- high- and low-ranking baboon males What can account for this apparent one. Additional physiological costs 22 Evolutionary Anthropology ARTICLES associated with high testosterone in- from Kanyawara support this predic- findings and stand on the threshold of clude direct energetic costs resulting tion. Adult males at Kanyawara exhib- filling several critical gaps in under- from increased metabolic rate and im- ited significant increases in rates of standing. We conclude by reviewing munosuppression.148 aggression, particularly escalated ag- some topics that warrant more re- Testosterone influences multiple as- gression such as chases and attacks, search. pects of male reproductive physiol- on days when maximally swollen par- First, what explains intraspecific ogy, from the development of the male ous females were present.124 Males variations in gregariousness, associa- reproductive anatomy to the mainte- also exhibited significant increases in tion patterns, and range use? Ecolog- nance of both reproductive function mean testosterone levels during these ical factors such as the spatial and and motivation. It has also classically periods of intense competition.145 temporal distribution of food are fre- been associated with aggression. A Mean cortisol levels also increased quently invoked as explanatory vari- large body of research, mostly on significantly, which is consistent with ables.153–156 Comparative studies that birds, suggests that variation in circu- the known role of this hormone in provide systematic data on food avail- lating testosterone levels is associated mobilizing energy during periods of ability and feeding competition prom- primarily with male aggression in re- crisis. ise to provide answers in the case of productive contexts, rather than Estrous females are frequent targets chimpanzees. Demographic and so- changes in reproductive physiolo- of male aggression, particularly when cial factors that affect the description, gy.149 The extent to which this idea, high-ranking males are attempting to measurement, and analysis of associ- formally known as the challenge hy- maintain exclusive mating access to ations should also be examined to pothesis,149 applies to mammals is not them.25,151 Being mate-guarded is seek clues to intraspecific variations yet clear.138,150 Data from Kanyawara likely to be extremely stressful for fe- in gregariousness and range use. permit a preliminary test of this hy- males, a proposition that is supported Second, we require additional study pothesis in chimpanzees. by our preliminary data. We were able of the factors that account for differ- In birds, testosterone induces male to collect multiple samples from one ences in dispersal patterns across reproductive aggression at the ex- parous female who was intensively study sites. Why do female chimpan- pense of paternal care.149 Therefore, mate-guarded during her periovula- zees at Gombe frequently remain in cross-species correlations are pre- tory period. During this time the ag- their natal group while those at other dicted between basal levels of breed- gression she received from males sites almost always disperse? Feeding ing-season testosterone and mating more than doubled, and she showed competition that ultimately limits fe- system. Monogamous birds are ex- dramatic increases in urinary cortisol male reproduction has been invoked pected to maintain high testosterone levels. We also examined multiple as a causal factor19,116,129 and here too levels during territory formation and samples from a nulliparous female direct measures of food availability breeding, but to decrease testosterone during both swelling and nonswelling will be necessary to unlock the key to periods. Because nulliparous females production when providing paternal this puzzle. The flip side of female dis- are not as attractive to males as moth- care. They should also react strongly persal, male philopatry, constitutes an ers are, males do not mate-guard to challenges from conspecifics with associated problem seldom addressed them, and they do not appear to re- increased testosterone production. in discussions of chimpanzee behav- ceive as much aggression during the Polygynous birds, on the other hand, ior. We need a new body of theory and late follicular phase.152 Such was the engage in less paternal care, so they empirical research to address the evo- case with this female, and she did not should exhibit high levels of testoster- lutionary causes and proximate deter- exhibit the increases in urinary corti- one throughout the breeding season. minants of male philopatry, not only sol during maximal swelling that were They are not expected to show a in chimpanzees but other animals as observed in the parous female. Future heightened endocrine response to well.157 work in this area at both Kanyawara challenges because their testosterone Females have taken a back seat to and Gombe will help us to document levels are already close to the physio- males in the study of chimpanzee be- the physiological costs imposed on fe- logical maximum. These predictions havioral ecology and we require more males by male sexual aggression. are supported by data from a large and better observations of them and number of avian species.149 their behavior. These data will not Chimpanzees are not seasonal DIRECTIONS FOR FUTURE only furnish insights into the nature breeders and do not engage in direct of female competition, reproduction, paternal care, so they represent a spe- RESEARCH and social behavior,158 but also will cial case for the challenge hypothesis. This is an especially exciting time help to clarify aspects of male-female Because the availability of cycling fe- for those of us who study chimpan- social relationships and the signifi- males varies temporally and the pres- zees in the wild. Years of dedicated cance of male coercion in this spe- ence of maximally swollen females is field work by numerous researchers at cies.51 As we have noted, chimpanzee attended by high rates of male aggres- sites scattered throughout Africa have hunting and meat eating take on spe- sion, the challenge hypothesis sug- produced a rich body of information cial significance for anthropologists gests that male testosterone levels regarding the behavior of our closest who seek an understanding of human should increase during periods of re- living relatives. We are now in a behavioral evolution. The extent to productive competition. Our data unique position to build on previous which male chimpanzees cooperate ARTICLES Evolutionary Anthropology 23 during hunts remains an empirically possible. C. Boesch, J. Fleagle, K. 15 Kortlandt A. 1962. Chimpanzees in the wild. challenging area in need of future Hunley, S. Olatunde, and R. Wrang- Sci Am 206:128–138. 16 Goodall J. 1965. Chimpanzees of the Gombe study. While current evidence sug- ham made helpful suggestions on the Stream Reserve. In: DeVore I, editor. Primate gests that male chimpanzees share manuscript, while K. Langergraber behavior. New York: Holt, Rinehart, Winston. p meat selectively with others in order provided editorial assistance. Our re- 425–473. to curry their favor and support, alter- search on chimpanzees has been gener- 17 Reynolds V, Reynolds F. 1965. Chimpanzees of the Budongo forest. In: DeVore I, editor. Pri- nate hypotheses proposed to explain ously funded by grants from the Detroit mate behavior. New York: Holt, Rinehart, Win- meat sharing in humans, such as shar- Zoological Institute, the Harry Frank ston. p 368–424. ing to enhance status and tolerated Guggenheim Foundation, the L.S.B. 18 Pusey A. 1979. Intercommunity transfer of chimpanzees in Gombe National Park. In: Ham- theft are still to be invoked and test- Leakey Foundation, the National Geo- burg D, McCown E, editors. The great apes. ed.159,160 Technical advances that in- graphic Society, the U.S. National Sci- Menlo Park: Benjamin/Cummings. p 405–427. volve the measurement of genetic re- ence Foundation, and the Wenner-Gren 19 Williams J. 1999. Female strategies and the reasons for territoriality in chimpanzees: lessons latedness and steroid hormones Foundation for Anthropological Re- from three decades of research at Gombe. Ph.D. between and within individuals prom- search. We dedicate this review to Pro- thesis, University of Minnesota. ise to yield deeper understanding of fessors Toshisada Nishida and Richard 20 Halperin S. 1979. Temporary association pat- both the ultimate and proximate fac- Wrangham, pioneering chimpanzee re- terns in free ranging chimpanzees: an assessment of individual grouping preferences. In: tors affecting chimpanzee behavior. searchers, mentors, and friends. Hamburg D, McCown E, editors. The great apes. Finally, no easy answers regarding the Menlo Park: Benjamin/Cummings. p 491–499. evolutionary factors that contribute to 21 Wrangham R, Smuts B. 1980. Sex differences in the behavioral ecology of chimpanzees in the infanticide and cannibalism have REFERENCES Gombe National Park, Tanzania. J Reprod Fertil emerged, despite decades of study. 28 (suppl):13–31. Understanding here will require more 1 Horai S, Hayasaka K, Kondo R, Tsugane K, 22 Kawanaka K. 1984. Association, ranging, and observations of the contexts of these Takahata N. 1995. Recent African origin of mod- the social unit in chimpanzees of the Mahale ern humans revealed by complete sequences of Mountains, Tanzania. Int J Primatol 5:411–434. infrequent events. hominoid mitochondrial DNAs. Proc Natl Acad 23 Hasegawa T. 1990. Sex differences in ranging We would be remiss if we failed to Sci 92:532–536. patterns. In: Nishida T, editor. The chimpanzees note the critically endangered status 2 Ruvolo M. 1997. Molecular phylogeny of the of the Mahale Mountains. Tokyo: University of Tokyo Press. p 99–114. of chimpanzees in the wild. It has be- hominoids: inferences from multiple independent DNA sequence data sets. Mol Biol Evol 14: 24 Wrangham R, Clark A, Isabiryre-Basuta G. come depressingly common for field 248–265. 1992. Female social relationships and social or- workers to comment on how time is 3 Chen F, Li W. 2001. Genomic divergences be- ganization of Kibale Forest chimpanzees. In: running out to save these animals. tween humans and other hominoids and the ef- Nishida T, McGrew W, Marler P, Pickford M, fective population size of the common ancestor deWaal F, editors. Topics in primatology, vol. 1. This is a regrettable fact. We are for- of humans and chimpanzees. Am J Hum Genet Human origins. Tokyo: Tokyo University Press. p tunate to have been given an opportu- 68:444–456. 81–98. nity to share in the lives of these ani- 4 Goodall J. 1963. Feeding behaviour of wild 25 Goodall J. 1986. The chimpanzees of Gombe. chimpanzees: a preliminary report. Symp Zool Cambridge: Belknap Press. mals in the wild. Anyone who has Soc London 10:39–48. 26 Wrangham R. 1979. On the evolution of ape been similarly blessed instantly real- 5 Goodall J. 1968. The behaviour of free-living social systems. Soc Sci Information 18:335–368. izes the extraordinary nature of chim- chimpanzees in the Gombe Stream area. Anim 27 Wrangham R. 2000. Why are male chimpan- panzees and recognizes the void that Behav Monogr 1:161–311. zees more gregarious than mothers? A scramble 6 Nishida T. 1990. The chimpanzees of the Ma- competition hypothesis. In: Kappeler P, editor. will exist if we do not meet the chal- hale Mountains. Tokyo: University of Tokyo Primate males. Cambridge: Cambridge Univer- lenge by doing everything in our Press. sity Press. p 248–258. power to conserve these remarkable 7 Nishida T. 1968. The social group of wild chim- 28 Wrangham R. 1986. Ecology and social rela- panzees in the Mahale Mountains. Primates tionships in two species of chimpanzee. In: creatures and their habitat in the wild. 9:167–224. Wrangham R, Rubenstein D, editors. Ecological 8 Nishida T, Kawanaka K. 1972. Inter-unit-group aspects of social evolution: birds and mammals. Princeton: Princeton University Press. p ACKNOWLEDGMENTS relationships among wild chimpanzees of the Mahale mountains. Kyoto Univ Afr Stud 7:131– 352–378. We thank John Fleagle for inviting 169. 29 Sakura O. 1994. Factors affecting party size 9 Itani J. 1979. Distribution and adaptation of and composition of chimpanzees (Pan troglodytes us to prepare this review. Our field chimpanzees in an arid area. In: Hamburg D, verus) at Bossou, Guinea. Int J Primatol 15:167– work has been sponsored in Tanzania McCown E, editors. The great apes. Menlo Park: 183. and Uganda by the Makerere Univer- Benjamin/Cummings. p 55–71. 30 Sugiyama Y, Koman J. 1979. Social structure and dynamics of wild chimpanzees at Bossou, sity, the Serengeti Wildlife Research 10 Tutin C, Fernandez M, Rogers M, Williamson E, McGrew W. 1991. Foraging profiles of sympa- Guinea. Primates 20:323–339. Institute, the Tanzanian Commission tric lowland gorillas and chimpanzees in the 31 Ghiglieri M. 1984. The chimpanzees of Kibale for Science and Technology, the Tan- Lope´ Reserve, Gabon. Philos Trans R Soc (Lon- forest. New York: Columbia University Press. zanian National Parks, and the Ugan- don B) 334:179–186. 32 Boesch C. 1996. Social grouping in Taı¨ chim- 11 Baldwin P, McGrew W, Tutin C. 1982. Wide- panzees. In: McGrew W, Marchant L, Nishida T, dan National Parks. We are grateful to ranging chimpanzees at Mt. Assirik, Senegal. Int editors. Great ape societies. Cambridge: Cam- G.I. Basuta, J. Kasenene, E. Massawe, J Primatol 3:367–385. bridge University Press. p 101–113. and the staffs of the Makerere Univer- 12 Wrangham R, McGrew W, deWaal F, Heltne 33 Boesch C, Boesch-Achermann H. 2000. The P. 1994. Chimpanzee cultures. Cambridge: Har- chimpanzees of the Taı¨ forest. Oxford: Oxford sity Biological Field Station and Ma- vard University Press. University Press. hale Mountains National Park for pro- 13 Whiten A, Boesch C. 2001. The cultures of 34 Doran D. 1997. Influence of seasonality on viding logistical assistance in the field. chimpanzees. Sci Am 284:61–67. activity patterns, feeding behavior, ranging, and We thank our Tanzanian and Ugan- 14 Whiten A, Goodall J, McGrew W, Nishida T, grouping patterns in Taı¨ chimpanzees. Int J Pri- Reynolds V, Sugiyama Y, Tutin C, Wrangham R, matol 18:183–206. dan field assistants, without whom Boesch C. 1999. Cultures in chimpanzees. Nature 35 Mitani J, Watts D, Lwanga J. n.d. Ecological our research would not have been 399:682–685. and social correlates of chimpanzee party size 24 Evolutionary Anthropology ARTICLES

and composition. In: Boesch C, Hohmann G, 57 Strier K. 2000. Primate behavioral ecology. gies among adult male chimpanzees of the Ma- Marchant L, editors. Behavioral diversity in Boston: Allyn Bacon. hale Mountains, Tanzania. In: McGrew W, chimpanzees and bonobos. Cambridge: Cam- 58 Angier N. 1999. Woman: an intimate geogra- Marchant L, Nishida T, editors. Great ape soci- bridge University Press. In press. phy. Boston: Houghton Mifflin. eties. Cambridge: Cambridge University Press. p 114–134. 36 Chapman C, White F, Wrangham R. 1993. 59 Hawkes N. 1997. DNA shows chimps’ eye for Defining subgroup size in fission-fusion societies. monkey business. The Times May 22, 1997 edi- 78 Watts D. 2000. Grooming between male chim- Folia Primatol 61:31–34. tion. panzees at Ngogo, Kibale National Park. II. In- 37 Matsumoto-Oda A. 1999. Mahale chimpan- 60 Westneat D, Sherman P, Morton M. 1990. The fluence of male rank and possible competition zees: grouping patterns and cycling females. ecology and evolution of extra-pair copulation in for partners. Int J Primatol 21:211–238. Am J Primatol 47:197–207. birds. In: Power D, editor. Current ornithology. 79 Newton-Fisher N. 1999. Association by male 38 Whitehead H, Dufault S. 1999. Techniques for New York: Plenum Press. p 331–369. chimpanzees: a social tactic? Behaviour 136:705– analyzing vertebrate social structure using iden- 61 Sillero-Zubiri C, Gotelli D, MacDonald D. 730. tified individuals: review and recommendations. 1996. Male philopatry, extra-pack copulations 80 Wrangham R. 1999. Evolution of coalitionary Adv Stud Behav 28:33–74. and inbreeding avoidance in Ethiopian wolves killing. Yearbook Phys Anthropol 42:1–30. 39 Pepper J, Mitani J, Watts D. 1999. General (Canis simensis). Behav Ecol Sociobiol 38:331– 81 Mitani J, Watts D. 2001. Why do chimpanzees gregariousness and specific social preferences 340. hunt and share meat? Anim Behav 61:915–924. among wild chimpanzees. Int J Primatol 20:613– 62 Keane B, Dittus W, Melnick D. 1997. Pater- 82 Riss D, Goodall J. 1977. The recent rise to the 632. nity assessment in wild groups of toque ma- alpha rank in a population of free-living chim- 40 Cairns S, Schwager S. 1987. A comparison of caques Macaca sinica at Polonnaruwa, Sri panzees. Folia Primatol 27:134–151. association indices. Anim Behav 35:1454–1469. Lanka, using molecular markers. Mol Ecol 83 Goldberg T, Wrangham R. 1997. Genetic cor- 6:267–282. 41 Nishida T, Takasaki H, Takahata Y. 1990. De- relates of social behaviour in wild chimpanzees: mography and reproductive profiles. In: Nishida 63 Petrie M, Kempenaers B. 1998. Extra-pair pa- evidence from mitochondrial DNA. Anim Behav T, editor. The chimpanzees of the Mahale Moun- ternity in birds: explaining variation between 54:559–570. tains. Tokyo: University of Tokyo Press. p 63–97. species and populations. Trends Ecol Evol 13:52– 84 Mitani J, Merriwether DA, Zhang C. 2000. 58. 42 Sugiyama Y. 1999. Socioecological factors of Male affiliation, cooperation, and kinship in wild male chimpanzee migration at Bossou, Guinea. 64 Sugiyama Y, Kawamoto S, Takenaka O, chimpanzees. Anim Behav 59:885–893. Primates 40:61–68. Kumazaki K, Miwa N. 1993. Paternity discrimi- 85 deWaal F. 1982. Chimpanzee politics. New 43 Morin P, Moore J, Chakraborty R, Jin L, nation and inter-group relationships of chimpan- York: Harper and Row. Goodall J, Woodruff D. 1994. Kin selection, so- zees at Bossou. Primates 34:545–552. cial structure, gene flow, and the evolution of 65 Morin P, Wallis J, Moore J, Woodruff D. 1994. 86 Mitani J, Watts D, Pepper J, Merriwether DA. chimpanzees. Science 265:1193–1201. Paternity exclusion in a community of wild chim- n.d. Demographic and social constraints on male chimpanzee behaviour. submitted to Anim Be- 44 Goldberg T, Ruvolo M. 1997. The geographic panzees using hypervariable simple sequence re- hav. apportionment of mitochondrial genetic diver- peats. Mol Ecol 3:469–478. sity in East African chimpanzees, Pan troglodytes 66 Constable J, Ashley M, Goodall J, Pusey A. 87 Altmann J. 1979. Age cohorts as paternal sib- schweinfurthii. Mol Biol Evol 14:976–984. 2001. Noninvasive paternity assignment in ships. Behav Ecol Sociobiol 6:161–164. 45 Gagneux P, Wills C, Gerloff U, Tautz D, Morin Gombe chimpanzees. Mol Ecol 10:1279–1300. 88 Ghiglieri M. 1987. Sociobiology of the great P, Boesch C, Fruth B, Hohmann G, Ryder O, 67 Vigilant L, Hofreiter, Siedel H, Boesch C. apes and the hominid ancestor. J Hum Evol 16: Woodruff D. 1999. Mitochondrial sequences 2001. Paternity and relatedness in wild chimpan- 319–357. show diverse evolutionary histories of African zee communities. Proc Natl Acad Sci 98:12890– 89 Wrangham R. 1987. The significance of Afri- hominoids. Proc Natl Acad Sci 96:5077–5082. 12895. can apes for reconstructing human social evolu- 46 Queller D, Strassmann J, Hughes C. 1993. 68 Edwards A, Civitello A, Hammond H, Caskey tion. In: Kinzey W, editor. The evolution of hu- Microsatellites and kinship. Trends Ecol Evol C. 1991. DNA typing and genetic mapping with man behavior: primate models. Albany: State 8:285–288. trimeric and tetrameric tandem repeats. Am J University of New York Press. p 51–71. 47 Kohn M, Wayne R. 1997. Facts from feces Hum Genet 49:746–756. 90 Boesch H, Boesch C. 1994. Hominization in revisited. Trends Ecol Evol 12:223–227. 69 Taberlet P, Waits LP, Luikart G. 1999. Nonin- the rainforest: the chimpanzee’s piece of the puz- 48 Tutin C. 1979. Mating patterns and reproduc- vasive genetic sampling: look before you leap. zle. Evol Anthropol 3:171–178. tive strategies in a community of wild chimpan- Trends Ecol Evol 14:323–327. 91 Stanford C. 1995. Chimpanzee hunting be- zees (Pan troglodytes schweinfurthii). Behav Ecol 70 Taberlet P, Griffin S, Goosens B, Questiau S, havior and human evolution. Am Sci 83:256– Sociobiol 6:29–38. Manceau V, Escaravage N, Waits LP, Bouvet J. 261. 49 Hasegawa T, Hiraiwa-Hasegawa M. 1983. Op- 1996. Reliable genotyping of samples with very 92 Stanford C. 1996. The hunting ecology of wild portunistic and restrictive matings among wild low DNA quantities using PCR. Nucl Acids Res chimpanzees: implications for the evolutionary chimpanzees in Mahale, Tanzania. J Ethol 1:75– 24:3189–3194. ecology of Pliocene hominids. Am Anthropol 98: 85. 71 Silk J. 2001. Ties that bind: the role of kinship 96–113. 50 Watts D. 1998. Coalitionary mate-guarding by in primate societies. In: Stone L, editor. New 93 Stanford C. 1999. The hunting apes. Prince- male chimpanzees at Ngogo, Kibale National directions in anthropological kinship. Lanham: ton: Princeton University Press. Rowman & Littlefield Publishers. p 71–92. Park, Uganda. Behav Ecol Sociobiol 44:43–55. 94 Stanford C. 2001. The ape’s gift: meat-eating, 51 Smuts B, Smuts R. 1993. Male aggression and 72 Nishida T, Hiraiwa-Hasegawa M, Hasegawa meat-sharing, and human evolution. In: deWaal sexual coercion of females in nonhuman pri- T, Takahata Y. 1985. Group extinction and fe- F, editor. The tree of origin. Cambridge: Harvard mates and other mammals: evidence and theo- male transfer in wild chimpanzees in the Mahale University Press. p 95–117. retical implications. Adv Stud Behav 22:1–63. Mountains National Park, Tanzania. Z Tierpsy- 95 Busse C. 1978. Do chimpanzees hunt cooper- chol 67:281–301. 52 Wallis J. 1997. A survey of reproductive pa- atively? Am Nat 112:767–770. rameters in the free-ranging chimpanzees of 73 Watts D, Mitani J. 2001. Boundary patrols 96 Boesch C. 1994. Chimpanzees-red colobus Gombe National Park. J Reprod Fertil 109:297– and intergroup encounters among wild chimpan- monkeys: a predator-prey system. Anim Behav 307. zees. Behaviour 138:299–327. 47:1135–1148. 53 Hasegawa T, Hiraiwa-Hasegawa M. 1990. 74 Nishida T. 1983. Alpha status and agonistic 97 Boesch C. 1994. Cooperative hunting in wild Sperm competition and mating behavior. In: alliance in wild chimpanzees (Pan troglodytes Nishida T, editor. The chimpanzees of the Ma- schweinfurthii). Primates 24:318–336. chimpanzees. Anim Behav 48:653–667. hale Mountains. Tokyo: University of Tokyo 75 Simpson M. 1973. Social grooming of male 98 Stanford C. 1998. Chimpanzee and red colo- Press. p 115–132. chimpanzees. In: Crook J, Michael R, editors. bus: the ecology of predator and prey. Cam- 54 Gagneux P, Woodruff D, Boesch C. 1997. Fur- Comparative ecology and behaviour of primates. bridge: Harvard University Press. tive mating in female chimpanzees. Nature 387: London: Academic Press. p 411–505. 99 Nishida T, Uehara S, Nyundo R. 1979. Pred- 358–359. 76 Goodall J, Bandora A, Bergmann E, Busse C, atory behavior among wild chimpanzees of the 55 Gagneux P, Boesch C, Woodruff D. 1999. Fe- Matama H, Mpongo E, Pierce A, Riss D. 1979. Mahale Mountains. Primates 20:1–20. male reproductive strategies, paternity and com- Intercommunity interactions in the chimpanzee 100 Anderson J, Williamson E, Carter J. 1983. munity structure in wild West African chimpan- population of the Gombe National Park. In: Chimpanzees of Sapo Forest, Liberia: density, zees. Anim Behav 57:19–32. Hamburg D, McCown E, editors. The great apes. nests, tools and meat-eating. Primates 24:594– 56 Wrangham R. 1997. Subtle, secret female Menlo Park: Benjamin/Cummings. p 13–54. 601. chimpanzees. Science 277:774–775. 77 Nishida T, Hosaka K. 1996. Coalition strate- 101 Boesch C, Boesch H. 1989. Hunting behav- ARTICLES Evolutionary Anthropology 25

ior of wild chimpanzees in the Taı¨ National Park. Nishida T, McGrew W, Marler P, Pickford M, among wild baboons. Psychoneuroendocrinol- Am J Phys Anthropol 78:547–573. deWaal F, editors. Topics in primatology. Vol- ogy 17:701–709. 102 Alp R. 1993. Meat eating and ant dipping by ume 1. Human origins. Tokyo: Tokyo University 142 Sapolsky R. 1982. The endocrine stress re- wild chimpanzees in Sierra Leone. Primates 34: Press. p 159–174. sponse and social status in the wild baboon. 463–468. 123 Bygott D. 1979. Agonistic behavior and dom- Horm Behav 16:279–292. 103 Basabose K, Yamagiwa J. 1997. Predation inance among wild chimpanzees. In: Hamburg 143 Sapolsky R. 1993. Endocrinology alfresco: on mammals by the chimpanzees in the montane D, McCown E, editors. The great apes. Menlo psychoendocrine studies of wild baboons. Rec forest of Kahuzi, Zaire. Primates 38:275–299. Park: Benjamin/Cummings. p 405–427. Prog Horm Res 48:437–468. 104 Mitani J, Watts D. 1999. Demographic influ- 124 Muller M. n.d. Agonistic relations among 144 Bygott D. 1974. Agonistic behaviour and ences on the hunting behavior of chimpanzees. Kanyawara chimpanzees. In: Boesch C, Hohm- dominance in wild chimpanzees. Ph.D. thesis. Am J Phys Anthropol 109:439–454. ann G, Marchant L, editors. Behavioral diversity Cambridge University. in chimpanzees and bonobos. Cambridge: Cam- 105 Uehara S. 1997. Predation on mammals by bridge University Press. In press. 145 Muller M. 2002. Endocrine aspects of ag- the chimpanzee (Pan troglodytes). Primates 38: gression and dominance in chimpanzees of the 193–214. 125 Bygott D. 1972. Cannibalism among wild Kibale Forest. Ph.D. thesis. University of South- chimpanzees. Nature 238:410–411. 106 Stanford C, Wallis J, Mpongo E, Goodall J. ern California. 126 Goodall J. 1977. Infant-killing and cannibal- 1994. Hunting decisions in wild chimpanzees. 146 Sapolsky R. 1991. Testicular function, social ism in free-living chimpanzees. Folia Primatol Behaviour 131:1–18. rank and personality among wild baboons. Psy- 28:259–282. 107 Mitani J, Watts D. 2001. Why do chimpan- choneuroendocrinology 16:281–293. 127 Newton-Fisher N. 1999. Infant killers of Bu- zees hunt and share meat? Anim Behav 61:915– 147 Whitten P. 2000. Evolutionary endocrinol- dongo. Folia Primatol 70:167–169. 924. ogy of the cercopithecoids. In: Whitehead P, Jolly 108 Busse C. 1977. Chimpanzee predation as a 128 Watts D, Mitani J. 2000. Infanticide and can- C, editors. Old world monkeys. Cambridge: Cam- possible factor in the evolution of red colobus nibalism by male chimpanzees at Ngogo, Kibale bridge University Press. p 269–297. National Park, Uganda. Primates 41:357–365. monkey social organization. Evolution 31:907– 148 Wingfield J, Jacobs J, Hillgarth N. 1997. Eco- 911. 129 Pusey A. 2001. Of genes and apes: chimpan- logical constraints and the evolution of hormone- 109 Stanford C. 1995. The influence of chimpanzee social organization and reproduction. In: de- behavior interrelationships. Ann NY Acad Sci zee predation on group size and anti-predator Waal F, editor. The tree of origin. Cambridge: 807:22–41. Harvard University Press, p 9–37. behaviour in red colobus monkeys. Anim Behav 149 Wingfield J, Hegner R, Dufty A, Ball G. 1990. 49:577–587. 130 Manson J, Wrangham R. 1991. Intergroup The “challenge hypothesis”: theoretical implica- aggression in chimpanzees and humans. Curr 110 Teleki G. 1973. The predatory behavior of tions for patterns of testosterone secretion, mat- Anthropol 32:369–390. wild chimpanzees. Lewisburg: Bucknell Univer- ing systems, and breeding strategies. Am Nat sity Press. 131 Wilson M, Hauser M, Wrangham R. 2001. 136:829–846. Does participation in intergroup conflict depend 111 Wrangham R, Conklin-Brittain NL, Hunt K. 150 Creel S, Wildt D, Monfort S. 1993. Aggres- on numerical assessment, range location, or rank 1998. Dietary response of chimpanzees and cer- sion, reproduction, and androgens in wild dwarf for wild chimpanzees? Anim Behav 61:1203– copithecines to seasonal variation in fruit abun- mongooses—a test of the challenge hypothesis. 1216. dance. I. Antifeedants. Int J Primatol 19:949– Am Nat 141:816–825. 132 Arcadi A, Wrangham R. 1999. Infanticide in 970. 151 Matsumoto-Oda A. 1998. Injuries to the sex- chimpanzees: review of cases and a new within- 112 Watts D, Mitani J. n.d. Hunting behavior of ual skin of female chimpanzees at Mahale and group observation from the Kanyawara study chimpanzees at Ngogo, Kibale National Park, their effect on behaviour. Folia Primatol 69:400– group in Kibale National Park. Primates 40:337– Uganda. Int J Primatol. In press. 404. 351. 113 Gorner P. 1995. Female chimpanzees extract 152 Wrangham R. n.d. The cost of sexual attrac- 133 van Schaik C. 2000. Infanticide by male pri- pound of flesh in return for sex. The Chicago mates: the sexual selection hypothesis revisited. tion: is there a tradeoff in female Pan between Tribune April 1, 1995 edition. In: van Schaik C, Janson C, editors. Infanticide sex appeal and received coercion? In: Boesch C, 114 Kopytoff V. 1995. Meat viewed as staple of by males and its implications. Cambridge: Cam- Hohmann G, Marchant L, editors. Behavioral di- chimp diet and mores. The New York Times June bridge University Press. p 27–71. versity in chimpanzees and bonobos. Cambridge: 27, 1995 edition. Cambridge University Press. In press. 134 Hamai M, Nishida T, Takasaki H, Turner L. 115 McGrew W. 1992. Chimpanzee material cul- 1992. New records of within-group infanticide in 153 Trivers R. 1972. Parental investment and ture: implications for human evolution. Cam- wild chimpanzees. Primates 33:151–162. sexual selection. In: Campell B, editor. Sexual bridge: Cambridge University Press. selection and the descent of man. Chicago: Al- 135 Whitten P, Brockman D, Stavisky R. 1998. dine de Gruyter. p 136–179. 116 Pusey A, Williams J, Goodall J. 1997. The Recent advances in noninvasive techniques to influence of dominance rank on the reproductive monitor hormone-behavior interactions. Year- 154 Bradbury J, Verhencamp S. 1977. Social or- success of female chimpanzees. Science book Phys Anthropol 41:1–23. ganization and foraging in emballonurid bats. III. Mating systems. Behav Ecol Sociobiol 2:1– 277:828–831. 136 Robbins M, Czekala N. 1997. A preliminary 17. 117 Hosaka K, Nishida T, Hamai M, Matsumoto- investigation of urinary testosterone and cortisol Oda A, Uehara S. 2001. Predation of mammals by levels in wild male mountain gorillas. Am J Pri- 155 Emlen S, Oring L. 1977. Ecology, sexual se- the chimpanzees of the Mahale Mountains, Tan- matol 43:51–64. lection and the evolution of mating systems. Sci- ence 197:215–223. zania. In: Galdikas B, Briggs N, Sheeran L, Sha- 137 Strier K, Ziegler T, Wittwer D. 1999. Sea- piro G, Goodall J, editors. All apes great and sonal and social correlates of fecal testosterone 156 Wrangham R. 1980. An ecological model of small, vol. 1. Chimpanzees, bonobos, and goril- and cortisol levels in wild male muriquis female-bonded primate groups. Behaviour las. New York: Kluwer Academic Publishers. p (Brachyteles arachnoides). Horm Behav 35:125– 75:262–300. 107–130. 134. 157 Pope T. 2000. The evolution of male philo- 118 Chapman C, Wrangham R, Chapman L. 138 Cavigelli S, Pereira ME. 2000. Mating season patry in neotropical monkeys. In: Kappeler P, 1995. Ecological constraints on group size: an aggression and fecal testosterone levels in male editor. Primate males. Cambridge: Cambridge analysis of spider monkey and chimpanzee sub- ring-tailed lemurs (Lemur catta). Horm Behav University Press. p 219–235. groups. Behav Ecol Sociobiol 36:59–70. 37:246–255. 158 Williams J, Pusey A, Carlis J, Farm B, Good- 119 Wrangham R. 1975. Behavioural ecology of 139 Brockman D, Whitten P, Richard A, all J. n.d. Female competition and male territo- chimpanzees in Gombe National Park. Ph.D. the- Benander B. 2001. Birth season testosterone lev- rial behaviour influence female chimpanzees’ sis, University of Cambridge. els in male Verreaux’s sifaka, Propithecus ver- ranging patterns. Anim Behav. In press. 120 Packer C, Ruttan L. 1988. The evolution of reauxi: insights into socio-demographic factors 159 Blurton-Jones N. 1987. Tolerated theft, sug- cooperative hunting. Am Nat 132:159–198. mediating seasonal testicular function. Behav gestions about the ecology and evolution of shar- 121 Stanford C, Wallis J, Matama H, Goodall J. Ecol Sociobiol 49:117–127. ing, hoarding and scrounging. Soc Sci Informa- 1994. Patterns of predation by chimpanzees on 140 Muller M, Wrangham R. 2001. The repro- tion 26:31–54. red colobus monkeys in Gombe National Park, ductive ecology of male hominoids. In: Ellison P, 160 Hawkes K. 1991. Showing off—tests of an 1982–1991. Am J Phys Anthropol 94:213–228. editor. Reproductive ecology and human evolu- hypothesis about men’s foraging goals. Ethol So- 122 Nishida T, Hasegawa T, Hayaki H, Takahata tion. New York: Aldine de Gruyter. p 397–427. ciobiol 12:29–54. Y, Uehara S. 1992. Meat-sharing as a coalition 141 Sapolsky R. 1992. Cortisol concentrations © 2002 Wiley-Liss, Inc. strategy by an alpha male chimpanzee? In: and the social significance of rank instability DOI 10.1002/evan.10008