Dominance, aggression and testosterone in wild chimpanzees: a test of the ‘challenge hypothesis’

The ‘challenge hypothesis’ posits that variation in male testosterone levels is more closely associated with aggression in reproductive contexts than it is with changes in reproductive physiology. Numerous bird studies support this idea, but few tests have been conducted with primates. We conducted behavioural observations and noninvasive hormone sampling of 11 male chimpanzees, Pan troglodytes schweinfurthii, in the Kanyawara study site, Kibale National Park, to test predictions of the challenge hypothesis. Results indicated that adult male chimpanzees showed significant testosterone increases during periods when parous females showed maximally tumescent sexual swellings. These periods were also marked by increased rates of male aggression. Male testosterone levels did not increase in the presence of maximally tumescent nulliparous females. Such females are less attractive to males: they are not mate-guarded, nor do rates of male aggression increase when they are swelling. Male chimpanzees copulate with parous and nulliparous females at similar rates, however, suggesting that testosterone increases in the presence of cycling parous females are associated with aggression rather than sexual behaviour. High-ranking chimpanzees were more aggressive than low-ranking males and produced higher levels of urinary testosterone. Thus, the predictions of the challenge hypothesis were generally upheld. This suggests that the hypothesis may have wider applicability among primates, including humans.     

Fatal Chimpanzee Attack in Loango National Park,Gabon

In some populations, chimpanzees engage in lethal aggression within and between social units. We report a fatal attack on an adult male chimpanzee at a new research site in Loango National Park, Gabon. We found a fresh corpse of an adult male chimpanzee only a few hundred meters from the research camp, after noting numerous vocalizations and chimpanzee movements the previous evening. Previous contacts with chimpanzees and fresh tracks in the area around the corpse suggest that 2 communities of chimpanzees range where the attack occurred and that members of the neighboring community killed the chimpanzee. To support the conclusion, we conducted genetic analysis for 13 Y-chromosome loci and 9 microsatellite loci of fecal samples from the dead individual, 5 possible attackers, and 2 members of the other community Though we cannot exclude the possibility of an intracommunity killing, the combined observational and genetic evidence suggest an intercommunity attack. The case study adds to the growing evidence that intercommunity killings are a rare but widespread phenomenon among chimpanzees and not an artifact of human provisioning or habituation.     

Genetic Sampling of Unhabituated Chimpanzees (Pan troglodytes schweinfurthii) in Gishwati Forest Reserve, an Isolated Forest Fragment in Western Rwanda

Many primate populations currently live in forest fragments. These populations are often unhabituated, elusive, and contain few individuals, making them difficult to study through direct observation. Noninvasive genetic methods are useful for surveying these unhabituated populations to infer the number and sex of individuals and the genetic diversity of the population. We conducted genetic analysis on 70 fecal samples from eastern chimpanzees (Pan troglodytes schweinfurthii) in Gishwati Forest Reserve, a forest fragment in western Rwanda. We genotyped all but two of these samples using 12 autosomal and 13 Y-chromosome microsatellite markers previously used in analyses of other chimpanzee populations. The genetic data show that these samples represent a minimum of 19 individuals (7 females, 12 males). However, because we may not have sampled all individuals in the population, we also performed mark-recapture analysis with the genetic data and found that the entire population likely numbers between 19 and 29 individuals. These results are consistent with opportunistic observations of at least 19 individual chimpanzees. Levels of variation at the Y-chromosome microsatellites were similar to those observed in other chimpanzee communities, suggesting that the chimpanzees in this forest are members of a single community. These results provide a baseline count of the number of male and female chimpanzees in the Gishwati Forest Reserve, and the data provide the potential for follow-up studies aimed at tracking individuals over time, thus aiding conservation management of this unhabituated population.     

Nonbreeding season pairing behavior and the annual cycle of testosterone in male and female downy woodpeckers, Picoides pubescens

Studies in birds show that testosterone (T) concentrations vary over the annual cycle depending on mating system and life history traits. Socially monogamous species show pairing behavior throughout the year and low levels of male-male aggression and are underrepresented in these studies, yet the function of testosterone could be particularly important for sexual and social interactions occurring outside the breeding season. We measured fecal T concentrations over the annual cycle and the frequency of interactions between male and female downy woodpeckers (Picoides pubescens) from late fall through early spring. We validated the fecal assay by collecting blood in conjunction with a subsample of our fecal samples: fecal T correlated with circulating levels in the blood. The annual peak level of T in males was relatively low and short-lived, similar to that of other bird species with low levels of male-male aggression and high paternal care. The annual cycle of female T resembled the male pattern, and the ratio of male T to female T was close to 1.0. Likewise, the frequency of aggression among females was similar to the frequency among males. Overall, testosterone levels in both sexes were variable, even in winter. In other bird species, sexual behavior during nonbreeding periods correlates with circulating levels of T in males. Based on this observation, we tested the hypothesis that T in winter was positively related to the frequency of interaction between mated downy woodpeckers. The results showed no such relationship. We discuss this finding and further relate the annual cycle of T in both males and females to behaviors that appear to facilitate mate choice and retention of the pair bond during conspecific challenge.     

Evidence for the consumption of arboreal,diurnal primates by bonobos (Pan paniscus)

We present evidence for the consumption of a diurnal, arboreal, group living primate by bonobos. The digit of an immature black mangabey (Lophocebus aterrimus) was found in the fresh feces of a bonobo (Pan paniscus) at the Lui Kotale study site, Democratic Republic of Congo. In close proximity to the fecal sample containing the remains of the digit, we also found a large part of the pelt of a black mangabey. Evidence suggests that the Lui Kotale bonobos consume more meat than other bonobo populations and have greater variation in the mammalian species exploited than previously thought [Hohmann & Fruth, Folia primatologica 79:103–110]. The current finding supports Stanford's argument [Current Anthropology 39:399–420] that some differences in the diet and behavior between chimpanzees (P. troglodytes) and bonobos are an artefact of the limited number of bonobo study populations. If bonobos did obtain the monkey by active hunting, this would challenge current evolutionary models relating the intra‐specific aggression and violence seen in chimpanzees and humans to hunting and meat consumption [Wrangham, Yearbook of Physical Anthropology 42:1–30]. Am. J. Primatol. 71:171–174, 2009. © 2008 Wiley‐Liss, Inc.     

Diet composition of chimpanzees inhabiting the montane forest of Kahuzi,Democratic Republic of Congo

The diet of chimpanzees was investigated by direct observations, feeding remains, and fecal analysis from January 1994 to December 2000 in the montane forest of Kahuzi-Biega National Park. A total of 171 food items were identified, among which 156 items were plant materials belonging to 114 species from 57 taxonomic families. Chimpanzees consumed 66 species of fruits (62 species of pulps and four species of seeds). Results of fecal analysis showed that fig fruits were the most frequently eaten. Their seeds occurred in 92% of a total of 7212 chimpanzee fecal samples. The chimpanzees changed their diet according to seasonal and annual variations in both abundance and diversity of fruit species. However, they are very selective frugivores. Only a few pulp-fruit species are regularly identified in their fecal samples. During the rainy season, when ripe fruit was scarce, chimpanzees relied heavily on piths and leaves. They swallowed leaves of two species of Commelinaceae without chewing, probably for medical purposes. Animal foods were eaten infrequently. The montane forest of Kahuzi, where chimpanzees range up to 2600 m above sea level, may be the highest altitudinal limit ever recorded for their distribution. Compared to other chimpanzee habitats, Kahuzi has a low diversity of fruit species and the availability of a few pulp-fruit species may be critical to the survival of Kahuzi chimpanzees.     

The geographic apportionment of mitochondrial genetic diversity in east African chimpanzees, Pan troglodytes schweinfurthii

This study is a geographically systematic genetic survey of the easternmost subspecies of chimpanzee, Pan troglodytes schweinfurthii. DNA was noninvasively collected in the form of shed hair from chimpanzees of known origin in Uganda, Rwanda, Tanzania, and Zaire. Two hundred sixty-two DNA sequences from hypervariable region 1 of which of the mitochondrial control region were generated. Eastern chimpanzees display levels of mitochondrial genetic variation which are low and which are similar to levels observed in humans (Homo sapiens). Also like humans, between 80% and 90% of the genetic variability within the eastern chimpanzees is apportioned within populations. Spatial autocorrelation analysis shows that genetic similarity between eastern chimpanzees decreases clinically with distance, in a pattern remarkably similar to one seen for humans separated by equivalent geographic distances. Eastern chimpanzee mismatch distributions (frequency distributions of pairwise genetic differences between individuals) are similar in shape to those for humans, implying similar population histories of recent demographic expansion. The overall pattern of genetic variability in eastern chimpanzees is consistent with the hypothesis that the subject has responded demographically to paleoclimatically driven changes in the distribution of eastern African forests during the recent Pleistocene.      

Year-to-year patterns of circulating levels of testosterone and corticosterone in relation to breeding density, experience, and reproductive success of the polygynous red-winged blackbird.

To determine across-year patterns in plasma testosterone (T) and corticosterone (B) levels in free-living birds, we took blood samples in the same 2-week breeding period during 4 consecutive years from territorial male red-winged blackbirds. We used our data to test predictions of the "challenge hypothesis" of T secretion and also to examine hormonal correlations with age, breeding experience, breeding density, and reproductive success. Average T and B levels across years were not significantly different. Within individuals, T levels between years were, in general, highly variable. T levels of males with territories in high-density breeding areas were significantly higher than those of males with territories in low-density areas. T levels were positively correlated with harem size and there was a trend for males with high T levels to fledge more offspring. We found no relationships between T and B levels and male age or breeding experience. Some results of the study were consistent with the predictions of the challenge hypothesis. This study constitutes one of the few examinations of across-year patterns in circulating hormone levels in wild populations.     

Lethal intergroup aggression by chimpanzees in Kibale National Park, Uganda

Chimpanzees (Pan troglodytes) have hostile intergroup relations throughout most or all of their geographic range. Hostilities include aggressive encounters between members of neighboring communities during foraging and during patrols in which members of one community search for neighbors near territory boundaries. Attacks on neighbors involve coalitions of adult males, and are sometimes fatal. Targets include members of all age/sex classes, but the risk of lethal intergroup coalitionary aggression is highest for adult males and infants, and lowest for sexually swollen females. The best-supported adaptive explanation for such behavior is that fission-fusion sociality allows opportunities for low-cost attacks that, when successful, enhance the food supply for members of the attackers' community, improve survivorship, and increase female fertility. We add to the database on intergroup coalitionary aggression in chimpanzees by describing three fatal attacks on adult males, plus a fourth attack on an adult male and an attack on a juvenile that were almost certainly fatal. Observers saw four of these attacks and inferred the fifth from forensic and behavioral evidence. The attackers were males in two habituated, unprovisioned communities (Ngogo and Kanyawara) in Kibale National Park, Uganda. We also summarize data on other intercommunity attacks at Ngogo. Our observations are consistent with the "imbalance of power" hypothesis [Manson & Wrangham, Current Anthropology 32:369-390, 1991] and support the argument that lethal coalitionary intergroup aggression by male chimpanzees is part of an evolved behavioral strategy.     

Infanticide and cannibalism by male chimpanzees at Ngogo, Kibale National Park, Uganda

Researchers have documented infanticide by adult males in four wild chimpanzee populations. Males in three of these have killed infants from outside of their own communities, but most infanticides, including one from Kanyawara, in Kibale National Park, Uganda, took place within communities. Here we report two new cases of infanticide by male chimpanzees at a second Kibale site, Ngogo, where the recently habituated chimpanzee community is the largest yet known. Both infanticides happended during boundary patrols, which occur at a high frequency there. Patrolling males attacked solitary females who were unable to defend their infants successfully. The victims were almost certainly not members of the Ngogo community. Males cannibalized both infants and completely consumed their carcasses. These observations show that infanticide by males is widespread in the Kibale population and that between-community infanticide also happens there. We discuss our observations in the context of the sexual selection hypothesis and other proposed explanations for infanticide by male chimpanzees. The observations support the arguments that infanticide has been an important selective force in chimpanzee social evolution and that females with dependent infants can be at great risk near range boundaries, but why male chimpanzees kill infants is still uncertain.     

Stress, social behaviour, and secondary sexual traits in a male primate

We examined variation in glucocorticoid levels in the mandrill, a brightly coloured primate species, to identify major social influences on stress hormones, and investigate relationships among glucocorticoid levels, testosterone and secondary sexual ornamentation. We collected a total of 317 fecal samples for 16 adult male mandrills over 13 months, including mating and non-mating periods and periods of both dominance rank stability and instability, and compared fecal glucocorticoid levels with dominance rank, rank stability, presence of receptive females, gastro-intestinal parasite infection, fecal testosterone and facial red coloration. Glucocorticoid levels did not vary systematically with dominance rank, but increased when the dominance hierarchy was unstable, and increased in the presence of receptive females. The relationship between dominance rank and glucocorticoid levels changed direction according to the stability of the dominance hierarchy: glucocorticoid levels were higher in subordinate males under stable conditions, but under conditions of instability higher ranking males had higher glucocorticoid levels. The influence of dominance rank also interacted with the presence of receptive females: glucocorticoids were higher in dominant males than in subordinates, but only during mating periods, suggesting that dominant males are more stressed than subordinates during such periods. These findings support previous studies showing that the relationship between glucocorticoids and dominance rank in male baboons is dependent on the social environment. We also found that males with higher glucocorticoids suffered a higher diversity of gastrointestinal parasite infection, in line with evidence that glucocorticoids suppress the immune system in other species. However, we found no support for the stress-mediated immunocompetence handicap hypothesis for the evolution of condition-dependent ornaments: glucocorticoid and testosterone levels were positively related, rather than the negative relationship predicted by the hypothesis, and we found no relationship between red colour and glucocorticoid levels, suggesting that glucocorticoids do not play a role in translating social conditions or physical health into ornament expression in this species.     

Fruit availability, chimpanzee diet, and grouping patterns on Rubondo Island, Tanzania

We examined seasonal patterns of fruit availability, dietary quality, and group size in the descendants of an introduced chimpanzee population on Rubondo Island, Tanzania. The site has supported a free-ranging population without provisioning for 40 years. Our goals were to determine whether Rubondo chimpanzees experience periods of fruit shortage, and whether they respond to changes in fruit availability similarly to chimpanzees at endemic sites. We indexed the fruit availability of tree and liana species on transects stratified across three chimpanzee ranging areas. We used fecal analyses to evaluate seasonal changes in diet, and used data on party size and nesting group size to examine seasonal patterns of grouping. Tree fruit availability was positively correlated with rainfall, with a period of relative tree fruit scarcity corresponding with the long dry season. Liana fruit availability was not related to rainfall, and lianas exhibited less variable fruiting patterns across seasons. Fruits made up the majority of the chimpanzee diet, with lianas accounting for 35% of dietary fruit species. Fruits of the liana Saba comorensis were available during all months of phenological monitoring, but they were consumed more when tree fruit was scarce, suggesting that Saba comorensis fruits may be a fallback food for Rubondo chimpanzees. There were no increases in consumption of lower-quality plant parts between seasons, and there were no changes in group size between seasons. These results contrast with evidence from several endemic chimpanzee study sites, and indicate that Rubondo chimpanzees may have access to abundant and high-quality foods year round.     

No evidence of short-term exchange of meat for sex among chimpanzees

The meat-for-sex hypothesis posits that male chimpanzees (Pan troglodytes) trade meat with estrous females in exchange for short-term mating access. This notion is widely cited in the anthropological literature and has been used to construct scenarios about human evolution. Here we review the theoretical and empirical basis for the meat-for-sex hypothesis. We argue that chimpanzee behavioral ecology does not favor the evolution of such exchanges because 1) female chimpanzees show low mate selectivity and require little or no material incentive to mate, violating existing models of commodity exchange; and 2) meat-for-sex exchanges are unlikely to provide reproductive benefits to either partner. We also present new analyses of 28 years of data from two East African chimpanzee study sites (Gombe National Park, Tanzania; Kanyawara, Kibale National Park, Uganda) and discuss the results of previously published studies. In at least three chimpanzee communities, 1) the presence of sexually receptive females did not increase hunting probability, 2) males did not share preferentially with sexually receptive females, and 3) sharing with females did not increase a male's short-term mating success. We acknowledge that systematic meat sharing by male chimpanzees in expectation of, or in return for, immediate copulations might be discovered in future studies. However, current data indicate that such exchanges are so rare, and so different in nature from exchanges among humans, that with respect to chimpanzees, sexual bartering in humans should be regarded as a derived trait with no known antecedents in the behavior of wild chimpanzees.     

Human Cells Display Reduced Apoptotic Function Relative to Chimpanzee Cells

Previously published gene expression analyses suggested that apoptotic function may be reduced in humans relative to chimpanzees and led to the hypothesis that this difference may contribute to the relatively larger size of the human brain and the increased propensity of humans to develop cancer. In this study, we sought to further test the hypothesis that humans maintain a reduced apoptotic function relative to chimpanzees by conducting a series of apoptotic function assays on human, chimpanzee and macaque primary fibroblastic cells. Human cells consistently displayed significantly reduced apoptotic function relative to the chimpanzee and macaque cells. These results are consistent with earlier findings indicating that apoptotic function is reduced in humans relative to chimpanzees.     

Chimpanzee feeding ecology and fallback food use in the montane forest of Nyungwe National Park,Rwanda

Almost all primates experience seasonal fluctuations in the availability of key food sources. However, the degree to which this fluctuation impacts foraging behavior varies considerably. Eastern chimpanzees (Pan troglodytes schweinfurthii) in Nyungwe National Park, Rwanda, live in a montane forest environment characterized by lower primary productivity and resource diversity than low‐elevation forests. Little is known about chimpanzee feeding ecology in montane forests, and research to date predominantly relies on indirect methods such as fecal analyses. This study is the first to use mostly observational data to examine how seasonal food availability impacts the feeding ecology of montane forest chimpanzees. We examine seasonal changes in chimpanzee diet and fallback foods (FBFs) using instantaneous scan samples and fecal analyses, supported by inspection of feeding remains. Chimpanzee fruit abundance peaked during the major dry season, with a consequent change in chimpanzee diet reflecting the abundance and diversity of key fruit species. Terrestrial herbaceous vegetation was consumed throughout the year and is defined as a “filler” FBF. In contrast to studies conducted in lower‐elevation chimpanzee sites, figs (especially Ficus lutea) were preferred resources, flowers were consumed at seasonally high rates and the proportion of non‐fig fruits in the diet were relatively low in the current study. These divergences likely result from the comparatively low environmental diversity and productivity in higher‐elevation environments.     

Violent coalitionary attacks and intraspecific killing in wild white-faced capuchin monkeys (<Emphasis Type="Italic">Cebus capucinus</Emphasis>)

During 12 years of observation, we have observed three confirmed and two inferred lethal coalitionary attacks on adult male white-faced capuchins (Cebus capucinus) by members of two habituated social groups at Lomas Barbudal Biological Reserve, Costa Rica. In one case, an alpha male was badly wounded and evicted from his group, and when later found by his former groupmates he was attacked by several of them and died less than 24 h later. In two other cases, lone extra-group males were mobbed by adult and immature males of a bisexual group. One victim's abdomen was torn open and he died less than 24 h later. A second victim was quite badly bitten but may have escaped. The fourth and fifth cases resulted from intergroup encounters. One victim lost the use of both arms but may have survived, whereas the other died of unknown causes within an hour of the attack. The observed death rate from coalitionary aggression at our site is approximately the same as that reported for eastern chimpanzees. Because at least three of the five observed incidents involved large coalitions attacking lone victims, they support the general hypothesis that imbalances of power contribute to intraspecific killing in primates. However, the occurrence of lethal coalitional attacks in a species lacking fission–fusion social organization poses a challenge to the more specific version of the imbalance-of-power hypothesis proposed by Manson and Wrangham in 1991 to explain chimpanzee and human intergroup aggression.     

Chimpanzee (Pan troglodytes) Seed Dispersal in an Afromontane Forest: Microhabitat Influences on the Postdispersal Fate of Large Seeds1

We examined the postdispersal fate of large seeds (≥5 mm) dispersed by chimpanzees in an afromontane forest to evaluate aspects of the effectiveness of seed dispersal by chimpanzees, Pan troglodytes. We assessed the influence of six microhabitat characteristics on seed persistence and germination in seeds dispersed in chimpanzee feces and “wadges.” A total of 257 fecal samples and 56 wadges were located over a 4‐mo period by tracking a semi‐habituated chimpanzee community on day follows. Forty‐nine (19.1%) of the fecal samples contained large seeds from five different tree species. The majority of fecal samples with seeds contained seeds from the mature forest tree Olea capensis (Oleaceae) (83.7%). Forty‐two wadges (75%) contained seeds from the mature forest tree Syzygium guineense (Myrtaceae). Seeds were monitored at their deposition site for removal and germination up to 49 d following deposition. We collected data on the microhabitat surrounding each fecal and wadge sample. Multivariate analyses indicated that while fecal and wadge samples were not clustered into particular microhabitats, there was little overlap in the microhabitats in which wadges and fecal samples were deposited. Significantly more seeds persisted over 49 d in wadges (67.9%) than in feces (30.3%). Elevation was the only microhabitat variable determined to have a significant influence on seed persistence, whereas slope was determined to have a significant influence on germination.     

The influence of age on fecal steroid hormone levels in male Budongo Forest chimpanzees (Pan troglodytes schweinfurthii)

Potential interactions between age and endocrinological functioning have been understudied in wild ape populations. Therefore, we examined the relationship between age and the secretion of androgens and glucocorticoids in 15 juvenile, subadult, and adult male chimpanzees (Pan troglodytes schweinfurthii) free ranging in the Budongo Forest of Uganda. One hundred and nine fecal samples were opportunistically collected, between 07:30 and 13:30 hr, during the wet season. Fecal samples were preserved, by oven drying, and steroid content extracted before radioimmunoassay for dehydroepiandrosterone-sulfate (DHEA-S), testosterone (TEST), cortisol (CORT), and corticosterone (CCT). Employing indexes of age as predictive factors, linear mixed-effects modeling and non-parametric statistical comparisons of fecal steroid levels were conducted. Age was observed to significantly influence the production of both glucocorticoids and androgens in male Budongo Forest chimpanzees. Basically, whereas TEST and CORT increased, DHEA-S and CCT levels slightly declined as animals matured.     

Transmission of human TT virus of genotype 1a to chimpanzees with fecal supernatant or serum from patients with acute TTV infection

Fecal supernatant or serum containing TT virus (TTV) of genotype 1a (10(5) copies/ml) from patients with acute TTV infection was inoculated intravenously into two naive chimpanzees. Serum samples were obtained weekly and tested for TTV DNA by genotype 1-specific polymerase chain reaction. TTV DNA was detected in chimpanzee 228 at weeks 5-15 after inoculation with 0.5 ml of serum, and in chimpanzee 234 at weeks 7-19 after inoculation with 1 ml of fecal supernatant. The TTV DNA titer peaked at weeks 12 and 13 in chimpanzee 228 and at weeks 14-16 in chimpanzee 234. Mild biochemical and histological changes in biopsied liver samples were observed in both chimpanzees in association with the reduction in TTV titer. TTV DNA was transient in chimpanzee 228, but in chimpanzee 234 it reappeared at week 21 and persisted through week 30. These results indicate that TTV in feces is infectious and suggest that TTV has hepatitis-inducing capacity.     

Exposure to perceived male rivals raises men's testosterone on fertile relative to nonfertile days of their partner's ovulatory cycle

The challenge hypothesis posits that male testosterone levels increase in the presence of fertile females to facilitate mating and increase further in the presence of male rivals to facilitate male–male competition. This hypothesis has been supported in a number of nonhuman animal species. We conducted an experiment to test the challenge hypothesis in men. Thirty-four men were randomly assigned to view high-competitive or low-competitive male rivals at high and low fertility within their partner's ovulatory cycle (confirmed by luteinizing hormone tests). Testosterone was measured upon arrival to the lab and before and after the manipulation. Based on the challenge hypothesis, we predicted that a) men's baseline testosterone would be higher at high relative to low fertility within their partner's cycle, and b) men's testosterone would be higher in response to high-competitive rivals, but not in response to low-competitive rivals, at high relative to low fertility within their partner's cycle. Contrary to the first prediction, men's baseline testosterone levels did not differ across high and low fertility. However, consistent with the second prediction, men exposed to high-competitive rivals showed significantly higher post-test testosterone levels at high relative to low fertility, controlling for pre-test testosterone levels. Men exposed to low-competitive rivals showed no such pattern (though the fertility by competition condition interaction fell short of statistical significance). This preliminary support for the challenge hypothesis in men builds on a growing empirical literature suggesting that men possess mating adaptations sensitive to fertility cues emitted by their female partners.