The role learning plays in the acquisition of communicative gestures by wild chimpanzees (Pan troglodytes) is unclear. We aimed to evaluate the likelihood that social experience influences the structure of chimpanzee buttress drumming displays by examining whether individuals differed in the way they used their hands and feet to strike trees. We analyzed digital video recordings of 245 bouts by 9 adult males from Gombe National Park, Tanzania, frame by frame in conjunction with acoustic analysis. We investigated 1) how limb sequences used to approach drumming trees influenced limb use during drumming, 2) the relative use of hands vs. feet in drumming, and 3) the relative amplitude of beats produced by hands vs. feet. We found that the chimpanzees most often approached trees at a gallop and usually initiated drumming bouts with limb sequences that were identical to gait limb sequences. All individuals produced more beats with their feet than with their hands, and foot beats were higher in relative amplitude than hand beats. In only one instance did an individual produce a bout with hands only, whereas in three of nine observations of drumming on resonant camp equipment, the individuals primarily used their hands rather than their feet. We suggest that although chimpanzees may, by observing others, learn to use buttresses as tools to generate loud sounds, it is unlikely that learning influences the structure of displays because they result from innately determined gait patterns deployed to generate sound from comparatively nonresonant substrates.

Chimpanzees exhibit cultural variation, yet examples of successful cultural transmission between wild communities are lacking. Here we provide the first account of tool-assisted predation ("ant fishing") on Camponotus ants by the Kasekela and Mitumba communities of Gombe National Park. We then consider three hypotheses for the appearance and spread of this behavior in Kasekela: (1) changes in prey availability or other environmental factors, (2) innovation, and (3) introduction. Ant fishing was recognized as habitual in the Mitumba community by 1992, soon after their habituation began. Apart from one session in 1978, Camponotus predation (typically with tools) was documented in the Kasekela community beginning only in 1994, despite decades of prior observation. By February 2010, ant fishing was customary in Kasekela and with one exception was practiced exclusively by chimpanzees born after 1981 and immigrant females. We hypothesize that changes in insect prey availability over time and/or the characteristics of one popular ant-fishing site may have influenced the establishment of ant fishing. Though innovation cannot be completely ruled out, the circumstantial evidence suggests that a Mitumba immigrant introduced ant fishing to Kasekela. We submit that this report represents the first documented case of successful transmission of a novel cultural behavior between wild chimpanzee communities.

Chimpanzees exhibit cultural variation, yet examples of successful cultural transmission between wild communities are lacking. Here we provide the first account of tool-assisted predation ("ant fishing") on Camponotus ants by the Kasekela and Mitumba communities of Gombe National Park. We then consider three hypotheses for the appearance and spread of this behavior in Kasekela: (1) changes in prey availability or other environmental factors, (2) innovation, and (3) introduction. Ant fishing was recognized as habitual in the Mitumba community by 1992, soon after their habituation began. Apart from one session in 1978, Camponotus predation (typically with tools) was documented in the Kasekela community beginning only in 1994, despite decades of prior observation. By February 2010, ant fishing was customary in Kasekela and with one exception was practiced exclusively by chimpanzees born after 1981 and immigrant females. We hypothesize that changes in insect prey availability over time and/or the characteristics of one popular ant-fishing site may have influenced the establishment of ant fishing. Though innovation cannot be completely ruled out, the circumstantial evidence suggests that a Mitumba immigrant introduced ant fishing to Kasekela. We submit that this report represents the first documented case of successful transmission of a novel cultural behavior between wild chimpanzee communities.

We examined lateral biases in the asymmetrical gallop gaits of wild chimpanzees (Pan troglodytes) approaching trees to drum in Gombe National Park, Tanzania. This is the first study of chimpanzee fast gaits under natural conditions. Analyzing digital video recordings collected over a 12-year period, we were able to determine lateral bias in 153 gallop bouts for the eight most frequently sampled males. For 89 of these bouts, symmetry and duty factor measurements were also possible for at least one stride cycle. Seven of eight males were lateralized, and the eighth was ambi-preferent. The degree of lateralization was comparable to that reported for chimpanzee hand preference during complex, hi-manual object manipulation, and similar to that reported for galloping in domestic horses. No group-level directional bias was found. Although little is known about lateral biases in the asymmetrical gaits of mammals in the wild, we speculate that predator-prey dynamics would select against group-level lateral biases but for the ability to switch leads. The absence of lead switching by chimpanzees in this study may reflect selection for maximum speed to reach arboreal escape routes. We discuss the possibility that locomotor lateralization could constrain the emergence of group-level biases in hand preference in chimpanzees, and could be implicated in the development of limb long bone asymmetries.

In 1998, four chimpanzees in the Mahale Mountains National Park, Tanzania, were observed wiping their mouths with non-detached leaves or stalks of grass, or rubbing their mouths with a tree trunk or branch, especially while eating lemons. The number of mouth-wiping/rubbing individuals increased to 18 in 1999. By 2005, 29 chimpanzees were documented wiping/rubbing their muzzles in this way. Although it is difficult to determine whether the chimpanzees acquired this behavior as a result of trial and error or social learning, the fact that chimpanzees at other sites perform this behavior with detached leaves or leafy twigs much more often than with intact items suggests the possibility that cleaning with intact plant parts at Mahale spread via social learning.

Aggression is generally more severe between males than between females because males gain greater payoffs from escalated aggression. Males that successfully defeat rivals may greatly increase their access to fertile females. Because female reproductive success depends on long-term access to resources, competition between females is often sustained but low key because no single interaction leads to a high payoff. Nonetheless, escalated aggression can sometimes immediately improve a female’s reproductive success. Resisting new immigrants can reduce feeding competition, and infanticide of other females’ young can increase a female’s access to resources. East African chimpanzees live in fission-fusion communities in which females occupy overlapping core areas. Growing evidence indicates that reproductive success correlates with core area quality, and that females compete for long-term access to core areas. Here we document 5 new cases of severe female aggression in the context of such competition: 2 attacks by resident females on an immigrant female, a probable intracommunity infanticide, and 2 attacks on a female and her successive newborn infants by females whose core areas overlapped hers. The cases provide further evidence that females are occasionally as aggressive as males. Factors influencing the likelihood and severity of such attacks include rank and size differences and the presence of dependable allies. Counterstrategies to the threat of female aggression include withdrawing from others around the time of parturition and seeking male protection. We also discuss an unusual case of a female taking the newborn infant of another, possibly to protect it from a potentially infanticidal female.

Despite considerable attention to chimpanzee intergroup violence, the number of observed cases remains small. We report 4 cases of intergroup violence that occurred in Gombe National Park, Tanzania, between 1993 and 2002. We observed (3 cases) or inferred (1 case) males from the Kasekela community to attack members of their 2 neighboring communities: Mitumba and Kalande. In 1993, Kasekela males killed and ate a female infant from Mitumba. In 1998, Kasekela males captured 2 infants (sex unknown) from Kalande, one of which escaped and the other was killed and eaten. Also in 1998, Kasekela males attacked an adolescent male from Kalande. The victim was alive but severely injured by the end of the attack. The intensity and duration of the attack are comparable to other attacks that resulted in fatal injuries. In 2002, observers found the body of an adolescent male from Mitumba following an incursion by Kasekela males into the area. The injuries inflicted on the Mitumba male together with circumstantial evidence suggest that Kasekela males killed him. The attacks support the view that intergroup violence is a persistent feature of chimpanzee societies and that the primary benefit attackers gain from them is reduced competition for resources.

Play in nonhuman animals has generally been viewed as being uniform among study sites. No studies have examined whether there are local variations in play. In this work we report an apparently locality‐specific form of play that is basically solo locomotor play, but also has aspects of object play and social play. We describe this unusual “leaf‐pile pulling” (LPL) pattern based on video footage of the chimpanzees of Mahale, Tanzania. Typically, when a party of chimpanzees moves in a procession down a slope in the dry season, a youngster will turn around and walk backward while raking many dry leaves with both hands. This activity accumulates many dry leaves while producing a lot of sound. After the player walks 1–15 m, he/she either turns around and walks forward or moves in a somersaulting fashion. The performer usually faces an individual that is immediately following him/her in the procession. The age of the performers ranges from 2 to 22 years, but 3–10 years are most typical. Compared to younger (<8 years) individuals, older (≥8 years) individuals tend to cover longer distances during play, and to be more likely to play only on sloped surfaces and during travel. One of the authors (W.W.) has only seen the behavior exhibited by five different individuals on 10–15 occasions at Gombe. The behavior is exhibited less often at Gombe than at Mahale. To date, LPL has not been reported elsewhere. The absence of LPL at other sites may or may not be explained by environmental differences, such as differences in the availability of dead leaves and sloped terrain. Although the possibility that LPL is acquired by individual learning cannot be completely ruled out, the hypothesis that it is a tradition of the Mahale study group is more likely. This suggests that play may warrant more careful consideration in studies of nonhuman primate culture. Am. J. Primatol. 60:167–173, 2003. © 2003 Wiley‐Liss, Inc.