Nonhuman Primates: Between-Group Conflicts
Aggressive interactions between two or multiple conspecifics from different social groups in primate species usually lead to the death of at least one of the contenders and to fitness advantages to victors.
Observations of between-group aggression and killing exist for a wide range of primate and nonprimate mammals. Such phenomena occur in different forms, such as coalitional efforts, or independent and individual attacks (i.e., dyadic aggression), and targets may be individuals of any age. This entry will review the current literature on lethal intergroup aggression in nonhuman primates, describing evidences of adaptive value for aggressive interactions and especially killing between groups, in contrast to hypotheses that rely on the concepts of environmental mismatch, trait neutrality, or maladaptation to explain such phenomena. Even though this review will cover reports of lethal aggression between groups which overall also support adaptive hypotheses in multiple species, special attention will be given to chimpanzees, considering the more intense research effort for this clade and an abundance of aggression descriptions, which permit detailed analyses of evolutionary predictions.
Early theories considered intergroup killings to be exclusive of human societies living in complex political systems such as nation-states. However, observations with chimpanzees (Pan troglodytes) during the 1970s at Gombe, Tanzania (Wilson 2013), indicated that lethal attacks across communities were not uncommon. It is now estimated that intergroup killings in chimpanzees exhibit rates that are comparable to those in small-scale human societies, and higher than in modern, industrialized societies (Wrangham et al. 2006). The implications of these reports for understanding the origins of violence and warfare in human societies were met with both skepticism and opposition. According to some researchers, intergroup killings should not be considered a natural trait of the species, but rather a pathological social behavior associated with human interference factors such as ecological disturbance, the size of the protected area, and food provisioning in field sites. Such view is now currently known as the human impact hypothesis (HIH; Wilson et al. 2014). Another concern is that human “warfare” is a term sometimes arbitrarily reserved for violent conflicts killing at least 1,000 individuals, and chimpanzee conflicts have never been observed to achieve that dimension.
Alternatively, although the adaptive strategies hypothesis (ASH) acknowledged that human activities could not adequately explain the occurrence of lethal encounters, factors such as the number of males in a community and population density were seen as better predictors than human disruption. It is important to note that the HIH does not lead to the prediction that consistent differences would exist in between-group warfare based on the location of the communities (e.g., East–West), the age and sex of attackers and victims, or any variation across species (e.g., bonobos–chimpanzees), beyond the direct effects of human impact. The ASH, on the other hand, leads to the prediction that natural tendencies for aggressive interactions are exhibited in different levels in different populations, sexes, and life history stages in consistent ways. Fifty years of observations, with 152 cases of intergroup killings reported for chimpanzees and bonobos across East and West Africa, suggest that a model presenting predictors associated with HIH was around seven times weaker compared to models accounting for the number of males and population density, both of which are predictors that are part of ASH (Wrangham et al. 2006; Wilson et al. 2014). From these studies, it is possible to conclude that these intergroup killings (1) exhibit design features that indicate a nonrandom distribution for characteristics of aggressors and victims, (2) occur in a predictable and organized fashion, (3) are enacted by individuals in sociodemographic categories which are most prepared for, and which would mostly benefit from, intense competition, as is detailed below. Importantly, victims are mostly unrelated to the attackers, in line with expectations from the Inclusive Fitness theory; and the victims and the attackers are mostly males, in line with sexual selection hypotheses.
The inherent costs associated with lethally attacking another individual pose the problem of how this behavior is maintained across chimpanzee communities. According to the imbalance of power hypothesis, intergroup killings would be expected under conditions favoring a maximum cost for the victim and a minimum risk for the attackers. As a supporting example, in the Taï Forest, Côte d’Ivoire, small parties of 1–3 males tend to avoid any direct engagement with adversaries, medium parties containing 4–6 males investigate more actively the territory, and larger parties with 7–9 males pursue physical contact with victims (Boesch 2003). Across chimpanzee communities, intergroup attackers and killers outnumber the targets by a range of 3–9 aggressors per target.
Playback experiments have also demonstrated that chimpanzees in the company of larger parties tend to approach speakers playing the vocalizations of rival males (Wilson 2013). The strategic nature of this behavior has also been demonstrated based on differences between patrolling (collective travel to the fringes of the territory) and other activities (foraging). Raiding parties spend less time feeding during patrols and more time traveling. Moreover, at the time of the incursion, invaders combine long travel sprints with short rests, periods where they are attentive to cues from members of the rival community. Such organized behaviors further suggest that attacks are not spurious phenomena but rather part of the behavioral repertoire of the species.
Some of the hypothesized functions of intergroup attacks include acquisition of females, territorial expansion, and defense of resources. Supporting this, for example, female chimpanzees in the Taï forest sometime copulate with the attackers within the fringes or their territory (Boesch 2003) or permanently disperse into the group of the attackers (Wilson 2013). Lethal intergroup aggression presents additional fitness benefits to the invaders. Females from invading groups exhibit heavier body mass and shorter interbirth intervals, demonstrating that raiding males improve the reproductive conditions of females and consequently their own, not only by defending the territory but by expanding it (Williams et al. 2004).
Furthermore, intergroup killings in chimpanzee present a median of 271 per 100,000 individuals per year, a rate within the range estimated for human intergroup conflict in small-scale societies (foragers: 164 per 100,000 per year; horticulturists and pastoralists: 595 per 100,000 per year; Wrangham et al. 2006). Thus, by systematically eliminating lone individuals, attackers are able to reduce the size of rival groups, making them more vulnerable to future attacks. The disappearance of communities such as Kahama (Gombe) or the K-group (Mahale) demonstrates the cumulative effect of such raids (Wilson 2013).
As in other forms of cooperation, the risk of free-riding is a potential destabilizer for collective, organized lethal aggression. Hence, by refusing to participate in any raid, freeloading males may obtain fitness benefits though the group without incurring significant costs. However, analyses with males at Kibale detected a positive relationship between the likelihood of participation, mating success and the relationship closeness between participants, as measured by the time spent grooming (Watts and Mitani 2001), suggesting that compensating individual fitness benefits exist for those who engage in these intergroup aggressive efforts.
Territorial species living in societies displaying fission-fusion dynamics are more prone to intergroup lethal aggression (Aureli et al. 2006). However, the degree of cohesion and dispersion should not be solely considered as a risk factor. Due to a lower intensity of within-group competition and a higher risk of predation, groups living in the Taï forest have been found to be more cohesive with respect to Eastern communities. This socioecological context provides a protective factor for potential victims. Thus, by decreasing the distance between individuals, lone targets are able to be rescued by other members of the group that may be close to the attack site.
Primates living in fission-fusion societies, such as spider monkeys (Ateles geoffroyi; Aureli et al. 2006), have also been observed to form small parties of 3–4 males and invade the territory of competing groups. Moreover, intruders encountering rivals (range 1–3 victims) have been reported to chase the victims. However, there are no reports of intergroup killings as with chimpanzees. Importantly, other primate species exhibit both coalitional and dyadic lethal aggression against individuals from other groups. Red colobus monkeys (Procolobus badius) and diana monkeys (Cercopithecus diana; Watts et al. 2006) for example, can form small coalitions and kill rivals.
Recently in the Mawas Reserve, Indonesia, a coalitional attack by a male and a female orangutan led to the death of another female, despite the interventions of another male during the attack (Marzec et al. 2016). Only rare evidence of coalitional killings has been described in mountain gorillas (Gorilla beringei). In Karisoke, Rwanda, roving males sometimes clash with resident males, and even though the vast majority of these interactions are nonfatal, rare killings have been described (Rosenbaum et al. 2016). In the case of white-handed gibbons (Hylobates lar; Palombit 1993), an invading individual was seen to be lethally wounded by another male. These observations, however, do not approach the mortality rates observed in between-group conflicts in chimpanzees, suggesting that characteristics particular to the socioecology of this species have selected for stronger intergroup competition.
Coalitional lethal aggression against vulnerable lone individuals has also been described in nonprimate species (Watts and Mitani 2001; Watts et al. 2006), including grey wolves (Canis lupus), African lions (Panthera leo), cheetahs (Acinonyx jubatus), and spotted hyenas (Crocuta crocuta). In some species such as grey wolves, the mortality rate due to intergroup conflict may reach considerable proportions, demonstrating significant parallelisms with lethal interactions present in nonhuman primates.
Intergroup lethal conflict can be classified as either coalitional or dyadic aggression. Chimpanzee data clearly demonstrate that intergroup killings are not an outcome of human disruption, but rather an emergent phenomenon based on socioecological features including population density, and the number of males per community, with adaptive consequences. Similarly, intercommunity killings display consistent sociodemographic patterns in that males are more likely to be both aggressors and victims, Eastern communities have higher intergroup lethality rates than Western populations, and chimpanzees have higher intergroup lethality rates than bonobos. Patterns of patrolling and lethal attacks have also been found to be strategic, suggesting adaptation. Raiding individuals spend less time foraging and more time travelling into the territory of rival communities. Similarly, lethal attacks occur based on an attackers/victims ratio. Invaders obtain fitness benefits, including attracting new females, inducing shorter interbirth intervals, and acquiring resources.
Several other anthropoid primate species have also been observed to engage in intergroup lethal aggression, however the mortality rate due to conflict is considerably lower than that of chimpanzees. Still, these occurrences as well as observations in various carnivoran species disconfirm previous hypotheses considering intergroup lethal aggression to be either adaptively neutral or maladaptive, or a contemporary phenomenon exclusive of complex human societies.
- Boesch, C. (2003). Complex cooperation among Taï chimpanzees. In F. B. M. de Waal & P. L. Tyak (Eds.), Animal social complexity (pp. 93–110). Harvard: Harvard University Press.Google Scholar
- Marzec, A. M., Kunz, J. A., Falkner, S., Atmoko, S. S. U., Alavi, S. E., Moldawer, A. M., … & van Noordwijk, M. A. (2016). The dark side of the red ape: Male-mediated lethal female competition in Bornean orangutans. Behavioral Ecology and Sociobiology, 70, 459–466.Google Scholar
- Wilson, M. L., Boesch, C., Fruth, B., Furuichi, T., Gilby, I. C., Hashimoto, C., … & Lloyd, J. N. (2014). Lethal aggression in Pan is better explained by adaptive strategies than human impacts. Nature, 513, 414–417.Google Scholar