Here, we report the first observations of two lethal coalitionary attacks of chimpanzees on another hominid species, gorillas. In both events, the chimpanzees considerably outnumbered the gorillas, however in the second event, the lethal attack started when the silverback had abandoned his group. In both events, the victims were gorilla infants, but the consumption of the victim was observed in one event only.
Recent studies59 were able to distinguish genetically distinct gorilla groups within the study area; when overlaid against the Rekambo chimpanzee community home range there was clear overlap with seven distinct gorilla groups (see Supplementary Fig. S1 in the Supplementary information). However, further data are needed to clarify whether our rare observations are due to lack of data or indeed mirror true frequencies of interspecies interactions in the study area.
In the following paragraphs, we will present and discuss several possible explanations that may account for the two lethal coalitionary encounters observed.
One explanation may be that the observed events represent cases of predation with the chimpanzees hunting and opportunistically targeting the smaller-bodied gorilla infants as prey. Although differences in behaviours accompanying hunting and hunting patrol patterns have been observed across sites, the behaviors observed at Loango were similar to the patterns reported for Taï41,60 and Ngogo38. For instance, the chimpanzees showed conspicuous behaviours prior to hunting such as being extremely attentive to any arboreal movements, scanning, changing directions several times without vocalizing, and performing specific call types—hunting calls38,45. Post-hunting behaviour is characterized by the prevalence of high-ranking males as the primary prey possessors and consumers, high levels of attention, arousal and excitement of party members, as well as begging and food sharing13,40,41,41,61. However, the behaviours observed during the two events were very different to those reported during hunting: The chimpanzees were noisy, emitted alarm barks and screams and performed displays long before the infants were killed. The excitement levels dropped immediately following the death of the infant gorillas. In addition, the observed feeding behaviours during the two events also differed from patterns expected during conventional hunting for the purpose of gaining nutritional benefits through the consumption of prey27. In the first encounter no feeding behaviour was observed, and in the second event the gorilla infant was almost entirely consumed by a single adult female. In contrast to species-typical hunts, in the second event the majority of individuals present, including adult males, showed almost no interest in the carcass, and only small amounts of meat were exchanged between low ranking individuals.
Another explanation may be that the two cases are the product of interspecific competition such as IGP and IK. So far, studies investigating interspecific competition in gorillas and chimpanzees have provided evidence for dietary niche differentiation and mutual avoidance to limit competition e.g.,62,63,64,65,66,67. All previous accounts of interspecies interactions as well as co-feeding events have been reported as peaceful despite a relatively high potential for feeding competition concerning key resources or during certain periods e.g.,68,69,70. Thus far, aggressive interference competition, including infanticide, has been observed between monkey species (e.g., Cercopithecus nicitans stampflii, Cercopithecus diana diana71; Ateles hybridus, Alouatta seniculus72) but not between chimpanzees and gorillas. Such interactions are however frequent in carnivore species and have been suggested as key determinants of their abundance and distribution33,73 (but see for an overview of other taxa31). As in the lethal interactions discussed here, carnivores tend to attack their closest dietary competitors31, most agonistic encounters occur in seasonal environments when food is scarce27, and killings decrease abruptly when dietary overlap is reduced73. Gorillas and chimpanzees show considerable dietary overlap and have a relatively high potential for dietary competition45,74. Across study communities, the degree of dietary overlap ranges between: 50% Kahuzi-Biega; Gorilla beringei graueri, Pan troglodytes schweinfurthii75 and 60–80% Loango, Lopé, and Ndoki; Gorilla g. gorilla, Pan t. troglodytes66,74,76. The two lethal encounters we observed occurred at times characterized by food scarcity and a period of high dietary overlap (for fruit resources)45,74—February and December 2019. In contrast, the two previously observed peaceful co-feeding events took place in April, a month characterized by relatively low dietary overlap between the two species45,74.
Furthermore, age, size and patterns of grouping seem to play a significant role in the outcome of IGP’s and IK’s (see e.g.,27). While relative body size of the opponents is the primary determinant of lethal interactions and results in favour of the larger species, in interactions involving adults, smaller species frequently kill the young of larger species27,73. There are cases where smaller species were able to kill or deter larger species such as wolves (Canis lupus) killing adult black bears (Ursus americanus)77 and hyenas (Crocuta Crocuta) killing lions (Panthera leo)27, however, these outcomes were only possible when individuals of the smaller species formed coalitions27,77. The grouping style of a species was found to strongly influence the outcome of IGP’s resulting largely in favour of species that form groups78. This is in line with our current observations, where the chimpanzees were at an advantage even against the larger gorilla species, given their ability to cooperate. Additionally, specific adaptations to prey-capture also influence the outcome of IGP’s, resulting in favour of species more adapted for vertebrate predation73 where the successful species, here, the chimpanzee, has adaptations to vertebrate predation13,39,41. Hence, as in IGP food webs (with specific emphasis of species classification) portrayed by Arim and Marquet79, the two reported killings may represent cases of IGP and IK between an intermediate omnivorous species (i.e. broad diets comprising both animal and plant foods80), the chimpanzee, and a herbivorous species (feeding mainly on plant foods81), the gorilla.
Lastly, both of the lethal encounters reported here also showed similarities to behaviours observed during chimpanzee intercommunity encounters. For instance, similar to territorial patrols, where chimpanzees move to the periphery and beyond their territorial boundaries to search for neighbours e.g.,11,13,82,83, the observed events took place in the peripheries of the territory before and during territorial patrols. In both events, infants were targeted and adult males were the main attackers and played the most active roles. Similarly, in lethal chimpanzee intercommunity encounters, infanticide is common and adult males are the main participants11,83,84,,83,84 (but see for female roles51). It has been proposed that in chimpanzees, adult males may kill infants of other communities to reduce competition for food by inducing foreign females to avoid contested regions84. The observed interspecies killings of gorilla infants by chimpanzees could have similar motivations85. We also observed behaviours before and during the encounters characteristic to coalitionary intercommunity encounters such as aggression (e.g., charges, chases, threatening displays, contact aggression), high levels of arousal and the use of loud vocalizations13,14,15,51. The imbalance-of-power hypothesis postulates that the function of unprovoked intercommunity aggression (such as deep incursions into other chimpanzee communities’ territory and coalitionary attacks) is a drive for dominance over neighbours resulting in fitness benefits for the attackers through improved access to resources such as food, females, or safety6,13. Two conditions are proposed to be required to trigger coalitional killing of neighbours: (i) a state of intergroup hostility, and (ii) sufficient imbalances of power between interacting parties resulting in impunity from aggressors. Thus, it may be possible that at Loango, which is characterized by relatively high dietary food overlap in specific months45,74, gorillas are perceived as competitors, for both space and resource use, similar to members of other chimpanzee communities. Lastly, we cannot rule out that the presence of human observers, in both events, may have had an effect on the unhabituated silverback’s departure and may have tilted the imbalance of power in favour of the habituated chimpanzees.
In sum, the observed events show similarities to patterns reported in IGP’s, IK’s and intraspecies agonistic encounters. Ultimately, additional observations in combination with isochronous assessments of fruit availability and dietary overlap are needed to differentiate whether coalitionary attacks are indeed the output of interspecific predation spurred by opportunistic hunting, interspecies competition for food resources or whether these interactions are merely a non-adaptive by-product of the “xenophobic nature” of chimpanzees. Finally, analyses of long-term phenological data could aid in investigating if potential high levels of feeding competition may be a more recent phenomenon caused by a collapse in fruit availability as observed in other tropical forests in Gabon86.