Chest beats as an honest signal of body size in male mountain gorillas: Positive correlations among male body size, dominance rank and reproductive success

Chest beats as an honest signal of body size in male mountain gorillas (Gorilla beringei beringei). Edward Wright, Sven Grawunder, Eric Ndayishimiye, Jordi Galbany, Shannon C. McFarlin, Tara S. Stoinski & Martha M. Robbins. Scientific Reports volume 11, Article number: 6879. Apr 8 2021. https://www.nature.com/articles/s41598-021-86261-8

Abstract: Acoustic signals that reliably indicate body size, which usually determines competitive ability, are of particular interest for understanding how animals assess rivals and choose mates. Whereas body size tends to be negatively associated with formant dispersion in animal vocalizations, non-vocal signals have received little attention. Among the most emblematic sounds in the animal kingdom is the chest beat of gorillas, a non-vocal signal that is thought to be important in intra and inter-sexual competition, yet it is unclear whether it reliably indicates body size. We examined the relationship among body size (back breadth), peak frequency, and three temporal characteristics of the chest beat: duration, number of beats and beat rate from sound recordings of wild adult male mountain gorillas. Using linear mixed models, we found that larger males had significantly lower peak frequencies than smaller ones, but we found no consistent relationship between body size and the temporal characteristics measured. Taken together with earlier findings of positive correlations among male body size, dominance rank and reproductive success, we conclude that the gorilla chest beat is an honest signal of competitive ability. These results emphasize the potential of non-vocal signals to convey important information in mammal communication.

Discussion

Our results indicate that mountain gorilla chest beats reliably convey information about the body size of the sender. Larger males consistently emitted chest beats with lower median peak frequencies than smaller males. This finding is an important contribution to the growing literature on honest signaling of body size in acoustic communication, which has predominantly focused on vocalizations5,10,11,12,13,14,15,16,17,18. This is one of a few studies in mammals demonstrating that body size is reliably encoded in a non-vocal acoustic signal. In eland bulls, body size (skeletal measures and muscle mass) was shown to be negatively correlated with the peak frequency of knee clicks21. In adult male gorillas, body size (a composite measure combining crest height and back breadth) is thought to reflect competitive ability because it correlates with dominance rank in multi-male groups and reproductive success31. Additionally, silverbacks chest beat more frequently on days when females are in estrous, presumably as a courtship display28. Taken together, this strongly suggests that the chest beat is an extremely important signal in intrasexual competition and intersexual mate choice in gorillas. Moreover, given that different forms of drumming behaviour, incorporating substrates other than the chest or body, are surprisingly common in a wide range of animals19,20, it is likely that this understudied non-vocal acoustic mode of communication functions to reliably indicate competitive ability in many other species as well.

Our measure of body size, back breadth, likely correlates with a range of morphological traits, including chest volume, pulmonary capacity and hand size. Therefore it is unclear which specific trait or traits are responsible for driving the inverse relationship between back breadth and peak frequency. Moreover, gorillas like other non-human primates possess laryngeal air sacs which are thought to act as resonators, enhancing acoustic signals4,11,24. Indeed, gorillas appear to use laryngeal air sacs during growling vocalizations which often accompany chest beating24. The volume of laryngeal air sacs is likely to be directly correlated with body size, at least within-species (which in orangutans (Pongo) can reach a massive 6 L38). Thus larger males are expected to have larger laryngeal air sacs than smaller males, further lowering the resonating non-vocal frequencies produced whilst chest beating. However, our knowledge of the size and function of laryngeal air sacs in primates and other taxa remains poor39.

Both dominant and subordinate male gorillas emitted chest beats. In general, gorilla males likely chest beat to attract estrous females and intimidate rivals22,28,31. However, younger subordinate males may also chest beat as a means to fine tune this signal and acquire social feedback from conspecifics. The importance of practice is evident as infants as young as one year of age commonly start emitting chest beats during social play22,40. Interestingly, the chest beat rate (number of chest beats per unit time) in the current study (2014–2016) was over three times higher than what was previously reported (1968–1969)23. We are unsure why this is the case, but it could be due to a number of different factors, including a higher number of estrous females per group, younger males, or more intergroup encounters over time41.

Even though we have demonstrated that chest beats reliably convey the body size of the emitter, future studies need to show that receivers actually attend to this information. Gorilla chest beats are thought to play a key role in male–male competition allowing individuals to assess the fighting ability of competitors and thus influence whether they should initiate, escalate or retreat in intra- and intergroup contests22,31,42. Similarly, male gelada baboons (Theropithecus gelada) assess the competitive ability of rivals through vocalizations and compare it to their own, governing how they respond in contests43. Intense contact aggression between males is infrequent in gorillas, which is presumed to reflect the high costs of aggression and their ability to resolve conflicts without resorting to this high risk behaviour (within-group31,42; between-group44). We expect that chest beats to also play a critical role in mate choice22,28,45, providing females with information about the size of the males in their own group and in neighbouring groups, which may influence their decision to transfer to another group. Larger alpha males lead groups with more adult females than smaller males, strongly suggesting that females actively chose to transfer into groups with large alpha males29,30,31. Gorillas may be similar to red deer hinds in their ability to discriminate between the acoustic signals of their current harem-holder stag and those of neighbouring stags46. Lastly, because chest beats can be heard over long-distances, we predict that both male size and the number of different males emitting chest beats are two important factors influencing group movement. Recent work in Bwindi mountain gorillas speculated that one of the functions of chest beats is to mediate how groups use space, with smaller groups with fewer adult males likely avoiding larger ones with more males, which would help to explain their findings that larger groups having more exclusive home ranges and core areas than smaller groups47.

We found no support for body size to influence the duration of chest beats, the number of beats, or the beat rate. Acoustic signals are thought to be energetically expensive to produce34,35,36,37 and we expected chest beats to be as well, with anecdotal accounts of gorillas that emit a high frequency of chest beats showing signs of exhaustion (personal observation). This is in contrast to studies of savannah baboons (Papio ursinus), showing that males with higher competitive ability produce longer vocalizations than weaker ones48,49. It is possible that the duration of chest beats (and the beat rate) decreases over time during periods of high chest beating frequency, and this decrease may be stronger in smaller males. In general the relationship between body size and the duration of vocalizations and other acoustic sounds has been understudied in mammals4,50.

In addition to conveying information on body size (and other phenotypic traits), we would expect it to be important for chest beats to be individual-specific, thereby allowing receivers to discriminate the identity of the emitter. Further study is needed to determine if there are individual signatures to the chest beats. Interestingly, we found smaller within-individual than between-individual coefficients of variation, particularly for chest beat duration and number of beats (39.0 vs. 85% and 31.6 vs. 67.9%, respectively). Notably, several temporal aspects of non-vocal drumming displays by chimpanzees (Pan troglodytes), ruffed grouse (Bonasa umbellus) and great spotted woodpeckers (Dendrocopos major) show significant individual variation, similar to many vocalizations in a wide range of species19,51,52. For example, the buttress drumming of individual chimpanzees significantly differ in the mean duration and the mean number of beats52,53.

The gorilla chest beat has both an acoustic and visual component, making it a multimodal signal. Individuals in visual proximity can benefit from seeing and hearing the gorilla emitting the chest beat, whereas individuals further away rely on the acoustic component. Researchers have been interested in determining whether the different components of multimodal signals convey the same (redundant signal or backup hypothesis) or different information (multiple messages hypotheses)54. Gorillas live in tropical forests with dense vegetation, meaning that it is often difficult to see conspecifics even if they are close by. Therefore, we argue that the evolution of the chest beat as a multimodal signal is at least in part to enhance signal transmission in an environment with limited visibility. We would expect the same messages to be transmitted in both visual and acoustic modalities, which would provide support for the redundant signal hypothesis. However, these two hypotheses are not mutually exclusive, as the chest beat signal may transmit additional information, other than body size, which is then repeated in the visual and acoustic modalities, providing support for both hypotheses.