Social complexity parallels vocal complexity: a comparison of three non-human primate species

Abstract

Social factors play a key role in the structuring of vocal repertoires at the individual level, notably in non-human primates. Some authors suggested that, at the species level too, social life may have driven the evolution of communicative complexity, but this has rarely been empirically tested. Here, we use a comparative approach to address this issue. We investigated vocal variability, at both the call type and the repertoire levels, in three forest-dwelling species of Cercopithecinae presenting striking differences in their social systems, in terms of social organization as well as social structure. We collected female call recordings from twelve De Brazza's monkeys (Cercopithecus neglectus), six Campbell's monkeys (Cercopithecus campbelli) and seven red-capped mangabeys (Cercocebus torquatus) housed in similar conditions. First, we noted that the level of acoustic variability and individual distinctiveness found in several call types was related to their importance in social functioning. Contact calls, essential to intra-group cohesion, were the most individually distinctive regardless of the species, while threat calls were more structurally variable in mangabeys, the most "despotic" of our three species. Second, we found a parallel between the degree of complexity of the species' social structure and the size, diversity, and usage of its vocal repertoire. Mangabeys (most complex social structure) called twice as often as guenons and displayed the largest and most complex repertoire. De Brazza's monkeys (simplest social structure) displayed the smallest and simplest repertoire. Campbell's monkeys displayed an intermediate pattern. Providing evidence of higher levels of vocal variability in species presenting a more complex social system, our results are in line with the theory of a social-vocal coevolution of communicative abilities, opening new perspectives for comparative research on the evolution of communication systems in different animal taxa.

Keywords: Cercocebus sp.; Cercopithecus sp.; acoustic individual distinctiveness; acoustic variability; evolution of communication; social system; vocal repertoire.

Figure 1

Sonograms of contact, threat and alarm calls in De Brazza's monkeys, Campbell's monkeys, and red-capped mangabeys. Sonograms: X-axis: duration (ms), Y-axis: frequency (kHz). The call type name is given above the sonogram call types are labeled as in previous publications (De Brazza's monkeys: Bouchet et al., ; Campbell's monkeys: Lemasson and Hausberger, ; red-capped mangabeys: Bouchet et al., 2010). Each sonogram is labeled according to the composition in units of the call (units are named using onomatopoeia), and the unit on which measurements have been performed is set in bold type. In the three species, alarm calls display a noisy pattern. In guenons, alarm calls (“Tek+” and “RRA”) consist of repeated pulses, while red-capped mangabey alarm calls (“WaHoo”) are systematically composed of two units of different types. Guenon threat calls (“Wrr+” and “RRC”) display an acoustic pattern similar to that of their alarm calls, whereas red-capped mangabey threat calls (“Un+(Uh)”) are composed of one or several short low-pitched tonal units, sometimes combined to an additional “Uh” unit. In the three species, contact calls are composed of one short, low-pitched, tonal unit which is uttered mostly singularly in De Brazza's monkeys (“On”), combined to a second arched-shape high-pitched unit in Campbell's monkeys (“CH”), and either singularly or repeated in red-capped mangabeys (“Ro+”).

Figure 2

Acoustic intra-individual variability and individual distinctiveness in contact, threat and alarm calls in each of three species of primate. CVmean_intra: call type intra-individual coefficient of variation (%) and Friedman test result (^*P < 0.05; results of the post-hoc multiple comparisons are illustrated by letters symbolizing homogeneous subsets). Hs: information content value (see the Statistical Analysis section for definitions). Dcall, call duration (ms); #units, number of units; Dunit, unit duration (ms); Fbase, unit base frequency (Hz); and Fpeak, unit peak frequency (Hz).

Figure 3

Hourly call rates in each of three species of primate. Number of calls/h: $\overline{X}\pm SE$. Results of the post-hoc multiple comparisons following the Kruskall-Wallis test performed after removing the outlier (one De Brazza's monkey female). The one-tailed pairwise comparisons were chosen according to the “vocal grooming” hypothesis (H1: “De Brazza's monkeys” < “Campbell's monkeys,” H1: “De Brazza's monkeys” < “red-capped mangabeys,” and H1: “Campbell's monkeys” < “red-capped mangabeys”) (^**P < 0.025, ^*P < 0.05, NS, not significant).

Figure 4

Relative proportion of the different “unit assembling patterns” produced by each of three species of primate. Below the species' name: number and name of unit types. Unit types are labeled as in previous publications for De Brazza's monkeys (Bouchet et al., 2012b) and red-capped mangabeys (Bouchet et al., 2010). Campbell's monkeys' repertoire, however, has been revised according to Bouchet et al. (2010) multi-level approach; the correspondence between this nomenclature and the call type names used in Lemasson and Hausberger (2011) is as follow: “On” = SH; “I” = ST; “Ra, RaRa, RaRaRa …” = RRA; “Wi” = SA; “RuRu, RuRuRu …” = RRC; “KekKek, KekKekKek …” = RSA; “OnHi” = CH; “IOn” = CT; “UnI” = RST. On the right-hand side: all the unit assembling patterns observed, grouped as “single”, “repeated”, “combined” (see the Data Analysis section for definitions). Below each diagram: DI, diversity index (Shannon and Weaver, 1949). The smaller the value of DI, the less diverse the repertoire.

Figure 5

Hourly call rate of “single”, “repeated” and “combined” unit assembling patterns in each of three species of primate. Number of calls/h: $\overline{X}\pm SE$. Results of the post-hoc multiple comparisons following the Kruskall-Wallis test (^*P < 0.05, NS, not significant).