THE RAFFLES BULLETIN OF ZOOLOGY 2011 THE RAFFLES BULLETIN OF ZOOLOGY 2011 59(2): 319–323 Date of Publication: 31 Aug.2011 © National University of Singapore PROBOSCIS MONKEYS (NASALIS LARVATUS (WURMB, 1787)) HAVE UNUSUALLY HIGH-PITCHED VOCALIZATIONS Amrita Srivathsan Department of Biological Sciences, National University of Singapore, Singapore Rudolf Meier Department of Biological Sciences, National University of Singapore, Singapore Email: [email protected] (corresponding author) ABSTRACT. – According to qualitative descriptions, proboscis monkeys have high frequency vocalizations, but no technical analysis of the calls is available because the species is rare, diffi cult to study in the fi eld, and few zoological institutions have breeding populations. Here we analyse the high pitch vocalizations of captive proboscis monkeys in the Singapore Zoological Gardens. The monkeys produce very high frequency tonal vocalizations ranging from 1.4 kHz to 6.8 kHz. The calls are produced during acts of aggression or used as alarm calls and the pitch is higher than expected based on a comparative study for such a large primate. We discuss that the high frequency may be an adaptation for predator avoidance (crocodiles) and/or communication in dense vegetation. KEY WORDS. – Primate calls; vocal communication; communication; Colobine; spectrograms; high frequency. INTRODUCTION could yield valuable contributions to the fi eld of primate communication. Given the problems with studying leaf The vocalizations of a large number of primate species have monkey vocalizations in the fi eld, using captive populations been studied since the early 1960s. Most major clades of is an alternative that can provide useful information about primates have been well covered and species such as Saimiri the species. This approach has the additional advantage of sciureus (Linnaeus) (Biben & Symmes, 1986; Symmes & yielding high-quality recordings that can be analysed using Goedeking, 1988; Boinski & Mitchell, 1997), Callithrix quantitative techniques. pygmaea (Spix) (Pola & Snowdon, 1975; de la Torre & Snowdon, 2009), Macaca fuscata (Blyth) (Masataka, 1983; Proboscis monkeys (Nasalis larvatus) are arboreal leaf- Inoue, 1988), Macaca mulatta (Zimmermann) (Rowell, 1962; eating monkeys that are found in Borneo, Brunei Kalimantan Hauser, 1993a), Chlorocebus aethiops (Linnaeus) (Owren (Indonesia), Sabah, and Sarawak (Malaysia) where they & Bernacki, 1988; Brown et al., 1992) and Lemur catta inhabit swamp and mangrove forests (Bennet & Sebastian, Linnaeus (Macedonia, 1986; Macedonia, 1993), are very well 1988; Sha et al., 2008) (IUCN Red list 2008). A fi rst and studied. However, the calls of the species in the species-rich brief qualitative description of the vocalizations can be clade of leaf monkeys (>32 species) remain poorly known, found in Kawabe & Mano (1972). The study discusses with detailed studies only available for four species, i.e., several vocalizations as being of high frequency. However, Trachypithecus johnii (Fischer), Semnopithecus entellus a quantitative description of the vocalizations has yet to be (Dufresne), Presbytis thomasi (Collett), and Rhinopithecus published and proboscis monkeys have thus been excluded roxellana (Milne-Edwards) (Gurmaya, 1986; Hohmann, from comparative studies of primate vocalizations. If 1988; Tenaza et al., 1988; Wich et al., 2008). In particular, confi rmed, high frequency calls would be intriguing because the odd-nosed leaf monkeys (Sterner et al., 2006) comprising among primates Hauser (1993b) documented a negative nine species in four genera (IUCN Red List 2008) have been correlation between primate body mass and the frequency largely ignored (except for Rhinopithecus roxellana: Tenaza of vocalizations. Nasalis larvatus has a body mass of over et al., 1988; Clarke, 1990). Often rare and shy, studying 20 kg and is the largest species of colobine monkeys (Fig. these leaf monkeys in the fi eld presents major challenges. 1; Murai, 2006; Smith & Cheverud, 2002). Here we describe Since their ecology makes it likely that they rely heavily the vocalizations of captive proboscis monkeys from the on vocal communication, studying their vocalizations Singapore Zoological Gardens. Specifi cally, we analyzed 319 Srivathsan & Meier: Vocalizations of Nasalis larvatus the tonal calls, i.e., the vocalizations in which harmonics are fundamental frequency values, we used PRAAT (Boersma, integral multiples of the fundamental frequency. 2001) which includes two algorithms (auto-correlation, and cross correlation) for drawing the contours for the fundamental frequency of a call in a spectrogram. Both methods were METHODS tested for a subset of the data, and it was found that the autocorrelation algorithm detected the pitch of the calls better, Recording of Vocalizations. – The study subjects were and thus it was used henceforth. The parameters for the a family of two adult proboscis monkeys (one male and spectrogram analysis were as follows: window length 0.005 one female) and two juveniles kept in a single cage in the s, dynamic range 40 dB and frequency range 0–20kHz. The Singapore Zoological Gardens. The vocalizations were minimum, maximum and mean fundamental frequencies for recorded during a two-month period (Jan – Feb.2010) each call were recorded. For noisy and complex vocalizations between 1000–1700 hours using SongMeter 2.0 (Wildlife with strong tonal components, we characterized the lowest Acoustics Inc., Concord, MA) with 16-bit precision and a dominant frequency band to determine the pitch. sampling rate of 48 kHz. The recording device was placed 2 m away from the cage. In order to distinguish the calls Jackknifi ng. – In our study we obtained 101 calls. Since most studies of primate vocalizations are based on a larger sample of the proboscis monkeys from ambient noise, the timing size [1000–10,000 calls, eg. Lagothrix lagotricha (Humboldt) of each call was noted. This enabled the identifi cation of (Casamitjana, 2002), Rhinopithecus roxellana (Clarke, 1990), the calls upon playback to the observer. This was done as Cercopithecus campbelli Waterhouse (Ouattara et al., 2009)], the location of the captive setting was such that the calls we were concerned whether our study yielded suffi ciently could be mixed with the calls from the neighboring primate robust data for obtaining reliable results. We thus performed cages, or bird calls. We noted the behavioral context of the a jackknife analysis (100 replicates). The premise behind such vocalizations of interest. The aspects noted included whether an approach is that, should a subset of the data set yield same the individual responded to a threat/surprise (“alarm”), results as the whole data set, no additional sampling would whether the individual was engaged in play (“aggression”), change the overall results. For each jackknife replicate we whether the call was food associated or any other behavior randomly removed e-1 (37%) of the sample as recommended that may have been associated with the call. Other calls, for jackknife analyses in phylogenetics (Farris et al., 1996; such as low frequency calls, and associated behaviors were Table 2). For each replicate we recorded the means, minima excluded from this study. and maxima of vocalization frequencies. We used a t-test to test for comparing the distribution of each jackknife replicate Analysis of the recordings. – SongScope 3.0 (Wildlife with the distribution of original dataset. Acoustics Inc., Concord, MA) was used to analyse and identify the vocalizations. In order to measure the RESULTS AND DISCUSSION Twenty-eight hours of recording generated 211 vocalizations with complete spectrographic structures including 101 high pitch calls. These calls were made by the adults and juveniles. However, our conclusions pertain mainly to adult females and juveniles because they were the source of most calls. The high-pitched calls were similar to the “shriek” and certain other nasal vocalizations described by Kawabe & Mano (1972). They are associated with two types of situations encountered in captivity, i.e., “alarm” or “aggression”. We considered the calls as “alarm calls” when they occurred in response to the sudden appearance of a human or a loud noise. This noise could originate from zoo visitors or animals (e.g., clanking sound from a nearby primate cage). These calls were also heard when the monkeys were excited during play and/or when an individual saw another primate or human in his vicinity and responded by approaching the cage boundary (“aggressive behaviour”). The mean pitch of these vocalizations range from 2.3 kHz to 6.1 kHz. The maximum frequency of the vocalizations varied from 2.7 kHz to 6.8 kHz (Table 1; see spectrogram in Fig. 2). These results are stable as revealed by the jackknife resampling regardless of whether all replicates were used or the extremes were removed (ca 6.5–6.7 kHz, see Table 2). None of the 100 jackknife replicates yielded a distribution that was signifi cantly different from the original distribution based Fig. 1. A male proboscis monkey in the Singapore Zoological on all calls (p>0.05). Gardens. 320 THE RAFFLES BULLETIN OF ZOOLOGY 2011 Table 1. Characteristics of the high-pitched vocalizations of Nasalis larvatus. Feature Low frequency limit High frequency limit Mean frequency of spectrogram Average (Hz) 3307 4708 4020 Minimum (Hz) 1402 2681 2271 Maximum (Hz) 5666 6835 6145 Table 2. Jackknife analysis of the vocalizations: Low and high frequency limits of high-pitched vocalizations based on all replicates (a) and after removal of replicates with the 5% (b) or 10% (c) most extreme readings. Characteristic of vocalization (a) all replicates (b) 95% cut-off (c) 90% cut-off Low frequency limit (Hz)* 1678 1668 1555 High frequency limit (Hz)* 6495 6743 6743 Approximately 10% of the 101 vocalizations had a maxima of the fundamental frequency of vocalization and body size >6 kHz and 35% had maxima >5 kHz (Fig. 3). These results across nonhuman primate species found a negative correlation support the previous, qualitative observation by Kawabe and between high frequency limit and body size (Hauser, 1993b). Mano (1972) that the proboscis monkeys can vocalize at high Based on this correlation, a primate species with the body frequencies. The high frequency limit of the vocalizations of mass of N. larvatus (15.1 kg: Milton and May, 1976) would proboscis monkeys is particularly interesting given its body be expected to have a maximum frequency of vocalization size. A comparative study analysing the relationship between of ca. 2 kHz. Even when the female body mass (10 kg) is considered, the expected maximum would be 2.3 kHz. Note, however, Hauser (1993b) used only adult vocalizations for his correlation study while we include juvenile vocalizations in our analysis because we lack information on which calls came from which individual. Fortunately, this does not affect our conclusion because all calls (including those of adults) were >2 kHz and most >3.5 kHz (Fig. 3). Future studies should nevertheless try to analyse the calls of adults and juveniles separately and focus on the high-pitched calls of the adult males. The finding of unusually high-pitched vocalizations raises the question of their adaptive signifi cance. We here consider two possible functions; i.e., predator avoidance and communication in a dense forest environment. Predator avoidance is a potential explanation especially if the frequencies can be shown to lie beyond the frequency range of the best hearing of relevant predators (Snowdon & Hodun, 1981). One of the main predators of proboscis monkeys is crocodiles, Tomistoma schlegelii (Müller) (Galdikas, 1985). Proboscis monkeys are known to sleep near river banks and Fig. 2. A sample spectrogram of a high frequency vocalization of the proboscis monkey, Nasalis larvatus, that shows the harmonic structure of these calls. The fundamental frequency of this call ranges Fig. 3. Cumulative curve of mean frequencies of calls in the study. from 3.4–5.4 kHz and the mean frequency is 4.9 kHz. Graph omits the one call that has a frequency <2.7 kHz. 321 Srivathsan & Meier: Vocalizations of Nasalis larvatus occasionally venture into the river (Galdikas, 1985); thus, ACKNOWLEDGEMENTS they are exposed to predation by crocodiles. One defense mechanism are group movements across the river (Yeager, We would like to thank the Singapore Zoological Gardens, 1991) but another may be the use of high-pitch vocalizations Wildlife Reserves Singapore, for granting access to primates which allows the monkeys to warn kin of danger without and facilitating the recordings. We would also like to alerting crocodiles whose maximum auditory sensitivity lies express our gratitude to John Sha for facilitating the study between 0.3–2 kHz (Wever, 1971). Of course, this comes at and providing the photograph of the proboscis monkey. We the expense of not being able to use these vocalizations to acknowledge funding from a grant by the Ah Meng Memorial harass the predator which, however, would not be adaptive Conservation Fund (N-154-001-036-001). while being within striking range of crocodiles. Another major predator of proboscis monkeys is the predominantly nocturnal clouded leopard Neofelis diardi (Cuvier) (Matsuda REFERENCES et al., 2008) and behavioural adaptation such as sleeping near river banks has been discussed in the literature as a Bennet, E. L. & A. C. Sebastian, 1988. Social organization and predator avoidance mechanism (e.g. Matsuda et al., 2008). ecology of proboscis monkeys (Nasalis larvatus) in mixed Unfortunately, no information is currently available for the coastal forest in Sarawak. International Journal of Primatology, 9: 233–255. hearing limits of this species. We are thus unable to determine whether the vocalizations of the proboscis monkey are outside Biben, M. & D. Symmes, 1986. Play vocalizations of squirrel of the frequency range of this predator. monkeys (Saimiri sciureus). Folia Primatologica, 46: 173– 182. A second factor that may explain the high frequency Boersma, O., 2001. Praat, a system for doing phonetics by computer. vocalizations in proboscis monkeys is their habitat, as acoustic Glot International, 5: 341–345. communication can be challenging in forest environments Boinski, S. & C. L. Mitchell, 1997. Chuck vocalizations of wild (Waser & Waser, 1977). High frequency vocalizations have female squirrel monkeys (Saimiri sciureus) contain information the advantage of being distinct from ambient environmental on caller identity and foraging activity. International Journal sounds (Snowdon & Hodun, 1981) and this can be particularly of Primatology, 18: 975–993. useful in dense vegetation, where high frequency tonal Brown, M. M., N. A. Kreiter, J. T. Maple & J. M. Sinnott, 1992. vocalizations retain their discernable nature. Because high Silhouettes elicit alarm calls from captive vervet monkeys frequency sounds attenuate at much shorter distances than (Cercopithecus aethiops). Journal of Comparative Psychology, 106: 350–359. low frequency sounds, animals which require communication over long distances generally use low-frequency vocalizations Casamitjana, J., 2002. The vocal repertoire of the Woolly Monkey (Fletcher, 2010). But high-pitched calls can be used as alarm Lagothrix lagothricha. Bioacoustics, 13: 1–19. calls because the distances between monkeys are generally Clarke, A. S., 1990. Vocal communication in captive golden small. For long-distance communication, other types of calls monkeys (Rhinopithecus roxellanae). Primates, 31: 601–606. may be used as suggested by certain low frequency calls by de la Torre, S. & C. T. Snowdon, 2009. Dialects in pygmy the males (see Fig. S1). marmosets? Population variation in call structure. American Journal of Primatology, 71: 333–342. Now that the high frequency nature of captive proboscis Farris, J. S., V. A. Albert, M. Kallersjo, D. Lipscomb & A. G. monkey vocalization has been confirmed in a technical Kluge, 1996. Parsimony jackknifi ng outperforms neighbor- analysis, it would be important to conduct further field joining. Cladistics, 12: 99–124. and zoo research that could reveal the precise use of the Fletcher, N.H., 2010. A frequency scaling rule in mammalian vocalizations. For example, in order to better characterise vocalization. In: Muller, C. P. & B. L. Jacobs (eds.), Handbook of the vocalizations and understand the functions of these high Behavioral Neuroscience, (Vol 19). Academic Press, Burlington. frequency vocalizations, playback studies should carried out. Pp. 51–56. It is also important to note that calls in captive conditions Galdikas, B. M. F., 1985. Crocodile predation on a proboscis may not be fully representative of calls used in wild. Future monkey in Borneo. Primates, 26: 495–496. studies should also use a larger number of individuals in Gurmaya, K. J., 1986. Ecology and behavior of Presbytis thomasi order to address age- and sex-specifi c differences and their in Northern Sumatra. Primates, 27: 151–172. relationships to size. Hauser, M. D., 1993a. Food-associated calls in rhesus macaques (Macaca mulatta): I. Socioecological factors. Behavioral Ecology, 4: 194–205. Hauser, M. D., 1993b. The evolution of nonhuman primate vocalizations: effects of phylogeny, body weight, and social context. The American Naturalist, 142: 528–542. Hohmann, G., 1988. Analysis of loud calls provides new evidence for hybridization between two Asian leaf monkeys (Presbytis johnii, Presbytis entellus). Folia Primatologica, 51: 209–213. Inoue, M., 1988., Age gradations in vocalization and body weight Fig. S1. Spectrogram of Call 2: Low frequency vocalization in Japanese monkeys (Macaca fuscata). Folia Primatologica produced by adult male. 51: 76–86. 322 THE RAFFLES BULLETIN OF ZOOLOGY 2011 IUCN, 2008. IUCN Red List of Threatened Species. http://www. Sha, J. C. M., H. Bernard & S. Nathan, 2008. Status and conservation iucnredlist.org/. (Accessed 1 Sep.2010) of proboscis monkeys (Nasalis larvatus) in Sabah, East Malaysia. Primate Conservation, 23: 107–120. Kawabe, M. & T. Mano, 1972. Ecology and behavior of the wild proboscis monkey, Nasalis larvatus (Wurmb), in Sabah, Smith, R. J. & J. M. Cheverud, 2002. Scaling of sexual dimorphism in Malaysia. Primates, 13: 213–227. body mass: A phylogenetic analysis of Rensch’s rule in primates. International Journal of Primatology, 23: 1095–1135. Macedonia, J. M., 1986. Individuality in a contact call of the ringtailed lemur (Lemur catta). American Journal of Snowdon, C. T. & A. Hodun, 1981. Acoustic adaptations in pygmy Primatology, 11: 163–179. marmoset contact calls: Locational cues vary with distances between conspecifi cs. Behavioral Ecology and Sociobiology, Macedonia, J. M., 1993. The vocal repertoire of the ringtailed lemur 9: 295–300. (Lemur catta). Folia Primatologica, 61: 186–217. Sterner, K. N., R. L. Raaum, Y. P. Zhang, C. B. Stewart & T. Masataka, N., 1983. Psycholinguistic analyses of alarm calls of R. Disotell, 2006. Mitochondrial data support an odd-nosed Japanese monkeys (Macaca fuscata fuscata). American Journal colobine clade. Molecular Phylogenetics and Evolution, 40: of Primatology, 5: 111–125. 1–7. Matsuda, I., A. Tuuga & S. Higashi, 2008. Clouded leopard (Neofelis Symmes, D. & P. Goedeking, 1988. Nocturnal vocalizations by diardi) predation on proboscis monkey (Nasalis larvatus) in squirrel monkeys (Saimiri sciureus). Folia Primatologica, 51: Sabah, Malaysia. Primates, 49: 227–231. 143–148. Milton, K. & M. L. May, 1976. Body weight, diet and home range Tenaza, R. R., H. M. Fitch & D. G. Lindburg, 1988. Vocal behavior area in primates. Nature, 259: 459–462. of captive Sichuan golden monkeys (Rhinopithecus r. roxellana). Murai, T., 2006. Mating behaviors of the proboscis monkey (Nasalis American Journal of Primatology, 14: 1–9. larvatus). American Journal of Primatology, 68: 832–837. Waser, P. M. & M. S. Waser, 1977. Experimental studies of primate Ouattara, K., A. Lemasson & K. Zuberbuhler, 2009. Campbell’s vocalization: Specializations for long-distance propagation. monkeys use affi xation to alter call meaning. PLoS ONE, 4: Zeitschrift für Tierpsychologie, 43: 239–263. e7808. Wever, E.G., 1971. Hearing in the crocodilia. Proceedings of the Owren, M. J. & R. H. Bernacki, 1988. The acoustic features of National Academy of Sciences USA, 68: 1498–1500. vervet monkey alarm calls (Cercopithecus aethiops). Journal Wich, S. A., A. M. Schel & H. de Vries, 2008. Geographic variation of the Acoustical Society of America, 5: 1927–1935. in Thomas langur (Presbytis thomasi) loud calls. American Pola, Y. V. & C. T. Snowdon, 1975. The vocalizations of pygmy Journal of Primatology, 70: 566–574. marmosets (Cebuella pygmaea). Animal Behaviour, 23: Yeager, C.P., 1991. Possible antipredator behavior associated 826–842. with river crossings by proboscis monkeys (Nasalis larvatus). Rowell, T. E., 1962. Agonistic noises of the rhesus monkey (Macaca American Journal of Primatology, 24: 61–66. mulatta). Symposia of the Zoological Society of London, 8: 91–96. 323