Correspondence to: Revised
P.O. Box 59-157
Fax: 07-5253623
Email: agoram@mail.nsysu.edu.tw
1 Department of Biological Sciences,
2 Tajen Institute of Technology, Yanpu, Pingtung 907,
* To whom correspondence and reprint requests should be addressed. Tel: 886-75252000 Ext 3623. Fax: 886-75253623. Email: agoram@mail.nsysu.edu.tw
Formosan macaques are endemic to the
This paper for the first time describes the physical structures of vocal patterns obtained from wild Formosan macaques. A total of 25 classes of vocal patterns have been distinguished and analyzed by sonograms. The frequency of use in each vocal pattern is illustrated. Data on social behavior were also collected to highlight the context in which each vocalization was given. Finally, we have compared the vocal repertoire of Formosan macaques with that of other species in the genus Macaca.
A long-term field study to investigate the demographic parameters and social behavior of Formosan macaques at
Behavioral data and sound recordings were gathered simultaneously from all study groups for a total of 375 hrs, using scan sampling (55 hrs), all occurrences sampling and focal animal (320 hrs) sampling methods (Lehner 1996). Sound recordings were mainly obtained (332 hrs, 88.5%) from the habituated group C. Data collection commenced when the appearance of the observer no longer caused any obvious change in the group activity. In addition to the recordings of vocalizations, age/sex of the vocalizer and behavioral data on social interactions were recorded. The distance between the observer and the animal varied from 1 to10 m.
Sound recordings were made using a Sony professional cassette recorder (TC-D5 PRO II) and a directional microphone (Sennheiser MKH 460). Spectrographic analyses were carried out on a Macintosh computer (G3/350KHz) using the Canary software package (version 1.2.4). Sounds were digitized with 16 bit accuracy from audio tapes and sampling at a rate of 44.1 kHz, using a 1024-pt fast fourier transform (FFT, with Hamming window), this setting provided a frequency resolution of 43 Hz, and a temporal resolution of 3 ms. The classification of the repertoire was based on measurement of the physical parameters time, frequency and energy distribution following Hohmann (1989), and Moody et al. (1990). In addition, the age and sex of the vocalizing individuals and the context of the utterance were also considered. All study animals mentioned in this paper are identifiable individually due to the availability of long-term demographic data and genealogical information. We follow the age classification previously described by Hsu and Lin (2001).
We followed the terminology of Struhsaker (1967) to describe the acoustic properties of the basic patterns, which contain the following elements: unit (tonal, non-tonal, compound, mixed), phrase and bout. Unit: The unit is the basic element of a monkey sound or call, and is represented as a continuous tracing along the temporal (horizontal) axis of the sonogram. Phrase: The phrase is a group of units that is separated from other similar groups by a time interval greater than any time interval separating the units within a phrase. Bout: A bout is a grouping of one or more phrases separated from other similar groupings by a time interval greater than that separating any of phrases within a bout (Struhsaker 1967). In addition, the classification of the repertoire was based on measurement of the acoustic parameters, context of emission and comparison of previous references. According to the acoustic parameters, context of emission and comparison of previous references were used for the classification of the vocal repertoire in Formosan macaques (Table 1). In addition, the age and sex of the vocalizing individuals and the context of the utterance were also considered.
Previous qualitative studies of vocal repertoire (Green 1975, Hohmann and Herzog 1985, Hohmann 1989) indicated that from the perspective of both production and perception, fundamental frequency was typically the most salient acoustic feature (Moody et al. 1990). Therefore a suite of acoustic features was selected for analysis. The spectral measurements were: Maximal fundamental frequency (MXF), Minimal fundamental frequency (MNF), Highest frequency (HF), Lowest frequency (LF), Peak frequency (PF) and Duration of each unit (Fig. 1). Peak frequency was the frequency at which the highest amplitude in a unit, and some physical parameters below were obtained by extrapolation, including median fundamental frequency [FF = (MXF+MNF)/2], modulation of fundamental frequency (MFF = MXF-MNF) and total range of frequency (RF = HF-LF).
All statistical analyses were conducted with Statistical Analysis System software (SAS Institute, 2000). All mean values are presented as ± 1 standard deviation. Duncan’s multiple range tests were used to test the similarity among four vocal types of coos and among alarm call, grunt and threat rattles in five acoustics variables (FF, MFF, RF, PF and duration) followed ANOVA (Analysis of variance) for each parameter. In addition, the significant differences of some acoustic parameters between different vocal types were tested through Wilcoxon rank tests. Canonical discriminant analysis was used to measure the squared distances between four types of coos. Chi-square test was used to test the difference of the occurrence of male and female copulation calls.
RESULTS
A total of 25 vocal patterns with 35 repertoires and behavioral contexts of vocalizations has been distinguished (Table 1) and analyzed by sonograms from 7997 units recorded from focal animal sampling (320.4 hrs). Basic features including structure, range of unit, FF, MFF, RF, PF and duration of 35 vocal repertoires are described in Table 2. Among all vocal repertoires of Formosan macaques, coo call was the most common vocalization (relative frequency 18.59 %), followed by hack (17.78 %) and grunt (11.85 %).
Coo call
These were the most commonly uttered calls in the vocal repertoire. According to the structural features and behavioral context of the utterance, at least four types were recognized.
(i) Contact coo: Adult females and both sexes of sub-adults, juveniles and infants emitted contact coos (Fig. 2A-D) in three different contexts. (a) When infants were exploring alone, they emitted this call apparently in order to maintain contacts with mothers. (b) They were associated with initiation of group procession─ usually one animal initiated the call and some peers responded with contact coo calls, which resulted in troop movement. (c) Expectation of food: When monkeys heard or saw a familiar person who used to provide fruits or bread to them, troop members jointed together and emitted this type of calls with excitement, followed by grunt and/or food yells.
(ii) Isolation coo: Only infants and juveniles uttered this type of calls when they lost physical contact with their kin. The youngest one recorded was a five-week-old infant. Infants and juveniles emitted this call when they were ignored or weaned by their mothers. It was usually accompanied by hacks, clucks and weeping.
(iii) Long distance coo: These calls are tonal or compound units characterized by high, rise-shape, well-modulated fundamental frequency, and high sound pressure (Fig. 2GH). All age/sex classes with the exception of adults uttered this call. They were uttered when immature animals lost visual contact with their mothers or the main group. It usually accompanied with visual orientation and locomotive activity.
(iv) Cohesion coo: There calls are mixed units characterized by low, extremely close harmonic frequency bands superimposed by a noisy portion (Fig. 2IJ). All age/sex classes with the exception of infants uttered this call, mainly during arousal states such as the appearance of expected feeder, inter-group agonistic encounter, helping isolated infants to search for mothers, and when lost contact with group members. During group dispersal, dominant females emitted this call apparently to increase group cohesion.
According to the Univariate test statistics, the acoustic features of these four types of coos were significant different (p<0.001) in FF (F3, 578= 198.48), MFF (F3, 578= 64.67), RF (F3, 578= 165.32), PF (F3, 578 = 84.87) and duration (F3, 578= 87.49). The average PF of contact coo was 0.44 kHz (± 0.17), significantly lower than that of the other three types (0.72-0.80 kHz, Table 2). The modulation of fundamental frequency of long distance coo (0.33 ± 0.13 kHz) was significantly higher than other types of coo calls (
Canonical discriminant analysis of the remaining principal components indicated 3 functions differentiating four types of coos. The first 2 canonical functions accounted for 95.5% of eigenvalue cumulative proportion. The plot of the first 2 canonical functions indicated that although all types of coos presented statistically significant differences, this could not be ascribed to a simple pattern of 1 function (plots of the remaining functions followed the same general pattern in Fig. 3). The squared Mahalanobis distance betweens any two types varied from 11.82 (isolation coo and cohesion coo) to 3.88 (contact coos and isolation coos); all distances were significantly different (p<0.001). The squared distances between cohesion coo and the other two types (isolation coo and long distance coo) were relatively far away, respectively 11.82 and 7.71.
Greeting
All age/sex classes uttered this type of call. Three types of greeting calls were distinguished. The Atonal greeting was given by females who approaching other females with infants (Fig. 4A). The Tonal greeting was uttered by all age/sex classes except infants (Fig. 4B). Subordinate individuals emitted tonal greeting when approaching dominant peers for interactions such as playing, moving, and foraging. The Girney greeting was a variable, but predominantly tonal sound of low frequency (less than 2 kHz) that was superimposed by short, atonal segments resulting from tongue or lip movement (Fig. 4C). This call was uttered when monkeys approached a dominant individual. It was followed by submissive behaviors such as huddling, embracing, grooming, mounting, eyebrow flashing or lip smacking.
Female copulation call
This call is a rapid succession of phrase, from 4 to 69 units, (24.5 ± 17.5 units, n=23), emitted exclusively from adult and sub-adult females during copulation. Each unit was a narrow noise beam (the structure of non-tonal vertical striation) in which energy emphasized at the baseline as well as at 3.47 ± 0.58 kHz (n=556, Fig. 5A). It's typical of temporal organization of a phrase, which started quietly and separated by greater intervals and then become louder rhythmical pulsed in rapid succession.
Not all females emitted this type of call during copulation. During 1999, 73 copulations were observed while this call occurred in 51 cases. The emission rate was close to 70 %. They often emitted this call in the last part of the copulation and usually lasted longer than the physical contact with the male companions.
Male copulation call
The male copulation calls with tonal units were seldom uttered, but unique to adult males during copulation. The whistle-like sounds uttered in irregular trains or well-structured phrases containing up to 4 pulses. The units consisted of 4~7 discrete frequency bands (from 1.5 to 19.5 kHz) with low degrees of modulation (Fig. 5B). In contrast to females, males emitted copulation calls significantly less frequent (X2 = 54.36, df = 1, p<0.001); the emission rate was 6.8 % (n=73). During mating, the head of the male was forward with its face towards the female. At this stage, the female often turned upward and/or pulled the shoulder of the male by exposing her tooth. The male's face occasionally was too close to the female that would lead its mouth touched her face. During emitting, its lips were parted but the jaws were always closed.
Mounting grunt
These grunts were irregular trains of short and noisy pulses, which with an addition frequency band at the approximately 1.2 kHz and/or 3 kHz (Fig. 6). Only adult males produced these calls in connection with affinitive mounting of another adult/sub-adult male. Typically, after a dominant male revealed affiliative expression (i.e. gave grunt, lip smacking) toward a subordinate male, it would induce the production of atonal greeting calls from the subordinate and the occurrence of an affiliative mounting of the male peer. At this time, a train of mounting grunts was uttered.
Alarm call
A single unit of alarm call is composed of 2~3 narrow non-tonal segments. The segmental structure contained discrete, irregularly modulated frequency bands and was preceded by a comparatively short non-tonal element (Fig. 7). The alarm call was produced by all age classes─ the youngest animal seen to produce it was a 3-month old male infant. This call is highly intensive, consisting of plosive utterances with a wide noise-like frequency range.
The alarm call was given for various reasons such as: (1) while being chased or attacked by stray dogs, (2) when solders in military uniform approached (parts of the study site is still a restricted military base where solders occasionally disturb monkeys), (3) when visitors approached with walking sticks, slingshots, and stones (people occasionally harass monkeys by chasing, throwing stones or using sling shots), (4) when people carried large objects such as bicycles, construction materials, etc., and (5) when monkeys sighted raptors in the area. As soon as the alarm call was produced, the group reacted by fleeing or freezing all activities. They scanned the area and the direction of the alarm call to locate possible causes of the threat. Alarm calls were also produced during inter-group conflicts, especially when intruder males were seen near social groups.
Grunt
All age/sex classes commonly produced these calls. Each unit was composed of 2~4 non-tonal or mixed segments with energy concentrated at the base (<600 Hz). Due to different contexts, the modulation of frequency range in each segments was from baseline of the sonogram to 2~16 kHz. These calls were used in four contexts. (1) When groups started to move, monkeys exchanged these calls while visually scanning the area and group members (Fig. 8A). (2) This call was given while approaching or being approached by an unrelated group member. Also, high-ranking individuals emitted this call to gain physical contact, mainly grooming, from low-ranking peers─ low-ranking individuals responded with grunt or greeting calls and rushed to initiate physical contact. ((3) This call was given during female-infant (non-mother) interactions. When a female attempted to reach an infant in order to groom, hold and care for, they produced this type of calls toward the infant who in close contact with its mother (Fig. 8BCE). (4) This call was given during artificial feeding.
Threat rattle
These rattles consist of temporally segmented of 3-6 non-tonal or mixed units with a longer-period amplitude modulation, which appears as burst or pulse (Fig. 9). The frequency range of each segment was from baseline of sonogram to 8~16 kHz and with a duration of 0.4~0.45s (Fig. 9). Moreover, some contained discrete, irregularly modulated frequency bands energy emphasis at approximately 2.82 kHz (± 0.7, n=21) and showed little base-line energy. All age/sex classes except infants less than 7 months old uttered this type of call during intra- and inter-group agonistic interactions as well as during inter-species interaction involving people and birds. The calls were accompanied by facial expressions of open mouth threat. With increasing intensity, the spectrographic changed in loss of microscopic pulsation and the proportion of energy at the baseline became greater. When the intensity and probability of attack increased, the fine segmentations diminish and the energy at baseline would be predominant, forming growl.
In contrast to grunt, threat rattles are characterized by their longer duration, wider frequency range and higher PF. There are statistically significant differences in FF, MFF, PF and duration among those of alarm calls, grunt and threat rattles (
Growl
Growls consisted of non-tonal or mixed units uttered predominantly by high-ranking females and males during agonistic inter- and intra- group encounter. In some cases, the calls were uttered in rapid succession and forming phrase. The characteristic of this call was predominant baseline energy. The modulation of frequency range each unit was from 1~16 kHz and duration with 0.1~0.2s (Fig. 10). This vocalization was never recorded in infants less than 7 months old. Whereas threat rattles were given during mild and moderate threats and within stand-offs, growls accompanied the intense threats or attack. Dominant individual emitted growls while lunging or chasing. The opponents involved were chased and mauled when caught. Even lasted to an attack finished, dominant individuals were seen staring and producing growls and/or rattle towards the subordinate individuals on occasions.
Vibrato growl
The vibrato growl is composed of several close, narrow frequency bands superimposed by a noisy overlay (frequency range from 1~16 kHz) with duration of 0.04~0.52 s (Fig. 11). This call was uttered by all age and sex classes during agonistic intra- /inter-group and interspecies encounters. When tourists harassed the monkeys or when other monkeys (of the same or a neighboring group) harassed them, these calls were produced accompanied by open mouth threat. Growls and/or threat rattles preceded these calls, and the spectrogram revealed transitional forms betweens these three basic patterns. Therefore, the vibrato growls were used in the beginning when an animal that produced the call approached the target (human or monkey) to attack. The duration of bout of the calls depended on the reaction of the receivers.
Roar
Units of roar were similar to growls, but given in a series with sequentially decreasing intensity and intervening sounds of inspiration (Fig. 12). Adult/sub-adult males and adult females roar in agonistic threats with intra-group interaction or with inter-group encounter. Roaring accompanied the tree-shaking display was typical to an adult male who was engaged in fights, disturbed or interrupted during sexual solicitations.
Bark
Barks are intense, plosive and non-tonal units with prolonged energy emphasis at relatively low of 1.18 kHz (± 0.37, n = 20, Fig. 13). Adult and sub-adults of both sexes barked during inter-species aggressive conflicts with dogs or birds and also during inter-/intra-group agonistic encounters. They were uttered separately or in short phrases by adult/sub-adult males and females predominantly. The bark was uttered during inter-species aggressive conflict (i.e. dogs, birds) and during inter-/intra-group agonistic encounters. On five occasions when monkeys saw Himalayan tree pie (Dendrocitta formosae) near the resting or feeding area, adult females emitted barks and growls toward the birds, as a result, the bird flew away. However, the average peak frequency of bark was significantly higher than that of growl (Wilcoxon rank test, p<0.005).
Squeal
Squeals are tonal, compound or mixed units with well-modulated, harmonics frequency bands (Fig. 14). The unit of squeal was composed of well-modulated, harmonics frequency bands with a range of 4 ~ 16 kHz and superimposed by /or attached to noise beams during long duration range from 0.2 ~ 1.1s (Fig. 14). All age/sex classes produced this call during intra-/inter-group and inter-species agonistic interactions, i.e. in response to the threat from high-ranking peers, monkeys from other groups and dogs. In addition, infants emitted this call with violent spasmodic motions and then ran to its mother for protection. Moreover, infants and young juveniles gave this call when they had lost contact with their mothers and/or main group. It frequently combined with several coo calls and even transited to scream.
Noisy and undulated screams
These calls were recorded from all sex and age class individuals in response to threat, attack or maltreatment by dominant or older individuals. The energy distribution and duration of Noise screams varied considerably, ranging from narrow beams with plosive onsets to prolonged segments with gradually increasing energy (Fig. 15AB). The main energy was often concentrated on medium frequency ranges (4.04 ± 1.89 kHz, n=129).
The units of Undulated screams were characterized by discrete, irregular and slightly modulated harmonics (the modulation of harmonics, mean = 0.24 kHz). As in noise screams, duration and energy distribution varied to large degree (Fig. 15C & D). In a number of samples, the tonal arrangement was marginal, the discrete harmonics were fading and frequency distributed more evenly. Usually, the sound lasted until the victim fled the area. The transition emerged gradually (frequently between onset and offset of a single unit), leading from the noise scream to more tonal undulated scream.
Tonal Scream
Tonal screams revealed the intergradations between squeals and tonal squeaks. In contrast to squeal, tonal scream consisted of 2~3 harmonics frequency bands with rarely a noisy portion (Fig. 16), of comparatively longer duration (from 0.3~0.6 s) than tonal squeaks. Sometimes, it was composed of 2 or 3 tonal squeaks that were combined with a short gap or a collapse. All age and sex groups except adult males uttered this call during inter-and intra-group agonistic encounters as well as when losing contact with the mother or the main group. However, these calls were also uttered when peers reunited, or sighted a feeder.
Squeak
Squeaks are plosive, tonal or compound sound of high pitch (above 4 kHz) and short duration (less 0.3 s), emitted during social arousal and in defensive withdrawal (Fig. 17). Two types of squeaks were distinguished─ Tonal squeak is composed of discrete, well-modulated frequency bands that follow a noise burst and Compound squeak with discrete, well-modulated frequency bands and superimposed/attached by a noisy portion. All age/sex classes emitted these two types of squeaks except no tonal squeak was recorded from adult males. During agonistic interactions, victims usually withdrew by producing a phrase of compound squeak and/or compound hack, and then might be followed by tonal squeaks. When subordinate monkeys were supplanted by dominant ones or by approaching people, they uttered the squeak, showing teeth and crouching on the ground or moving away. In mating seasons, adult females produced squeak call when dominant males directly approached for mating. Infants also uttered this call when they were separated from their mothers, or when infants approached their mothers who were preoccupied with their mates; the males either pushed or threatened the infants to keep them away.
The acoustic structures and utilization of these submissive calls (including noise and undulated screams, squeals, tonal screams and squeaks) were depended on the existence of physical contact during aggressive interactions. Noise and undulated screams were mostly given in physical aggressive interactions (68%, 36 out of 53 cases) whereas squeals, tonal screams and squeaks were mostly given in non-physical aggressive interactions (76%, 123 out of 163 cases).
Chuckle
The structure of chuckle was similar to that of food yells, but with lower pitch (Fig. 18). The units of chuckles consist of discrete, ascending modulated frequency bands that are superimposed by noisy segments of low energy. All age/sex classes uttered this call in connection with agonistic intra- and inter-group interactions. The behavioral sequences involving chuckles were characterized by the alternation of agonistic signals (vocal and visual) regularly, directed against the opponent and submissive signals with producing chuckles toward high-ranking group members, similar to the protect threat described by Kummer (1967). Chuckle often occurred in combination with noise and undulation scream.
Harmonic arch
Based on the acoustic structure and context of emission, three types could be identified. (1) Food yells (Fig. 19A- D), these calls consisted of harmonic frequency bands of arch-shape, which superimposed by noise portion. These calls only occurred during appearance of feeder. (2) Oui (Fig. 19E), this type showed such an abrupt upward shift of pitch that tracing is displaced vertically with an apparent discontinuity. In some case, it was superimposed by noise segment. These calls were heard in the appearance of feeders as well as the reunite of main group or the mother. In addition, it was also heard from a one-year-old female rushing to reach its mother in response to the contact coo of mother during group movement. (3) Harmonic arch (Fig. 19F), this call exhibited a rising in pitch. The additional energy is evidenced either as a burst of noise crossing the tonal band or sudden enrichment of overtone structure. This call occurred not only in reunite as well as the appearance of feeders but also during inter- and intra-group agnostic interaction.
Hack
Based on the structural features, two types of hacks were recognized. The Tonal hack consists of discrete frequency bands with a sharp negative slope (Fig. 20A-D). The fundamental frequency was characteristic at 0.7~1.3 kHz and with an extremely short duration (approximate 0.03 s). The Compound hack is a tonal element superimposed, attached or masked by noise. In some samples, a narrow noise beam preceded the compound unit (Fig. 20E-H).
Hacks were produced by all age/sex classes but mostly by infants. When mothers rejected them, infants produced this call to resume physical contact. Subordinate individuals produced these calls in response to rejection attempts to achieve physical contact, or maltreatment from dominant peers. In states of high arousal, spasmodic motions and specific facial expression (i.e. bared-teeth gecker face) accompanied. Tonal and compound hacks were mostly used in the contexts of young weaned, rejected, isolated and ignored by mothers (48.3%, 85 out of 176 cases), followed by intra-group physical aggressive interactions (12.5%) and non-physical aggressive interactions (11.4%).
Squawk
Squawk contains a predominance of harmonic frequency bands, which is a plosive, bark-like vocalization, but with a higher fundamental frequency range 0.8~1.5 kHz (Fig. 21). These calls were used during defense, i.e. when an animal was supplanted or withdrew from the approach of a dominant monkey or cringes at its approach. Also, these calls were emitted when macaques were ignored during food begging. This call was emitted predominantly by Infants, followed by juveniles, occasionally was heard from adult females.
Cluck
Clucks are short duration (0.04~0.06s) tonal or compound units composed of 1~4 slightly modulated harmonics frequency bands ranging from 0.5~5.8 kHz (Fig. 22). Clucks were predominantly emitted from Infants (85.7%), few from juveniles, when isolated from the group, ignored or rejected by their mothers. Clucks were also recorded during group movements since infants often had difficulties in keeping up the movement of the troop. Those mothers sat in proximity to infants but disregarded them until infants’ emitted gecker and scream, and then the mothers rushed to carry them. Cluck was emitted while food begging was ignored, and ceased after monkeys receiving food. Moreover, in some cases, it was combined with noise scream or coo calls.
Gecker
The units of gecker were of poor tonal quality, uttered in staccato-like phrase and exclusively by infants with one exception from juveniles. A unit of gecker call consisted of 2~4 modulated frequency bands with descending tendency (Fig. 23). In contrast to tonal hack, the lower fundamental frequency of gecker was characterized at 0.45~0.6 kHz. Geckers were predominantly emitted from Infants (90.1%), few from juveniles.
Infants usually uttered Gecker call while their mothers neglected or rejected their attempts to gain physical contact. It also responds to its mother’s sudden locomotive activities and didn’t hold infant and as a result, the infant was about to lose physical contact with its mother. At that time, the infant produced the gecker call and that resulted in its mother to hold it well immediately. Gecker usually stopped after the mother resumed physical contact. Moreover, infants also gave the call in response to rough pull from other females or other unspecific disturbances. In addition, this vocalization was accompanied with violent spasmodic motions and submission facial expression (e.g. bared teeth).
Whine
Each unit of whine is composed of several warbled harmonic frequency bands of rich modulation and occasionally superimposed by/or attached to noisy segments (Fig. 24). The duration range of a unit was from 0.12 ~0.7s and the maximal frequency band was approximate from 2~4 kHz (Fig. 24).
These calls were recorded from infants II when aroused and agitated, particularly after weaning, or from female juveniles who were lost or abandoned by their mother. It was given as an infant once again came into affiliative contact with its mother, i.e. at mother‘s chest.
Weep
The isolation coo was segmented by lapses of vocal output, forming a weeping call. Only infants and juveniles uttered this type of calls. The number of segments range from 2~5. Each unit was with a range of 0.6~1.2 kHz during 0.4~0.6s (Fig. 25A). The tonal units of weeping are emitted by infants (class I and II) to signal mission attention or physical contact as is typical of the stage of weaning. It often followed with isolation coos, hacks and/or squeaks in this period. Frequently, a cluck or tonal hack was combined with a weeping call.
A 6-month-old female infant wept and looked towards its mother when the mother kept sleeping and rejected its approaches for several times. In addition, she also produced gecker, squeaks and squeals together with the weeping. And that also induced another adult female to approach.
Babble
Babbles are soft, mild tonal units occasionally uttered, exclusively from infants of both sexes (Fig. 25B). This call was connected with the process of independent exploration. It was also directed towards mothers or other member when they were attempting to leave the infants. Or when infants saw an adult male appeared in the area.
DISCUSSION
Formosan macaques use 25 basic patterns of vocalizations in different context and they are comparable to the reported vocal patterns of other species in the genus Macaca (Green 1975, Gouzoules et al. 1984 1998, Hohmann and Herzog 1985, Dittus 1988, Hohmann 1989, Bauers and de Waal 1991, Hauser 1991, Palombit 1992, Fisher et al. 1998). Assessment of the vocalizations of Formosan macaques demonstrated a high symmetry early in life. Except for the sex-specific copulation calls and the mounting grunts, male and female Formosan macaques almost shared the entire repertoire.
Communication systems can be distinguished on the basis of whether signals within the repertoire are discrete or graded (Marler 1977, Hammerschmidt and Fischer 1998, Fischer and Hammerschmidt 2002). The discrete signals of Formosan macaques include calls such as coo, greeting, grunt, mounting grunt, male and female copulation call, weep, whine and babble, where alternations of the structural features were restricted and transitional forms between call types were rare or absent. The graded signals were aggressive calls (threat rattle, growl, bark, and alarm), submissive calls (screams, squeals, and squeak), and distress calls (squawk, hack, cluck and gecker). Green (1981) reported for lion-tailed macaques a vocal repertoire composed of discrete and intergraded patterns. Similarly, in bonnet macaques, Hohmann (1989) recorded vocalizations with variations in physical parameters on large scale that led to structural intergradations between different elements of repertoire.
Vocal repertories composed of intergraded signals may be expected to have evolved among those species with ready visual access to each other (Marler 1976 1977). The intergradations were supposed to be more frequent in species less dependent on audible signals for their social regulation (Fischer et al 1995, Fischer and Hammerschmidt 2002). However, most primates repertoires are neither completely graded nor consist completely of discrete signals, but characterized as mixture of graded and discrete signal structures (Green and Marler 1979). Therefore, structural criteria alone are insufficient for a comprehensive assessment and characterization of the vocal repertoire. Earlier studies of primate vocalizations showed minor differences in certain parameters (Green 1975, Snowdon and Pola 1978, Masataka 1983, Dittus 1984 1988) which may serve to transmit different type of information, either about the internal state of vocalizer or external factors such as food or predator (Hohmann 1985).
Looking at the correlations between acoustic signals and behaviour context, some of the calls, namely, copulation calls and alarm call were clearly related to particular interactions or specific external stimuli. Concerning the context of emission, the coo calls, greetings and Grunts, can be summarized as affinitive contact calls for reduction and/or maintenance of close proximity between group members. The growls, threat rattles and vibrato growls were uttered by dominant animals to intimidate sub-dominate group mates, members of other groups or other species.
In spite of the significant structural differences in the submissive calls, correlations between acoustic structure and social context were less clear. Playback studies of rhesus macaques revealed that distinctively structured screams reflected the level of arousal of the caller and were significantly associated with specific social factors (Gouzoules et al. 1984 1998). In the present study, the noisy screams and undulated screams were given during severe agonistic interactions leading to physical aggression, and squeal, tonal scream or squeak was given during agonistic interactions without physical aggression.
When infant and juvenile of Formosan macaques were in trouble, they produced screams, squeals, squeaks and alarm calls that attracted the attention of kin-related individuals. In many species of primates, mothers can identify their offspring using vocal cues and vocal recognition of infants and juveniles, especially their distress calls, is advantageous in primate groups when mothers can reduce the risk faced by them (Maestripieri, 1994, Cheney and Seyfarth 1990). In Formosan macaques, screams and squeals appeared to acquire or gain support from group mates to form agonistic alliance and distinctly structural submission calls may function in ally recruitment as the key element of social tactics rather than the emotional state of the vocalizer as reported among the rhesus macaques (Gouzoules et al. 1984).
On the other hand, the acoustic structure was obviously influenced by the context of emission. The fundamental frequency of the call used in submissive context and in aggressive context the fearful individual produced significantly higher fundamental frequency than the aggressive individual. The data support the motivation-structure rule suggested by Morton (1977) using published materials on birds and mammals.
Alarm calls were emitted mostly by sub-adult males and females. Interestingly, alpha males never gave the alarm call. Conspicuous usage differences in both sexes also occurred in relation to the appearance of a feeder. Adult females and juveniles usually produced tonal screams, food yells, coo calls, and grunts in response to feeder, whereas adult males exclusively gave grunts. On the other hand, adult males did not use tonal scream, tonal squeak and contact coo, but adult females gave these calls. Moreover, certain sounds associated with appeasement and/or submission (e.g. scream, squeal and squeak) were not produced by adult males.
Age-specific patterns of utterance of vocalizations are distinct in Formosan macaques. Some calls were produced initially by infants, but later vanished from individuals’ repertoires, these included cluck weep, gecker and babble. Although weep and gecker were heard from young juveniles whose mothers didn’t have dependent infants, this might be due to the juveniles’ relapse into infancy. Although they were given in the same agonistic context, the tonal vibrato growl used by infants differed from both the harsher, noisy call (i.e. growl) of the older individuals, as well as the threat rattle also given by infants. These differences might be associated with the ontogeny of physical development and vocal usage (Hauser 1993).
The most noticeable differences can be found in the coo calls.
This study indicates that Formosan macaques employ a complex vocal system composed of discrete sound classes as well as acoustic categories connected by intermediate gradations. Age specific differences in the vocal behavior are more pronounced than asymmetries between sexes. Certain vocalizations were produced or ceased by certain sex/age class and these differences might be associated with social organization, sexual selection, social suppression, and experience. The vocal repertoire of Formosan macaques revealed similarities to other species within the genus Macaca, especially to the closely related Japanese macaques and rhesus macaques suggesting the influence of phylogeny, habitat distribution and social organization in the evolution of vocal communication. The data presented in this paper on the vocal communication in Formosan macaques could be a valuable tool for further understanding not only the evolutionary function of vocal communication but also the complex social behavior in this less known primate species.
REFERENCES
Bauers KA. 1993. A functional analysis of staccato grunt vocalizations in the stumptailed macaque (Macaca arctoides). Ethology 94: 147-161.
Bauers KA, de Waal FBM. 1991. “Coo” vocalizations in stumptailed macaques: A controlled functional analysis. Behaviour 119: 143-159.
Ceugniet M, A Izumi. 2003. Vocal individual discrimination in Japanese monkeys. Primates 45: 119-128.
Cheney DL, RM Seyfarth. 1990. How monkeys see the world. Inside the mind of another species.
Cronie JE, Cann R, Sarich VM. 1980. Molecular evolution and systematic of the genus Macaca. In DG Lindburg, ed. The Macaques: Studies in Ecology, Behavior and Evolution.
Dittus WPG. 1984. Toque monkey food calls: semantic communication concern food distribution in the environment. Anim. Behav. 32: 470-477.
Dittus WPG. 1988. An analysis of toque macaque cohesion calls from an ecological perspective. In D Todt, P Goedeking, D Symmes, eds. Primate vocal communication.
Fischer J, K Hammerschmidt, D Todt. 1995. Factors affecting acoustic variation in
Fischer J, K Hammerschmidt. 2002. An overview of the
Fooden J. 1980. Classification and distribution of living macaques. In DG Lindburg, ed. The Macaques: Studies in Ecology, Behavior and Evolution.
Gouzoules S, H Gouzoules, P Marler. 1984. Rhesus monkey (Macaca mulatta) sreams: representational signaling in recruitment of agonistic aid. Anim. Behav. 32: 182-193.
Gouzoules H, S Gouzoules, M Tomaszycki. 1998. Agonistic screams and the classification of dominance relationships: are monkeys fuzzy logicians? Anim. Behav. 55: 51-60.
Green S, P Marler. 1979. The analysis of animal communication. In P Marler, JG Vandenbergh, eds. Handbook of behavioral neurobiology, social behavior and communication.
Green S. 1975. Variation in vocal pattern with social situation in the Japanese monkey (Macaca fuscata): a field study. In A Rosenblum, ed. Primate behavior: Developments in field and laboratory research,
Green S. 1981. Sex differences and age graduations in vocalizations of Japanese and liontailed monkeys (Macaca fuscata and Macaca silenus). Am. Zool. 21: 165-183.
Hammerschmidt K, J Fischer. 1998. The vocal repertoire of
Hauser MD, P Marler. 1993. Food-associated calls in rhesus macaques (Macaca mulatta). I. Socioecological factors influencing call production. Behav. Ecol. 4: 194-205.
Hauser MD. 1991. Sources of acoustic variation in rhesus macaque (Macaca mulatta) vocalizations. Ethology 89: 29-46.
Hauser MD. 1993. The evolution of nonhuman primate vocalizations: effects of phylogeny, body weight, and social context. Am. Nat. 142: 528-542.
Hohmann GM, MO Herzog. 1985. Vocal communication in lion-tailed macaques (Macaca silenus). Folia Primatol. 45: 148-178.
Hohmann GM. 1989. Vocal communication of wild bonnet macaques (Macaca radiata). Primates 30: 325-345.
Hohmann GM. 1991. Comparative analyses of age- and sex-specific patterns of vocal behaviour in four species of old world monkeys. Folia Primatol. 56: 133-156.
Hsu MJ, G Agoramoorthy, JF Lin. 2001.Birth seasonality and interbirth intervals in free-ranging Formosan macaques, Macaca cyclopis, at Mt. Longevity, Taiwan. Primates 42: 15-25.
Hsu MJ, JF Lin, LM Chen, G Agoramoorthy. 2002. Copulation calls in male Formosan macaques: Honest signals of male quality? Folia Primatol. 73: 220-223.
Hsu MJ, JF Lin. 2001. Troop size and Structure in Free-range Formosan macaques (Macaca cyclopis) at
Hsu MJ, J Moore, JF Lin, G Agoramoorthy. 2000. High incidence of supernumerary nipples and twins in wild Formosan macaques at
Kummer H. 1967. Tripartite relations in hamadryas baboons. In SA Altmann, ed. Social communication among primates.
Lehner, P. N. 1996. Handbook of Ethological Methods.
Marler P. 1976. Social organization, communication and graded signals: the chimpanzee and the gorilla. In PP Bateson, RA Hinde, eds. Growing points in ethology.
Marler P. 1977. The structure of animal communication sounds. In TH Bullock, ed. Recognition of complex acoustic signals.
Masataka N. 1983. Categorical responses to natural and synthesized alarm calls in Goeldi’s monkeys (Callimico goeldii). Primates 24: 40-51.
Melnik DT, KK Kidd. 1985. Genetic and evolutionary relationships among Asian macaques. Int. J. Primatol. 6: 123-160.
Mitani JC. 1985. Gibbon song duets and intergroup spacing. Behaviour 92: 59-96.
Moody DB, WC Stebbins, BJ May. 1990. Auditory perception of communication signals by Japanese monkeys. In WC Stebbins, MA Berkley, eds. Comparative perception: Complex perception, vol. 2, Complex Signals.
Morton ES. 1977. On the occurrence and significance of motivation- structural rules in some bird and mammal sounds. Am. Nat., 111: 855-869.
Muroyama Y, B. Thierry. 1998. Species differences of male loud calls and their perception in
Palombit RA. 1992. A preliminary study of vocal communication in wild long-tailed macaques (Macaca fascisularis). I. Vocal repertoire and call emission. Int. J. Primatol. 13: 143-182.
Rendall D, PS Rodman, RE Emond. 1996. Vocal recognition of individuals and kin in free-ranging rhesus monkeys. Animal Behavior 51: 1007-1015.
Rowell TA, RA Hinde. 1962. Vocal communication by rhesus monkey (Macaca mulatta). Proc. Zool. Soc. Lond. 128: 279-294.
SAS Institiute. 2000. SAS/ETS software: Changes and enhancements. Release 8.1.
Struhsaker TT. 1967. Auditory communication among vervet monkeys (Cercopithecus aethiops). In
van Schaik CP, MA van Noordwijk. 1985. Evolutionary effects of the absentce of felids on the social organization of the macaques of the
Vogel, C. 1973. Acoustical communication among free-ranging common Indian langurs (Presbytis entellus) in two different habitats of
Waser P. 1977. Individual recognition, intragroup cohesion, and intergroup spacing: evidence from sound playback to forest monkeys. Behaviour 60: 28-74.
Wiley RH, DG Richards. 1978. Physical constraints on acoustic communication in the atmosphere: implications for the evolution of animal vocalizations. Behav. Ecol. Sociobiol. 3: 69-94.
Wheatley BP. 1999. The vocal repertoire of Macaca fascicularis at the monkey forest at Padangtegal. In: The Sacred Monkeys of Bali,
沒有留言:
張貼留言