Auditory communication

Elephants use a variety of sounds, described as trumpets, roars, barks, snorts, growls, and rumbles, for communication at close range to long distances. While most of the sounds produced by elephants are true vocalizations (i.e., are generated from the vocal cords within the larynx), there are some that emanate from the trunk that may also convey a signal.

The sound language of elephants, as audible to humans, was described in subjective terms by several observers. The more recent discovery and characterization of "infrasound," along with the possibilities of playback experiments using modern gadgetry, however, has opened new avenues to study acoustic communication among elephants. Problems with studying vocalization are to first identify the individual making the call, coin an appropriate term to describe it, and then interpret the meaning of the call.

A repertoire of "basic sounds," the modification of their properties (such as amplitude and resonance), the resulting sound, and its interpretation for wild Asian elephants was given by George McKay. Several of these had also been described by other observers, including G. P. Sanderson and M. Krish-nan. I have taken the liberty of reproducing McKay's repertoire with my own suggestions (table 4.2). For the wild African elephant, Joyce Poole gives a more comprehensive list of calls based on their context, such as group dynamics, distress, and sexual signaling. She observed that, of 26 calls made by adult elephants, 19 are made only by females, 3 by both sexes, and only 4 exclusively by males. Female elephants thus seem to possess a much richer repertoire of acoustic signals compared to males. This may be related to the more

Table 4.2

Vocalizations and their possible functions in Asian elephants.

Table 4.2

Vocalizations and their possible functions in Asian elephants.


Basic Sound








Short-distance contact


Resonate in trunk


Mild arousal


Increase amplitude


Long-distance contact


Low frequency


Infrasound in different contexts






Multiple short squeaks


Conflict, nervousness


Lengthen, increase amplitude


Extreme arousal





Change in activity


Increase amplitude


Mild-to-strong arousal


Same as no. 2 plus bounce trunk tip on ground


Threat display

Source: Modified from McKay (1973).

Source: Modified from McKay (1973).

frequent and more diverse communications in which females have to engage, both within the family and across family groups encompassing the larger clan.

Several elephant observers recognized a call barely audible to humans. G. P. Sanderson, writing about the southern Indian elephants in 1878, briefly noted "an almost inaudible purring sound from the throat" (p. 49) of an elephant he interpreted as expressing pleasure. The Indian naturalist M. Krishnan clearly recognized in a 1972 article that wild Asian elephants communicate at sound frequencies not fully audible to humans. He described these as "throaty, hardly audible," "low-pitched but clearly audible from a distance," and "a throbbing purr." The earliest experiments on the hearing abilities of the elephant were probably those by Rickye and Henry Heffner on a 7-year-old female Asian elephant at the Mitchell Zoo in Independence, Kansas. Using a pair of loudspeakers to broadcast varying sound frequencies this study, published in 1980, determined that the hearing range of an elephant extended from 17 Hz (hertz or cycles per second) to a maximum of 10.5 kHz (kilohertz). By comparison, the auditory range of humans is about 30Hz-20kHz. Thus, elephants can hear low frequency sounds that humans cannot, while humans can hear high frequency sounds that elephants cannot. This was not really surprising because it was known that hearing ability in mammals was related to body size, with smaller mammals better able to hear high frequency sounds. It was the first time this hypothesis was tested on a very large mammal. More precisely, the ability to hear high frequency sound varies inversely with the distance between the ears (interaural distance); smaller mammals with close-set ears are therefore better equipped to hear higher frequencies. In contrast, the elephant had better ability to hear low frequency sound than any other mammal tested.

Judith Berg recorded sounds made by captive African elephants at San Diego Wild Animal Park in California. She described some "low intensity" calls and inferred the existence of sound frequencies between 18 and 28 Hz, although the dominant frequency was above 90 Hz, which is well within the human audible range.

The first complete description of infrasonic calls in elephants came from a 1986 paper by Katharine Payne, William Langbauer Jr., and Elizabeth Thomas. Katharine Payne was observing Asian elephants at the Washington Park Zoo in Portland, Oregon, when she sensed "a palpable throbbing in the air like distant thunder, yet all around [her] was silent" (1989, p. 266). She realized that the elephants could be communicating at low sound frequencies below the threshold of human hearing. Using a tape recorder capable of registering low-frequency sounds, Payne's team found that the elephants were calling at fundamental frequencies ranging from 14 to 24 Hz with a duration of 10-15 seconds and sound pressure levels of 85-90 dB (decibels) at a distance of 5 m. Because low-frequency sound at high pressure can pass through vegetation with little attenuation, this provides a means for communication over distances greater than several hundred meters. A few years later, from playback trials of prerecorded low-frequency calls, they inferred that captive African elephants responded to what seemed to be recognition of biologically meaningful signals (fig. 4.11).

The next step was to catalog the repertoire of infrasonic calls, assign a meaning to each, and carry out field trails to document the extent to which infrasounds are used by elephants in various social contexts in the wild. For this, Katharine Payne and her associates collaborated with Joyce Poole and Cynthia Moss at Amboseli (Kenya) and with Rowan Martin and Ferrel Osborn at Sengwa (Zimbabwe) and Etosha (Namibia). Using the considerable expertise of Joyce Poole and Cynthia Moss in the study of elephant social behavior, the team obtained spectrograms of eight distinct "call types" with fundamental frequencies of 14-35 Hz. Only about a third of the rumbles recorded were actually audible to one of the observers in the field. The rest emerged using techniques such as speeding up the tape or through examination of the spectrograms (table 4.3).

While a catalog of infrasonic calls made by elephants was a basic first step in studying their role in communication, this did not by itself prove that elephants in the wild were indeed responding to such calls or show the extent to which these were important in long-distance communication. At Sengwa, several female elephants were fitted in 1990 with radiotransmitters to track their movements and to relate these to infrasonic calls they made. Although the study had to be abandoned after a year's work (because of an elephant cull at Sengwa), this did indicate that some family groups coordinate their movements through infrasound contact calls. Rowan Martin's unpublished earlier study of elephant movement there had strongly suggested such coordination; at that time, the mechanism was not known.

The team, this time led by William Langbauer, had earlier carried out field trails using playbacks in Etosha National Park, Namibia. Using an array of

Figure 4.11

Infrasound calls in elephants. Sound spectrogram of estrous call and of musth rumble by African elephants. (Courtesy of Katharine Payne.)

Figure 4.11

Infrasound calls in elephants. Sound spectrogram of estrous call and of musth rumble by African elephants. (Courtesy of Katharine Payne.)

Table 4.3

Characteristics of infrasounds and their biological context recorded at Amboseli.

Sound Pressure

Levels at 5 m Fundamental Call Type (dB SPL) Frequencies Description of Social Context

1. Greeting rumble 92

2. Contact call

3. Contact answer 103

5. Musth rumble

6. Female chorus

7. Postcopulatory call 102

8. Mating pandemonium 100

18-25 Hz* This call is typically used by adult females within the same family or bond group when they meet after separation for several hours.

18 Hz A relatively soft, unmodu lated sound accompanied by steady ear flapping. The contact call and the answer seem to be used when elephants are separated by up to 2 km during feeding.

18 Hz In response to the contact call, the answer starts loudly and then becomes softer at the end.

15 Hz As the term suggests, this soft rumble is used by a female to herd the members of the group before moving to a different location.

14 Hz A male in musth gives out a low-frequency pulsated sound (the "motorcycle") several times an hour.

15-24 Hz* Several adult females may answer a musth rumble with a low-frequency modulated chorus.

18-35 Hz* An estrous female makes a series of loud calls for up to 30 minutes; more common after mating.

— This indicates much excite ment among members of a mated female's family. They indulge in a frenzy of activity.

Source: After Poole et al. (1988). Reproduced with the permission of Springer-Verlag GmbH & Co. KG.

Note: All sound pressure estimates are based on measurements at a particular distance and extrapolation to 5 m (SD = ±3 in all cases).

*These calls are modulations from the lower frequency to the higher one and back to the lower one.

microphones, recorders, two video cameras, and a loudspeaker mounted on a van, they played back infrasonic calls to elephants that were 1.2 km and 2.0 km from the speaker. The calls were broadcast at only half the sound pressure levels of the strongest calls recorded because of a technical limitation of the loudspeaker. Simultaneously, they made audio and video recordings of the elephants' behavior from a tower overlooking a waterhole. The playback experiments were directed toward both family and bull groups. The infrasonic calls selected for the playbacks included a variety of female social contact and postcopulatory calls.

The results showed that both family and male groups responded, through changes in their behavior, more after the playback than before it at the two distances tested. In statistical terms, the overall response scores were significant for male groups at 1.2 km and 2.0 km, while for family groups they were significant only at 1.2 km. The responses obviously varied with respect to the specific behavior of the elephants and their change in distance from the loudspeaker after the playback. In a typical full response, one or more elephants "would lift their heads within a few seconds of the onset of the playback and raise, spread and stiffen their ears, and then freeze, apparently listening. Simultaneously or shortly after this, one or more animals vocalized. The animals would then 'scan,' slowly swinging their heads from side to side, orient towards and move towards the loudspeaker" (Langbauer et al. 1991, p. 42). One notable difference in the responses by family groups and males was a greater tendency in the former to vocalize and in the latter to move longer distances toward the loudspeaker. This could be related to the need for family groups to maintain regular contact for coordinating their movements and for males to locate an estrous female (which has signaled) quickly to take advantage of a mating opportunity. Because the sound pressure levels of the playbacks were about half (or 6 dB less than) the strongest calls recorded, William Langbauer and the team calculated that the effective range of infrasound could be at least 4 km, or twice the broadcast distance. This translates roughly to a 50-km area over which a strong infrasonic call could be audible.

Although there is still much to learn about auditory communication in elephants, the discovery of infrasound is proving to be a powerful new tool to investigate aspects of elephant social behavior and ecology. Indeed, recent research by Karen McComb in association with Cynthia Moss and her team at Amboseli points to infrasound as the medium for a well-developed network of individual recognition among elephants. Using recordings of female contact calls from several individually identified family groups, they conducted a series of playback experiments to test the ability of adult females to recognize the calls of other females socially related to them to varying degrees. Social familiarity was determined from long-term records of associations among various female-led family groups. Behaviors such as contact calling in response, approaching the loudspeaker, listening with ears extended, bunching together in agitation, and avoidance by moving away from the loudspeaker were used to interpret the subjects' reactions.

The first two positive behaviors were exclusively performed in reaction to playbacks from members of the same family unit or bond group (see section 4.5.2 for definitions of social units). In all but one instance (11 of 12 playbacks), listening only was directed toward members of high-association families. The opposite was true of playbacks from low-association family members beyond the bond group. Playbacks from such members elicited the more negative behaviors, such as bunching and avoidance (in 11 of 12 cases). Interestingly, a female's contact call played back 3 months after her death elicited a positive response from her family members. These experiments confirm the ability of elephants to recognize individuals through discrimination of their vocalizations. McComb and her associates calculated that an adult female elephant at Amboseli would be familiar with the calls of roughly 14 families comprising about 100 adult females in the population, an unusually extensive network of vocal recognition for a mammalian species.

It is well known that the propagation of sound is influenced by environmental factors such as atmospheric conditions (wind, temperature, turbulence, moisture, etc.), topography, ground reflection, and attenuation by vegetation in addition to the physical properties of the sound itself. David Larom and Michael Garstang, along with their associates, have researched the influence of atmospheric conditions over the southern African savannas, using empirical data and computer modeling, on propagation of animal vocalizations. The area of sound propagation changed substantially over a 24-hour period. Strong atmospheric temperature inversions prevail close to the ground before sunset and decay with sunrise. Calm winds may accompany this temperature inversion during the early evening. Under such conditions of clear, cold, and calm nights, about 1-2 hours after sunset, the conditions are optimum for sound propagation. At this time, the spread of a 15-Hz vocalization may be over 10 km and "calling areas" over 200 km2, as opposed to daytime figures of under 50 km2 (the calling area is calculated from within the contour of the sound pressure dropping by 67 dB from the source).

One obvious characteristic of low-frequency sound is that it can travel over longer distances with little attenuation by vegetation. From this consideration, I tentatively suggested in 1994 that, in theory, infrasound could be expected to be even more useful for forest-dwelling elephants compared to those in the savannas. Recent recordings of infrasound in African rain forest habitat seem to support this expectation. Several recordings of the forest elephant (Loxodonta africana cyclotis) made in Central Africa by Katharine Payne and Stephen Gulick (unpublished 2001) show strong calls at much lower frequencies of 5 Hz (the lowest frequency recorded for L. a. africana in the savannas is 14 Hz).

The inverse relationship between body size (or interaural distance) and frequencies of auditory communication in mammals is well known. At the same time, there is a broad positive correlation between body size and home range area in terrestrial mammals. The larger mammals thus need to communicate over longer distances, which is facilitated by infrasound.

The diurnal variation in sound propagation (of not just infrasound, but a range of frequencies) through the atmosphere also suggests that it would be adaptive for animals to link their communications to the more favorable periods. The collared elephants at Sengwa also used loud, infrasonic calls most frequently during the late afternoon when they trekked to a waterhole. I have observed a high frequency of "roaring" by Asian elephants during evening and early night hours.

Sound propagation characteristics would also vary in the short term from one day to another and with the longer seasonal climatic changes. This brings up the possibility of more complex linkages among auditory communication, movement patterns, social organization, and behavior among elephants. The social dynamics of elephants, from maintaining contact while foraging during the day, gathering at a waterhole in the evening or making sudden seasonal movements, could all be related to the windows of communication opportunity.

Malan Lindeque has suggested, for instance, that elephants may make early seasonal movements to areas receiving rains by detecting infrasounds generated by thunderstorms. This could be in addition to or superior to their ability to smell moisture, as suggested by other observers. Clearly, there are many opportunities for research into the complex world of elephant communication.

0 0

Post a comment