How do animals distinguish between their own vocalisations and those of another?

 by Niamh McCormack

Introduction

Many animals produce a large variety of sounds which play an important role in a complex communication system. Differences in acoustic structure may convey information about individual, sexuality and group identity, body size and reproductive status (Ghazanfar & Hauser, 2001).

Vocal recognition plays a huge role in parent-offspring recognition. Individual and kin recognition have important implications for the evolution of social behaviour (Ghazanfar & Hauser, 2001). So far, individually distinctive, learned signals for recognition and communication have only be found in humans but recent studies suggest that the same maybe true of several animals including dolphins and some species of bats.

The production of sounds by an animal, present the auditory system of the caller with two fundamental problems: the first is discriminating between self-generated and external sounds and secondly, preventing desensitisation (Poulet & Hedwig, 2002). The auditory system is stimulated by the animal’s self-vocalized sound. This self-stimulation is important in controlling the vocalisation, as evidenced by the interference with normal development of songs in deaf birds or in the irregularity of speech sounds from deaf persons (Suga & Shimozawa, 1974). Acoustic self-stimulation would be unnecessarily intense for monitoring vocalisations if there were no mechanisms to attenuate the amount of self-stimulation (Suga & Shimozawa, 1974). Such attenuation occurs at both the receptors and the brain and these mechanisms vary within the animal kingdom. In humans, bats and other mammals, the muscles of the middle ear contract synchronously with vocalisation to attenuate self-stimulation (Suga & Shimozawa, 1974).

I.                Bats

A particularly interesting group of animals to study are bats (order Chiroptera) as these animals are highly dependent on their auditory system to perceive the world around them. Contrary to popular belief bats are not blind, but in many species sight is limited. Most bats emit high-frequency sounds through their mouth or nose and perceive their nocturnal environment by listening to the echo’s from objects that surround them. This phenomenon termed ‘echolocation’ has succeeded in making bats independent of sunlight as a medium for perceiving their world (Neuweiler, 2000). Echolocation is not unique to bats it is also found in the Cetaceans (whales and dolphins), the insectivores (shrews and tenrecs) and some species of birds such as the South American oilbird or the Indo-Australian cave swiftlet. Echolocating animals have an anti-jamming system, which allows them to recognize and pick up their own sound, and not cross echoes with other animals.

The origin of echolocation in bats has two very different hypothesis, the first is that echolocation evolved due to the need to navigate in caves; the second is that echolocation evolved to aid in capturing insect prey (Neuweiler, 2000). Both hypothesis propose that echolocation was needed to occupy an ecological niche that was unexploited by others.

The order Chiroptera consists of two suborders; Megachiroptera and Microchiroptera, almost seventy percent of all bats species belong to the Microchipteran suborder. Not all bats can echolocate, the majority of the Megachiroptera rely mainly on vision and olfaction, which are highly developed, and as a result the need to echolocate is reduced.

The only Megachiroptera that echolocate, are flying foxes of the genus Rousettus. Unlike the Microchiroptera, Rousettus produces calls with its tongue. When the tongue is raised from the floor of the mouth it produces a double-click as it is passed over each lip. Echolocation is only used by Rousettus to navigate in caves where they roost; otherwise they use their large eyes for orientation.

The sounds produced for echolocation in Microchiroptera are generated by the larynx, released through the mouth or nose, and echo is received through large specialised ears. The elaborate nose-leaves seen in many species functions as a megaphone. Because air absorbs energy contained in sound waves, the bat must emit high intensity sounds in order to receive a faint echo. Not all echolocating calls are at the same pitch. Although low-frequency sound travels further than high-frequency sound, small ultrasonic sound waves provide more accurate information on size, distance, speed and direction of prey. The high frequency sounds range from 20-200 kHz or more, humans cannot hear sounds above 20kH.

Figure 1: Sonar pulses of an echolocating bat.

 Echolocation calls are produced at constant frequencies (CF calls) and varying frequencies that are frequently modulated (FM calls). Most bats produce a complicated sequence of calls, combining CF and FM components. Target velocity is carried by the difference between the emitted pulse and the returning echo CF (that is, the Doppler shift) and distance information is carried by a difference in projected and returning FM sounds (that is, the echo delay), (Suga & Yan, 1996). The nuclei of the lateral lemniscus (tract of axons ascending the brainstem) play an important role in processing timing information that is essential for target range determination in echolocation (Huffman, Argeles & Covey, 1998).

To increase the limited range of sounds, bats produce high intensity calls. The loudness can vary from as low as 50dB and as high as 120 dB, which is louder than a smoke detector 10cm from your ear. The little brown bat Myotis lucifugus can emit such intense sound. If these sounds directly stimulated the ears, hearing would be impaired (Suga & Shimozawa, 1974).

Vocalisation is initiated by activity in the forebrain. Sound is generated by larynx and funnelled through the mouth or nostrils, thus most of the sound energy is concentrated in the direction in which the bat is flying. The bat has several mechanisms, which protect it from deafening itself when emitting such high intensity sounds. (i) The stapedius (middle ear muscle) contracts to separate the three bones, the malleus, incus and stapes, reducing the hearing sensitivity. This contraction occurs about 6ms before the crycothyroid (larynx muscles) begin to contract. The middle ear muscle relaxes 2-8ms later, and then the ear is ready to receive an echo. This mechanism attenuates self-stimulation by 20-25dB. (ii) The inner ear is loosely attached to the skull to reduce the amount of bone conduction. (iii) Sensitivity for sound frequencies emitted during flight is low but sensitivity is greater for sounds a few kHz higher than emitted sounds (i.e. an echo shifted by the Doppler effect) (Suga & Schlegel, 1972). (iv) There is indication of neural attenuating mechanism operating synchronously with vocalisation (Suga & Shimozawa, 1974). This neural attenuating mechanism was found by measuring the responses of the primary auditory neurons. The mechanisms for neural attenuation are not yet known but it is known that it takes place between the cochlear nerve and the inferior colliculus. The attenuation by neural events is thought to be as much as 10-15dB (Suga & Shimozawa, 1974). It is thought that the duration of the inhibitory period is very short because anything longer than a few milliseconds would interfere with echo detection.

The moustached bat (Pteronotus parnelli rubiginous) emits a vocal sound which consists of a CF component and a FM component, each component is made up of four harmonics therefore each sound emitted comprises of eight harmonics. The emitted sound overlaps with the echo of the previous call; as a result biosonar information must be extracted from a complex sound with up to 16 components (Suga & O’Neill, 1979). With many conspecific bats echolocating in a confined space their many sounds and echo’s would impair ‘echo-imaging’ unless some mechanism protected the system from jamming (Suga & O’Neill, 1979). The moustached bat overcomes this interference by suppressing the first harmonic of its call so that it is much weaker than the other harmonics and that other bats are unable to hear it. However the bat hears its own harmonic directly by bone conduction between the vocal chords and cochlea. The bat does not respond to the weak first harmonics of other bats and is therefore not confused by their presence (Altringham, 1996).

II. Insects

Like birds or mammals, some insects use elaborate acoustic signals for intraspecific communication. Insects often sing in aggregations, some species avoid interference by singing either in alteration or in synchrony with others (Pollack, 2000). Grasshoppers rub their hindlegs against the wings resulting in a form of sound production called stridulation. During sound production grasshoppers stimulate their own ears, which are placed on the abdomen. The self-stimulation of the grasshopper’s song is so loud it masks the afferent and interneuronal response to external stimuli (Hedwig, 2003). As a result, grasshoppers are unable to respond to external stimuli during stridulation.

Crickets have developed a more sophisticated system for processing self-generated auditory signals. The calling song of singing cricket, Gryllus bimaculatus is generated by rhythmically rubbing the forewings together. Parts of the wings act as a resonator and the sound emitted is restricted to a narrow frequency range. This calling song can be produced for many hours with a sound intensity greater than 100dB sound pressure level (SPL), (Poulet & Hedwig, 2002). The cricket’s ears are placed on the forelegs and are fully exposed to the self-generated sounds. Many animals reduce the sensitivity of the auditory system during sound production, but crickets do not. Experiments have shown that singing crickets respond to surrounding sounds during chirp intervals (Poulet & Hedwig, 2002).

To distinguish between reafferent and external stimuli, the two types of sensory information must be treated differently. During stridulation in the cricket, a ‘corollary discharge’ (efferent signal) mechanism pre-synaptically inhibits the auditory afferents and post-synaptically inhibits central auditory interneurons (Hedwig, 2003). The response of an identified interneuron (the Omega 1 neuron-ON1) to the self-generated sound is reduced and the neuron remains sensitive to external sounds in the chirp interval (Poulet & Hedwig, 2002). This inhibitory mechanism works in synchrony with sound production and prevents desensitisation of the auditory pathway.

Insects hear their calling songs exactly as they sound because there is no bone conduction of sounds to the tympanic membrane. Insect songs (with rare exception) lack the ‘melodies’ that characterize most vertebrate communication signals, songs consist of relatively analogous sound pulses (Pollack, 2000). The ears of many insects are capable of spectral analysis, the biological information of importance is carried in the temporal structure of the song (i.e. the durations, shapes of sound pulses and the spacing between them). The grasshopper knows its own song because when stridulating it hears only self-produced sound, the cricket on the other hand can hear while stridulating but sensitivity is reduced to its own sound so, it can process external sounds simultaneously.

III Frogs

The croaking sound is produced by air passing over the vocal cords in the larynx of the throat, so it continues a true voice (Dickerson, 1969). The loud call is produced usually by males during the breeding season. Females generally produce quieter calls and do not show external vocal pouches. The sounds can also be produced underwater in some species. The air enters through the nostrils and passes back from the mouth to the vocal cords. The Salienta possess internal sacs in the throat region or on each side at the shoulder. The vocal sacs are filled with air and act as resonators (Dickerson, 1969).

Figure 2: Inflated vocal sacs of a Green Tree Frog.

Each species has a typical call or song with a definite pitch and quality of tone. During spring and early summer, many frogs call in a chorus. Similar to insects, sounds can be timed to fit in the brief moments when nearest neighbours are taking a break. Large frogs produce deep calls at low frequency and small frogs emit high frequency chirps.

The frogs auditory system has an unusual adaptation in that there is an unbroken air link from the eardrums to the lungs that is thought to function in sound location and also to protect the eardrums from desensitisation. Frog calls are extremely loud, each male calls his loudest to drown out the others and attract a distant female. Some species such as Eleutherodactylus coqui call at 90-95dB. The intensity of the call in the eardrum of these frogs is somewhat louder, reaching levels of 120dB SPL, which is potentially damaging (Narins, 1995).

The frog’s lungs protect his ears by equalizing pressures between the inner and outer surfaces of the eardrum. When a male calls, the vocal sac bulges out, pulling the membrane tight. The sounds then impinge on both the inner and, outer surface of the eardrums after being radiated from the vocal sac (Narins, 1995). If the sounds arrive at the same time so that periods of high pressure coincide on both sides, the eardrum will not vibrate much in response. The frog knows when it is vocalising as it hears itself, but the response of the eardrums is attenuated to prevent desensitisation.

Many fish also use their lungs to hear, they have a lung-like bladder, specialized for sound reception. Sound travels underwater, vibrates the air sac, which in turn vibrates the fish’s inner ear.

IV Dolphins

Vocal communication in dolphins is very complex, they make many types of sounds both underwater and in air. Sounds made in air generally communicate distress through whistling. The sounds that are made underwater serve to locate food, and are used to communicate and navigate. Dolphins lack vocal cords but it is hypothesised that sounds are produced in three pairs of air sacs located underneath the blowhole. When a breath is taken, the blowhole is shut and air returns from the lungs into an air sac, inflating it. As air is forced out of the air sac, it passes over a nasal plug at the opening of each air sac producing various noises (DRC, 2003).

Of the sounds made underwater by dolphins, three that are well known are whistles, clicks and burst pulses. Burst pulse sounds are low frequency sounds emitted only under emotional duress.

Dolphins produce many types of whistles but each dolphin has an individually distinctive signature whistle. The signature whistle is developed in the first six months of life and is thought to function in individual recognition and group cohesion (Janik, 2000). The frequencies of whistles range from 5-15 kHz. Dolphins are capable of vocal learning and imitation, they have been recorded making signature whistles of other dolphins, possibly to address them (DRC, 2003). Although vocal learning is common in birds, bottlenose dolphins and some bat species are the only mammals (except humans), which interact with learned signals (Janik, 2000).

Echolocation is also used by dolphins. Clicks emitted by sphincter muscles in the blowhole range in frequency from 0.25-220 kHz. Dolphins have a waxy, lens-shaped structure in their forehead called the melon that focuses the clicks into a tight beam forward (DRC, 2003). When scanning an object dolphins will move their head from side to side to direct the beam back and forth over the object. The echoes are picked up through the lower jaw, it is also thought that the arrangement of teeth in the jaw acts as an antenna to focus the incoming sound. Similar to bats, dolphins have an anti-jamming ability associated with their echolocation. Each dolphin has the ability to pick up its own echo in a group of echolocating dolphins without getting confused. Dolphins recognise their own calls because like birds and humans, each voice is individually distinctive.

V.              Birds

Vocal communication in birds is extremely well developed. Sound is produced as air is forced out over the syrinx, which is a membrane over each of the two bronchi. Birds lack a larynx but this structure allows allows separate sound sources, one in each bronchus. The ear bone structure also differs in birds, a single bone; the columella replaces the three ear ossicles found in mammals. The frequency range of hearing in birds varies from species to species but the maximum sensitivity is similar to man, between 1-4 kHz (Catchpole, 1979). The bone structure of birds differs slightly in that bones are pneumatic, this may reduce bone conduction of vibrations to the middle ear when the bird is vocalising hence reducing self-induced deafening. Although no information available suggests that neural attenuation occurs in the bird group it is thought that this also prevents desensitisation of the middle ear.

Birds have a great variety of vocal sounds, these range from short, monosyllabic calls to complex sequences, their songs. The order Passeriformes (songbirds) learn to sing during a sensitive period shortly after birth (Brainard & Doupe, 2002). Birds develop a distinctive song or ‘voice’ that distinguishes it from other birds, although dialects are often found among bird species. A bird’s song is based on the sounds it hears during an early sensory learning period, which is usually the song of parents and neighbouring birds (Brainard & Doupe, 2002). Songbirds must also be able to hear themselves in order to learn to vocalise normally. Birds use auditory feedback to compare their developing vocalizations with the template, and guide their song modification using this comparison (Brainard & Doupe, 2002). The importance of auditory feedback is revealed by the irregularity of song in deaf birds (Suga & Shimozawa, 1974). The capacity for such hearing-dependent vocal learning is not widespread, as vocal learning is found only in humans, birds and dolphins (Doupe & Kuhl, 1999). Information to be communicated from one animal to another is often present in the temporal structure of the acoustic signal (Alder & Rose, 1998). Songbirds have one of the most complex sensory neuron systems known (Brainard & Doupe, 2002). Neurons have been recorded that respond selectively to the sound of the birds own song, these neurons are tuned in for particular pairs of notes/syllables that are presented with a particular temporal order and spacing (Alder & Rose, 1998). This proves that the bird recognizes its own song distinctly from others.

Conclusion

Bats and Dolphins recognize their own vocalisations with help from an anti-jamming mechanism, which ensures they do not pick-up other bats echoes. This mechanism has not been studied much in detail in dolphins. In the moustached bat (Pteronoyus parnelli rubiginosus) the mechanism involves suppressing the first harmonic of the call so that only the bat producing the sound can hear it. It is reasonable to hypothesise that a similar mechanism operates in echolocating dolphins.

The grasshopper can distinguish between its sounds and those of another because when stridulating it hears only its own sound. The cricket on the other hand is capable of hearing itself and is aware of external sounds at the same time. To the human ear crickets calls sound analogous but within the sound pulses the information is carried in the temporal structure. Crickets are aware of external acoustic signals during stridulation and can modify the sound in the interval between chirps, changing the song to convey different information; therefore calls are not always analogous. The cricket hears the response he generates from his environment and changes his sound accordingly, it can distinguish self-produced sounds from others due to a neural inhibitory mechanism. This mechanism ensures maximal sensitivity of the auditory system to external sounds by greatly reducing sensitivity to self-produced sounds. The frog is capable of both hearing and vocalizing simultaneously. The frog hears the self-produced call, therefore it can distinguish between its own call and those of others. The intensity of the frog call can reach very high levels so attenuation is achieved by incorporating the lungs into the auditory system to prevent damaging the eardrums. This is not a unique adaptation as many fish also use the lungs for hearing.

Bird vocalisations are individually distinctive, as are human and dolphin sounds. Each bird knows what it sounds like because it can hear its own song while singing, when the bird ceases singing it can be sure that what it hears is not self-produced. It is suspected that neural attenuating mechanisms similar to those found in mammals and insects also prevent self-induced desensitisation in birds as well as the bone structure of the skull and middle ear.

Vocalisation is initiated in the brain of all species and it appears that small animals hear themselves vocalize to one extent or another. In some species (especially those with high intensity calls), the sound they hear is greatly attenuated to what they actually emit. Auditory feedback plays an important role in vocalisation but continuous stimulation may lead to self-induced desensitisation. There are many different mechanisms to prevent desensitisation of the ear, the most common mechanism found in this study is neural attenuation but others include contraction of middle ear muscles and, use of the lung in the auditory system. The auditory neurons of humans and other vertebrates have been recorded to be less responsive during vocalisation, but the nature and source of the inhibition has never been characterized (Poulet & Hedwig, 2002).

References

Alder, T.B. & Rose, G.J. (1998). Long-term temporal integration in the anuran auditory system. Nature Neuroscience Vol. 1, No. 6, pp. 519-523.

Altringham, J.D. (1998). Bats biology and behaviour. Oxford University Press.

Bennu, D. (2001). The night is Alive with the Sound of Echos. University of Washington http://students.washington.edu/~nyeve/bats.html.

Brainard, M.S. & Doupe, A.J. (2002). What songbirds teach us about learning. Nature 417: 351-358.

Catchpole, C.K. (1979). Vocal Communication in Birds. Edward Arnold Publishers Limited.

 Dickerson, M.C. (1969). The Frog Book. Dover Publications, Inc. New York.

Dolphin Research Centre-DRC (2003). http://dolphins.org/drc.htm.

Ghazanfar, A.A. & Hauser, M.D. (2001). The auditory Behaviour of Primates: a Neuroethological Prospective. Current Opinion in Neurobiology Vol. 11, Issue 6, pp.712-720.

Hedwig, B. (2003). Neurobiology of insect Communication. Department of Zoology, University of Cambridge. www.zoo.cam.ac.uk/zoostaff/hedwig/.

Huffman, R.F., Argeles, P.C. & Covey E. (1998). Processing of sinusoidally amplitude modulated signals in the nuclei of the lateral lemniscus of the big brown bat, Eptesicus fuscus. Hearing Research Vol. 126, Issues 1-2, pp. 181-200.

Janik, V.M. (2000). Whistle Matching in Wild Bottlenose dolphins (Tursiops truncatus). Science 289: 1355-1357.

Narins, P.M. (1995). Frog Communication. Scientific American 273: 78-83.

Neuweiler, G. (2000). The Biology of bats. Oxford University Press pp. 140-209.

Pollack, G. (2000). Who, What, Where? Recognition and localization of acoustic signals by insects. Current Opinion in Neurobiology Vol. 10, Issue 6, pp. 763-767.

Poulet, J.A. & Hedwig, B. (2002). A corollary discharge maintains auditory sensitivity during sound production. Nature 418: 872-876.

Suga, N. & O’Neill, W.E. (1979). Neural Axis Representing Target Range in the Auditory Cortex of the Moustache Bat. Science 206: 351-353.

Suga, N. & Schlegel, P. (1972). Neural Attenuation of Responses to Emitted Sounds in Echolocating Bats. Science 177: 82-84.

Suga, N. & Shimozawa, T. (1974). Site of Neural Attenuation of Responses to Self-Vocalized sounds in Echolocating Bats. Science 183: 1211-1213.

Suga, N. & Yan, J. (1996). Corticofugal Modulation of Time-Domain Processing of Biosonar Information in Bats. Science 273: 110-1103.