How does echolocation work in odontocetes?

Seeing with your ears. The idea long belonged to science fiction before it was demonstrated in bats in 1938 by Donald Griffin, then in dolphins in 1947 by Arthur McBride at Marine Studios in Florida. Since then, seventy-five years of research have revealed in odontocetes one of the most sophisticated biological detection systems evolution has ever produced. A common bottlenose dolphin can detect a steel ball the size of a grape at more than a hundred metres, tell whether a coin is copper or aluminium with its eyes closed, and reconstruct an acoustic scene in three dimensions as fast as a human reconstructs its own visually.

To understand how this is possible, one has to enter the specialised anatomy of a group that has, over fifty million years, profoundly reshaped its braincase, its nasal passages and its fatty tissues in order to exploit the underwater acoustic channel.

A discovery in several stages

Lazzaro Spallanzani, at the end of the 18^th century, had already grasped that bats navigate in the dark by some sense other than sight. But it was only in 1938 that Donald Griffin and Robert Galambos, at Harvard, demonstrated experimentally that bats emit ultrasound and perceive its echoes. Griffin then coined the word echolocation.

In cetaceans, the same phenomenon was suspected from the 1940s onwards. Arthur McBride, a researcher at Marine Studios in Florida, observed in 1947 that his bottlenose dolphins avoided nets in complete darkness. He recorded the observation in a laboratory notebook, which was published only posthumously, in 1956, by William Schevill and Barbara Lawrence. Between 1960 and 1990, Whitlow Au, working with the United States Navy, ran in Hawaii a series of experiments still regarded as the reference: discrimination ability, detection range, resolution, target memorisation. Those results remain strikingly current today.

How the sound is produced

Contrary to natural intuition, dolphins do not “speak” through the mouth. The sound is generated in the nasal complex, just beneath the blowhole. It holds two pairs of pliable muscular structures, the phonic lips or monkey lips, so named because they vaguely recall a simian pout. When the animal drives a stream of pressurised air towards these lips, they clap against one another extremely fast, producing a click of barely a few microseconds. The compression and decompression generates a brief but intense acoustic pulse, rich in harmonic frequencies.

This primary click is then injected into the melon, the fatty organ that forms the bulging forehead characteristic of odontocetes. The melon is not a simple mass of fat: its lipid composition varies gradually, creating a gradient in the speed of sound. At the core of the melon, sound travels more slowly than at the periphery. This gradation acts as a converging acoustic lens which collimates the beam and focuses it forwards, into a narrow cone. Without the melon, the dolphin would emit in every direction at once and would lose all power of localisation.

Frequencies that vary with the species

Depending on the species, the acoustic profile of the click varies considerably. The sperm whale, a deep diver hunting squid in total darkness, emits clicks around 5 to 25 kHz that are unusually intense: they reach 230 dB re 1 µPa, the underwater acoustic equivalent of a rocket launch. That power allows it to probe the ocean over several kilometres and to detect a squid 500 metres away.

The common bottlenose dolphin typically emits between 40 and 150 kHz, in a broad band suited to the fine detection of fish at short range. The harbour porpoise is a special case: it produces clicks that are extremely narrow in frequency (NBHF, narrow-band high-frequency) centred on 130 kHz. This very high frequency is inaudible to killer whales, their main predator, and probably represents an evolutionary strategy of acoustic evasion. Cuvier’s beaked whale, for its part, deploys a different repertoire again, suited to hunting cephalopods more than two kilometres down.

The ear is not in the ear

The echo reflected by a target comes back to the dolphin. But unlike a land mammal, the animal does not pick it up through an outer ear (the external auditory canals of cetaceans are rudimentary, and may be closed altogether). It is the lower jaw that acts as the receiving antenna. It contains a fatty cavity, the acoustic fat body, which conducts sound without attenuation to the inner ear, isolated from the rest of the skull by air sinuses. That isolation allows the animal to tell precisely whether sounds come from the left or from the right, a capacity essential to three-dimensional localisation.

Once in the inner ear, the signal is transmitted to the brain by a cochlear nerve five times as large as that of comparable land mammals. The auditory cortex, itself very extensive, processes dozens of parameters at once: intensity, frequency, delay, Doppler shift (which indicates the relative speed of the target), spectral shape. It is this massively parallel processing that makes echolocation both fast and precise.

A resolution beyond imagining

Whitlow Au and his colleagues systematically tested the limits of this perception. A few of the results have become classics:

• A common bottlenose dolphin can tell apart two steel balls whose diameters differ by only 1 mm at a distance of 3 metres;
• Blindfolded, it distinguishes cylinders of copper, brass, steel or aluminium of identical size (the acoustic signature varies with the material);
• It detects a 5 cm fish at more than 100 metres in turbid water;
• More impressive still, it can recognise by echo alone an object learned visually, evidence of a sensory cross-modality rare in the animal kingdom.

The terminal “buzz”

Echolocation is not a continuous signal. The animal emits isolated clicks at a rate of one or two a second while searching, then speeds up progressively as it closes on a target. In the last few hundred milliseconds before a fish is caught, the rate rises to more than 500 clicks a second. This burst, known as a buzz, sounds in recordings like a continuous crackle. It is the equivalent of an acoustic zoom: the closer the target, the faster the image has to be refreshed. Hunters hear the buzzes of their sonar the way a trader hears price movements, by ear.

And the mysticetes?

One question comes up often: do baleen whales echolocate as well? The current answer is no. They have neither a fatty melon, nor phonic lips, nor the corresponding neural architecture. Their vocalisations are very low (10-200 Hz) and suited to communication over very long distances, not to short-range probing. A few observations have suggested that humpback whales might produce “clicks” at 30-300 Hz useful for detecting fish shoals, but the hypothesis remains a minority one. Mysticetes probe their surroundings mainly by sight and through ambient acoustic cues, not by an active sonar.

Convergent evolution and conservation

Cetacean echolocation developed independently of that of bats, a spectacular case of convergent evolution. The anatomical solutions differ: bats vocalise through the larynx and receive through the external ears; cetaceans vocalise through the nasal passages and receive through the jaw. Yet the end structures in the brain, an enlarged auditory cortex, specialised neurons, converge to a striking degree (Madsen & Surlykke, 2013).

This extreme acoustic specialisation makes toothed whales particularly vulnerable to anthropogenic noise: military sonar, seismic surveying, shipping traffic. See our feature on the impact of ocean noise and the one on military sonar. To understand echolocation is also to understand why an ocean that is too noisy condemns these animals to fly blind.

Sources

• Au, W.W.L. (1993). The Sonar of Dolphins. Springer-Verlag, New York.
• Cranford, T.W., Amundin, M. & Norris, K.S. (1996). “Functional morphology and homology in the odontocete nasal complex: implications for sound generation”. Journal of Morphology, 228(3), 223-285.
• Madsen, P.T. & Surlykke, A. (2013). “Functional convergence in bat and toothed whale biosonars”. Physiology, 28(5), 276-283.
• Schevill, W.E. & Lawrence, B. (1956). “Food-finding by a captive porpoise (Tursiops truncatus)”. Breviora, 53.
• Griffin, D.R. (1958). Listening in the Dark. Yale University Press.
• NOAA Fisheries, Marine mammal echolocation, fact sheet.

---

Going further