Hunting by Sound: The Bat's Sonar

AThe phrase most often used of a hunting bat, that it sees with its ears, is a compliment which quietly misleads. Echolocation is indeed an extraordinary instrument, able to resolve a target thinner than a pencil line, and the picture that has settled in popular accounts is of an animal for which darkness poses no difficulty whatever. What that picture leaves out is distance. Sound spreads on the way to a target and spreads again on the way back, so the energy returning from a moth falls away with the fourth power of the range, and even a bat shouting at the top of its voice picks up an insect of ordinary size at no more than about five metres. A hunting bat is therefore working with a description of the world that extends barely two wingspans ahead of it. Vision in a nocturnal bird reaches very much further, and this discrepancy, long known and long treated as a curiosity, has lately become the centre of an argument about what the sense is really doing.

BThe constraints are not merely a matter of reach. A call loud enough to return a usable echo is also loud enough to injure the animal producing it, and the interval between shout and echo, at these distances, is a fraction of a millisecond, so the ear must recover almost at once. In most species a small muscle contracts just before each call and relaxes immediately afterwards, cutting the sensitivity of the ear by roughly forty decibels and restoring it within four milliseconds. Horseshoe bats have taken a different route, separating call from echo in frequency rather than in time: they lower the pitch of the outgoing call by exactly the amount their own flight would raise the returning one, holding the echo inside a narrow band to which the ear is finely tuned. The engineering is remarkable. It is also costly to run, which is part of what makes the next question awkward.

CHanne Vestergaard, a sensory ecologist at the Cranmoor Laboratory, maintains that a sense of such limited reach cannot be doing the navigating, and that bats in familiar country are flying largely from memory. In experiments first reported in 2014 her group flew greater mouse-eared bats along a corridor they had used for weeks and then, between one flight and the next, hung a fine net across it. On 62 of 80 subsequent passes the animals flew straight into the obstruction, although recordings showed they were calling normally and the net lay well inside their detection range. Bats, she argues, treat the echo not as a picture but as a check upon a map they already hold, and a check is disregarded when the map is confident. She concedes that the effect appears only along routes flown repeatedly, since animals released into unfamiliar rooms avoid the same net without any difficulty at all.

DRafael Otoya, of the Delham Institute of Animal Behaviour, is sceptical of conclusions drawn inside buildings. Corridors, he contends, teach an animal that nothing ever changes, and a laboratory which has trained its subjects to expect an empty tunnel has manufactured the very carelessness it then proceeds to measure. Between 2018 and 2021 his group fitted miniature tags to 140 bats hunting over open water and recorded the acoustic signature of a successful capture on the first pass through airspace the individual had never entered before. If memory were doing the work, he insists, such captures ought not to happen at all. Otoya accepts two weaknesses. Open water is the least cluttered setting in which a bat can hunt, so his animals are being tested where echolocation performs best rather than where it ordinarily operates, and the tags weigh enough to be carried only by the heavier species, which leaves the small forest bats that dominate the family unsampled.

EA second dispute concerns the insects. Many moths answer an approaching bat with a burst of ultrasonic clicks, and the textbook explanation is that these clicks jam the sonar by returning false echoes at the moment the hunter most needs its own. Vestergaard is unconvinced. In her view the clicks are a warning, comparable to the stripes of a wasp, announcing that the moth is unpalatable, and she notes that the species which click most persistently are also those whose bodies carry the strongest toxins. Otoya replies that both accounts may hold, so long as the distinction is drawn by timing: clicks made early in an approach are too distant to interfere and can only be advertising, whereas the dense trains delivered in the last two hundred milliseconds arrive exactly when interference would do most harm. Trials in 2019 found that inexperienced bats avoided clicking moths whatever their toxicity, which suits the warning account, while trained bats were disrupted only by the late trains.

FWhat emerges is less tidy than the phrase about seeing with the ears. Echolocation is best described not as a substitute for vision but as a short-range correction applied to information the animal has gathered in other ways, and its weight in that mixture shifts with the habitat: high over open water, where nothing is remembered and nothing obstructs, and much lower along a hedgerow flown every night for a season. Neither researcher disputes the description in outline. They disagree about which case is typical, and the disagreement persists because the two of them work in the two habitats that give opposite answers, so each is generalising from the setting that suits his or her own claim. Tags light enough for a forest bat would settle a good deal of it. Nobody, so far, has built one.