Introduction

In 1793, Lazzaro Spallanzani, an Italian Catholic priest, biologist and physiologist, demonstrated that bats were able to avoid obstacles without the aid of vision. He stretched thin wires with small bells attached across a completely darkened room and observed that bats were able to fully navigate between them without causing the bells to ring. Blinding the bats also did not impair their ability to manoeuvre around them. Meanwhile, a Swiss zoologist, Charles Jurine, revealed that blocking one of the ears of a bat spoiled its navigating abilities, a finding that Spallanzani then pursued. A series of experiments which involved blocking the ears or gluing the muzzle closed led him to conclude that while bats did not have much use for their eyes, any interference with their ears that adversely affected hearing was disastrous, resulting in them colliding with objects they could usually avoid and being unable to forage for prey. He concluded that 'The ear of the bat serves more efficiently [than the eye] for seeing, or at least for measuring distance'. At the time, Spallanzani's findings were met by his fellow scientists with ridicule and scepticism as bats were believed incapable of producing any sound and therefore such results defied logic.

Nearly 150 years after Spallanzani's work, Donald R. Griffin, while an undergraduate at Harvard University in the 1930s, took an interest in Spallanzani's 'bat problem'. New advances in technology allowed him to use a 'sonic receiver', designed and built by Harvard physics professor George Washington Pierce. This device captured high-frequency sounds that were beyond the range of human hearing, and reduced the pitch to an audible level. For the first time, it became apparent that bats emit short, loud, ultrasonic clicking sounds. Along with a fellow student, Robert Galambos, who was an expert in auditory physiology, Griffin designed a set of further experiments which showed that bats were avoiding obstacles by hearing the echoes of their ultrasonic cries. Further experimentation revealed that bats were able to adjust the structure of their calls for prey search and capture and for collision avoidance. Griffin named this acoustic orienting behaviour 'echolocation'.

A bat's echolocation system is highly sophisticated. By emitting short high-frequency pulses of sound from their mouths or noses, bats are able to use the information contained within the echoes returned from a solid object to construct a 'sound picture' of their environment. Not only are they able to identify the size, position and speed of objects within three-dimensional space, they are also able to differentiate forms and surface textures. However, as there is no single signal form that is optimal for all purposes, bats have evolved a large number of signal types. This diversity of echolocation signals is likely to reflect adaptations to the wide range of ecological niches occupied by different bat species. For example, in the UK, the noctule, which largely forages high over parkland, pasture and woodland in an uncluttered environment, tends to produce extremely loud low- frequency calls of relatively long duration, narrow bandwidth and low repetition rate. Conversely, Bechstein's bat, which often forages very close to or within woodland vegetation in a very cluttered environment, usually produces relatively quiet, very broadband calls of short duration with a high repetition rate. Thus the calls of different bat species are shaped by the habitats in which they usually forage and the resulting different call types can often be used to separate species in the field. However, echolocation call shape is not fixed for a species and shows a certain degree of plasticity depending on the habitat within which an individual is currently located. In addition, although habitat is a significant factor determining the 'shape' of bat echolocation calls, they may also vary with sex, age and body size, geographic location and presence of conspecifics. Finally, species that occupy similar niches may use similar echolocation call types and there is often significant overlap in calls between species. An understanding of these different levels of variation both within and between individuals and species is essential to the successful use of echolocation calls for bat species identification.

Social calls produced by bats are often more structurally complex than echolocation calls used for orientation. Social calls are used to communicate with other bats, and for many species they consist of a wide variety of trills and harmonics, comparable in many respects to bird song. It is often surprising to discover just how musical these creatures sound. Social calls may have a number of functions. Some are used to defend patches of insects against other bats or to sustain territorial boundaries. Others function in attracting a mate or, in the case of distress calls, to initiate a mobbing response. Perhaps the most astounding are the isolation calls emitted by young bats, which allow their mothers to identify them. At Bracken Cave in Texas, for example, millions of Mexican free-tailed bats (Tadarida brasilensis) cluster in a large maternity colony. After the mothers have given birth the walls of the cave are literally covered with young bats packed tightly together. Each of these young bats produces an individually specific call that is in some way different from that of all the other young bats. These variations enable a returning mother to distinguish her own offspring amongst all the others.

Since Griffin's discovery, a number of techniques have been developed to allow us to listen to the ultrasonic vocalisations of bats. These range from relatively cheap 'heterodyne' detectors which convert a narrow range of frequencies into an audible signal in the field to 'real time' recording that has become possible through the development of high-speed analogue to digital converters built into or connected to computers or solid-state recorders. These high-tech devices utilise a sufficiently high sample rate to enable the ultrasound to be captured digitally without any subsequent processing and allow later processing and analysis of recordings. More recently, bat enthusiasts and researchers have been taking advantage of the explosion in the availability of smartphones. Currently, these devices can be used as recording devices when connected to a bat detector that converts the ultrasound into the audible range. However, with the very recent development of small inexpensive USB ultrasonic microphones, the possibility of turning a smartphone into a 'real time' ultrasound recording device is not far off. Incorporation of classification algorithms (a small number of which are available or under development and which assign these calls to species) into such equipment will one day provide instant identification of bat species in the field in a readily available, cost-effective hand- held device.

Donald Griffin referred to his discovery of echolocation as 'opening a magic well' from which scientists have been extracting knowledge ever since. Echolocation provides a window into the lives of bats, giving us access to a previously unknown...