Auditory specialisations for Doppler shift analysis

All echolocating bats need to be able to analyse sound frequencies accurately. With FM signals, fine frequency analysis yields accurate range measurement and detailed description of the target, and with CF signals it yields an accurate measure of the Doppler shift. Frequency analysis is certainly excellent in bats that use only FM signals, such as Myotis, but their abilities are not so very different from those of other mammals. However, in bats that use long CF signals, such as Rhinolophus, frequency resolution is quite exceptional and greatly exceeds the abilities of non-echolocating mammals.

Specialisations for fine-frequency analysis begin at the inner ear, with the basilar membrane, which is the accessory structure that couples the sound stimulus to the receptor cells (hair cells) in the cochlea (see Fig. 6.10). Mammalian hair cells are tuned to particular frequencies according to their position on the basilar membrane, typically with equal lengths of the membrane being devoted to each octave. But in the greater horseshoe bat, the representation of frequencies from 80 to 86 kHz is greatly expanded, and the greatest expansion is found at the reference frequency of 83 kHz. This expanded representation on the basilar membrane is reflected in the number of first-order neurons that innervate the hair cells. Compared to frequencies below 70 kHz, which are not involved in echolocation, frequencies of the expanded region are over-represented approximately 10 times, with the result that 21 per cent of all first-order auditory neurons represent the frequency range from 80 to 86 kHz (Fig. 6.14a). By analogy with the visual system, this expanded representation is termed an acoustic fovea.

Figure 6.14 The acoustic fovea in the greater horseshoe bat (Rhinolophus ferrumequinum). (a) A histogram of frequency representation among firstorder auditory neurons, showing the great over-representation of frequencies around 83 kHz. (b) The sharpness of tuning, expressed as the Q10 dB value, for neurons with different best frequencies in the cochlear nucleus, showing the exceptional sharpness of tuning in neurons with best frequencies around 83 kHz. See also Fig. 6.11c. (a redrawn after Bruns & Schmieszek, 1980; b redrawn after Suga, Neuweiler & Moller, 1976.)

Figure 6.14 The acoustic fovea in the greater horseshoe bat (Rhinolophus ferrumequinum). (a) A histogram of frequency representation among firstorder auditory neurons, showing the great over-representation of frequencies around 83 kHz. (b) The sharpness of tuning, expressed as the Q10 dB value, for neurons with different best frequencies in the cochlear nucleus, showing the exceptional sharpness of tuning in neurons with best frequencies around 83 kHz. See also Fig. 6.11c. (a redrawn after Bruns & Schmieszek, 1980; b redrawn after Suga, Neuweiler & Moller, 1976.)

The fovea is almost certainly related to structural specialisations in the basal part of the basilar membrane, where the highest frequencies are represented. These structural peculiarities abruptly disappear at a distance of 4.5 mm from the oval window, and the expanded frequency region is located immediately beyond this critical point. As a result, the basilar membrane acts as a mechanical filter, which tunes a disproportionate length of the membrane to a narrow frequency band (Vater, Feng & Betz, 1985). A consequence of this arrangement is that the first-order neurons innervating the hair cells within the foveal region are each extremely sharply tuned. A measure of the sharpness of tuning is provided by the Q10 dB value, which is the neuron's best frequency divided by the bandwidth of its threshold curve 10 dB above the minimum threshold. Very high Q10 dB values of between 50 and 200 are found in the foveal region and, in the region of greatest expansion around 83 kHz, even values of over 400 are found (Fig. 6.14b). For frequencies below 70 kHz, the Q10 dB values fall below 20, which is within the range found in other mammals.

This over-representation and sharp tuning are conserved at all higher levels of the auditory pathway. During echolocation, Doppler-shift compensation serves to clamp the echo of the CF component within this expanded frequency range. In effect, this behavioural response creates a constant carrier frequency, on which the small frequency modulations produced by the wing beats of flying insects are superimposed, so enabling them to be analysed by the sharply tuned neurons. A similar combination of peripheral tuning and Doppler-shift compensation has evolved independently of the horseshoe bats in the moustached bat and this combination therefore probably represents a general strategy for Doppler shift analysis.

The kinds of echo modulation produced by flying insects have been examined by the simple expedient of echolocating them with a simulated horseshoe bat, consisting of a loudspeaker broadcasting a pure tone of 80 kHz and a microphone. These tests show that the fluttering wings of an insect produce a strong echo or acoustic glint only when they are approximately perpendicular to the impinging sound waves, which happens for a short moment in each wing-beat cycle, but there are no acoustic glints from non-flying insects. A glint consists of a momentary increase in echo amplitude and a concomitant broadening of echo frequency, which represents Doppler shifts caused by the movement of the wings with respect to the sound source.

In echoes returning from a flying insect, glints modulate the echo at a rate corresponding to the wing-beat frequency of the insect, and this is termed the modulation frequency. The extent of frequency modulation involved is termed the modulation depth and is generally between 1 and 2 kHz above or below the carrier frequency. By perceiving these glints, a horseshoe bat should be able to distinguish with certainty between echoes from a fluttering insect and echoes from inanimate objects. Neurons specialised to encode the acoustic glints produced by flying insects are found among those processing the CF echo (foveal) frequencies in all the major staging posts of the auditory pathway.

At the level of the inferior colliculus, the over-representation of the CF echo frequencies is actually enhanced, being approximately 24 times that for frequencies below 70 kHz. Consequently, the normal tonotopic arrangement is distorted by the substantial block of interneurons devoted to the CF echo frequency range. Within this block, the neurons are very sharply tuned and the great majority are extremely sensitive to small frequency modulations. When stimulated with sinusoidal frequency modulations that sweep as little as ± 10 Hz around the 83 kHz carrier frequency, these neurons respond with discharges that are phase-locked to the frequency modulation. The response remains phase-locked at all modulation frequencies up to about 500 Hz, but some neurons show a preference for rates between 20 and 100 Hz. The ability of these collicular neurons to follow the complex modulations produced by real wing beats is confirmed by using the recorded echoes from a flying moth as stimuli in the experiments. With this natural stimulus, the neurons reliably encode the wing-beat frequency of the moth, as well as more subtle features of the echo (Pollak & Schuller, 1981; Schuller, 1984).

The encoded information is passed on to the tonotopic area of the auditory cortex, within which neurons devoted to the CF echo frequencies form a disproportionately large block, often termed the CF area. Neurons in the CF area respond to sinusoidal frequency modulations in much the same way as the collicular neurons but are more selective in the range of modulation to which their response is phase-locked. Most of them prefer modulation depths around ± 1 kHz and modulation frequencies of 100 Hz or less, with a strong preference for frequencies between 40 and 70 Hz. Among the nocturnal moths that are potential prey for horseshoe bats, many species have wing-beat frequencies around 40 to 60 Hz, and such moths produce acoustical glints with a modulation depth of about 1 kHz. The phase-locking of the cortical neurons is thus tuned to a behaviourally relevant range, in contrast to the wide range of responses found in the collicular neurons.

Behavioural observations confirm that horseshoe bats do in fact detect their natural prey by means of these modulations of the CF echo. Newly caged horseshoe bats only pursue insects that are beating their wings and ignore stationary insects or those walking on the sides of the cage. The caged bats will take dead, tethered insects when these are associated with an artificial wing-beat simulator placed nearby. The crucial role of the CF

signal is also shown by observations on horseshoe bats foraging in flycatcher style in natural forests: while hanging on twigs, they scan the surrounding area for flying insects and take off on a catching flight only after the prey has been detected. During this stationary scanning for prey, the bats emit long CF pulses without any FM component, indicating that prey detection is based on the CF signal alone (Link, Marimuthu & Neuweiler, 1986; Neuweiler et al., 1987).

Essentials of Human Physiology

Essentials of Human Physiology

This ebook provides an introductory explanation of the workings of the human body, with an effort to draw connections between the body systems and explain their interdependencies. A framework for the book is homeostasis and how the body maintains balance within each system. This is intended as a first introduction to physiology for a college-level course.

Get My Free Ebook


Post a comment