Shelley A. Kick

Target detection by the echolocating bat, Eptesicus fuscus

SummaryThe long-range echo-detection capabilities of echolocating bats (Eptesicus fuscus) were studied in a two-choice psychophysical procedure.E. fuscus can detect 4.8 mm diameter spheres at a distance of 2.9 m, and 19.1 mm diameter spheres at a distance of 5.1 m. The threshold of echo-detection corresponds to the distance at which a target returns an echo amplitude in the region of 0 dB SPL. The results demonstrate that the maximum effective range of bat sonar is greater than previously indicated by obstacleavoidance and target-interception tasks.

Acuity of horizontal angle discrimination by the echolocating bat,Eptesicus fuscus

SummaryEcholocating bats of the speciesEptesicus fuscus were trained to discriminate between two arrays of vertical rods differing in the size of the angle in the horizontal plane separating the rods in each array. With both two-rod arrays and fiverod arrays the bats were able to distinguish angular differences as small as 1.5 °(75% correct-response threshold). The similarity of the results obtained with both array sizes shows that the angular separation of adjacent rods, and not the width of the array as a whole in the case of five-rod arrays, is the source of cues for the discrimination. Acoustic control experiments were carried out to determine whether echoes reflecting laterally from one rod to another contributed cues to the bat. The results indicate that discrimination was based on the bats perception of the horizontal angular position of each rod relative to its neighbors, and not upon acoustic interactions among the rods.The acuity of horizontal-angle discrimination measured at 1.5 ° may represent the acuity with which the bats sonar system processes echoes to display one targets azimuth relative to the azimuth of another target. The results thus may represent the acuity of discrimination of target location by the binaural mechanisms of localization of sound. Although bats are considered to have heads that are too small for producing adequate interaural arrival-time cues for sound localization, the extremely broad bandwidth of FM echolocation signals and echoes may compensate for small interaural separation in bats. It is, therefore, most provocative that the observed performance ofEptesicus fuscus at horizontal-angle discrimination matches the performance to be expected if the 0.5 μs acuity of perception of the time-of-occurrence of sonar echoes measured in echo-jitter experiments can be applied to binaural echo arrivaltime perception. It cannot be ruled out that the FM bat compares the arrival time of echoes at the two ears and reconstructs in some manner the relative phase of these sounds to determine target direction in the horizontal plane.

Echolocation and hearing in the mouse-tailed bat,Rhinopoma hardwickei: acoustic evolution of echolocation in bats

Summary1.Mouse-tailed bats (Rhinopoma hardwickei) use short-duration, multiple-harmonic, constant-frequency (CF) or slightly frequency-modulated (FM) signals for echolocation of flying insect prey. The frequency components are 18–20 kHz (first harmonic), 36–40 kHz (second harmonic), 56–60 kHz (third harmonic), and 75–80 kHz (fourth harmonic). The second harmonic is the strongest component, with the third harmonic 2 to 10 dB weaker, and the first and fourth harmonics about 10 to 20 dB weaker than the second.2.The bats hearing, as indicated by N4 auditory evoked potentials, is moderately sharply tuned to the second harmonic, broadly sensitive to the first harmonic and to lower frequencies, and moderately sensitive to the third and fourth harmonics. The degree of tuning is sufficient to indicate some specialized function for the second harmonic, perhaps in the task of detecting targets at maximum range, since this is the lowest of the strong harmonics and least affected by atmospheric attenuation. These bats probably do not perform the Doppler compensation response even though their hearing is tuned.3.The pattern of emission of sonar sounds (repetition rate, duration) during interception of prey is similar to that observed in most other species of bats. The signals are emitted at repetition rates of 10–20/s during the search stage, 20–40/s during the approach stage, and about 100/s during the terminal stage of the pursuit process. The sonar sounds emitted during the search stage are 6–10 ms long, during the approach stage they are 2–4 ms long, and during the terminal stage they are less than 1 ms long.4.The principal acoustic dimension of echolocation sounds that relates to perception of targets is signal bandwidth, whichRhinopoma hardwickei manipulates throughout the pursuit process by shortening the duration of its sonar sounds and also by slightly broadening the FM sweeps in its terminal-stage sounds. The onset-time of the long, search-stage sounds and the longer, early approach-stage sounds is more abrupt than for the shorter, late approach-stage and the short terminal-stage sounds. This apparently deliberate transient beginning broadens the signals bandwidth at the onset relative to the narrower bandwidth prevailing for the rest of the CF harmonic structure which follows. In the shorter sounds, which are emitted during the late approach stage and terminal stage, the onset is more gradual, but the duration (the envelope) and the offset-time are now short enough that the bandwidth of the signal as a whole is increased. The bandwidth is manipulated primarily by these changes in signal envelope and secondarily by the increased FM sweep-width in the terminal stage.5.Except for duration these signals are relatively inflexible and suggestive of a primitive kind of echolocation in which only one dimension is changed to achieve qualities which most other species of bats obtain by changing a variety of signal dimensions simultaneously. The abrupt signal onsets may be an indication that multiple-harmonic CF and FM echolocation sounds evolved from click-like echolocation sounds of the type emitted byRousettus and some terrestrial mammals.

Interception of Flying Insects by Bats

Many species of insectivorous bats (Chiroptera) use echolocation to find their prey and to avoid obstacles to their flight (Griffin 1958). An echolocating bat emits sonar sounds in the frequency range from 10 to over 100 kHz and perceives objects from the echoes of these sounds returning to its ears (Novick 1977, Schnitzler and Henson 1980). In the process of avoiding collisions with obstacles or capturing flying insects, the bat must gather various different kinds of information about targets, and the nature of the sonar signals which the bat emits at different stages in these behaviors reflects the kinds of information that the bat is interested in gathering (Simmons et al. 1975, 1979, Simmons and Stein 1980). For example, the sonar sounds emitted by a bat searching for a flying insect are often quite different from the sounds emitted to determine the location or identity of a target that already has been detected.

Localization with Biosonar Signals in Bats

Bats use a kind of biological sonar, called echolocation, for perception of objects in the environment. They belong to the mammalian order Chiroptera (“wing-handed”), and echolocating bats comprise the suborder Microchiroptera (“little” bats), which contains about 700 living species. About 550 of these species are insectivorous, capturing their prey in flight or on such surfaces as vegetation and the ground. About 150 other species have more diverse habits, feeding on small mammals, birds, frogs, lizards, fish, the blood of larger animals, or on fruit and pollen-and-nectar. Insectivorous bats generally use their sonar to detect, identify and track their prey to a successful capture, and they use it also for such tasks as avoiding obstacles to flight. To some extent insectivorous bats probably also hear the sounds made by their prey. Bats with other habits probably use a mixture of sonar, passive hearing, vision, and olfaction to find food (Griffin, 1958; Novick, 1977; Schnitzler and Henson, 1980).

Neural Mechanisms for Target Ranging in an Echolocating Bat Eptesicus fuscus

We have concluded a series of behavioral, anatomical and physiological experiments in order to gain a better understanding of the neural basis of target ranging in Eptesicus fuscus. Behaviorally, we were interested in learning whether or not binaural integration is essential for ranging. We trained two bats to perform range discrimination using two-choice paradigms (Simmons, 1973) until they reached a criterion of 85% correct responses. We then observed their performance under intact conditions and when one ear was occluded. Occlusion was effected by inserting silicone grease to fill the ear canal of the selected ear. The performance of these two bats under monaural and binaural conditions is shown in Table 1. We found that the performance under monaural conditions was not significantly different from that of intact conditions (p<0.05, Hoel test for the difference of two means). Thus, bilateral interactions, which are essential for encoding directional information about the target, are not required for ranging in this species. Instead, time or intensity information from target echoes is more important for ranging.

Echolocation and hearing in the mouse‐tailed bat (Rhinopoma hardwickei)

Rhinopoma produces four‐harmonic, constant frequency orientation sounds. The 7‐ to 8‐ms‐long search and 2‐ to 4‐ms‐long approach signals rise to full amplitude in two to three waves of the fundamental (18 kHz), producing an onset transient. This click‐CF signal appears unique to Rhinopoma and may link the complex orientation sounds of the Microchiroptera with the sonar clicks of shrews. The N4 audiogram shows sensitivity peaks corresponding to the harmonic peaks of the signal spectrum. Sensitivity is particularly great for the second harmonic, the strongest component of the signal. This sharp tuning around the second harmonic suggests that primary auditory neurons may have Q‐10 dB values as high as 20 to 50. The sharp roll‐off on the low frequency side of this sensitivity peak further suggests Rhinopoma may perform the Doppler‐compensation response.

Time‐domain processing of target range information by central auditory neurons in the echolocating bat, Eptesicus fuscus

Echolocating bats (Eptesicus fuscus) perceive either monaurally or binaurally the range of sonar targets with an acuity of about 1 cm, or a time delay acuity of about 60 μsec. Single‐unit recordings from cortical and midbrain auditory neurons in unanesthetized bats indicate response properties to stimuli consisting of pairs of simulated echolocation signals (three‐harmonic descending FM sweeps with energy from 23 to 100 kHz) with naturally occurring transmission‐echo time delays and intensity differences. While many neurons respond to both transmissions and echoes given sufficient intensities and long echo delays, some neurons respond exclusively to the pair when the time delay is within a narrow range. Their temporal response selectivity increases with decreasing overall absolute transmission‐echo intensity but is independent of intensity differences between transmissions and echoes. These results demonstrate “tuning” of central auditory neurons to specific echo delays, providing the neural basis for tar...