Constant Frequency-frequency Research Articles

Jamming avoidance reveals segregation of processing for detection and ranging in echolocating bats

Big brown bats emit frequency-modulated (FM) biosonar sounds that sweep from 50 to 23 kHz in the first harmonic. When searching for targets in the open or when detecting targets in wideband noise, bats lengthen their sounds and allow the FM sweep to tail off into a nearly constant-frequency (CF) segment, which boosts the energy in the sound over a narrow band of 24–28 kHz. When detecting a target in CF jamming sounds, bats shift their emitted frequencies up or down to move this terminal CF component from the interference when the CF frequency is 23–27 kHz. Other researchers have found a population of neurons in the bats inferior colliculus that is sharply tuned (Q10dB=30–40) to frequencies of 24–30 kHz and may be used to detect echoes. However, these same frequencies also contribute incrementally to determining echo-delay accuracy for target ranging, and a parallel population of neurons has only moderately sharp tuning (Q10dB <5–10). Sharper tuning in the cochlea might facilitate detection but could compromise time constants for reception, so these bats use neuronal inhibition in the inferior colliculus to create two parallel receivers at frequencies of 24–30 kHz. [Work supported by ONR and NIH.]

Evidence for a mechanical filter in the cochlea of the ‘constant frequency’ bats, Rhinolophus rouxi and Pteronotus parnellii

To investigate the function of basilar membrane (BM) thickenings in the cochlea of bats which use constant frequency (CF) echolocation calls, acoustic distortion products were measured while placing the primary stimuli f 1 and f 2, at frequencies which are represented in the thickened BM regions. In Rhinolophus, for primary stimuli between about 80–100 kHz, pronounced maxima of the level of distortion products (2/ f 1– f 2, 3 f 1–2 f 2, 4 f 1–3 f 2) can be measured if the frequency separation between the two primary tones is chosen so that the resulting distortion frequency matches the dominant CF frequency (resting frequency, RF). The distortion maxima extend from the individual RF down to frequencies which are 2–4 kHz lower. The data indicate that the thickened BM region in the basal halfturn of the cochlea strongly oscillates at the bats' RF and slightly below. The hearing threshold, however, is at a maximum at the RF (see Kössl 1993). Therefore, the RF oscillations are thought to be involved in absorbing the respective frequency such that the more apically located frequency place of the RF is insensitive and a mechanical notch filter is established. In Pteronotus, there are maxima of the level of 2 f 1– f 2 at distortion frequencies around the RF of about 61.5 kHz up to the frequency of a strong stimulus-frequency otoacoustic emission (SFOAE) which is a few hundred Hz higher. Pronounced distortions in the RF range can only be elicited when the stimulus frequencies are between about 62 to 72 kHz. Similar to the situation in Rhinolophus, this frequency band is represented on a stretch of thickened BM.

Echolocation, flight morphology and foraging strategies of some West African hipposiderid bats

We studied echolocation call structure, flight morphology and feeding behaviour of three hipposiderid bats (Asellia tridens, Hipposideros caffer and H. ruber: Chiroptera: Hipposideridae) in The Gambia during the wet season (July‐August). All three species emitted brief CF/FM echolocation calls. In A. tridens CF (constant frequency) frequencies between 108 and 122 kHz were recorded. This variation was caused mainly by sex and age difference in call frequencies: juveniles used lower frequencies than did adults, and males were lower in frequency than females. Among adults, CF frequency was related to forearm length in a polynomial manner. Asellia tridens is unusual for a microchiropteran in that males are larger than females: nevertheless this species follows the typical trend for bats using CF components in their calls in that males call at lower frequencies than do females. The spread in frequencies noted for A. tridens in group flight was not caused primarily by individual shifting their frequencies when flying in a group compared with when flying alone. Three bats were flown separately and then together in a small room. The hypothesis that bats shift emitted frequencies in group flight to minimize confusing their own echoes with those from conspecifies was not supported. Rather, we suggest each bat has a personal CF frequency determined by sex, age and size, and that the variation created by these factors reduces confusion with other bats' echoes during group flight. Bats of the genus Hipposideros were separated into two groups by discriminant analysis on flight morphology. The two groups showed little overlap of CF frequencies used in echolocation. Bats identified as H. ruber called with CF resting frequencies between 121 and 136 kHz, while H. caffer used 128–153 kHz. Our results support the classification of these bats into two sibling species. We compared the flight performance of A. tridens and H. ruber by flying bats through obstacle courses: H. ruber had a lower wing loading and was able to negotiate more complex arrays of obstacles than could A. tridens. In the wild, H. ruber fed in more cluttered situations than did A. tridens. Both species were feeding mainly on Coleoptera during the study period, and in H. ruber individuals with higher aspect ratios and longer wingspans tended to eat more months than did bats with broader and shorter wings.

Biosonar signals and cerebellar auditory neurons of the mustached bat.

In the vermis (VIp, VIIa, VIIp, and VIII), crus, and paraflocculus of unanesthetized mustached bats Pteronotus parnellii parnellii, responses of single neurons to acoustic stimuli were studied. The stimuli delivered were constant-frequency (CF) tones, frequency-modulated (FM) sounds, noise bursts (NBs), and sounds similar to the orientation sounds (pulses) of the species and echoes. The effect of ablation of the cerebellar cortex on vocalization was also investigated to explore whether the cerebellum was involved in sound emission. In the cerebellum of the mustached bat, auditory neurons are predominantly tuned to frequencies within the bands between 23 and 30, 55 and 63, or 85 and 94 kHz, which are found in the first, second, and third harmonics of bat's biosonar signals, respectively. The first harmonic is represented in the paraflocculus. The second harmonic is represented in vermis VIp and VIIa and crus I and IIa. The third harmonic is mainly represented in vermis VIIp and crus IIp. Different lobules represent different frequencies, but there is no systematic tonotopic representation in each lobule. The resting frequency of the CF component of the second harmonic (CF2) of the pulse differs among bats. The majority of auditory neurons in vermis VIp and VIIa and crus IIa are tuned to the CF2 frequency of the bat's own pulse. The frequency-tuning curves of cerebellar neurons are broader than those of peripheral neurons, reflected in significantly lower quality factors of Q-10, -30, and -50 dBs. In vermis VIp and VIIa, there are tiny clusters of FM-FM and CF/CF combination-sensitive neurons. They show strong facilitation of responses when two FM or CF sounds are delivered with particular relationships in the frequency, amplitude, and time domains. Because the clusters of these combination-sensitive neurons in the cerebellum are so small, we found no sign of a systematic representation of certain acoustic parameters, unlike that found in the auditory cortex. In vermis VIp and VIIa, there is a large cluster of NB-sensitive neurons that are more sensitive to NBs than to CF tones. The wider the bandwidth of the NBs, the better are the responses of these NB-sensitive neurons. The ablation of the vermis (VIp, VIIa, and VIIp), crus, and paraflocculus increases the variation of the CF frequency of the pulse. The ablation of the crus and paraflocculus causes a clear increase in the variation of CF frequency. The ablation of vermis (VIp, VIIa, and VIIp) has only a small effect on the variation. Any of the above ablations has little effect on the repetition rate of the pulse emission and the duration of pulses.(ABSTRACT TRUNCATED AT 400 WORDS)

Echolocation calls of emballonurid bats from Panama

Saccopteryx bilineata, Cormura brevirostris and two other species tentatively identified as emballonurid bats, were recorded in the field in Panama. The search and approach phase calls of all the species began with a rise in frequency followed by a constant frequency (CF) portion and a terminal drop in frequency.S. bilineata and one of the unknown species produced pairs of calls, the first 2–3 kHz lower in CF frequency than the second. This has been suggested as associated with Doppler shifts of echoes caused by the bats' flight and bats' beaming calls towards different targets. However, a second unidentified species produced call triplets with the total difference in CF frequencies too great to be explained by this hypothesis. Increased bandwidth and integration of echo information, or social communication may be alternative functions. Terminal phase calls of three of the species were typical frequency modulated signals.C. brevirostris terminal calls, however, were CF in nature, a form not believed to be suitable for tasks requiring high resolution such as prey pursuit. The characteristics of these bats' signals are more complex than previously suggested and bring into question their supposed ‘primitive’ nature.