CF-FM Bat Research Articles

Single-Cell Landscape of the Cochlea Revealed Cell-Type-Specific Diversification in Hipposideros armiger Based on PacBio Long-Read Sequencing

Echolocation represents one of the most rapid adaptive sensorimotor modulation behaviors observed in mammals, establishing bats as one of the most evolutionarily successful mammals. Bats rely on high-frequency hearing for survival, but our understanding of its cellular molecular basis is scattered and segmented. Herein, we constructed the first single-cell transcriptomic landscape of the cochlea in Hipposideros armiger, a CF-FM bat, using a PacBio-optimized genome and compared it with the results obtained from unoptimized original genomes. Sixteen distinct cell types were distributed across five spatial regions of the cochlea. Notably, through hematoxylin and eosin staining and fluorescence in situ hybridization, we identified new types of spiral ganglion neuron (SGN) cells in the cochlea of H. armiger. These SGN cells are likely critical for auditory perception and may have driven the adaptive evolution of high-frequency hearing in this species. Furthermore, we uncovered the differentiation relationships of among specific cell types, such as the transition from supporting cells to hair cells. Using the cochlear cell atlas as a reference, cell types susceptible to deafness-associated genes (in the human) were also identified. In summary, this study provides novel insights into the cellular and molecular mechanisms underlying the adaptive high-frequency hearing in bats and highlights potential candidate cell types and genes for therapeutic interventions in hearing loss.

Background noise responding neurons in the inferior colliculus of the CF-FM bat, Hipposideros pratti

The Lombard effect, referring to an involuntary rise in vocal intensity, is a widespread vertebrate mechanism that aims to maintain signal efficiency in response to ambient noise. Previous studies showed that the Lombard effect could be sufficiently implemented at subcortical levels and operated by continuously monitoring background noise, requiring some subcortical auditory sensitive neurons to have continuous responses to background noise. However, such neurons have not been well characterized. The inferior colliculus (IC) is a major auditory integration center under the auditory cortex and provides projections to the putative vocal pattern generator in the brainstem. Thus, it is reasonable to speculate that the IC is a likely auditory nucleus candidate having background noise responding neurons (BNR neurons). In the present study, we isolated 183 sound-sensitive IC neurons in a constant frequency-frequency modulation bat, Hipposideros pratti, and found that around 19% of these IC neurons are BNR neurons when stimulated with 70 dB SPL background white noise. Their firing rates in response to noise increased with increasing noise intensity and could be suppressed by sound stimulation. Furthermore, compared to neurons with similar best frequencies, the BNR neurons had smaller Q10-dB values and lower noise-induced minimal threshold change, indicating that BNR neurons received fewer inhibitory inputs. These results suggested that the BNR neurons are ideal candidates for collecting information about background noise. We proposed that the BNR neurons synapsed with neurons in vocal-pattern-generating networks in the brainstem and initiated the Lombard effect by a feed-forward loop.

Echolocation and flight behavior of the bat Hipposideros armiger terasensis in a structured corridor.

In this study, the echolocation and flight behaviors of the Taiwanese leaf-nosed bat (Hipposideros armiger terasensis), which uses constant-frequency (CF) biosonar signals combined with a frequency-modulated (FM) sweep, are compared with those of the big brown bat (Eptesicus fuscus), which uses FM signals alone. The CF-FM bat flew through a corridor bounded by vertical poles on either side, and the inter-pole spacing of the walls was manipulated to create different echo flow conditions. The bat's flight trajectories and echolocation behaviors across corridor conditions were analyzed. Like the big brown bat, the Taiwanese leaf-nosed bat centered its flight trajectory within the corridor when the pole spacing was the same on the two walls. However, the two species showed different flight behaviors when the pole spacing differed on the two walls. While the big brown bat deviated from the corridor center towards the wall with sparse pole spacing, the Taiwanese leaf-nosed bat did not. Further, in comparison to E. fuscus, H. a. terasensis utilized different echolocation patterns showing a prevalence of grouping sounds into clusters of three. These findings indicate that the two species' distinct sonar signal designs contribute to their differences in flight trajectories in a structured corridor.

Recovery cycles of inferior collicular neurons in the leaf-nosed bat, Hipposideros armiger

When stimulated with biologically relevant constant frequency–frequency modulation (CF–FM) sounds, the inferior collicular neurons of the CF–FM bat, Hipposideros armiger, either only discharged impulses to the onset (76%, single-on neurons) of the CF-FM sounds or to the onset of both CF and FM components of CF–FM sounds (24%, double-on neuron). Some neurons were single-on responders at low sound amplitude but become double-on responders at high sound amplitude. Single-on responders had longer latency and recovery cycle than double-on responders. While most neurons did not respond to the second sound when the paired CF-FM sounds overlapped, 3 single-on and 7 double-on neurons did such that they had “cyclic” recovery cycles with inter-pulse intervals. The different response latency and dynamic variation in the recovery cycle of these two types of neurons suggest they may serve as the neural basis underlying a bat's ability to perform echo ranging throughout different phases of hunting.

Beam-width control of echolocation pulses in Japanese CF-FM bats (Rhinolophus ferrumequinum nippon) during aerial prey hunting.

Beam-width control of echolocation pulses in Japanese CF-FM bats (Rhinolophus ferrumequinum nippon) during aerial prey hunting.

Recovery cycles of single-on and double-on neurons in the inferior colliculus of the leaf-nosed bat, Hipposideros armiger

Our previous study showed that when stimulated with constant frequency–frequency modulation (CF–FM) sounds, neurons in the central nucleus of the inferior colliculus of the CF–FM bat, Hipposideros armiger, either only discharged impulses to the onset of CF–FM sounds (76%, single-on neurons) or to the onset of both CF and FM components of CF–FM sounds (24%, double-on neuron) (Fu et al., 2010). The present paper reports the recovery cycles of these two types of neurons using paired CF, FM and CF–FM sounds as stimuli. Both types of neurons had similar recovery cycle for CF sounds but had the shortest recovery cycle for FM sounds. Whereas single-on neurons had similar recovery cycle for CF and CF–FM sounds, double-on neurons had longer recovery cycle for CF sounds than for CF–FM sounds. In addition, double-on neurons had significantly shorter recovery cycles than single-on neurons for FM and CF–FM sounds. Most neurons did not respond to the second sound when each pair of sounds overlapped. However, when stimulated with paired CF–FM sounds, 3 single-on and 7 double-on neurons discharged to the second sound even when both sounds overlapped. As such, they had “cyclic” recovery cycles that varied between maximum and minimum with inter-pulse intervals. Possible mechanisms underlying the different recovery cycles of these neurons are proposed. Possible biological significance of these neurons in relation to responding to varied pulse repetition rate during hunting is discussed.

Recovery Cycle of Inferior Collicular Neurons Determine Pulse Following Rate in CF-FM Bat*

The characteristics of recovery cycles in inferior collicular (IC) neurons of leaf-nosed bat (Hipposideros armiger) and effect of the recovery cycle on the following pulse repetition rate were studied using mimic CF-FM sound stimuli emitted by free flying bat. Recovery cycle of 93 IC neurons were obtained from IC of five bats with normal hearing. These neurons were classified into three types, i.e. long recovery (LR, 47.4%), moderate recovery (MR, 35.1%), and short recovery (SR, 17.5%), according to their inter pulse interval (IPI) (ms) of 50% recovery under two CF-FM sound stimulation condition. Each type of the neurons could also be categorized into different sub-types according to changes induced by IPI increasing such as single-IPI response area neurons, multi-IPI response area neurons, and monotonic-IPI response neurons. Mean IPIs of 50% recovery of LR, MR, and SR neurons were (64.0 ± 24.8), (19.6 ± 5.8), and (7.1 ± 2.4) ms, respectively (P 0.001). The calculated theoretically following pulse repetition rate (pulse per second, Hz) of LR, MR, and SR neurons by mean IPI of 50% recovery for each type were (18.2 ± 7.0), (55.4 ± 15.7), and (171.3 ± 102.9) Hz, respectively (P 0.001). These three types of IC neurons were well corresponding to their three hunting phases, search, approach, and catch phases. The sub-types of single-IPI response area neurons and multi-IPI response area neurons had hunting phase selectivity, and sub-type of monotonic-IPI response neurons had well sensitivity to IPI change, but their hunting phase selectivity was of a sort. These results demonstrated that recovery cycle of IC neurons determined the ability to follow pulse repetition rate and matched this bat's echolocation behavior.

The use of Doppler in Sonar-based mobile robot navigation: inspirations from biology

So-called CF–FM bats are highly mobile creatures who emit long calls in which much of the energy is concentrated in a single frequency. These bats face sensor interpretation problems very similar to those of mobile robots provided with ultrasonic sensors, while navigating in cluttered environments. This paper presents biologically-inspired engineering on the use of narrowband Sonar in mobile robotics. It replicates, using robotics as a modelling medium, methods CF–FM bats use to exploit Doppler-shifts––a rich source of information not used by commercial robotic ultrasonic range sensors––in different tasks. The experimental platform for the work is RoBat, a 6 DOF biomimetic sonarhead mounted on a commercial 3 DOF mobile platform. The platform is provided with signal processing capabilities inspired by the bat's auditory system. The CF–FM bat modifies––increasing or decreasing––the carrier frequency of its own calls, compensating the Doppler-shift produced when the bat, the reflector or both are moving. This echolocating behaviour, called Doppler-shift compensation, is successfully implemented in RoBat. Inspired by this behaviour, a convoy navigation controller following a set of simple Doppler-dependent rules is successfully devised. The performance of the controller is satisfactory despite low Doppler-shift resolution caused by the lower velocity of the robot when compared to real bats. Finally, Müller's hypothesis on the use of acoustic flow by CF–FM bats for obstacle avoidance is also implemented in RoBat, resulting in a crude estimation of the target's passing distance at small bearing angles, which improves as the angle increases, nevertheless sufficing for avoiding the two reflectors of the experiment.

Auditory properties of the superior colliculus in the horseshoe bat, Rhinolophus rouxi

Auditory response properties were studied in the superior colliculus (SC) of the echolocating horseshoe bat Rhinolophus rouxi, a long CF-FM bat, by the use of stationary, dichotic stimuli. The most striking finding in the horseshoe bat was an enormous overrepresentation of neurons with best frequencies in the range of the constant frequency component of the species specific echolocation call (72% of the auditory neurons). These neurons had response thresholds as low as 0 dB SPL and were narrowly tuned with Q10dB--values up to 400, just as in the nuclei of the primary auditory pathway in this species. This overrepresentation may suggest the importance of the superior colliculus in the context of echolocation behavior. While noise stimuli were not particularly effective, other auditory response properties were similar to those described in other mammals. 65% of the SC neurons in the horseshoe bat responded only to monaural stimulation of one ear, primarily the contralateral one. 32% of the neurons received monaural input from both ears. The proportion of neurons responsive to ipsilateral stimulation (41%) was rather high. Mean response latency was 8.9 ms for contralateral stimulation. A tonotopic organization is lacking, but high-frequency neurons are less frequent in rostral SC.

The influence of the auditory cortex on acoustically evoked cerebellar responses in the CF-FM bat, Rhinolophus pearsonic chinesis

1. Acoustically evoked responses of 284 neurons isolated from the cerebellar vermis, hemispheres and paraflocculus of Rhinolophus pearsonic chinesis were studied under free field acoustic stimulation conditions. 2. The BFs of these cerebellar auditory neurons ranged from 24 to 76 kHz but they mostly fall either between 48 and 64 kHz or between 65 and 76 kHz. However, the BF distribution varies among vermal, hemispheric and parafloccular neurons. 3. Threshold curves of cerebellar neurons are generally broad but those tuned to the frequency of the predominant CF component are extremely narrow. 4. Response latencies of cerebellar neurons ranged from 2 to 48 ms suggesting multiple auditory cerebellar pathways. The latency distribution also varies among vermal, hemispheric and parafloccular neurons. 5. Although both the vermis and hemispheres contain a disproportionate number of 65-74 kHz neurons, the response latencies of those neurons isolated from the vermis are scattered over a wide range of 2.2-28 ms while those neurons isolated from the hemispheres are generally stabilized between 5 and 12 ms. 6. Electrical stimulation of the auditory cortex evokes discharges from a recorded cerebellar auditory neuron. Cortical stimulation also facilitates the response of an acoustically evoked cerebellar neuron by increasing its number of impulses. The degree of facilitation is dependent upon the amplitude of the acoustic stimulus. 7. For a given electrical and acoustic stimulation condition, the facilitative latency and the degree of facilitation varied with the interstimulus interval. Among 23 neurons studied, most of them (19 neurons, 82.6%) had a maximal facilitative latency between 2 and 10 ms. 8. By examining the difference in the facilitative effect in each isolated cerebellar auditory neuron before and after a topical application of local anesthetic, procaine, onto the point of electrical stimulation in the auditory cortex, we found that the facilitative pathways to vermal and hemispheric neurons may be different from the pathway to parafloccular neurons. 9. Possible auditory pathways to different parts of the cerebellum are discussed in relation to the wide range of recorded response latencies. 10. The facilitative influence of the auditory cortex on the cerebellar auditory neurons is assumed to enhance the cerebellar role in acoustic motor orientation.

Ontogenesis of tonotopy in inferior colliculus of a hipposiderid bat reveals postnatal shift in frequency-place code.

The postnatal development of midbrain tonotopy was investigated in the inferior colliculus (IC) of the south Indian CF-FM bat Hipposideros speoris. The developmental progress of the three-dimensional frequency representation was determined by systematic stereotaxic recordings of multiunit clusters from the 1st up to the 7th postnatal week. Additional developmental measures included the tuning characteristics of single units (Figs. 3f; 4f; 5f), the analysis of the vocalised pulse repertoire (Figs. 3e, 4e, 5e), and morphometric reconstructions of the brains of all experimental animals (Fig. 1). The maturation of auditory processing could be divided into two distinct, possibly overlapping developmental periods: First, up to the 5th week, the orderly tonotopy in the IC developed, beginning with the low frequency representation and progressively adding the high frequency representation. With regard to the topology of isofrequency sheets within the IC, maturation progresses from dorsolateral to ventromedial (Figs. 3c, 4c). At the end of this phase the entire IC becomes specialised for narrowly tuned and sensitive frequency processing. This includes the establishment of the 'auditory fovea', i.e. the extensive spatial representation of a narrow band of behaviorally relevant frequencies in the ventromedial part of the IC. In the 5th postnatal week the auditory fovea is concerned with frequencies from 100-118 kHz (Fig. 4c, d). During subsequent development, the frequency tuning of the auditory fovea increases by 20-25 kHz and finally attains the adult range of ca. 125-140 kHz. During this process, neither the bandwidth of the auditory fovea (15-20 kHz) nor the absolute sensitivity of its units (ca. 50 dB SPL) were changed. Further maturation occurred at the single unit level: the sharpness of frequency tuning increased from the 5th to the 7th postnatal weeks (Q-10-dB-values up to 30-60), and upper thresholds emerged (Figs. 4f, 5f). Although in the adult the frequency of the auditory fovea matches that of the vocalised pulses, none of the juvenile bats tested from the 5th to the 7th weeks showed such a frequency match between vocalisation and audition (Figs. 4e, 5e). The results show that postnatal maturation of audition in hipposiderid bats cannot be described by a model based on a single developmental parameter.

Auditory self-stimulation by vocalization in the CF-FM bat,Rhinolophus rouxi

1. Cochlear microphonic potentials (CM) were recorded from the cochlear aqueduct ofRhinolophus rouxi in response to vocalizations and artificial stimuli. 2. Vocal self-stimulation was evaluated by comparing the amplitude of the vocalization-induced CM to the calibration curve of the CM determined with artificial stimuli. The equivalent sound pressure level for self-stimulation at the resting frequency of the bat ranged between 78 and 83 dB SPL. 3. The same value of self-stimulation was obtained when we measured the maximum amplitude of beats produced by interference between acoustic self-stimulation and a slightly frequency shifted (500 Hz) playback signal. 4. Middle ear muscle (MEM) contractions occurred in synchrony with vocalizations even though the vocal self-stimulation level was below the threshold for the acoustic MEM reflex. The vocalization-induced MEM contractions had little effect on energy transfer to the inner ear for frequencies around the resting frequency but led to a reduction of the response at lower frequencies. 5. The importance of the vocal self-stimulation for different behavioural conditions of echolocation in CF-FM bats is discussed.

Natural ultrasonic echoes from wing beating insects are encoded by collicular neurons in the CF-FM bat,Rhinolophus ferrumequinum

1. Acoustic reflections from a wing beating moth to an 80 kHz ultrasonic signal were recorded from six different incident angles and analyzed in spectral and time domains. The recorded echoes as well as independent components of amplitude and frequency modulations of the echoes were employed as acoustic stimuli during single unit studies. 2. The responses of single inferior colliculus neurons to these stimuli were recorded from four horseshoe bats,Rhinolophus ferrumequinum, a species which uses a long constant frequency (CF) sound with a final frequency modulated (FM) sweep during echolocation. All neurons responding to wing beat echoes reliably encoded the fundamental wing beat frequency as well as the more refined frequency and amplitude modulations. 3. These neurons may provide the bat a neural mechanism to detect periodically moving targets against a cluttered background and also to discriminate various insect species on the basis of their wing beat patterns.

Responses of cerebellar neurons of the CF-FM bat, Pteronotus parnellii to acoustic stimuli

Single units (125) which faithfully discharged action potentials to acoustic stimuli (35 ms in duration with 0.5 ms rise and decay times) were recorded in the cerebellar vermis and hemispheres of the CF-FM bat, Pteronotus parnellii. These units had response latencies between 1.5 and 27 ms and minimum thresholds between 2 and 83.5 dB SPL. Best frequencies (BFs) of these units ranged from 30.32 to 79.28 kHz, but more than half (64 units, 51.2%) were between 59.73 and 63.32 kHz. While most tuning curves of these units were either broad or irregular, those curves with BFs tuned at around 61 kHz which is the frequency of the predominant CF component of the bat's echolocation signals were extremely narrow with Q-10-dB values as high as 153. Those units (29) with BFs tuned near the 61 kHz also showed off-responses. These data indicate that auditory specialization for processing of species-specific orientation signals also exists in the cerebellum of this bat.

Analysis of orientation signals emitted by the CF-FM bat,Pteronotus p. parnellii and the FM bat,Eptesicus fuscus during avoidance of moving and stationary obstacles

1. The echolocative skills ofPteronotus parnellii parnellii andEptesicus fuscus were studied by measuring their ability in avoiding stationary and moving obstacles. 2. The frequency, repetition rate, duration and amplitude of the orientation signals emitted by the bat during three phases of negotiation of obstacles were studied. 3. During the avoidance of obstacles, both species of bats systematically shorten the duration, increase the repetition rate, and decrease the amplitude of their emitted orientation signals as they approach, negotiate and pass the obstacles. 4. Flying at different speeds inside the flight room during the avoidance of obstacles,Pteronotus parnellii parnellii appropriately adjust their emitted CF frequency according to the flight speed to accurately compensate for the positively Doppler-shifted echoes. 5. The frequency of the FM signals emitted at high repetition rate byEptesicus fuscus shifts downward as the bat enters into the final phase of negotiation of the obstacles. 6. The significance of the change in parameters of emitted signals in relation to echolocation is discussed. Presumably, the increase in the repetition rate of sound emission and the shortening of the sound duration is to monitor rapid changes in obstacle positions. The decrease in sound amplitude is either due to the difficulty in producing loud short sounds at the end of a long-held breath or to appropriately adjust the echo amplitude into the optimal range of sensitivity of the bat's ears.

Laryngeal nerve activity during pulse emission in the CF-FM bat,Rhinolophus ferrumequinum

The activity of the recurrent laryngeal nerve (RLN) was recorded in the greater horseshoe bat,Rhinolophus ferrumequinum. Respiration, vocalization and nerve discharges were monitored while vocalizations were elicted by stimulation of the central gray matter. This stimulation evoked either expiration or expiration plus vocalization depending on the stimulus strength. When vocalization occurred it always took place during expiration. Recordings from the RLN during respiration showed activity during the inspiration phase, but when vocalization occurred there was activity during inspiration and expiration. These results are consistent with the view that the RLN innervates muscles which control the opening and closing of the glottis. During vocalization the vocal folds are closely approximated and the discharge patterns of the nerve suggests that it controls the muscles which start and end each pulse.

Laryngeal nerve activity during pulse emission in the CF-FM bat,Rhinolophus ferrumequinum

The activity of the external (motor) branch of the superior laryngeal nerve (SLN), innervating the cricothyroid muscle, was recorded in the greater horseshoe bat,Rhinolophus ferrumequinum. The bats were induced to change the frequency of the constant frequency (CF) component of their echolocation signals by presenting artificial signals for which they Doppler shift compensated. The data show that the SLN discharge rate and the frequency of the emitted CF are correlated in a linear manner.

Multiple specializations in the peripheral auditory system of the CF-FM bat,Pteronotus parnellii

Cochlear microphonic (CM) and evoked neural potentials (N1) were recorded from the cochlear aqueduct of awakePteronotus parnellii. The CM audiograms obtained with continuous sounds had more or less uniform thresholds except for a sharp threshold notch at about 60 kHz (Fig. 1). When brief tone bursts were presented, the envelopes of the CM responses were always similar to the envelopes of the applied signals except when tone bursts having frequencies at or close to the frequency of the tuned sensitivity notch were presented (i.e., 59–63 kHz). The CM rise-decay times for frequencies around 60kHz were much longer than those of the presented signals (Fig. 2). The prolonged decay times are thought to be due to the ringing of the basilar membrane resulting from a mechanical resonance in the cochlea. The evoked neural potential audiograms (N1-on and N1-off responses) differed considerably from the CM audiogram. Of particular importance is the N1-off audiogram which exhibited very sharp tuning in four frequency regions: 31–33 kHz, 60–63 kHz, 71–73 kHz, and 91–92 kHz (Fig. 5). The frequencies evoking the lowest thresholds of the CM and N1-off (in the 60 kHz region) were either identical or differed by only 100–400 Hz. The sharp tuning in the 60 kHz region of both the CM and N1 audiograms could be eliminated by presenting 90–100 dB continuous sounds for one min but only if the signal frequency was equal to the tuned frequency of the CM audiogram (Figs. 8–13). Presenting intense sounds having frequencies above or below the tuned 60kHz region had no effect on the audiogram. The overstimulation procedure had remarkably specific effects on the CM and N1-off audiograms causing the greatest threshold increases at the 60 kHz tuned frequency and progressively smaller threshold changes on the slopes of the tuned notch. Assuming that the sharp changes of the N1-off thresholds reflect some important underlying mechanism, the N1-off audiograms demonstrate multiple specializations in the peripheral auditory system ofPteronotus with the bat possessing at least three and possibly four sharply tuned regions. With regard to mechanism, the tuned notch in the CM audiogram, the curious CM rise-decay times evoked by tone bursts, and the ease with which the 60 kHz sensitivity notch can be eliminated all argue strongly in favor of a mechanical resonance in the cochlea which is responsible for the sharp tuning around 60 kHz. On the other hand, the absence of tuned notches in the 30 kHz and 90 kHz regions of the CM audiogram together with the absence of any discernable ringing of the CM potentials evoked by 30 kHz and 90 kHz tone bursts both argue against a resonance mechanism for the tuning at these harmonically related frequency regions. Finally, the fact that overstimulating the 60 kHz region had no discernable effect on the N1-off tuning at 30 kHz and 90 kHz demonstrates that the mechanism responsible for the tuned regions at 30 kHz and 90 kHz are independent of the resonance feature of the cochlea at 60 kHz.

Response patterns to pure tones of cochlear nucleus units in the CF-FM bat,Rhinolophus ferrumequinum

1. In horseshoe bats temporal response patterns to pure tone stimuli (10–100 kHz, 20 ms duration, 0.5 ms rise/fall-time) of 149 cochlear nucleus units (DCN and PVCN) have been recorded. 2. Distribution of the units' Best Frequencies (BF): low frequency neurons 26% (BF 10–65 kHz); FM-frequency neurons 20% (BF 65–81 kHz, i.e. frequencies occurring in the FM-part of the bat's echo signal); filter frequency neurons 52% (BF 81–88 kHz, i.e. frequencies occurring in the CF-part of the bat's echo signal); high frequency neurons 2% (BF > 88 kHz) (Table 1). 3. According to PST-histograms the neurons were classified as: sustained responders (28%, Fig. 1D, E); transient responders (51%, Fig. 1A–C); negative responders (4%, Fig. 1F) and complex responders (17%, Fig. 2–4). In the latter class response patterns drastically change with stimulus frequency and intensity. These units have suppressory sidebands on one or both sides of the excitatory field, sometimes overlapping and enclosing the excitatory area (Fig. 2 and 4). Frequently excitatory response patterns display simultaneous inhibitory processes the latency and duration of which depend on stimulus parameters. 4. In a few complex responders two or more excitatory areas exist, the BF of which may be harmonically related (Fig. 3 and 4). 5. Tuning curves of four auditory nerve fibers are reported showing two separate excitatory areas: a broad less sensitive one from 37 to 79 kHz (low frequency tail) and a narrow, more sensitive one from 82 to 90 kHz (Fig. 5).

Directional coding by binaural brainstem units of the CF-FM bat,Rhinolophus ferrumequinum

1. Monaural and binaural neurophysiological data were obtained from 264 units in the medullary and mesencephalic levels of the central auditory system in the CF-FM bat,Rhinolophus ferrumequinum, using earphones for acoustically separated stimulation of the two ears. 2. The units responded to pure tones delivered at their best frequencies with typical basic patterns (PST-histograms) which were classified according to the literature. 3. The latencies of the units' responses changed systematically with the recording depth from about 8 to 1.5ms when passing from the inferior colliculus at the brain's surface through the nucleus of the lateral lemniscus down to the superior olivary complex. 4. Best frequencies covered almost the whole hearing range from 2kHz up to 96kHz, but 50% of all units lay between 75 and 85kHz, the animal's best hearing range, the emission frequency of orientation sounds being between 82 and 84kHz (CF-part). 5. All lowest thresholds of single units put together demonstrate an almost perfect fitting with the behavioural hearing curve. 6. Monaural excitatory impulse count functions were mostly monotonic but many were very non-monotonic and intermediate. 7. The binaural classification (Table 1) confirmed the classical concept of the pathway and the theory of binaural interaction. The most frequently observed types were E/I (ipsilateral inhibitory, contralateral excitatory; 53 %) and E/E (both sides excitatory; 22%). 8. In all cases tested so far, best frequencies and tuning curves were the same for both sides regardless of the kind of binaural interaction. 9. The most interesting units found (13% of the total) showed together with binaural inhibition, strong facilitation in certain ranges of their dynamic range thus reinforcing the response to particular stimulus combinations. These units may be called directional specialists, useful for angular sensitivity. 10. The excitatory impulse count functions were usually steep covering a best dynamic range of 10–30dB, whereas inhibitory curves were much flatter but covered a wide dynamic range of 30–100dB. 11. Since therefore the excitatory influence is always predominant in any unit tested, coding of the sound direction is ambiguous when the absolute intensity is altered. Hence the response is not only a function of the interaural intensity difference as postulated by the theory, but also dependent upon absolute intensity levels. Neural mechanisms to compensate such ambiguities in presumably higher centers are postulated and discussed.