Seismic sensitivity and bone conduction mechanisms enable extratympanic hearing in salamanders

The tympanic middle ear is an adaptive sensory novelty that evolved multiple times in all the major terrestrial tetrapod groups to overcome the impedance mismatch generated when aerial sound encounters the air-skin boundary. Many extant tetrapod species have lost their tympanic middle ears, yet they retain the ability to detect airborne sound. In the absence of a functional tympanic ear, extratympanic hearing may occur via the resonant qualities of air-filled body cavities, sensitivity to seismic vibration, and/or bone conduction pathways to transmit sound from the environment to the ear. We used auditory brainstem response recording and laser vibrometry to assess the contributions of these extratympanic pathways for airborne sound in atympanic salamanders. We measured auditory sensitivity thresholds in eight species and observed sensitivity to low-frequency sound and vibration from 0.05-1.2 kHz and 0.02-1.2 kHz, respectively. We determined that sensitivity to airborne sound is not facilitated by the vibrational responsiveness of the lungs or mouth cavity. We further observed that, although seismic sensitivity probably contributes to sound detection under naturalistic scenarios, airborne sound stimuli presented under experimental conditions did not produce vibrations detectable to the salamander ear. Instead, threshold-level sound pressure is sufficient to generate translational movements in the salamander head, and these sound-induced head vibrations are detectable by the acoustic sensors of the inner ear. This extratympanic hearing mechanism mediates low-frequency sensitivity in vertebrate ears that are unspecialized for the detection of aerial sound pressure, and may represent a common mechanism for terrestrial hearing across atympanic tetrapods.

Lung-to-ear sound transmission does not improve directional hearing in green treefrogs (Hyla cinerea)

Amphibians are unique among extant vertebrates in having middle ear cavities that are internally coupled to each other and to the lungs. In frogs, the lung-to-ear sound transmission pathway can influence the tympanum's inherent directionality, but what role such effects might play in directional hearing remains unclear. In this study of the American green treefrog (Hyla cinerea), we tested the hypothesis that the lung-to-ear sound transmission pathway functions to improve directional hearing, particularly in the context of intraspecific sexual communication. Using laser vibrometry, we measured the tympanum's vibration amplitude in females in response to a frequency modulated sweep presented from 12 sound incidence angles in azimuth. Tympanum directionality was determined across three states of lung inflation (inflated, deflated, reinflated) both for a single tympanum in the form of the vibration amplitude difference (VAD) and for binaural comparisons in the form of the interaural vibration amplitude difference (IVAD). The state of lung inflation had negligible effects (typically less than 0.5 dB) on both VADs and IVADs at frequencies emphasized in the advertisement calls produced by conspecific males (834 and 2730 Hz). Directionality at the peak resonance frequency of the lungs (1558 Hz) was improved by ∼3 dB for a single tympanum when the lungs were inflated versus deflated, but IVADs were not impacted by the state of lung inflation. Based on these results, we reject the hypothesis that the lung-to-ear sound transmission pathway functions to improve directional hearing in frogs.

Internally coupled ears in living mammals

It is generally held that the right and left middle ears of mammals are acoustically isolated from each other, such that mammals must rely on neural computation to derive sound localisation cues. There are, however, some unusual species in which the middle ear cavities intercommunicate, in which case each ear might be able to act as a pressure-difference receiver. This could improve sound localisation at lower frequencies. The platypus Ornithorhynchus is apparently unique among mammals in that its tympanic cavities are widely open to the pharynx, a morphology resembling that of some non-mammalian tetrapods. The right and left middle ear cavities of certain talpid and golden moles are connected through air passages within the basicranium; one experimental study on Talpa has shown that the middle ears are indeed acoustically coupled by these means. Having a basisphenoid component to the middle ear cavity walls could be an important prerequisite for the development of this form of interaural communication. Little is known about the hearing abilities of platypus, talpid and golden moles, but their audition may well be limited to relatively low frequencies. If so, these mammals could, in principle, benefit from the sound localisation cues available to them through internally coupled ears. Whether or not they actually do remains to be established experimentally.

Acoustic analysis of the frequency-dependent coupling between the frog's ears

The ears of anurans are coupled through the Eustachian tubes and mouth cavity. The degree of coupling varies with frequency showing a bandpass characteristic, but the characteristics differ between empirically measured data based on auditory nerve responses and tympanic membrane vibration. In the present study, the coupling was modeled acoustically as a tube connected with a side branch. This tube corresponds to the Eustachian tubes, whereas the side branch corresponds to the mouth cavity and nares. The analysis accounts for the frequency dependency shown by the empirical data and reconciles the differences observed between the coupling as measured by tympanic membrane vibration and auditory nerve responses.

Active control of ultrasonic hearing in frogs

Vertebrates can modulate the sound levels entering their inner ears in the face of intense external sound or during their own vocalizations. Middle ear muscle contractions restrain the motion of the middle ear ossicles, attenuating the transmission of low-frequency sound and thereby protecting the hair cells in the inner ear. Here we show that the Chinese concave-eared torrent frog, Odorrana tormota, can tune its ears dynamically by closing its normally open Eustachian tubes. Contrary to the belief that the middle ear in frogs permanently communicates with the mouth, O. tormota can close this connection by contraction of the submaxillary and petrohyoid muscles, drastically reducing the air volume behind the eardrums. Mathematical modeling and laser Doppler vibrometry revealed that the reduction of this air volume increases the middle ear impedance, resulting in an up to 20 dB gain in eardrum vibration at high frequencies (10-32 kHz) and 26 dB attenuation at low frequencies (3-10 kHz). Eustachian tube closure was observed in the field during calling and swallowing. Besides a potential role in protecting the inner ear from intense low-frequency sound and high buccal air pressure during calling, this previously unrecognized vertebrate mechanism may unmask the high-frequency calls of this species from the low-frequency stream noise which dominates the environment. This mechanism also protects the thin tympanic membranes from injury during swallowing of live arthropod prey.

Binaural interaction in the frog dorsal medullary nucleus

We have studied binaural and directional processing in cells in the frog dorsal medullary nucleus (DMN) stimulated with dichotic sound (couplers) and free-field sound. We present evidence that already at this stage of central processing the neural directionality is sharpened, probably by binaural interaction. Binaural interaction in the DMN was usually interpreted as inhibition, mostly driven from the contralateral side and dependent on a certain combination of interaural time differences (ITD) and interaural level differences (ILD). In free-field measurements, the strength and timing of the binaural inputs will depend on sound direction as processed by the auditory fibers. Thus, the directionality of DMN cells is caused by both monaural directional cues generated by acoustical coupling of the eardrums and non-tympanic pathways as well as binaural interaction. Most DMN cells show ovoidal directional characteristics and the directionality is sharpened compared to that of auditory nerve fibers. We suggest that the sharpening is due to the inhibitory interactions.

Mechanical leverage in the middle ear of the American bullfrog, Rana catesbeiana

Textbooks lump the middle ears of 'submammalian Tetrapoda' as being 'one-ossicle ears'. Conventionally the anuran middle ear is depicted with a shaft-like skeletal unit connecting the tympanic membrane to the inner ear. This shaft comprises mediad a long bony columella and laterad a short cartilaginous extracolumella. But dissection of Rana catesbeiana ears showed: the extracolumella, as long as the columella, is proximally expanded in the vertical plane, forming dorsal and ventral heads. The medio-dorsal head is movably jointed to the columella, between these two there is an obtuse angle ventrad; the extracolumellar medio-ventral head is anchored by a ligament to the middle-ear cavity ceiling. When the tympanic membrane moves outwards, pulling the extracolumella, the medio-dorsal head of the extracolumella must be forced inwards, rotating on the ventral anchorage, pushing the columella towards the inner ear. The ossicular chain thus includes a mechanical lever, possessing the magnitude of the ratio length:width of the extracolumella; this is additional to the lever known from the columellar footplate, which rotates on its firm ventral attachment. These levers are confirmed physiologically, by the difference between the inner-ear sensitivity (shown by isopotential audiograms of microphonic potentials) when stimulated by a vibrator first at the tympanic membrane, then at the proximal stump of the amputated columella. Perusal of the primary literature showed that this morphology is widespread among anuran ears.

Hearing in the frog: a neurophysiological study of the auditory response in the midbrain

The evoked response in the torus semicircularis of Rana temporaria and R. esculenta to acoustic stimulation has been investigated. The amplitude of the potential is maximal when the recording electrode is placed 1 mm below the tectal surface. We show that responses are evoked by tones over the frequency range 50 to 3500 Hz, but that the upper frequency limit can be extended beyond 5000 Hz by mechanical vibration of the middle ear. The projection from lower centres to the torus semicircularis is organized tonotopically; fibres responding to high frequencies project more rostrally than those responding to lower frequencies. Moreover, maximum response is evoked from each region in the torus semicircularis by incident sound from a particular direction. Best responses are obtained in the anterior region of the torus semicircularis for high frequencies presented rostrally and in the posterior region by lower frequencies presented at 90° to the rostrocaudal axis. In this way, a map of auditory space is represented in the midbrain.

Dynamic behavior of the middle ear based on sweep frequency tympanometry

A measuring apparatus was developed; its probe tip, which exhibits flat frequency characteristics, enables this apparatus to measure the absolute sound pressure and absolute phase variations versus both sweeping frequency and external auditory canal pressure. Although it is difficult to diagnose ossicular chain separation and fixation from the commonly used tympanograms with a low probe tone frequency (e.g.f = 220 Hz), the results obtained with this apparatus enable one to clearly distinguish patients with ossicular chain disorders from normal subjects. Therefore, it seems to be highly useful in the clinical diagnosis of ossicular chain disorders.

Tympanic and extratympanic sound transmission in the leopard frog

The inner ear of the leopard frog, Rana pipiens, receives sound via two separate pathways: the tympanic-columellar pathway and an extra-tympanic route. The relative efficiency of the two pathways was investigated. Laser interferometry measurements of tympanic vibration induced by free-field acoustic stimulation reveal a broadly tuned response with maximal vibration at 800 and 1500 Hz. Vibrational amplitude falls off rapidly above and below these frequencies so that above 2 kHz and below 300 Hz tympanic vibration is severely reduced. Electrophysiological measurements of the thresholds of single eighth cranial nerve fibers from both the amphibian and basilar papillae in response to pure tones were made in such a way that the relative efficiency of tympanic and extratympanic transmission could be assessed for each fiber. Thresholds for the two routes are very similar up to 1.0 kHz, above which tympanic transmission eventually becomes more efficient by 15-20 dB. By varying the relative phase of the two modes of stimulation, a reduction of the eighth nerve response can be achieved. When considered together, the measurements of tympanic vibration and the measurements of tympanic and extratympanic transmission thresholds suggest that under normal conditions in this species (1) below 300 Hz extratympanic sound transmission is the main source of inner ear stimulation; (2) for most of the basilar papilla frequency range (i.e., above 1.2 kHz) tympanic transmission is more important; and (3) both routes contribute to the stimulation of amphibian papilla fibers tuned between those points. Thus acoustic excitation of the an uran's inner ear depends on a complex interaction between tympanic and extratympanic sound transmission.

Auditory responses in the torus semicircularis of the cane toad, Bufo marinus. I. Field potential studies

Field potentials have been recorded in the torus semicircularis of the toad, Bufo marinus, in response to brief tones presented in the free field. The amplitude of the potentials varied with the frequency of the stimulus and location of the electrode along the rostro-caudal axis of the torus. All frequencies in the auditory range evoked largest potentials when the stimulus was located in the contralateral auditory field. Potentials evoked by low to mid frequencies were largest when the stimulus was located near the line orthogonal to the long axis of the animal. For progressively higher frequencies, the optimal stimulus position was progressively more anterior in the contralateral field. In animals in which one eighth nerve had been sectioned, field potentials evoked by tones of low to mid frequency were less sensitive to changes in stimulus direction than in normal animals. However, the directional sensitivity of field potentials evoked by mid to high frequencies was similar in monaural and normal animals. These observations suggest that binaural neural integration is important in determining the directional sensitivity of field potentials in the torus evoked by low to mid frequencies but not for potentials evoked by mid to high frequencies.

Mechanical properties of the frog ear: vibration measurements under free- and closed-field acoustic conditions

The acoustically induced motion of the eardrum of the frog was measured by an incoherent optical technique. When free-field sound stimulation was used, the eardrum vibration had a band-pass characteristic with maximum amplitude at 1-2.5 kHz. However, when the sound was presented in a closed-field acoustic coupler the response was low-pass (cut-off frequency about 2.5 kHz). We demonstrate that the motion is the result of the mechanical properties of the eardrum and the sound pressure acting upon it. The net pressure is due to a combination of sound incident directly on the front of the drum and of sound conducted to the rear via internal (resonant) pathways. The frog ear therefore acts as a pressure-gradient receiver at low frequency and a pressure receiver at high frequency. A model is proposed and analysed in terms of its electrical analogue. This model accounts for both our own experimental observations and those of previous studies.