In a recent paper in The Journal of Experimental Biology, Moir and colleagues did not find evidence for DPOAEs in the motion of the locust typanum (Moir et al., 2011).
Insects with tympanal organs produce sensitive DPOAEs (distortion-product otoacoustic emissions), as shown in locusts, several moth species and bushcrickets (for a review, see Kössl et al., 2008). Their characteristics are largely comparable to those reported from vertebrate ears. Insect DPOAEs are highly vulnerable to manipulations that affect the animal's physiology: hypoxia (Kössl and Boyan, 1998), ethyl ether (Kössl et al., 2007), electrical auditory nerve stimulation (Möckel et al., 2007), a neuroactive insecticide (Möckel et al., 2011) and acoustic suppression by third tones (Kössl and Coro, 2006). These findings strongly indicate a biological origin of acoustical two-tone distortions. Mechanical data from Michelsen (Michelsen, 1971) and Windmill et al. (Windmill et al., 2005) showed for locusts that the tympanum's thin part is maximally deflected by high frequency stimuli, and therefore scolopidia attached at the pyriform vesicle (PV) were most sensitive to that frequency range. Exclusion of these scolopidia via mechanical ablation specifically reduced the levels of high frequency DPOAEs evoked by stimuli above 15 kHz. Scolopidia were therefore suggested to be directly involved in frequency-specific DPOAE generation in insects (Möckel et al., 2007) [see also fig. 6 of Kössl et al. (Kössl et al., 2008)].
Moir et al. (Moir et al., 2011) aimed to record mechanical two-tone distortions in the movement of the tympanum and to relate these mechanical responses to the 2f1–f2 DPOAE frequency defined by the used acoustical stimuli. Several times, the authors termed their measurements ‘DPOAEs’, which in association with their chosen method (laser Doppler vibrometry) is not quite correct. DPOAEs are defined as sounds that can be recorded in the outer ear canal, resulting from cochlear mechanical non-linearity and depending on physiological activity (for a review, see Kemp, 2008). Likewise, in the case of locusts, they are acoustically recorded close to the tympanum, to which the auditory receptors are directly attached. The authors' second aim was to visualize the effect that hypoxia has on tympanum vibration. Within the limits of their experimental setup, as stated in the last paragraph of Moir et al. (Moir et al., 2011), both parts of their work failed to show positive results. From our extensive experience with insect DPOAEs, we feel we should comment on the present study.
Possible reasons for difficulties in measuring mechanical two-tone distortions
One important factor for measuring sensitive DPOAEs is the acoustical coupler and its tip. The coupler consists of separate tubes for two speakers for stimulation and a microphone to record the emission. To record sensitive DPOAEs in locusts, it is crucial that the microphone is aimed perpendicular to the tympanal membrane, and is positioned as close to it as possible without making direct contact. Additionally, to measure sensitive DPOAEs especially in the low frequency range, it is essential to seal the ear with the coupler in position and hence produce a closed system (see also Kemp, 2008). This creates a situation where a large portion of the sound detected by the microphone comes from the tympanum. The study of Moir et al. (Moir et al., 2011), however, worked with an open system, as required for laser Doppler vibrometry. The authors used the stimulation parameters given in Kössl and Boyan (Kössl and Boyan, 1998), which in that study had produced an emission of 10 dB SPL in a closed system, but did not measure DPOAEs in their preparation. They also calculated the expected mechanical tympanum vibration associated with a 10 dB SPL tone based on a planar source model in an open system. However, the actual DPOAE recording situation (see above) and the complex mechanics of the locust tympanum might not be sufficiently reflected in this simple simulation. In this respect, it would be advisable, if possible, to record DPOAEs and mechanical tympanum vibration with the same closed coupler system, as Dalhoff et al. (Dalhoff et al., 2007) did for the human tympanum. This coupler included an optical fiber for the laser. Of course, the locust tympanum with its directly attached sensory cells and the strong membrane inhomogeneities cannot easily be compared with the human tympanum, and parallel acoustic and mechanical measurements in the locust are needed.
The second part of Moir et al.'s study involved the influence of hypoxia, which reduces DPOAE amplitude (Kössl and Boyan, 1998). As we interpret their data, the authors did not measure mechanical (2f1–f2) distortions or their modulation during hypoxia, but found that mechanical vibrations evoked by f1 or f2 were not affected by CO2. Of course, during DPOAE recordings, we would not expect that hypoxia influences f1 or f2 responses but only the much smaller 2f1–f2 response, whose mechanical analogue Moir et al. could not measure.
Artefacts
In their Results section, Moir and colleagues report the presence of setup artefacts at the DPOAE frequency of about 32 dB SPL (0.8 mPa). The acoustical stimuli used may amount to ∼72 dB SPL for f1 (82 mPa) and to ∼63 dB SPL for f2 (28 mPa) and not the other way round as stated in their fig. 4 legend (which is obviously a typing error). This very high distortion level within their sound-producing system might have been caused by the use of a single loudspeaker to deliver both pure-tones. Using two loudspeakers would cause much lower setup distortion during DPOAE measurements and also could be beneficial for investigating their mechanical analogue on the tympanal membrane. It is a common praxis in our lab to use two loudspeakers for delivering the f1 and the f2 stimulus and also to keep the stimulation and DPOAE recording channels separated as far as the tip of the acoustical coupler, positioned close to the tympanum. In the original report on locust DPOAEs, setup-produced distortions appeared just above the background noise level (–15 to –25 dB SPL) at high stimulation levels of 83 to 91 dB SPL (Kössl and Boyan, 1998). Sensitive locust preparations allow DPOAE measurements at stimulation levels as low as 10–20 dB SPL, giving a safe distance of more than 60 dB to any setup-generated distortion. DPOAEs from tympanal organs reach their highest threshold sensitivity at frequencies defined by the hearing range of the investigated species (cf. Kössl and Boyan, 1998; Coro and Kössl, 1998; Kössl et al., 2008). The pronounced differences in DPOAE threshold characteristics between diverse animal groups (moths, locusts, bushcrickets and several mammalian species) rules out setup distortions whose characteristics would depend on the recording system and not the animal species.
