We thank Kössl and Möckel for their correspondence regarding our article (Moir et al., 2011) and the Editor for the opportunity to reply.
In general, the occurrence of distortion products (also called ‘intermodulation’ or ‘intermodulation distortion’) indicates that the system in question is non-linear. This non-linearity can really be anything: from active feedback to passive structural non-linearities. A single-frequency input into a non-linear system will result in a distorted signal that, spectrally, will contain a series of harmonics of the fundamental. Similarly, if two frequencies are input, then intermodulation will occur and sidebands at various arithmetic combinations of the two input frequencies will exist in the frequency spectrum.
A distortion-product otoacoustic emission (DPOAE) is thus a result of such an intermodulation, recorded as a sound emission from an ear. For sound to be emitted from any system, a structure within that system must be vibrating (as in a loudspeaker). In the case of locust ears, the DPOAE sounds are understood to originate from the tympanal organ. As a result, there should be a mechanical signature on the motion of the tympanal organ – something we aimed to measure. Over a large number of animals, at both ethologically relevant sound pressures and louder, we were unable to measure the largest distortion product (at 2f1–f2) in the mechanical motion of the tympanum using laser Doppler vibrometry. As pointed out by Kössl and Möckel, sensitivity of this effect to the physiology of the insect indicates a biological origin, and the mechanosensitive neurones (scolopidia), which attach to the tympanum, were the most likely candidate for the production of DPOAEs in insects (Kössl et al., 2008).
If the locust ear produces DPOAEs as part of its everyday sound transduction behaviour, then we should be able to measure the structure vibrating to generate them (within the limits of our measurement system), regardless of whether the system is closed or not. Therefore, if the tympanal membrane is that structure, and is receiving the correct two-tone sounds [as is clearly evident from measurements showing that the membrane vibrates at those tone frequencies; see fig. 2, Moir et al. (Moir et al., 2011)], such that DPOAEs are being produced by the same membrane (or the scolopidia attached to the membrane) then that tympanal membrane must also be vibrating at the 2f1–f2 distortion-product frequency. Such vibrations were not recorded (down to the laser vibrometer's 3 pm noise floor).
Kössl and Möckel suggest that the use of an acoustic coupler and its tip are important for measuring DPOAEs. The use of an acoustic coupler is simply a method by which faint sounds from a small source can be recorded by a microphone. While this is clearly beneficial for measuring very quiet sounds, it is irrelevant for our report. The existence of a DPOAE at a known frequency and sound pressure must correlate with the motion of a structure emitting the sound. Our investigation could not find evidence of the tympanal membrane vibrating in relation to known DPOAEs.
The reasonable concern of artefacts in the current experiments was also raised. When studying an effect that is derived from any non-linearity, artefacts must be carefully noted and avoided. However, we were very careful not to overdrive our single loudspeaker so as not to generate artefactual two-tone distortion; this was monitored using the microphone. It was only possible to produce distortion-products by overdriving the loudspeaker and not the animal, and we took this into account throughout our experiments [see fig. 4, Moir et al. (Moir et al., 2011)]. In any case, the presence or not of some acoustic artefacts in the setup used does not in any way relate to the absence of any evidence for measured vibrations at any distortion-product frequencies.
Finally, we are not questioning the existence of DPOAEs in locusts, or their physiological dependence, the evidence for which is strong. However, it is concerning that we were not able to measure a mechanical two-tone vibration on the locust tympanum with an experimental setup that is capable of doing so. Rather than put these results in the ‘file drawer’, we believe it is better that the scientific community is made aware of these results so that the question of where DPOAEs are generated can be resolved more quickly. We very sincerely hope that our work will be followed by that of others in order to continue to examine, and seek to explain, all the intricate workings of the locust (and other) insect ears.
