In humans, as in other vertebrates, audition is an active process that relies on the contractile mechanical properties of hair cells. Recent evidence gathered from research on mosquitoes demonstrates that comparable active auditory mechanics occurs in invertebrates. Mosquitoes hear with their antennae. In the large tropical species Toxorhynchites brevipalpis, the mechanical response of the antennal flagellum exhibits all the diagnostic features of active mechanics known from vertebrates.

(a) Vulnerable sensitivity and tuning. The sharp resonant peak that characterises the in vivo mechanical response of the antennal flagellum consistently vanishes post-mortem. In effect, the mechanical sensitivity at resonance frequency (quantified as the flagellar-tip displacement normalized to the acoustic air-particle displacement) decreases by a factor of 1.5. Also, tuning sharpness (measured as quality factor Q) is affected, decreasing 1.6-fold in both sexes. Hence, both sensitivity and tuning of mosquito flagella depend on the physiological condition of the animals.

(b) Vulnerable non-linearity. The in vivo mechanical response of male and female antennal flagella exhibits non-linear damping, such that the lower the intensity, the higher and sharper the resonance peak. This band-limited non-linearity is shown to disappear post-mortem, providing an indication that a physiological process is at work that enhances the resonant properties of mosquito flagella in an intensity- and frequency-dependent way.

(c) Sensitivity to hypoxia. In female mosquitoes, the transient exposure to carbon dioxide alters the sharpness (reducing Q factor) of the resonance peak, an effect that is completely reversible. In contrast, for males, the resonance peak reversibly shifts towards lower frequencies in the presence of CO₂ while, surprisingly, it increases in sharpness and amplitude. Notably, carbon dioxide exposure also elicits flagellar vibrations in the absence of acoustic stimulation, an effect that is seen in males only. Since hypoxia-induced vibrations in males are not related to external forces acting on the flagellum, vibrations must be autonomous and self-sustained, and generated by an internal motor.

(d) Autonomous activity. Autonomous flagellar vibrations (AV) were observed only occasionally in untreated males, yet they could be reliably induced in both sexes by the haemolymphatic injection of dimethylsufoxide (DMSO). DMSO-induced vibrations are long lasting (up to 90 min duration) and are characterised by large vibration amplitudes (up to ± 400 nm).AV is also related to the amplification process revealed by in vivo post-mortem comparisons. In effect, AV occurs only at frequencies at which amplification was shown to take place. Influenced by tones delivered at other frequencies, AV exhibits non-linear characteristics. As expected from the results in (c), carbon dioxide exposure transiently suppresses AV in females, but facilitates AV in males. Thus AV in mosquitoes present all the diagnostic properties of spontaneous otoacoustic emissions reported from vertebrates. Importantly, the examination of antennal deflection shapes for a broad range of excitation frequencies rules out the possibility that active auditory mechanics in mosquitoes originate from muscular activity. Antennal muscles are found only in the proximal segment (the scape), which, along with the auditory organ proper (Johnston’s organ) remains immobile during acoustic stimulation as well as AV. The auditory mechanoreceptor units in Johnston’s organ thus appear to serve as mechanical motors driving the flagellum. As revealed by compound potential of neural origin simultaneously recorded with mechanical activity, AV coincides with local electric activity inside the auditory organ. This finding strongly supports the idea that active auditory mechanics in mosquitoes is based on mechanoreceptor motility.

This work was supported by grants from the Swiss National Science Foundation, the Deutsche Akademie der Naturforscher Leopoldina, and the School of Biological Sciences, University of Bristol.

Figure 1. Build up of AV in an untreated male mosquito. Flagellar tip vibrations (top trace) and particle oscillations (bottom trace) were recorded during acoustic stimulation at 400 Hz. Gradually building up, AV had a frequency of 430 Hz and outlasted the last pulse of the series by 2.4 min.