Harmonic Convergence in the Love Songs of the Dengue Vector Mosquito

See allHide authors and affiliations

Science  20 Feb 2009:
Vol. 323, Issue 5917, pp. 1077-1079
DOI: 10.1126/science.1166541


The familiar buzz of flying mosquitoes is an important mating signal, with the fundamental frequency of the female's flight tone signaling her presence. In the yellow fever and dengue vector Aedes aegypti, both sexes interact acoustically by shifting their flight tones to match, resulting in a courtship duet. Matching is made not at the fundamental frequency of 400 hertz (female) or 600 hertz (male) but at a shared harmonic of 1200 hertz, which exceeds the previously known upper limit of hearing in mosquitoes. Physiological recordings from Johnston's organ (the mosquito's “ear”) reveal sensitivity up to 2000 hertz, consistent with our observed courtship behavior. These findings revise widely accepted limits of acoustic behavior in mosquitoes.

Mosquito-borne diseases such as malaria, yellow fever, and dengue continue to afflict millions, even after decades of work to control vector populations. Despite this effort, basic aspects of mosquito biology are not fully understood, including mating behavior, an important target for vector control. We describe investigations in Aedes aegypti that require revision of the current understanding of mosquito mating behavior. Since Johnston (1) first suggested in 1855 that mosquitoes could perceive sound, over 14 studies have been published on sound production and hearing in A. aegypti (217) (table S1). The buzz of a flying female mosquito acts as a mating signal, attracting males. Typically, the behaviorally salient frequency component of flight tone is the fundamental frequency of wing beat, which is between 300 to 600 Hz depending on species (8). However, mate attraction is not simply a matter of a male passively hearing and homing in on a 400-Hz tone. For example, males and females of the non–blood-feeding mosquito Toxorhynchites brevipalpis modulate their 300- to 500-Hz wing beat frequencies to match each other (18). Thus, acoustically mediated mate attraction involves active modulation by both sexes, creating a duet.

We show that males and females of the dengue and yellow fever vector A. aegypti also modulate their flight tones when brought within a few centimeters of each other. This modulation, however, does not match the fundamental wing beat frequency of around 400 Hz (female) or around 600 Hz (male) but a shared harmonic frequency of around 1200 Hz (Fig. 1). Consistent with this, a neurophysiological examination of the ears of A. aegypti shows response in both males and females up to 2000 Hz (Fig. 2). These results are unexpected because over 5 decades of behavioral and physiological studies had concluded that male mosquito ears (antennae and associated Johnston's organ) are tuned to 300 to 800 Hz and deaf to frequencies above 800 Hz (8, 19). The present study also directly addresses the issue of auditory competence in female mosquitoes. Acoustic duetting behavior in the nonvector mosquito T. brevipalpis (18) would seem to imply active audition in both sexes, and laser vibrometry studies of the Johnston's organ in that species and A. aegypti (16, 17) indicate that they respond mechanically to salient sounds. Moreover, female frog-biting mosquitoes are reported to be attracted by the sounds of their singing hosts (20). The auditory physiology on A. aegypti provides direct evidence that females can hear and puts to rest textbook wisdom that females are deaf (8, 9).

Fig. 1.

(A) An oscillogram from a sound clip of a tethered male and female duetting. (B) A spectrogram depicting the harmonic stack of the same sound clip. The male was held in a fixed position and sang continuously for nearly 2 min. The female was brought within 2 cm of the male on three separate occasions. (C) An expanded view of (B) showing the synchronization of the flight tone at the second harmonic of the male (blue) and the third of the female (red). At t = 45 s, the two tones converge to the extent that they cannot be readily distinguished. (D) A separate recording of a tethered male (blue) flying solo and synchronizing to a loud speaker stimulus consisting of a simulated female flight tone with a missing fundamental (red). (E) A recording of a tethered male flying solo (blue) modulating his flight tone to match the playback of a simulated third harmonic of a female. (F) A female flying solo (red) matches a simulated second harmonic of a male (blue) playback tone.

Fig. 2.

(A) Acoustically evoked field potentials recorded from Johnston's organ exhibit periodic oscillations (inset, F0 to F4) riding on top of a sustained deflection (SD). Shown are averages of 10 and 5 repetitions to 1200 and 400 Hz, respectively, in a male. Thoracic control recordings are in black; the stimulus envelope is at bottom. (B) Spectral analysis of the response to the 400-Hz tone in (A) shows substantial power at low frequencies (SD) during the stimulus (blue) compared with a prestimulus background period (black), as well as peaks at multiple harmonics (F1 to F4) of the stimulus's fundamental frequency (F0). (C) Averaged across 12 males (blue) and 15 females (red), the amplitude (mean ± SEM) of the sustained deflection (SD, solid squares) remains higher than that of prestimulus background noise (dashed line at zero) up to 2000 Hz, whereas the amplitudes of F0 and F1 (open and solid circles, respectively) are substantially smaller.

For behavioral experiments, we tethered each mosquito to the end of an insect pin. When suspended in midair, flies initiated bouts of wing-flapping flight. We recorded flight tones with a particle velocity microphone. Acoustic interaction was demonstrated by moving a tethered flying mosquito past a stationary tethered flying partner (movie S1 with audio). Females were brought in and out of the male hearing range (2 cm) for 10-s fly-bys. Recordings revealed acoustic interaction: In 14 of 21 (67%) pairs, both sexes altered their flight tones so that the male's second harmonic [fundamental (F0) = 636.7 ± 15.1, second harmonic (F1) = 1238.3 ± 31.0 (SEM) Hz] matched the female's third harmonic [F0 = 430.6 ± 10.8, F2 = 1356.2 ± 29.2 (SEM) Hz] (Fig. 1, A to C). The period of synchronization lasted an average of 9.71 ± 1.05 (SEM) s with the synchronization frequency averaging 1354.5 ± 31.5 (SEM) Hz. A. aegypti do not shift their flight tones in the absence of acoustic stimulation, as tested both by deafening the mosquitoes [and stimulating with tones, Fisher' exact test, males P = 0.02, females P = 0.04 (21)] and flying intact control subjects in silence [Fisher's exact test, males P = 0.02, females P = 0.04 (21)].

The presence of the fundamental frequency tone was not necessary for harmonic matching. We stimulated tethered mosquitoes with electronically generated pure sinusoidal tones as well as with harmonic combinations of pure tones lacking the fundamental frequency (Fig. 1, D to F). The intensity of the pure tones was set at a particle velocity of 0.024 mm/s corresponding to 54 dB sound pressure level (relative to 20 μPa) when played through an ear bud speaker positioned 1.5 cm in front of the test mosquito. This intensity is well within the response range of A. aegypti's Johnston's organ, as measured by Doppler vibrometry (17). Stimuli were played in 10- to 15-s bursts with 5- to 20-s recovery periods. Playback experiments with pure-tone combinations demonstrated that 11 of 28 (39%) males could synchronize their 1200-Hz second harmonic to the simulated female's 1200-Hz third harmonic in the absence of the 400-Hz fundamental tone of an actual female. Furthermore, 12 of 54 (22%) males could match a pure 1200-Hz tone (the third harmonic of female flight tone) in the absence of the fundamental and any other harmonic components. We also tested the ability of females to synchronize to playback of pure tones mimicking male sound. Six of 20 (30%) unmated females matched the second harmonic of their flight tone to a complete stack of tones (F0 to F3, 700 to 2800 Hz). When unmated females were stimulated with a pure 1400-Hz tone, 7 of 20 (35%) responded by matching with their third harmonic. In contrast, only 2 of 18 (11%) previously mated females (confirmed by the presence of sperm in the female's sperm storage organs) performed a frequency match to playback of complete male songs, suggesting that mating decreases sensitivity to male stimuli.

Physiological data from the mosquito's auditory organ were obtained by impaling Johnston's organ with tungsten electrodes and recording acoustically evoked field potentials. The Johnston's organ of both males and females responded to 0.5-s cosine-enveloped pure-tone pulses at all frequencies tested (125 to 2000 Hz), including the shared harmonic of their flight tones at 1200 Hz (Fig. 2). The response consists of a sustained voltage deflection, which is negative with respect to the thoracic ground electrode, and a concurrent periodic oscillation at the stimulation frequency and its harmonics, a form of neural encoding that bears resemblance to that seen in the mammalian cochlea (22). It is the amplitude of the sustained deflection that remains significantly higher than prestimulus background noise and thoracic control recordings up to 2000 Hz (t test, P < 0.001 in both males and females); the amplitude of the periodic oscillation is about an order of magnitude smaller and remains higher than background and controls only up to 1000 Hz, in the case of the stimulus's fundamental, and 700 Hz, for its second harmonic (t test, P < 0.01). These recordings confirm earlier studies that the mosquito Johnston's organ is sensitive to 100- to 500-Hz tones and extend the upper limit of hearing to at least 2000 Hz. Earlier physiological studies of Johnston's organ that failed to report a high-frequency response are likely due to filter bandwidths set at the time of recording. We recorded high-frequency responses only when we set the high-pass filter to 1 Hz or less, not the customary 100 Hz or higher used in extracellular recording.

Tone-matching behavior does not require that a tethered, flying mosquito hear a live partner. Either a male or a female A. aegypti can modulate its flight tone harmonics to match electronically generated pure-tone probes (Fig. 1). Frequency match is elicited to a probe that simulates a natural flight tone stripped of its fundamental frequency and even to a probe that contains only a single harmonic. Moreover, mosquitoes can modulate their flight tone harmonics to match the probe tone whether the probe frequency is set above or below the fly's actual flight tone. This directly demonstrates that both sexes of this species (and likely other mosquito species) can hear and respond to high-frequency tones alone. These behavioral findings are supported by sensory physiology (Fig. 2). Extracellular recordings of acoustically evoked field potentials from the Johnston's organ elicited clear responses not only to frequencies of the fundamental (400 to 600 Hz), as expected, but into the kilohertz range, where males and females perform active acoustic modulation of their flight tone harmonics. Thus, both behavioral and physiological experiments establish that A. aegypti signal to each other by using frequencies in the kilohertz range and can detect these with their auditory organs. Taken together, these data call for revision of our understanding of acoustically mediated mating behavior in mosquitoes and, in particular, removal of the long-accepted benchmark ceiling for hearing in mosquitoes.

Lastly, once mated, females are much less responsive to male flight tones and less likely to perform tone matching. These results are consistent with the observation that, once mated, A. aegypti females are not responsive to additional matings for the duration of one or more egg-laying cycles (23, 24). This implies that an initial mating depresses the likelihood of subsequent matings. Thus, releasing sterile male mosquitoes into the wild might adversely affect the reproductive potential of virgin females, providing a rationale for controlling these disease vectors through diminishing mating potential. Other vector and pest species of flies have been controlled by the release of sterile males (2527). Our investigation connects control to signal function. We hypothesize that the ability of males to modulate their flight tones to females is the result of sexual selection. Hence, harmonic convergence could be a measure of a male's reproductive fitness and the ability of lab-reared, sterilized or genetically modified males to modulate their flight tones could be a useful behavioral bioassay for the sterilization program (28, 29). At the very least, our findings open the door to a new understanding of their mating behavior, one that stresses acoustic interactivity between the sexes at frequencies thought previously to be beyond their range of hearing.

Supporting Online Material

Materials and Methods

Fig. S1

Table S1


Movie S1

References and Notes

View Abstract

Navigate This Article