Auditory behavior of the cricket

Summary

• 1. Essential and nonessential features of the male field-cricket (Gryllus campestris L.) calling song have been identified by varying acoustic parameters while testing for phonotactic tracking by females walking on the Kramer spherical treadmill.

• 2. In the natural calling song, a 5-kHz carrier is modulated into 4-syllable chirps, repeated 2–4 times per second; the syllable rate within each chirp is near 30 Hz. Surprisingly, however, the majority of females that are ‘good trackers’ of calling song also track pure ‘trill’ — continuous 30-Hz trains of syllables having no chirp structure — for many minutes. Trill having identical sound power but the wrong syllable rate is not tracked.

• 3. The ratio of syllable duration to syllable repetition interval — that is, the ‘duty cycle’, which is near 50% in nature — can be increased to as much as ca. 90% (intersyllable pauses as brief as 4–5 ms) without degrading tracking performance, as long as syllable repetition rate is near 30 Hz. Similarly, very short duty cycles — in the extreme case songs with syllables consisting of 200-μs ‘clicks’ — can elicit tracking of the sound source by the female.

• 4. Fine details of the temporal structure of the syllables evidently do not convey information in our conditions, because (i)G. campestris females track both the maleG. campestris and the somewhat similar maleG. bimaculatus calling song, and (ii) tracking ofcampestris calling song is indistinguishable from tracking of the same song played backward.

• 5. All the above results are compatible with the view that repetitive modulation near 30 Hz is the chief — and in some cases both necessary and sufficient — parameter of songs that trigger the neuronal ‘recognition’ of male song. Properties of alternative candidate mechanisms for such a process are described.

• 6. As one test that may distinguish among such mechanisms, the range of syllable rates over which recognition is elicited is measured by a paradigm that keeps chirp duration, chirp rate and chirp energy approximately constant. At 80 dB, tracking is observed reliably in the range 20–40 Hz, whereas syllable rates below 15 Hz and above 50 Hz are largely rejected. Multiple-peak structure within this response spectrum, predicted by certain of the recognition mechanisms considered, is — if present — not resolved by our methods.

• 7. Variation of the song parameter ‘carrier frequency’ to values above the natural ca. 5 kHz results in a striking phenomenon at high sound intensities. Maintained ‘tracking’ is observed, but at characteristic, erroneous angles with respect to the sound source. To compare this ‘anomalous phonotaxis’ with the negative phonotaxis reported in other gryllids during flight, we repeat Moiseff et al.'s (1978) ‘one-eared circling’ experiments on the Kramer treadmill, in three species of gryllids. InTeleogryllus, walking shows parallels to flight (reversal of one-eared circling in comparisons of 5 and 15 kHz) with respect to the carrier-frequency dependence of phonotaxis. WalkingGryllus andTeleogryllus evidently differ in this respect;Gryllus rarely exhibits negative phonotaxis (by one-ear criteria) at frequencies eliciting anomalous phonotaxis.

  Finally, we show that simple pressure-gradient effects associated with the ipsilateral spiracular pathways could in principle account for the directional hearing observed inGryllus, and that the same formulation predicts anomalous directions of tracking, qualitatively like those measured, at elevated carrier frequencies. Such trial explanations evidently require drastic modification, however, if they are to account for the behavioral effects of blocking single tympanal and spiracular inputs.