Auditory behavior of the cricket

Summary

• 1. In applying the recently developed methods of compensated walking to the study of cricket phonotaxis, we have first asked how such devices alter the dynamic body-substrate situation from that experienced in natural walking on a fixed substrate. Because a quantitative treatment has not been available, we derive certain general mechanical requirements of such compensation systems, and show in this context just what the Kramer treadmill accomplishes.

• 2. Mathematical predictions of the above analysis were confirmed by frame-by-frame TV analysis of crickets during natural walking and compensated walking on the Kramer treadmill. These are as follows. Body-substrate inertial effects can be made to resemble those occurring during fixed-substrate walking as long as the animal walks continuously and the compensation is highly damped. During starts and stops, however, body-substrate acceleration, and hence the associated leg forces, differ severely from those experienced in walking on a fixed substrate.

• 3. Despite the above effects, the presence of some bearing noise, and the elimination of all visual guidance, we show that a useful form of ‘blind auditory tracking’ can be obtained regularly. Loudspeakers playing the male calling song are accurately tracked by females, with meandering of direction about the target direction, via several characteristic modes of start-stop behavior. The tracking mode adopted at any instant can depend upon song attractiveness, song intensity, the history of stimuli presented, the individual, and factors unknown to us. The mode adopted also affects the nature of the meandering seen in the direction record.

• 4. Depending upon the season and the source of the crickets, one-quarter to one-half of adult females older than 3 to 4 weeks turn out to be ‘good trackers’ in our conditions, and to remain so for some weeks. Sound intensity has little effect upon tracking accuracy in the range 70–90 dB, as long as the song is an attractive one.

• 5. We have developed experimental paradigms which test for discrimination by the female between male calling song and various ‘wrong songs’. These employ two alternating loudspeakers to control against directional clues available from the apparatus, and allow comparison of response to different songs both sequentially and simultaneously (the choice situation). We show that in both kinds of presentation females can clearly and reproducibly discriminate between calling song and wrong songs having no syllable structure or the wrong syllable repetition rate.

• 6. If during the tracking of calling song the song is turned off, the female stops within a few seconds. If, instead, one replaces the calling song suddenly with a wrong song (rejected when presented alone), the wrong song is ‘tracked’ for 10 to 20 s or more. These effects have implications regarding operational definitions of ‘motivation’ and ‘context setting’; they may also lead to ways of distinguishing between the recognition and tracking processes.

• 7. Contrary to expectation based on earlier work withGryllus campestris, we find that increasing the number of syllables per chirp in the range 4–8 (with repetition interval 30 ms) increases, rather than decreases, attractiveness of the song in both our sequential and our choice tests.

• 8. The finding that attractiveness of a song increases with number of syllables suggests that inGryllus the key feature of sound signals that elicit ‘recognition and pursuit’ may be modulation power near 30 Hz, associated with the syllable repetition rate.

  This hypothesis predicts that natural ‘rivalry song’ (with chirps of 20 or more syllables near the 30-Hz rate) ought to be very attractive; we find that it is, and suggest a reinterpretation of its role in nature.

• 9. In a separate set of experiments, the callingsong sound was gated by the walking-velocity signal so that sound could be heard either (i) only during walking or (ii) only during stops. Only our two best trackers succeeded in our tracking tests if sound was presented strictly during walking; 8 good trackers failed. With sound available only during stops, many succeeded if they happened to employ a suitable tracking mode. Our impression is that ‘stopping to listen’ (or even to look) is ordinarily a significant part of the search behavior.