The influence of arctiid moth clicks on bat echolocation; jamming or warning?

Abstract

1. Many arctiid and ctenuchid moths produce clicking sounds in response to the ultrasonic cries of bats. Clicks were recorded from the two arctiid moth species Arctia caja, the garden tiger, and Phragmatobia fuliginosa, the ruby tiger. The threshold for eliciting clicks was around 60 to 75 dB pe SPL in both species. A. caja produced single clicks, and P. fuliginosa bursts of clicks. The maximum intensity of the clicks was 90 to 94 dB pe SPL at 5 cm for A. caja and 85 dB pe SPL at 5 cm for P. fuliginosa. The clicks contain most energy in the frequency range from 40 to 80 kHz (Figs. 2, 3). 2. Pipistrelle bats (Pipistrellus pipistrellus) were trained to sit on a platform and discriminate the difference in range, Δd, to two targets. The minimum Δd the bats could discriminate with more than 75% success rate was 1.5 cm. 3. The targets had built-in electrostatic loudspeakers through which different sounds could be played back to the bat. Playback of arctiid moth clicks from both targets did not disturb the bat's discrimination accuracy. The success rate did not decrease at any Δd, and the minimum Δ d in the presence of clicks was 1 cm. 4. The clicks played from both loudspeakers did not influence the acoustic behavior or discrimination behavior of the bats in any obvious way. In all trials the bats went through a period with long (3 ms) slowly repeated (12-15 pulses/s) cries, a period with shorter cries and increased PRR (20 pulses/s) in which the decision seemed to be made, and finally a period with very short cries (0.5 ms) repeated at rates of up to 150 pulses/s (Figs. 4 and 5). The cries were FM sweeps from 120 kHz to 55 kHz with a second harmonic, which was strongest in the short cries. 5. The bats' response to the playback of different sounds, such as noise and recorded bat cries, from either the left or right loudspeaker, suggested that the bats reacted to clicks as if they were noise. The playback of sounds from only one speaker at a time decreased the bats' success rate, since the bats were attracted to the sounds (Figs. 6 and 7). 6. A secretion from the cervical glands of A. caja, which contains choline ester, was given to a bat if it crawled towards a clicking target. Both bats tested in this way learned to associate the clicks with a noxious reward and avoided the clicks after just one or two trials (Fig. 8). 7. These results suggest that the function of the garden tiger and ruby tiger clicks in nature is to warn the bat of the moth's distastefulness, and not to 'jam' the bat's sonar system. © 1985 Springer-Verlag.