Summary
The ears of moths we tested in Canada and CSte d'Ivoire are most sensitive to sounds between 20 and 40 kHz, and much less sensitive to sound over 65 kHz. The insectivorous bats most commonly encountered in these (and other) locations use high intensity, frequency modulated echolocation calls with frequency components in the 20-40 kHz range, making them detectable by the most sensitive tympanate moths up to 40 m away. In Africa bats such as species in the Nycteridae, Megadermatidae, and some in the Hipposideridae, use low intensity calls with high frequency components, and these species are not detectable by moths at over 2 m. The hearing ability of moths may significantly influence the feeding efficiency of bats, and changes in the intensity and frequency components of bat echolocation calls can drastically reduce the range at which bats are detected, and thus the time available to the moths for evasive behaviour. The use of low intensity, high frequency echolocation calls may constitute a bat counter-maneuver against insects tuned to bat calls.
Methodology
We monitored the activity of echolocating bats at four field locations: the Queen's University Biology Station on Lake Opinicon in southeastern Ontario (44~ ; 79 ~ 5'W); the Lamto Ecology Station on the Bandama River in C6te d'Ivoire (6~ 5~ the Southwest Research Station near Portal, Arizona (31~ 109~ and in the Sengwa Wild Life Research Area in Rhodesia (18~ 28~ Recordings of bat echolocation calls were made at the Ontario, Arizona and CSte d'Ivoire sites, and neural preparations of moths in Ontario and C6te d'Ivoire. We recorded bats as they hunted for insects in the field in a variety of situations using broadband microphones (10-200 kHz; Simmons, Fenton, Ferguson, Jutting and Palin, in press) and a Lockheed Store 4D tape recorder operated at 76 cm/s. Tapes were played at reduced speeds (8 16 times) into a Princeton Applied Research Model 4513 FFT (Fast Fourier Transform) Real Time Spectrum Analyzer (2048 lines; flat weighting) to determine the power spectra of the bat calls. In the field we also used a zerocrossing period meter with the broadband microphone and an oscilloscope (Simmons, Fenton, Ferguson, Jutting and Palin, in press) to assess habitat use by bats with different echolocation call characteristics and frequency patterns. Bat activity was quantified by counting the number of passes past the microphone during periods of monitoring. In the field the feeding behaviour of bats was observed directly, sometimes with the aid of a night vision scope, in conjunction with the broadband microphone and period meter.