Spectral and temporal gating mechanisms enhance the clutter rejection in the echolocating bat, Rhinolophus rouxi.

Abstract

Doppler shift compensation behaviour in horseshoe bats, Rhinolophus rouxi, was used to test the interference of pure tones and narrow band noise with compensation performance. The distortions in Doppler shift compensation to sinusoidally frequency shifted echoes (modulation frequency: 0.1 Hz, maximum frequency shift: 3 kHz) consisted of a reduced compensation amplitude and/or a shift of the emitted frequency to lower frequencies (Fig 2). Pure tones at frequencies between 200 and 900 Hz above the bat's resting frequency (RF) disturbed the Doppler shift compensation, with a maximum of interference between 400 and 550 Hz (Fig. 1). Minimum duration of pure tones for interference was 20 ms and durations above 40 ms were most effective (Fig. 3). Interfering pure tones arriving later than about 10 ms after the onset of the echolocation call showed markedly reduced interference (Fig. 4). Doppler shift compensation was affected by pure tones at the optimum interfering frequency with sound pressure levels down to -48 dB rel the intensity level of the emitted call (Fig. 5, 6). Narrow bandwidth noise (bandwidth from +/- 100 Hz to +/- 800 Hz) disturbed Doppler shift compensation at carrier frequencies between -250 Hz below and 800 Hz above RF with a maximum of interference between 250 and 500 Hz above resting frequency (Fig. 7). The duration and delay of the noise had similar influences on interference with Doppler shift compensation as did pure tones (Fig. 8, 9). Intensity dependence for noise interference was more variable than for pure tones (-32 dB to -45 dB rel emitted sound pressure level Fig. 10).(ABSTRACT TRUNCATED AT 250 WORDS)