Faculty of Natural Sciences

Echolocating bats navigate and track their prey for capture via active sensing. To do so, they need to operate on a favorable echo-to-noise ratio (ENR) that allows them to discern target echoes from background noise and clutter. Early studies using wire experiments have demonstrated high resilience to masking in free-flying bats, but whether this is explained by spatial release from masking or complex auditory signal detection in noise remains unresolved. To test the hypothesis that bats in masking noise defend a certain ENR by increasing source levels, we trained four Daubenton’s bats to approach and land on a spherical hydrophone (TS of approx. -15 dB) that also acted as an omnidirectional noise
source at four different noise levels (NL, white noise from 20-90 kHz, 60-90 dB re: 20 μPa in 10 dB steps). Behind the hydrophone, an array of seven microphones was used to localize the bats during the approach and to estimate source level (SL) from received level (RL) and range. Compared to the no-noise treatment, bats substantially increased their SL by 6 dB at a NL of 60 dB re: 20 μPa, and more so at higher noise levels, ending up with a SL around 130
dB re: 20 µPa at NLs of 90 dB. We show that bats are sensitive to masking when they cannot employ spectral, temporal or spatial means to reduce such auditory interference and that they can produce very high source levels in the lab to defend ENRs good enough to perform echo-guided landings.