Passive sound localization of prey by the pallid bat (Antrozous p. pallidus).

Abstract

The pallid bat (Antrozous p. pallidus) uses passive sound localization to capture terrestrial prey. This study of captive pallid bats examined the roles of echolocation and passive sound localization in prey capture, and focused on their spectral requirements for accurate passive sound localization. Crickets were used as prey throughout these studies. All tests were conducted in dim, red light in an effort to preclude the use of vision. Hunting performance did not differ significantly in red light and total darkness, nor did it differ when visual contrast between the terrestrial prey and the substrate was varied, demonstrating that the bats did not use vision to locate prey. Our bats apparently used echolocation for general orientation, but not to locate prey. They did not increase their pulse emission rate prior to prey capture, suggesting that they were not actively scanning prey. Instead, they required prey-generated sounds for localization. The bats attended to the sound of walking crickets for localization, and also attacked small, inanimate objects dragged across the floor. Stationary and/or anesthetized crickets were ignored, as were crickets walking on substrates that greatly attenuated walking sounds. Cricket communication sounds were not used in prey localization; the bats never captured stationary, calling crickets. The accuracy of their passive sound localization was tested with an open-loop passive sound localization task that required them to land upon an anesthetized cricket tossed on the floor. The impact of a cricket produced a single 10-20 ms duration sound, yet with this information, the bats were able to land within 7.6 cm of the cricket from a maximum distance of 4.9 m. This performance suggests a sound localization accuracy of approximately +/- 1 degree in the horizontal and vertical dimensions of auditory space. The lower frequency limit for accurate sound localization was between 3-8 kHz. A physiological survey of frequency representation in the pallid bat inferior colliculus suggests that this lower frequency limit is around 5 kHz.