Sound localization by the barn owl (Tyto alba) measured with the search coil technique

Abstract

The dynamics and accuracy of sound localization by the barn owl (Tyto alba) were studied by exploiting the natural head-orienting response of the owl to novel sound stimuli. Head orientation and movement were measured using an adaptation of the search coil technique which provided continuous high resolution azimuthal and elevational information during the behavior. The owls responded to sound sources with a quick, stereotyped head saccade; the median latency of the response was 100 ms, and its maximum angular velocity was 790°/s. The head saccade terminated at a fixation point which was used to quantify the owl's sound localization accuracy. When the sound target was located frontally, the owl's localization error was less than 2° in azimuth and elevation. This accuracy is superior to that of all terrestrial animals tested to date, including man. When the owls were performing open-loop localization (stimulus off before response begins), their localization errors increased as the angular distance to the target increased. Under closed-loop conditions (stimulus on throughout response), the owls again committed their smallest errors when localizing frontal targets, but their errors increased only out to target angles of 30°. At target angles greater than 30°, the owl's localization errors were independent of target location. The owl possesses a frontal region wherein its auditory system has maximum angular acuity. This region is coincident with the owl's visual axis.