Sound localization in the bottlenose porpoise, Tursiops truncatus (Montagu)

Abstract

Behaviorally mediated, psychophysical measurements of passive sound localization acuity (minimum audible angle, MAA) have been determined in the bottlenose porpoise, Tursiops truncatus. A sub-adult male was tested in a pen acoustically open to Kaneohe Bay at the Naval Undersea Center facilities, Oahu, Hawaii. Localization thresholds were measured for pure tone pulse trains which varied as a function of 1) frequency (Horizontal, 6-100 kHz; Vertical, 20-100 kHz); 2) signal parameters (Horizontal, number of pulses, pulse duration and rise-decay times); and, 3) azimuth (Horizontal, 0°, 15°, 30°, 330° and 345° ; Vertical 0° and 15° ). Localization capability was also determined for click trains containing 166 clicks each of 35 usec duration, centered at 64.35 kHz at 0° azimuth in both the horizontal and vertical planes. Total signal duration was maintained at either 0.5 or 1.0 seconds. The discrimination task was of a spatial, forced choice, two alternative design. The MAAs for pure tones were determined at the 70th percentile on the psychometric function of localization using a modified up-down staircase method of stimulus presentation. Click trains were presented using the method of constants and MAAs again were determined at the 70th percentile of correct responses. Calculated MAAs were the mean of a minimum of seven sessions each containing an average of 60-80 trials (pure tones) or 90 trials (clicks). The animal maintained a fixed spatial relationship to two underwater transducers by biting a stationary bar (depth 77 cm) placed 18.6 meters from and midway between the transducers. Sound pressure levels measured at the bite bar were 20 dB: re 1 ubar. The animal indicated the source of the signal by swimming to his right or left and hitting a response paddle. The experimental results demonstrate that the bottlenose porpoise is capable of localizing spatial positions of sound sources within 2-3 degrees in both the horizontal and vertical planes. Angular resolution of pure tones was greatest for frequencies of 20-60 kHz (MAAs: 2.1° to 2.5° ); and, MAAs increased at both higher and lower frequencies. The MAA of 3.6° at 6 kHz suggests that the animal is using binaural differences in phase and time at lower frequencies and binaural intensity information at frequencies of 20 kHz and above as localization cues. It also suggests that the tissue overlying the ear and not the mandibular fat body is the main route of sound transmission to the middle ear at frequencies below 20-30 kHz. Pure tone thresholds were not significantly altered by changes in pulse duration (3 msec to 500 msec) or rise-decay times greater than 0.25 msec. Horizontal localization capability to a pulsed 40 kHz signal significantly improved when the porpoise positioned 15° to the right or left of the midline between the two transducers (MAAs: 1.7° and 1.4° respectively. Increase in acuity at these azimuths is possibly a function of differential, binaural intensity information which would be greater in this situation than when the animal is at 00 because of the importance of angle of incident sound on the lower jaw (sound is funneled to the middle ear by the mandibular fat body). Maximal MAAs of 5.3° and 5.2° were measured at horizontal azimuths of 30° and 330°. Greatest localization resolution occurred when the animal was presented click trains containing extremely fast rise times and a broadband spectrum of frequencies. The horizontal and vertical MAAs for click trains presented at 0° azimuth were 0.9° and 0.7° respectively. These low MAAs for click trains possibly result from the two transmission routes being used simultaneously in conjunction with binaural intensity comparisons and possible binaural comparisons of spectral characteristics in the central nervous system. The rigorous testing situation in pure tone experiments presented minimal auditory information to an immobile animal. This was necessary for determining what acoustic cues are most important for localization. The pulse durations and rise-decay times of the tested stimuli were much longer than that used in active sonar, and the information contained in broadband clicks was absent. The MAAs reported indicate the animal's capabilities only under restricted conditions; and therefore are conservative.