Direct measurements of subjective loudness in a bottlenose dolphin.

Abstract

Equal-loudness contours were measured in a bottlenose dolphin (it Tursiops truncatus) trained to perform a loudness comparison task. The subject was presented two sequential tones and whistled if the first tone was louder, and produced a burst pulse or “buzzed” if the second tone was louder. Approximately 70% of trials were “known” comparisons [e.g., tones of same frequency but different sound pressure levels (SPLs)] and allowed performance to be tracked within sessions. The remaining comparisons were probe trials, consisting of a 10-kHz standard tone with a fixed SPL (either 90, 105, or 115 dB re 1 μPa) and a comparison tone, whose frequency was fixed but whose SPL varied. Presentation order of probe trials was balanced. Eleven comparison frequencies ranged from 3.5 to 113.1 kHz. Logistic regression was used to derive curves relating the probability of the comparison tone being perceived louder. The 50% point represented the SPL at which the comparison and standard tones were equally loud. The data represent the first direct measurement of equal-loudness curves in any animal and show the relationship between the frequency and subjective loudness. Loudness contours may be more appropriate for assessing behavioral effects of sound, assuming behavioral reactions are more strongly related to loudness than SPL.