Direct Measurements of Subjective Loudness in a Bottlenose Dolphin

Abstract

For humans and terrestrial mammals, the variation in susceptibility to noise as a function of frequency is handled by “weighting” sound exposures to emphasize frequencies where auditory sensitivity is highest and lessen the importance of frequencies outside the audible range. This technique allows the use of single, weighted numeric values for impact or damage-risk criteria regardless of the sound frequency. Human weighting schemes were derived from measurements of equal-loudness curves obtained from subjective experiments where a listener compares the loudness of sounds at different frequencies. Previous terrestrial mammal data have shown that response latencies measured in the context of a simple acoustic-detection task may be used to construct equal-latency contours that are analogous to equal-loudness contours, albeit measured indirectly (Pfingst et al. 1975; Stebbins 1966). Until now, there were no empirical measures of equal-loudness curves or auditory weighting functions in marine mammals. This data gap became especially apparent following certain marine mammal experiments of temporary threshold shift (TTS). Limited data at 75 kHz (Schlundt et al. 2000) and more recent TTS data at frequencies up to 28 kHz (Finneran and Schlundt 2010; Finneran et al. 2007) have been compared with results of midfrequency data at 3 kHz (Finneran et al. 2010) and reveal substantial differences between onset TTS levels. Specifically, TTS will occur after lower exposure levels for these higher frequencies. Data at higher frequencies should be used to create more accurate frequency-dependent estimates for onset TTS (i.e., TTS weighting functions). Similarly, equal-loudness data would show the relationship between the frequency of sound and the subjective loudness of the sound. The objective of this effort was to develop auditory weighting functions for Tursiops truncatus. The weighting functions would be defined by measuring subjective loudness as a function of the sound frequency. Loudness contours may be more appropriate for assessing behavioral effects of sound, assuming behavioral reactions are more strongly related to loudness than to sound pressure level (SPL).