The rate of cochlear compression in a dolphin: a forward-masking evoked-potential study.

The "active" cochlear mechanism of hearing manifests in the cochlear compression. Investigations of compression in odontocetes help to determine the frequency limit of the active mechanism. The compression may be evaluated by comparison of low- and on-frequency masking. In a bottlenose dolphin, forward masking of auditory evoked potentials to tonal pips was investigated. Measurements were performed for test frequencies of 45 and 90 kHz. The low-frequency maskers were - 0.25 to - 0.75 oct relative the test. Masking efficiency was varied by masker-to-test delay variation from 2 to 20 ms, and masker levels at threshold (MLTs) were evaluated at each of the delays. It was assumed that low-frequency maskers were not subjected or little subjected to compression whereas on-frequency maskers were subjected equally to the test. Therefore, the compression rate was assessed as the slope of low-frequency MLT dependence on on-frequency MLT. For the 90-kHz test, the slopes were 0.63 and 0.18 dB/dB for masker of - 0.25 and - 0.5 oct, respectively. For the 45 kHz test, the slopes were 0.69 and 0.39 dB/dB for maskers of - 0.25 and - 0.5 oct. So, compression did not decay at the upper boundary of the hearing frequency range in the dolphin.

Adaptation in the auditory system of a beluga whale: effect of adapting sound parameters.

The effects of adapting sounds (pip trains or pure tones) on auditory evoked potentials (the rate following response, RFR) were investigated in a beluga whale. During RFR acquisition, adapting signals lasting 128 ms each were alternated with test signals lasting 16 ms each; the test signal levels varied randomly. Adapting signals were trains of cosine-enveloped tone pips or pure tones. Pip rate varied with the envelope cosine cycle maintained at 0.125 of pip intervals and the cosine rise-fall time maintained at 0.0625 of pip intervals. Adapting signals shifted the amplitude-level function upward compared to the baseline (no adapting signal) function. The higher the adapting signal level was, the bigger the shift in the amplitude-level function was. The slower the pips were in the adapting signal, the smaller the adaptation effect was. A train of pips with a 0.0625-ms rise-fall time and 125 dB SPL shifted the function by 35-40 dB, whereas a train of pips with a 1-ms rise-fall time or a pure tone with the same SPL shifted the function by approximately 15 dB. The difference between the "fast" and "slow" adapting signals is supposed to be associated with their abilities to stimulate the auditory system in odontocetes.

Level-dependent masking of the auditory evoked responses in a dolphin: manifestation of the compressive nonlinearity.

At suprathreshold sound levels, interactions between masking noise and sound signals are liable to compressive nonlinearity in the auditory system. The compressive nonlinearity is a property of the "active" cochlear mechanism. It is not known whether this mechanism is capable to function at frequencies close to or above 100 kHz that are available to odontocetes (toothed whales, dolphins, and porpoises). This question may be answered by the use of the frequency-specific masking. Auditory evoked potentials to sound stimuli in a bottlenose dolphin, Tursiops truncatus, were recorded in the presence of simultaneous maskers. Stimulus frequencies were 45, 64, or 90 kHz. Maskers were on-frequency bandlimited noise or low-frequency noise of frequencies 0.25-1 oct below the stimulus frequency. The stimuli provoked responses as a series of brain-potential waves following the pip-train rate. For the on-frequency masker, the masker level at threshold dependence on the signal level was 1.1 dB/dB. For maskers of 1 oct below the stimulus, the dependence was 0.53-0.57 dB/dB. The data considered evidence for the compressive nonlinearity of responses to stimuli, and therefore, are indicative of the functioning of the active mechanism at frequencies up to 90 kHz.

Temporary Threshold Shifts in Naïve and Experienced Belugas: Can Dampening of the Effects of Fatiguing Sounds Be Learned?

In belugas (Delphinapterus leucas), substantial (10-15 dB) differences in temporary threshold shifts (TTSs) were observed between the first and subsequent experimental sessions in the same subjects. In the first session (naïve subject state), the TTSs produced by exposure to fatiguing noises were larger than the TTSs produced in subsequent sessions (experienced subject state). After one to two sessions, the TTSs stabilized. The baseline hearing thresholds did not differ between the naïve and experienced states. One possible explanation for this effect is that the animals learned to dampen their hearing during exposure to fatiguing noises and thus mitigate the impact of those noises.

Is Sound Exposure Level a Convenient Metric to Characterize Fatiguing Sounds? A Study in Beluga Whales.

Both the level and duration of fatiguing sounds influence temporary threshold shifts (TTSs) in odontocetes. These two parameters were combined into a sound exposure level (SEL). In the beluga whale Delphinapterus leucas, TTSs were investigated at various sound pressure level (SPL)-to-duration ratios at a specific SEL. At low SPL-to-duration ratios, the dependence was positive: shorter high-level sounds produced greater TTSs than long low-level sounds of the same SEL. At high SPL-to-duration ratios, the dependence was negative: long low-level sounds produced greater TTSs than short high-level sounds of the same SEL. Thus, the validity of SEL as a metric for fatiguing sound efficiency is limited.

Auditory evoked potentials in the auditory system of a beluga whale Delphinapterus leucas to prolonged sound stimuli.

The effects of prolonged (up to 1500 s) sound stimuli (tone pip trains) on evoked potentials (the rate following response, RFR) were investigated in a beluga whale. The stimuli (rhythmic tone pips) were of frequencies of 45, 64, and 90 kHz at levels from 20 to 60 dB above threshold. Two experimental protocols were used: short- and long-duration. For the short-duration protocol, the stimuli were 500-ms-long pip trains that repeated at a rate of 0.4 trains/s. For the long-duration protocol, the stimuli were continuous pip successions lasting up to 1500 s. The RFR amplitude gradually decreased by three to seven times from 10 ms to 1500 s of stimulation. Decrease of response amplitude during stimulation was approximately proportional to initial (at the start of stimulation) response amplitude. Therefore, even for low stimulus level (down to 20 dB above the baseline threshold) the response was never suppressed completely. The RFR amplitude decay that occurred during stimulation could be satisfactorily approximated by a combination of two exponents with time constants of 30-80 ms and 3.1-17.6 s. The role of adaptation in the described effects and the impact of noise on the acoustic orientation of odontocetes are discussed.

Spectrum pattern resolution after noise exposure in a beluga whale, Delphinapterus leucas: Evoked potential study.

Temporary threshold shift (TTS) and the discrimination of spectrum patterns after fatiguing noise exposure (170 dB re 1 µPa, 10 min duration) was investigated in a beluga whale, Delphinapterus leucas, using the evoked potential technique. Thresholds were measured using rhythmic (1000/s) pip trains of varying levels and recording the rhythmic evoked responses. Discrimination of spectrum patterns was investigated using rippled-spectrum test stimuli of various levels and ripple densities, recording the rhythmic evoked responses to ripple phase reversals. Before noise exposure, the greatest responses to rippled-spectrum probes were evoked by stimuli with a low ripple density with a decrease in the response magnitude occurring with an increasing ripple density. After noise exposure, both a TTS and a reduction of the responses to rippled-spectrum probes appeared and recovered in parallel. The reduction of the responses to rippled-spectrum probes was maximal for high-magnitude responses at low ripple densities and was negligible for low-magnitude responses at high ripple densities. It is hypothesized that the impacts of fatiguing sounds are not limited by increased thresholds and decreased sensitivity results in reduced ability to discriminate fine spectral content with the greatest impact on the discrimination of spectrum content that may carry the most obvious information about stimulus properties.

The limits of applicability of the sound exposure level (SEL) metric to temporal threshold shifts (TTS) in beluga whales, Delphinapterus leucas.

The influence of fatiguing sound level and duration on post-exposure temporary threshold shift (TTS) was investigated in two beluga whales (Delphinapterus leucas). The fatiguing sound was half-octave noise with a center frequency of 22.5 kHz. TTS was measured at a test frequency of 32 kHz. Thresholds were measured by recording rhythmic evoked potentials (the envelope following response) to a test series of short (eight cycles) tone pips with a pip rate of 1000 s(-1). TTS increased approximately proportionally to the dB measure of both sound pressure (sound pressure level, SPL) and duration of the fatiguing noise, as a product of these two variables. In particular, when the noise parameters varied in a manner that maintained the product of squared sound pressure and time (sound exposure level, SEL, which is equivalent to the overall noise energy) at a constant level, TTS was not constant. Keeping SEL constant, the highest TTS appeared at an intermediate ratio of SPL to sound duration and decreased at both higher and lower ratios. Multiplication (SPL multiplied by log duration) better described the experimental data than an equal-energy (equal SEL) model. The use of SEL as a sole universal metric may result in an implausible assessment of the impact of a fatiguing sound on hearing thresholds in odontocetes, including under-evaluation of potential risks.

Frequency tuning of hearing in the beluga whale: discrimination of rippled spectra.

Frequency tuning was measured in the beluga whale (Delphinapterus leucas) using rippled-noise test stimuli in conjunction with an auditory evoked potential (AEP) technique. The test stimulus was a 2-octave-wide rippled noise with frequency-proportional ripple spacing. The rippled-noise signal contained either a single reversal or rhythmic (1-kHz rate) reversals of the ripple phase. Single or rhythmic phase reversals evoked, respectively, a single auditory brainstem response (ABR) or a rhythmic AEP sequence-the envelope following response (EFR). The response was considered as an indication of resolvability of the ripple pattern. The rhythmic phase-reversal test with EFR recording revealed higher resolution than the single phase-reversal test with single ABR recording. The limit of ripple-pattern resolution with the single phase-reversal test ranged from 17 ripples per octave (rpo) at 32 kHz to 24 rpo at 45 to 64 kHz; for the rhythmic phase-reversal test, the limit ranged from 20 to 32 rpo. An interaction model of a ripple spectrum with frequency-tuned filters suggests that the ripple-pattern resolution limit of 20 to 32 rpo requires a filter quality Q of 29 to 46. Possible causes of disagreement of these estimates with several previously published data are discussed.