7. Electrophysiological Techniques
319
above the curve minima. An alternative measure, frequently referred to as
B IO (the bandwidth 10dB above the curve minima) is a measure of the width
of the tuning curve. When measured using toneburst ABRs, the shape of
the tuning curves varied somewhat as a function of frequency. For probe
frequencies above approximately 64kHz the QIO values were roughly constant, ranging from 17.3 to 18.2. However, at lower frequencies the tuning
curves became quite broad with Q IO values reaching 4.7 at the 16-kHz probe
condition. In contrast, B IO was approximately constant (3.2 to 3.5 kHz) for
probe frequencies between 16 and 45 kHz; above this frequency the B IO
enlarged proportionally to the frequency. Thus, below 45 kHz QIO was proportional to frequency, whereas above 45 kHz QIO was constant.
The tone burst stimuli used in these experiments, being relatively broad
in frequency due to their abrupt rise times, would likely yield exaggerated
estimates of filter widths at the lower stimulus frequencies. Therefore, filter
widths for the lower frequencies have been re-estimated using the EFR by
Supin and Popov (1995c). Using amplitude-modulated tones, the Q IO values
for high-frequency probes (>60kHz) were similar to those obtained using
the transient-evoked ABR. However, at lower frequencies tuning curves
were much narrower than previously obtained, with QIO values greater than
12 at lower stimulus frequencies, down to 1102kHz. Because the authors
used modulation frequencies of 600 Hz it was not possible to get to lower
probe frequencies and these QIO values, and the sharpness of tuning at the
lower frequencies may still be underestimated.
This frequency tuning found in dolphins in much sharper than reported
in other mammals. For instance, the maximal values of QIO obtained using
evoked potential tuning curves in terrestrial mammals and humans were on
the order of 4 to 6 (e.g., Dallos and Cheatham 1976; Mitchell and Fowler
1980; Harrison, Aran, and Erre 1981). These values rival or exceed the Q IO
values obtained from single-unit studies in nonecholocating mammals, as
well as those of highly specialized constant frequency/frequency modulated
(i.e., CF/FM) bats.
3.2.2 Threshold Estimation and Audiograms
It is often of importance to determine the hearing status of an animal. This
is particularly true for stranded or rehabilitated animals. A hearing assessment may be accomplished with a threshold estimate approach using a
wideband stimulus such as a noise burst, or, for frequency-specific evaluation, by determining a pure-tone audiogram. An audiogram is a measure of
auditory sensitivity over a specified frequency range and is most often
obtained following a behavioral paradigm in which a pure tone at a given
frequency is presented to the subject a some fixed intensity. If the tone is
heard, the intensity is decreased; if the tone is determined to be inaudible,
the intensity is increased. Thus, using an iterative approach, the minimum
sound intensity at which a tone can be heard is determined. The audiogram
319
above the curve minima. An alternative measure, frequently referred to as
B IO (the bandwidth 10dB above the curve minima) is a measure of the width
of the tuning curve. When measured using toneburst ABRs, the shape of
the tuning curves varied somewhat as a function of frequency. For probe
frequencies above approximately 64kHz the QIO values were roughly constant, ranging from 17.3 to 18.2. However, at lower frequencies the tuning
curves became quite broad with Q IO values reaching 4.7 at the 16-kHz probe
condition. In contrast, B IO was approximately constant (3.2 to 3.5 kHz) for
probe frequencies between 16 and 45 kHz; above this frequency the B IO
enlarged proportionally to the frequency. Thus, below 45 kHz QIO was proportional to frequency, whereas above 45 kHz QIO was constant.
The tone burst stimuli used in these experiments, being relatively broad
in frequency due to their abrupt rise times, would likely yield exaggerated
estimates of filter widths at the lower stimulus frequencies. Therefore, filter
widths for the lower frequencies have been re-estimated using the EFR by
Supin and Popov (1995c). Using amplitude-modulated tones, the Q IO values
for high-frequency probes (>60kHz) were similar to those obtained using
the transient-evoked ABR. However, at lower frequencies tuning curves
were much narrower than previously obtained, with QIO values greater than
12 at lower stimulus frequencies, down to 1102kHz. Because the authors
used modulation frequencies of 600 Hz it was not possible to get to lower
probe frequencies and these QIO values, and the sharpness of tuning at the
lower frequencies may still be underestimated.
This frequency tuning found in dolphins in much sharper than reported
in other mammals. For instance, the maximal values of QIO obtained using
evoked potential tuning curves in terrestrial mammals and humans were on
the order of 4 to 6 (e.g., Dallos and Cheatham 1976; Mitchell and Fowler
1980; Harrison, Aran, and Erre 1981). These values rival or exceed the Q IO
values obtained from single-unit studies in nonecholocating mammals, as
well as those of highly specialized constant frequency/frequency modulated
(i.e., CF/FM) bats.
3.2.2 Threshold Estimation and Audiograms
It is often of importance to determine the hearing status of an animal. This
is particularly true for stranded or rehabilitated animals. A hearing assessment may be accomplished with a threshold estimate approach using a
wideband stimulus such as a noise burst, or, for frequency-specific evaluation, by determining a pure-tone audiogram. An audiogram is a measure of
auditory sensitivity over a specified frequency range and is most often
obtained following a behavioral paradigm in which a pure tone at a given
frequency is presented to the subject a some fixed intensity. If the tone is
heard, the intensity is decreased; if the tone is determined to be inaudible,
the intensity is increased. Thus, using an iterative approach, the minimum
sound intensity at which a tone can be heard is determined. The audiogram