7. Electrophysiological Techniques
317
the auditory nerve in a number of species have been lowpass or bandpass
in shape with corner frequencies in the range of 500 to 700Hz (e.g., Mf/lller
1976; Joris and Yin 1992). In the cochlear nucleus, neurons were found to
respond preferentially to sinusoidally amplitude-modulated frequencies
between 80 and 500Hz (Mf/lller 1974; Frisina et al. 1990; Dolphin and
Marangos 1997). Mf/lller and Rees (1986), in the rat, and Batra et al. (1989),
in the rabbit, have shown that neurons in the inferior colliculus exhibit a
preference for modulation frequencies of less than 120 Hz, with a mean best
amplitude modulated frequency of 87 Hz. Neurons in the medial geniculate
nuclei of the guinea pig (Creutzfeldt, et al. 1980) responded well to modulation frequencies up to 100Hz, while cortical cells failed to follow modulation rates beyond 20 Hz. These studies all support the suggestion that
the EFR arises in different generator sites dependent on the modulation
frequency. Responses to the lower modulation frequencies presumably are
more central in origin than responses to high-frequency modulation.
The latencies of the MTF obtained by Dolphin et al. (1995) in cetaceans
were consistent in direction with results of previous studies in other animals
(e.g., Kuwada et al. (1986) in humans; Dolphin and Mountain (1992) in
Mongolian gerbils) in that responses obtained to low modulation frequency
stimuli had considerably longer latencies than signals with high modulation
frequencies. However, the actual values obtained in cetaceans differed
significantly when one takes into consideration the relative head size of
the different species. In humans, Kuwada et al. found response latencies of
approximately 30ms with modulation frequencies below 55Hz, while high
modulation frequencies (100 to 400 Hz) had 7 to 9 ms latencies. In the gerbil,
three latency regions were identified, corresponding to low (10 to 50Hz,
12ms latencies), intermediate (50 to 200Hz, 6ms latencies), and high
(>200 Hz, inconsistent, very short, latencies) modulation frequencies. In
contrast, the cetaceans, having heads-and therefore neural pathlengthslarger than humans, exhibited latencies intermediate between the human
and gerbil. Hence, despite the relatively larger heads and the correspondingly longer neural pathways of the cetaceans, these animals had a
decreased response latency. These data strongly support the supposition of
specialization of the cetacea for very rapid processing of auditory information, not just in the auditory periphery and brain stem structures as indicated by the ABR latencies, but throughout the entire auditory pathway.
3.2 Measures of Auditory Frequency Processing
3.2.1 Tuning Curves and Auditory Filter Shape
A number of psychophysical studies have attempted to estimate frequency
tuning in cetaceans using measurements of critical bands and the critical
ratio (e.g., Johnson 1968; Johnson et al. 1989) as well as tone-on-tone
masking paradigms (e.g., Johnson 1971). However, such experiments are
317
the auditory nerve in a number of species have been lowpass or bandpass
in shape with corner frequencies in the range of 500 to 700Hz (e.g., Mf/lller
1976; Joris and Yin 1992). In the cochlear nucleus, neurons were found to
respond preferentially to sinusoidally amplitude-modulated frequencies
between 80 and 500Hz (Mf/lller 1974; Frisina et al. 1990; Dolphin and
Marangos 1997). Mf/lller and Rees (1986), in the rat, and Batra et al. (1989),
in the rabbit, have shown that neurons in the inferior colliculus exhibit a
preference for modulation frequencies of less than 120 Hz, with a mean best
amplitude modulated frequency of 87 Hz. Neurons in the medial geniculate
nuclei of the guinea pig (Creutzfeldt, et al. 1980) responded well to modulation frequencies up to 100Hz, while cortical cells failed to follow modulation rates beyond 20 Hz. These studies all support the suggestion that
the EFR arises in different generator sites dependent on the modulation
frequency. Responses to the lower modulation frequencies presumably are
more central in origin than responses to high-frequency modulation.
The latencies of the MTF obtained by Dolphin et al. (1995) in cetaceans
were consistent in direction with results of previous studies in other animals
(e.g., Kuwada et al. (1986) in humans; Dolphin and Mountain (1992) in
Mongolian gerbils) in that responses obtained to low modulation frequency
stimuli had considerably longer latencies than signals with high modulation
frequencies. However, the actual values obtained in cetaceans differed
significantly when one takes into consideration the relative head size of
the different species. In humans, Kuwada et al. found response latencies of
approximately 30ms with modulation frequencies below 55Hz, while high
modulation frequencies (100 to 400 Hz) had 7 to 9 ms latencies. In the gerbil,
three latency regions were identified, corresponding to low (10 to 50Hz,
12ms latencies), intermediate (50 to 200Hz, 6ms latencies), and high
(>200 Hz, inconsistent, very short, latencies) modulation frequencies. In
contrast, the cetaceans, having heads-and therefore neural pathlengthslarger than humans, exhibited latencies intermediate between the human
and gerbil. Hence, despite the relatively larger heads and the correspondingly longer neural pathways of the cetaceans, these animals had a
decreased response latency. These data strongly support the supposition of
specialization of the cetacea for very rapid processing of auditory information, not just in the auditory periphery and brain stem structures as indicated by the ABR latencies, but throughout the entire auditory pathway.
3.2 Measures of Auditory Frequency Processing
3.2.1 Tuning Curves and Auditory Filter Shape
A number of psychophysical studies have attempted to estimate frequency
tuning in cetaceans using measurements of critical bands and the critical
ratio (e.g., Johnson 1968; Johnson et al. 1989) as well as tone-on-tone
masking paradigms (e.g., Johnson 1971). However, such experiments are