2. Cetacean Ears
53
effort to communication signals (Tyack and Clark, Chapter 4; Herzing,
Chapter 5).
Audiograms are available currently for seven Type II delphinids, one
monodontid (beluga whales, Delphinapterus leucas) and two Type I species
(Amazonian boutu, I. geoffrensis, and the harbor porpoise, P phocoena)
(Nachtigall et aI., Chapter 8). There are no published audiograms for the
largest odontocetes, the sperm whales (Physeteridae) nor for any beaked
whale (Ziphiidae), and relatively little is known about their vocalizations;
they remain unclassified. Because much of the behavioral and electrophysiologic hearing data on cetaceans is covered in other chapters, only the
salient points related to peripheral auditory processing mechanisms and
anatomy are mentioned here.
The total hearing range, frequency resolution, localization, and acuity of
an ear are dictated primarily by peripheral auditory system anatomy.
Current data indicate that odontocetes have a 10 to 12 octave functional
hearing range, compared with eight to nine octaves in the majority of
mammals. Most have best sensitivities above 30 kHz, with some going as
high as 130kHz (Ml/lhl and Andersen 1973; Supin and Popov 1990). Peak
spectra of echolocation types are consistent with the audiometric curves;
that is, the signal peaks are near the best frequency of hearing in audiograms from individuals of the same species tested behaviorally. In addition
to good ultrasonic hearing, odontocetes have good frequency and angular
resolution. Target detection thresholds as small as 5em at 5 m have been
reported, implying an auditory angular resolution of 0.5° although 1° to 4°
for horizontal and vertical resolution are more commonly reported (Au
1993). Minimal intensity discrimination in Tursiops truncatus (bottlenose
dolphin) is 1dB, which equals the average human value. Frequency discrimination varies from 0.28 to 1.4% relative discrimination limens (rDL)
between 1 and 140 kHz; best values are found between 5 and 60 kHz
(Popper 1980). Angular resolution and frequency discrimination in P phocoena (0.5°-1°; 0.1 %-0.2% rDL) are similar to values in microchiropteran
bats and superior to those for T. truncatus and humans (Popper 1980; Kossl
and Vater 1995).
An important aspect of any sensory system is the ability to detect signals
in noise. Critical bands (CB) and critical ratios (CR) are two measures of
the ability to detect masked signals. Fletcher (1940) showed that as the
bandwidth of a masking noise narrows, the target suddenly becomes easier
to detect. If the ear's frequency resolution is relatively poor, there is a broad
skirt of frequencies around the target tone that will initiate a response, and
the CB is large. If the membrane is narrowly tuned, the ear responds only
to a narrow band of frequencies at each point, and the CB is narrow. Critical bands are thought to depend on stiffness variations in the inner ear. In
most mammals, including odontocetes, the critical bandwidths are relatively
constant at 0.25 to 0.35 octaves/mm of basilar membrane (Allen and Neeley
1992; Ketten 1992). Critical ratios are a related measure that are calculated
53
effort to communication signals (Tyack and Clark, Chapter 4; Herzing,
Chapter 5).
Audiograms are available currently for seven Type II delphinids, one
monodontid (beluga whales, Delphinapterus leucas) and two Type I species
(Amazonian boutu, I. geoffrensis, and the harbor porpoise, P phocoena)
(Nachtigall et aI., Chapter 8). There are no published audiograms for the
largest odontocetes, the sperm whales (Physeteridae) nor for any beaked
whale (Ziphiidae), and relatively little is known about their vocalizations;
they remain unclassified. Because much of the behavioral and electrophysiologic hearing data on cetaceans is covered in other chapters, only the
salient points related to peripheral auditory processing mechanisms and
anatomy are mentioned here.
The total hearing range, frequency resolution, localization, and acuity of
an ear are dictated primarily by peripheral auditory system anatomy.
Current data indicate that odontocetes have a 10 to 12 octave functional
hearing range, compared with eight to nine octaves in the majority of
mammals. Most have best sensitivities above 30 kHz, with some going as
high as 130kHz (Ml/lhl and Andersen 1973; Supin and Popov 1990). Peak
spectra of echolocation types are consistent with the audiometric curves;
that is, the signal peaks are near the best frequency of hearing in audiograms from individuals of the same species tested behaviorally. In addition
to good ultrasonic hearing, odontocetes have good frequency and angular
resolution. Target detection thresholds as small as 5em at 5 m have been
reported, implying an auditory angular resolution of 0.5° although 1° to 4°
for horizontal and vertical resolution are more commonly reported (Au
1993). Minimal intensity discrimination in Tursiops truncatus (bottlenose
dolphin) is 1dB, which equals the average human value. Frequency discrimination varies from 0.28 to 1.4% relative discrimination limens (rDL)
between 1 and 140 kHz; best values are found between 5 and 60 kHz
(Popper 1980). Angular resolution and frequency discrimination in P phocoena (0.5°-1°; 0.1 %-0.2% rDL) are similar to values in microchiropteran
bats and superior to those for T. truncatus and humans (Popper 1980; Kossl
and Vater 1995).
An important aspect of any sensory system is the ability to detect signals
in noise. Critical bands (CB) and critical ratios (CR) are two measures of
the ability to detect masked signals. Fletcher (1940) showed that as the
bandwidth of a masking noise narrows, the target suddenly becomes easier
to detect. If the ear's frequency resolution is relatively poor, there is a broad
skirt of frequencies around the target tone that will initiate a response, and
the CB is large. If the membrane is narrowly tuned, the ear responds only
to a narrow band of frequencies at each point, and the CB is narrow. Critical bands are thought to depend on stiffness variations in the inner ear. In
most mammals, including odontocetes, the critical bandwidths are relatively
constant at 0.25 to 0.35 octaves/mm of basilar membrane (Allen and Neeley
1992; Ketten 1992). Critical ratios are a related measure that are calculated