4. Acoustic Communication in Whales and Dolphins
181
above, this 50 to 75 kHz frequency range spans typical peak frequencies of
echolocation clicks recorded in captive settings, and is a little below the 120
to 130 kHz peaks recorded in the wild. Detecting faint echoes from clicks
is probably the auditory problem that is most signal-to-noise limited for
dolphins. The frequency range for most sensitive hearing thus appears to
have been selected to match the typical frequency range for echolocation,
while the frequency range for best frequency discrimination appears to
have been selected to match the frequency range of whistles.
Categorization of pulsed calls of dolphins is even more problematic than
tonal calls. The pulsed calls of dolphins are traditionally split into two categories: echolocation clicks and burst-pulsed calls. The term "burst-pulse"
appears to be an old Navy sonar term for sonar signals that are sent out in
a burst of pulsed sounds and then processed together as a packet. This
would create a reasonable contrast to the normal echolocation clicks of bottlenose dolphins, because these dolphins usually wait for the echo of one
click to return before they send out the next pulse. However, the category
of burst-pulse sounds in the dolphin vocal repertoire is usually described
as sounds used for social communication and emitted in stereotyped bursts
of pulses. This combination of a functional and an acoustic structural definition for burst-pulsed sounds creates lots of problems. For example, Wood
(1953) described bursts of pulses he called "rasps" and "hinges." These
sound relatively stereotyped and are associated with particular behavioral
contexts such as feeding. Rasps and hinges certainly sound like bursts to
our ears, in fact we often perceive the interval between these rapid pulses
as a tone (Watkins 1967). However, dolphins can perceive clicks as separate when they are separated by as little as 250J..ls (Au 1993). Rasps and
hinges are now thought to represent rapid series of echolocation clicks produced as a dolphin closes in on a target such as a fish. By contrast, the jaw
clap is one of the pulsed calls for which there is the best data on social
function. Jaw claps tend to be produced during fights as a threat display
(Overstrom 1983). Yet jaw claps are not a burst of multiple pulses, but rather
are best described as one relatively long low-frequency pulse.
3.1.2 Beluga Whales
Beluga whales, D. leucas, were traditionally called "sea canaries" for their
loquaciousness. They were the first cetaceans whose vocalizations were
recorded for a scientific publication (Schevill and Lawrence 1949, 1950).
Sjare and Smith (1986) analyzed the vocal repertoire of wild beluga whales
recorded in a shallow inlet. Whistles were the most common calls, comprising about two-thirds of the sample. Recchia (1994) conducted a systematic
quantitative analysis and categorization of the vocal repertoire of captive
beluga whales. She found a much lower proportion of tonal signals, about
20%. There appears to be a significant difference in the vocal repertoires of
wild versus captive belugas, although the authors used different definitions
181
above, this 50 to 75 kHz frequency range spans typical peak frequencies of
echolocation clicks recorded in captive settings, and is a little below the 120
to 130 kHz peaks recorded in the wild. Detecting faint echoes from clicks
is probably the auditory problem that is most signal-to-noise limited for
dolphins. The frequency range for most sensitive hearing thus appears to
have been selected to match the typical frequency range for echolocation,
while the frequency range for best frequency discrimination appears to
have been selected to match the frequency range of whistles.
Categorization of pulsed calls of dolphins is even more problematic than
tonal calls. The pulsed calls of dolphins are traditionally split into two categories: echolocation clicks and burst-pulsed calls. The term "burst-pulse"
appears to be an old Navy sonar term for sonar signals that are sent out in
a burst of pulsed sounds and then processed together as a packet. This
would create a reasonable contrast to the normal echolocation clicks of bottlenose dolphins, because these dolphins usually wait for the echo of one
click to return before they send out the next pulse. However, the category
of burst-pulse sounds in the dolphin vocal repertoire is usually described
as sounds used for social communication and emitted in stereotyped bursts
of pulses. This combination of a functional and an acoustic structural definition for burst-pulsed sounds creates lots of problems. For example, Wood
(1953) described bursts of pulses he called "rasps" and "hinges." These
sound relatively stereotyped and are associated with particular behavioral
contexts such as feeding. Rasps and hinges certainly sound like bursts to
our ears, in fact we often perceive the interval between these rapid pulses
as a tone (Watkins 1967). However, dolphins can perceive clicks as separate when they are separated by as little as 250J..ls (Au 1993). Rasps and
hinges are now thought to represent rapid series of echolocation clicks produced as a dolphin closes in on a target such as a fish. By contrast, the jaw
clap is one of the pulsed calls for which there is the best data on social
function. Jaw claps tend to be produced during fights as a threat display
(Overstrom 1983). Yet jaw claps are not a burst of multiple pulses, but rather
are best described as one relatively long low-frequency pulse.
3.1.2 Beluga Whales
Beluga whales, D. leucas, were traditionally called "sea canaries" for their
loquaciousness. They were the first cetaceans whose vocalizations were
recorded for a scientific publication (Schevill and Lawrence 1949, 1950).
Sjare and Smith (1986) analyzed the vocal repertoire of wild beluga whales
recorded in a shallow inlet. Whistles were the most common calls, comprising about two-thirds of the sample. Recchia (1994) conducted a systematic
quantitative analysis and categorization of the vocal repertoire of captive
beluga whales. She found a much lower proportion of tonal signals, about
20%. There appears to be a significant difference in the vocal repertoires of
wild versus captive belugas, although the authors used different definitions