210
P.L. Tyack and c.w. Clark
cavities. The sperm whale does not appear to fit the prey size prediction
very well, because it feeds on prey smaller than much prey of 0. orca and
no larger than the fish prey of P. crassidens, yet produces clicks with much
lower frequency emphases. Body size thus appears to be a better predictor
of click frequency than prey size in odontocetes. However, until an allometric analysis is performed, it is not possible to discriminate whether this
variation in click frequency simply reflects a correlation with body size, or
whether some animals appear to have specialized click frequencies that
deviate from this correlation. Our view is that these analyses are premature. We need much better data on the acoustic structure of echolocation
signals of odontocetes in the wild, coupled with analyses of the acoustic
properties of their targets, before we can test for specialized associations
between clicks and target properties.
5. Vocal Learning
Perhaps the most unusual auditory specialization in cetaceans involves the
evolution of vocal learning, or the ability to modify vocal output depending upon auditory input. Vocal learning requires neural pathways allowing
feedback between auditory input and neural centers responsible for motor
control ofthe vocal apparatus (Konishi 1970, 1985; Nottebohm 1991). While
vocal learning is critical for the development of language and music in the
human primate, there is very little evidence that nonhuman terrestrial
mammals, including nonhuman primates, are capable of vocal learning
(Janik and Slater 1997). By contrast, even though vocal development has
been little studied in marine mammals, there is strong evidence for vocal
learning in one species of baleen whale, one seal species, and several odontocete species. Some of the strongest evidence for vocal learning comes
from species that have been reported to imitate human-made sounds in
captivity. A few individual harbor seals, Phoca vitulina (Ralls et al. 1985),
and beluga whales, D. leucas (Eaton 1979; Ridgway et al. 1985) have been
reported to imitate the sounds of human speech. Many bottlenose dolphins,
T truncatus, have been shown to imitate human-made whistlelike sounds
(Caldwell and Caldwell 1972; Herman 1980; Richards et al.1984). The vocal
convergence at anyone time within a population of singing humpback
whales, M. novaeangliae, coupled with the rapid changes in the song over
time, provides evidence for vocal learning in these animals (Payne et al.
1983).
Evidence for vocal learning among seals (Ralls et al. 1985) is particularly
interesting from an evolutionary perspective, because the pinnipeds
evolved from a different terrestrial ancestor than the Cetacea. This suggests
that there were at least two independent origins of vocal learning among
marine mammals, perhaps even three, if vocal learning had independent
origins in the toothed and baleen whales. Since vocal learning is so rare
P.L. Tyack and c.w. Clark
cavities. The sperm whale does not appear to fit the prey size prediction
very well, because it feeds on prey smaller than much prey of 0. orca and
no larger than the fish prey of P. crassidens, yet produces clicks with much
lower frequency emphases. Body size thus appears to be a better predictor
of click frequency than prey size in odontocetes. However, until an allometric analysis is performed, it is not possible to discriminate whether this
variation in click frequency simply reflects a correlation with body size, or
whether some animals appear to have specialized click frequencies that
deviate from this correlation. Our view is that these analyses are premature. We need much better data on the acoustic structure of echolocation
signals of odontocetes in the wild, coupled with analyses of the acoustic
properties of their targets, before we can test for specialized associations
between clicks and target properties.
5. Vocal Learning
Perhaps the most unusual auditory specialization in cetaceans involves the
evolution of vocal learning, or the ability to modify vocal output depending upon auditory input. Vocal learning requires neural pathways allowing
feedback between auditory input and neural centers responsible for motor
control ofthe vocal apparatus (Konishi 1970, 1985; Nottebohm 1991). While
vocal learning is critical for the development of language and music in the
human primate, there is very little evidence that nonhuman terrestrial
mammals, including nonhuman primates, are capable of vocal learning
(Janik and Slater 1997). By contrast, even though vocal development has
been little studied in marine mammals, there is strong evidence for vocal
learning in one species of baleen whale, one seal species, and several odontocete species. Some of the strongest evidence for vocal learning comes
from species that have been reported to imitate human-made sounds in
captivity. A few individual harbor seals, Phoca vitulina (Ralls et al. 1985),
and beluga whales, D. leucas (Eaton 1979; Ridgway et al. 1985) have been
reported to imitate the sounds of human speech. Many bottlenose dolphins,
T truncatus, have been shown to imitate human-made whistlelike sounds
(Caldwell and Caldwell 1972; Herman 1980; Richards et al.1984). The vocal
convergence at anyone time within a population of singing humpback
whales, M. novaeangliae, coupled with the rapid changes in the song over
time, provides evidence for vocal learning in these animals (Payne et al.
1983).
Evidence for vocal learning among seals (Ralls et al. 1985) is particularly
interesting from an evolutionary perspective, because the pinnipeds
evolved from a different terrestrial ancestor than the Cetacea. This suggests
that there were at least two independent origins of vocal learning among
marine mammals, perhaps even three, if vocal learning had independent
origins in the toothed and baleen whales. Since vocal learning is so rare