and macrogeographic scales (Ding et al. 1995). Whistles from three subpopulations in the Gulf of Mexico differed. Comparisons among populations in the Gulf of Mexico, the Gulf of California, Argentina, Japan, and
Australia also indicated that these populations differed from one another.
Other studies failed to find such evidence (Evans and Dreher 1962; Graycar
1976; Steiner 1981). Additional work is necessary to determine the prevalence of dialects in dolphins.
A primary factor favoring vocal learning in dolphins may be individual
identification (Janik and Slater 1997; Janik 1999). In terrestrial animals,
differences in vocal tract morphology and body size often result in individually distinctive vocalizations. Dolphins may not be able to rely on these
by-product mechanisms because the shape of air sacs involved in whistle
production is influenced by pressure, which changes when they dive. Thus,
dive-induced variation may undermine the consistency of individual voice
characteristics (Tyack and Sayigh 1997). Janik (1999) suggests that this
constraint favored individual identification via distinctly different whistle
contours. He argues that vocal learning was therefore favored—individuals
could listen to social companions and develop a whistle contour that differentiated them. This idea is intriguing, but no data exist to test it directly.
Some of the information described above seems consistent with this
hypothesis, but the existence of shared whistles in dolphin repertoires
(McCowan and Reiss 1995a, 1995b) and the occasional matching of
mothers’ whistles by sons (Sayigh et al. 1990, 1995) contradict it. Further
work to test this hypothesis is warranted.
It is clear that dolphins can imitate other sounds and this imitation may
require both forms of vocal learning—learned acquisition and social
modification. What is unclear is how often they use imitation in nature, what
induces them to copy a particular individual or sound, and the functions of
such copying. Perhaps these questions are difficult to address because
whistles from known individuals can only be recorded in limited contexts,
and these contexts may not adequately elicit the full repertoire nor allow
identification of who serves as tutor from the potentially large pool of individuals. More thorough sampling of free-living dolphins, although logistically difficult, would be extremely useful for clarifying the pattern of sharing
among individuals in social groups, across populations, and the extent to
which signature whistles are individually specific. Captive studies are likely
to continue to provide important information on ontogeny, yet they cannot
adequately address some important questions with respect to vocal learning because of constraints in composition of social groups, the small number
of individuals in captivity, and the absence of ecological context.
2.1.3. Primates—Macaque Food and Contact Calls
The vocal behavior of several species of macaques (Macaca spp.) has been
studied to test whether vocalizations are learned. Most macaques live in
large, multimale social groups with strong dominance hierarchies among
164
J.W. Boughman and C.F. Moss
Australia also indicated that these populations differed from one another.
Other studies failed to find such evidence (Evans and Dreher 1962; Graycar
1976; Steiner 1981). Additional work is necessary to determine the prevalence of dialects in dolphins.
A primary factor favoring vocal learning in dolphins may be individual
identification (Janik and Slater 1997; Janik 1999). In terrestrial animals,
differences in vocal tract morphology and body size often result in individually distinctive vocalizations. Dolphins may not be able to rely on these
by-product mechanisms because the shape of air sacs involved in whistle
production is influenced by pressure, which changes when they dive. Thus,
dive-induced variation may undermine the consistency of individual voice
characteristics (Tyack and Sayigh 1997). Janik (1999) suggests that this
constraint favored individual identification via distinctly different whistle
contours. He argues that vocal learning was therefore favored—individuals
could listen to social companions and develop a whistle contour that differentiated them. This idea is intriguing, but no data exist to test it directly.
Some of the information described above seems consistent with this
hypothesis, but the existence of shared whistles in dolphin repertoires
(McCowan and Reiss 1995a, 1995b) and the occasional matching of
mothers’ whistles by sons (Sayigh et al. 1990, 1995) contradict it. Further
work to test this hypothesis is warranted.
It is clear that dolphins can imitate other sounds and this imitation may
require both forms of vocal learning—learned acquisition and social
modification. What is unclear is how often they use imitation in nature, what
induces them to copy a particular individual or sound, and the functions of
such copying. Perhaps these questions are difficult to address because
whistles from known individuals can only be recorded in limited contexts,
and these contexts may not adequately elicit the full repertoire nor allow
identification of who serves as tutor from the potentially large pool of individuals. More thorough sampling of free-living dolphins, although logistically difficult, would be extremely useful for clarifying the pattern of sharing
among individuals in social groups, across populations, and the extent to
which signature whistles are individually specific. Captive studies are likely
to continue to provide important information on ontogeny, yet they cannot
adequately address some important questions with respect to vocal learning because of constraints in composition of social groups, the small number
of individuals in captivity, and the absence of ecological context.
2.1.3. Primates—Macaque Food and Contact Calls
The vocal behavior of several species of macaques (Macaca spp.) has been
studied to test whether vocalizations are learned. Most macaques live in
large, multimale social groups with strong dominance hierarchies among
164
J.W. Boughman and C.F. Moss