a single breeding season. However, the song is not static; progressive
changes occur (Payne and Guinee 1983; Cato 1991). At the start of a breeding season, the song resembles that from the prior year. As the breeding
season progresses, new themes are added and old themes deleted. Thus, the
song at the end of a breeding season may bear little resemblance to that at
the beginning of that year (Payne and Payne 1985). These changes can be
extensive, but it is still sometimes possible to recognize songs from one year
as coming from the same population as songs several years later. Observed
temporal changes do not result from variation among individuals recorded
because individuals recorded over two breeding seasons alter their songs
in concert with others in the population (Guinee et al. 1983). The most striking finding with respect to vocal learning is that most individual males in
the population make parallel changes (Payne and Payne 1985). The general
consensus is that this pattern results from cultural transmission of specific
song characteristics, and indeed it is difficult to imagine how such shared
changes could occur without copying. The mechanism by which males copy
one another and the function of copying is, however, completely unknown.
Additional data supporting the role of social modification come from
studies of geographic variation in song. Three ocean basins—North Pacific,
North Atlantic, and South Pacific—exhibit distinctly different songs.
Recordings within those basins as far as 5,000 km apart show that songs
from different populations have very similar structure and the same annual
changes (Fig. 4.11; Winn et al. 1981). However, studies of several locations
within the North Pacific indicated that whales in Japan, Hawaii, and Mexico
share only a portion of their themes (Helweg et al. 1990). Vocal differences
correspond roughly to genetic differences in the North Pacific. A similar
pattern emerges for humpbacks recorded in the South Pacific in Tonga, New
Caledonia, New Zealand, and Australia, although most themes were shared
(Helweg et al. 1998), except between the east and west coasts of Australia
(Cato 1991).
These patterns suggest a degree of acoustic isolation between populations within ocean basins with some acoustic contact or migratory exchange
(Helweg et al. 1990, 1998). The extent of geographic distance, genetic differentiation, and song differentiation appear to be highly correlated. Thus,
dialects could arise because of genetic differences, social modification, or
both. Currently, we cannot tease apart their separate effects.
Cultural exchange of songs between populations could occur if males sing
occasionally on feeding grounds, which they may share with males from
other breeding populations (Payne and Guinee 1983). Winn et al. (1981)
discuss the possibility that geographic variation arises because populations
in one ocean basin are out of synchrony with those in another, even though
they may have similar song elements in their repertoires. If this hypothesis
is correct, then vocal sharing does not require learned acquisition but is
more reminiscent of whistle sharing in dolphins (Smolker and Pepper
1999). This hypothesis is certainly possible; a large number of elements and
194
J.W. Boughman and C.F. Moss
changes occur (Payne and Guinee 1983; Cato 1991). At the start of a breeding season, the song resembles that from the prior year. As the breeding
season progresses, new themes are added and old themes deleted. Thus, the
song at the end of a breeding season may bear little resemblance to that at
the beginning of that year (Payne and Payne 1985). These changes can be
extensive, but it is still sometimes possible to recognize songs from one year
as coming from the same population as songs several years later. Observed
temporal changes do not result from variation among individuals recorded
because individuals recorded over two breeding seasons alter their songs
in concert with others in the population (Guinee et al. 1983). The most striking finding with respect to vocal learning is that most individual males in
the population make parallel changes (Payne and Payne 1985). The general
consensus is that this pattern results from cultural transmission of specific
song characteristics, and indeed it is difficult to imagine how such shared
changes could occur without copying. The mechanism by which males copy
one another and the function of copying is, however, completely unknown.
Additional data supporting the role of social modification come from
studies of geographic variation in song. Three ocean basins—North Pacific,
North Atlantic, and South Pacific—exhibit distinctly different songs.
Recordings within those basins as far as 5,000 km apart show that songs
from different populations have very similar structure and the same annual
changes (Fig. 4.11; Winn et al. 1981). However, studies of several locations
within the North Pacific indicated that whales in Japan, Hawaii, and Mexico
share only a portion of their themes (Helweg et al. 1990). Vocal differences
correspond roughly to genetic differences in the North Pacific. A similar
pattern emerges for humpbacks recorded in the South Pacific in Tonga, New
Caledonia, New Zealand, and Australia, although most themes were shared
(Helweg et al. 1998), except between the east and west coasts of Australia
(Cato 1991).
These patterns suggest a degree of acoustic isolation between populations within ocean basins with some acoustic contact or migratory exchange
(Helweg et al. 1990, 1998). The extent of geographic distance, genetic differentiation, and song differentiation appear to be highly correlated. Thus,
dialects could arise because of genetic differences, social modification, or
both. Currently, we cannot tease apart their separate effects.
Cultural exchange of songs between populations could occur if males sing
occasionally on feeding grounds, which they may share with males from
other breeding populations (Payne and Guinee 1983). Winn et al. (1981)
discuss the possibility that geographic variation arises because populations
in one ocean basin are out of synchrony with those in another, even though
they may have similar song elements in their repertoires. If this hypothesis
is correct, then vocal sharing does not require learned acquisition but is
more reminiscent of whistle sharing in dolphins (Smolker and Pepper
1999). This hypothesis is certainly possible; a large number of elements and
194
J.W. Boughman and C.F. Moss