significant heritabilities for multiple features of calls that serve as individual signatures, especially in bats (Jones and Ransome 1993; Scherrer and
Wilkinson 1993; G.S. Wilkinson and D. Lill unpublished data). The distinctiveness in these species’ calls has a strong genetic basis. Indeed, Scherrer
and Wilkinson (1993) calculated that sufficient heritable variation exists in
the isolation calls of evening bat pups (Nycticeius humeralis) for more than
1,800 pups to produce unique calls; the largest colonies consist of 1,000 bats.
This is a huge amount of variation, none of which is due to learning, and
these calls indicate individual identity (100% correct classification in a discriminant analysis) and relatedness (significant heritability) quite well. A
strong inherited component will lead to acoustic similarity among relatives;
thus, when relatives need to identify each other, heritable individual signatures should be favored.
We expect vocal learning of individual signatures to play a role in the following cases: (1) determining which call types to include in an individual’s
repertoire, and (2) increasing distinctiveness of voices among a specific set
of individuals (for instance, a social group, set of relatives, or territorial
neighbors) to allow for greater precision in individual identification within
that group. In both cases, mechanisms to detect cheaters should be built
in. Determining which call types to produce, especially when the calls are
designed signatures, is likely to rely on learned acquisition. In contrast,
social modification is likely to be the predominant form of vocal learning
for increasing distinctiveness of individual voices. Group mates provide a
benchmark against which to gauge distinctiveness. Learned acquisition is
unlikely here because group mates are the most likely models, and copying
call types from them makes it difficult to achieve the desired outcome of
sounding different from social partners.
Although the specific function of repertoires is not clearly known for
many taxa, learning repertoires may provide different benefits than
learning a single, variable call type and may rely on different neural and
behavioral mechanisms. The form of selection that has molded designed
signatures is also likely to differ from that underlying individual voices
arising from by-product distinctiveness. Identifying the selective factors
that generate these different vocal systems promises to be a difficult but
exciting area of research.
In the case studies described below, we begin by describing the social
biology and the context in which calls are produced and identification of
individuals favored for each species. We present information on call patterns to demonstrate that calls are individually distinctive and are used by
conspecifics in identification. We follow with a discussion of the evidence
that vocal learning produces individual signatures and conclude with a consideration of alternative mechanisms. Little information is available on the
process of vocal learning for individual signatures, but we include it where
we can. In addition to these case studies, we briefly review other taxa where
vocal learning has been studied for calls that are individually distinctive
(Tables 4.1 and 4.2).
4. Comparative Vocal Learning
147
Wilkinson 1993; G.S. Wilkinson and D. Lill unpublished data). The distinctiveness in these species’ calls has a strong genetic basis. Indeed, Scherrer
and Wilkinson (1993) calculated that sufficient heritable variation exists in
the isolation calls of evening bat pups (Nycticeius humeralis) for more than
1,800 pups to produce unique calls; the largest colonies consist of 1,000 bats.
This is a huge amount of variation, none of which is due to learning, and
these calls indicate individual identity (100% correct classification in a discriminant analysis) and relatedness (significant heritability) quite well. A
strong inherited component will lead to acoustic similarity among relatives;
thus, when relatives need to identify each other, heritable individual signatures should be favored.
We expect vocal learning of individual signatures to play a role in the following cases: (1) determining which call types to include in an individual’s
repertoire, and (2) increasing distinctiveness of voices among a specific set
of individuals (for instance, a social group, set of relatives, or territorial
neighbors) to allow for greater precision in individual identification within
that group. In both cases, mechanisms to detect cheaters should be built
in. Determining which call types to produce, especially when the calls are
designed signatures, is likely to rely on learned acquisition. In contrast,
social modification is likely to be the predominant form of vocal learning
for increasing distinctiveness of individual voices. Group mates provide a
benchmark against which to gauge distinctiveness. Learned acquisition is
unlikely here because group mates are the most likely models, and copying
call types from them makes it difficult to achieve the desired outcome of
sounding different from social partners.
Although the specific function of repertoires is not clearly known for
many taxa, learning repertoires may provide different benefits than
learning a single, variable call type and may rely on different neural and
behavioral mechanisms. The form of selection that has molded designed
signatures is also likely to differ from that underlying individual voices
arising from by-product distinctiveness. Identifying the selective factors
that generate these different vocal systems promises to be a difficult but
exciting area of research.
In the case studies described below, we begin by describing the social
biology and the context in which calls are produced and identification of
individuals favored for each species. We present information on call patterns to demonstrate that calls are individually distinctive and are used by
conspecifics in identification. We follow with a discussion of the evidence
that vocal learning produces individual signatures and conclude with a consideration of alternative mechanisms. Little information is available on the
process of vocal learning for individual signatures, but we include it where
we can. In addition to these case studies, we briefly review other taxa where
vocal learning has been studied for calls that are individually distinctive
(Tables 4.1 and 4.2).
4. Comparative Vocal Learning
147