A number of researchers have explored the possibility that
signal–receiver systems maintain their functional integration through
sharing a central pattern generator. If central pattern generators control
signal–receiver variation, then hybrids should be intermediate relative to
the two parental species in both temporal signal properties and response
to these same properties. Studies of crickets and frogs both offered some
qualified support for this hypothesis. Hoy et al. (1977) and Doherty and
Gerhardt (1983) showed that in crickets and tree frogs hybrids tended to
have intermediate pulse rates relative to the two parental species. Furthermore, in some but not all of the possible comparisons, hybrid females preferred the hybrid calls over the calls of at least one of the parental species.
The interpretation of these studies has been a challenge (Boake 1991).
An alternative to the pleiotropy (central pattern generator) hypothesis is
that signals and receivers are under separate and quantitative genetic
control. If this were the case, one would still expect hybrids to be intermediate. This issue could be resolved through detailed quantitative trait locus
mapping studies (Lynch and Walsh 1998). Another criticism of the central
pattern generator hypothesis was offered by Bauer and Helverson (1987).
If there is a single central pattern generator controlling the signal and
receiver, then their response properties should be thermally linked. The
authors showed, however, that if the head and thorax of a grasshopper are
heated separately, the song and preference are decoupled.
One of the best-characterized genes controlling biological rhythms is the
period locus in Drosophila. This locus is implicated in controlling a variety
of rhythms, such as the circadian rhythms, as well as the fruit flies’ love song
(Hall 1994). The love song is produced by vibrating the wings, and mutants
at the period locus vary in song rhythm. Previous studies had shown that
mutant lines not only differed in love song pattern but in female preference
for the same pattern (Kyriacou and Hall 1986). These results suggested
that the female’s preference for song pattern was one more rhythm
under control of the period locus. More recent studies have suggested that
this is not true, however. Greenacre et al. (1993) examined Drosophila
melanogaster with mutations at the period locus, which predictably altered
the rhythmic pattern of the song. Female mutants retained a preference
for the wild type over the mutant song rhythm. This suggests that the song
rhythm and the preference for song rhythm are under separate genetic
control. Females from a period mutant stock that had been maintained for
over ten years, however, did show preference for the mutant song rhythm.
Thus, song and song preference are able to coevolve in Drosophila, but not
through the pleiotropic effects of the period locus.
6.2. Linkage Disequilibrium
Signals and receivers could also maintain their functional integration if
there were a statistical linkage between genetic variation influencing signal
5. Selection on Signals
251
signal–receiver systems maintain their functional integration through
sharing a central pattern generator. If central pattern generators control
signal–receiver variation, then hybrids should be intermediate relative to
the two parental species in both temporal signal properties and response
to these same properties. Studies of crickets and frogs both offered some
qualified support for this hypothesis. Hoy et al. (1977) and Doherty and
Gerhardt (1983) showed that in crickets and tree frogs hybrids tended to
have intermediate pulse rates relative to the two parental species. Furthermore, in some but not all of the possible comparisons, hybrid females preferred the hybrid calls over the calls of at least one of the parental species.
The interpretation of these studies has been a challenge (Boake 1991).
An alternative to the pleiotropy (central pattern generator) hypothesis is
that signals and receivers are under separate and quantitative genetic
control. If this were the case, one would still expect hybrids to be intermediate. This issue could be resolved through detailed quantitative trait locus
mapping studies (Lynch and Walsh 1998). Another criticism of the central
pattern generator hypothesis was offered by Bauer and Helverson (1987).
If there is a single central pattern generator controlling the signal and
receiver, then their response properties should be thermally linked. The
authors showed, however, that if the head and thorax of a grasshopper are
heated separately, the song and preference are decoupled.
One of the best-characterized genes controlling biological rhythms is the
period locus in Drosophila. This locus is implicated in controlling a variety
of rhythms, such as the circadian rhythms, as well as the fruit flies’ love song
(Hall 1994). The love song is produced by vibrating the wings, and mutants
at the period locus vary in song rhythm. Previous studies had shown that
mutant lines not only differed in love song pattern but in female preference
for the same pattern (Kyriacou and Hall 1986). These results suggested
that the female’s preference for song pattern was one more rhythm
under control of the period locus. More recent studies have suggested that
this is not true, however. Greenacre et al. (1993) examined Drosophila
melanogaster with mutations at the period locus, which predictably altered
the rhythmic pattern of the song. Female mutants retained a preference
for the wild type over the mutant song rhythm. This suggests that the song
rhythm and the preference for song rhythm are under separate genetic
control. Females from a period mutant stock that had been maintained for
over ten years, however, did show preference for the mutant song rhythm.
Thus, song and song preference are able to coevolve in Drosophila, but not
through the pleiotropic effects of the period locus.
6.2. Linkage Disequilibrium
Signals and receivers could also maintain their functional integration if
there were a statistical linkage between genetic variation influencing signal
5. Selection on Signals
251