The divergence of male courtship songs in closely related species raises several questions. First, is the genetic basis of variation in the song traits within species qualitatively similar to that between species, or does speciation involve novel genetic processes such as fixation of alleles with a large effect and genomic resetting? Second, what kinds of selection pressures have affected songs during their evolution? And third, what are the physiological constraints in sound production and reception and how do they restrict song evolution?
In the melanogaster group, variation both at the intra- and interspecific level in the song IPI seems to be mainly caused by autosomal genes, but the genes affecting interspecifc variation are different from those contributing to intraspecific variation (Gleason and Ritchie, 2004). Gleason and Ritchie (2004) proposed that the type I architecture of many small-effect genes (Templeton, 1981) combined with bidirectional allelic effects found for the species difference in IPI is most compatible with a history of gradual divergence without strong selection. The montana subgroup of the virilis group offers another kind of song evolution. Here an X chromosomal gene(s) plays a major role in determining the large species differences, e.g. in IPI and PL, and this gene also seems to control the action autosomal song genes (Paallysaho et al., 2003). It still remains to be studied whether this gene also affects song variation within the species.
Recent studies on song variation between populations in different Drosophila species show that songs may vary over a species' distribution range. Stabilising and directional selection leave different signs on the genetic architecture of song traits, and so it is possible to trace whether song differences have evolved as a consequence of genetic drift or whether songs have been affected by directional selection pressures (character displacement, sexual selection) during their evolution. There is no convincing evidence so far of the effects of character displacement on song evolution in Drosophila species, but there is some evidence on the effects of sexual selection. In D. montana a diallel study has revealed strong directional dominance for shorter sound pulses and a higher carrier frequency (Suvanto et al., 2000). Here the direction of dominance was the same as the direction of female preferences for the song, and so female song preferences might be a driving force in song evolution (Ritchie et al., 1998). The question of whether coevolution of male song traits and female preferences for these traits require coordinated changes in both sexes remains to be studied.
Evolution of male courtship songs can be hindered by the restricted ability of males to produce songs or of females to receive them and to recognise different song traits. Ewing (1979a) has argued that in the funebris group, the flies' inability to produce songs with much shorter than 10 msec IPI has caused a bottleneck in song evolution. Also, Gopfert and Robert (2002) have shown that arista tips of the antennae in D. melanogaster are moderately tuned to frequencies around 425 Hz. The restrictions can, however, be broken by evolving new ways to produce and perceive sounds, as has happened in Hawaiian picture-winged Drosophila species where the males can produce high-frequency sound pulses with 5 to 6 msec IPIs and the females can recognise the songs of up to 15,000 Hz (Hoikkala et al., 1989).
Was this article helpful?