Search

Supplementary information on samples and PCR markers, supplementary results containing pictures of agarose gels and phylogenetic trees.

Sex differences in dioecious animals are pervasive and result from gene expression differences. Elevated sexual selection has been predicted to increase the number and expression of male-biased genes, and experimentally imposing monogamy on Drosophila melanogaster has led to a relative feminisation of the transcriptome. Here, we test this hypothesis further by subjecting another polyandrous species, D. pseudoobscura, to 150 generations of experimental monogamy or elevated polyandry. We find that sex-biased genes do change in expression but, contrary to predictions, there is usually masculinisation of the transcriptome under monogamy, although this depends on tissue and sex. We also identify and describe gene expression changes following courtship experience. Courtship often influences gene expression, including patterns in sex-biased gene expression. Our results confirm that mating system manipulation disproportionately influences sex-biased gene expression but show that the direction of change is dynamic and unpredictable., Sexual selection on males is thought to favour male-biased gene expression. Here, Veltsos et al. experimentally evolve Drosophila pseudoobscura under different mating systems and, contrary to expectation, most often find masculinization of the transcriptome under monogamy rather than under elevated polyandry.

Sexual selection has the potential to contribute to population divergence and speciation. Most studies of sexual selection in Drosophila have concentrated on a single signaling modality, usually either courtship song or cuticular hydrocarbons (CHCs), which can act as contact pheromones. We have examined the relationship between both signal types and reproductive success using F1–3 offspring of wild- collected flies, raised in the lab. We used two populations of the Holarctic species Drosophila montana that represent different phylogeographic clades that have been separate for ca. 0.5 million years (MY), and differ to some extent in both traits. Here, we characterize the nature and identify the targets of sexual selection on song, CHCs, and both traits combined within the populations. Three measures of courtship outcome were used as fitness proxies. They were the probability of mating, mating latency, and the production of rejection song by females, and showed patterns of association with different traits that included both linear and quadratic selection. Courtship song predicted courtship outcome better than CHCs and the signal modalities acted in an additive rather than synergistic manner. Selection was generally consistent in direction and strength between the two populations and favored males that sang more vigorously. Sexual selection differed in the extent, strength, and nature on some of the traits between populations. However, the differences in the directionality of selection detected were not a good predictor of population differences. In addition, a character previously shown to be important for species recognition, interpulse interval, was found to be under sexual selection. Our results highlight the complexity of understanding the relationship between within-population sexual selection and population differences. Sexual selection alone cannot predict differences between populations. Publisher PDF

In the grasshopper Podisma pedestris, units of the ribosomal DNA (rDNA) multigene family are not identical, but comprise multiple genetic variants. We surveyed this variation using a novel pyrosequencing approach. The history of the study populations is well characterized as the pattern of colonization can be inferred from the distribution of two chromosomal races that invaded from different directions after the last glacial maximum and finally met to form a hybrid zone. This knowledge of the populations' ancestry allows us to draw inferences about the rate of change in rDNA composition. The rDNA data have, in turn, been revealing about the populations' ancestry, indicating a previously unsuspected route of postglacial colonization. The two chromosomal races were found to have genetically distinctive rDNA composition, demonstrating the persistence of differences for thousands of generations. It follows that the hybrid zone represents a natural experiment in which repeated crossing and backcrossing between these different rDNA lineages has occurred for over 8000 generations. The association between chromosomal race and rDNA composition has been broken down within the zone. It therefore appears that rDNA variants move freely across the zone and are not under opposing selection pressures in the two races, as had previously been suspected.