The once and future genomics of sexual selection

250px-PaonroueEvolutionary biology predates genetics. This is well known. One of the major problems with Charles Darwin’s original theory is that its lack of a plausible mechanism of inheritance meant that it was difficult for him to conceive of how heritable variation could be maintained over the generations. A “blending” model, where the offspring are a synthesis of the trait values of the parent, is intuitively appealing, but also implies that all the variation is going to be “mixed away” very quickly. In contrast, a Mendelian genetic framework, where traits are encoded by discrete and particulate units of inheritance, “genes,” illustrates simply how variation can be maintained despite mixing between lineages in sexual organisms. In short, each generation is simply a reconfiguration of the discrete elements of variation of the previous generation (see: Mendel’s laws).

255217Eventually R. A. Fisher fused what had been rival traditions, the Darwinian/biometrical and Mendelian genetical, into a single framework in his The Correlation between Relatives on the Supposition of Mendelian Inheritance. This was further extended in The Genetical Theory of Natural Selection, and elaborated in more detail by the broader coterie of population geneticists in the early to middle decades of the 20th century, culminating in the Neo-Darwinian Synthesis.

The fusion of genetics with evolutionary biology allowed for a deeper investigation of the dynamics of evolutionary process. This is because of the fact that genes are concrete and definable units of evolutionary bookkeeping (the reason that economics is the most prominent of the social sciences also is grounded in the existence of transparent currencies which mediate exchanges). Though there are models of evolutionary processes which do not rely explicitly on genes, when possible genetic models are optimal. The broader population genetic worldview conceives of evolution as change in allele frequencies over time, and as such made evolution measurable in a very concrete sense via genetic analysis. The emergence of molecular methodologies in the 1960s, and genomics in the 2000s, has resulted in progressively more power to understand how evolutionary change affects the distribution of genetic variation amongst organisms. With genome-wide analyses now available researchers can ascertain the power of selection (positive, negative, background, and balancing) within natural populations.

But one area of evolutionary biology that has been relatively untouched by the genomics revolution is that of the study of sexual selection. This is a major gap, because sexual selection is an intuitively appealing idea which often serves as a deus ex machina when you have no other explanation on hand. So it was of great interest to me to see this review paper, The locus of sexual selection: moving sexual selection studies into the post-genomics era, in the Journal of Evolutionary Biology. There are several major issues with genomics and sexual selection which are highlighted in this review. First, it seems that many sexually dimorphic traits which are being driven by sexual selection vary due to differences in gene expression across the sexes, possibly due to modifications on regulatory elements or alternative splicing. Simple sequence level analyses then may not be good at capturing these sorts of dynamics. Second, sexual selection can leave different signatures because in some cases there are antagonistic pressures between the two sexes. Additionally, sexual selection is often frequency dependent, rather than a simple positive sweep toward fixation (as noted in the paper, a simple sweep would result in exhaustion of variation, meaning sexual selection is a very ephemeral phenomenon). Finally, there is extensive discussion of the utilization of GWAS to discover loci associated with mating fitness. Much of this work has already been done in Drosophila.

Which brings me to the point that from reading this review I have a hard time believing that sexual selection is a strong force for humans for most of history. The reason being that our reproductive skew is just not that notable in comparison to the experimental models cited within the paper. But it seems to me that a better understanding of the relationship between sequence level and regulatory variations in humans could get at this question indirectly, since there are still live debates as to the long term nature of human mating patterns. Presumably if sexual selection was copious then there’d be more extant regulatory variation, perhaps maintained by balancing selection.

Citation: Wilkinson, Gerald S., et al. “The locus of sexual selection: moving sexual selection studies into the postā€genomics era.” Journal of Evolutionary Biology (2015).

Posted in Uncategorized

Comments are closed.