Year: 2013
Type: dissertation
Abstract: Much of evolutionary biology attempts to explain why the phenotypes of local populations have diverged from a common ancestral type. It is often tempting to explain the observed pattern of differentiation by natural selection, but this often lacks scientific justification, because random genetic drift i.e. the gradual change of allelic frequencies due to random sampling of alleles from parents to offspring can cause striking phenotypic differentiation, being arguably a simpler and a more eloquent explanation. Random genetic drift needs to be ruled out as an explanation to argue for a scenario of phenotypic differentiation based on natural selection. The rate of random drift in a study system can be measured by using neutral DNA markers, such as microsatellites. Subsequently, these measurements can be compared with the observed pattern of phenotypic differentiation. This is usually done by using so-called QST-FST comparisons. However, these comparisons suffer from a number of conceptual limitations, and their power to detect signals of natural selection is relatively low. This thesis presents a re-thinking of the neutral DNA-phenotype comparison. To this end, the Chapters of this thesis discuss the evolution of randomly drifting quantitative traits, resulting in a probability distribution over the evolutionary process. I use this distribution to construct statistical tests that can be used to detect signals of natural selection in quantitative traits. The method of this thesis is derived from the first principles of quantitative genetics (assuming diploidism and additive genetic architecture), and it is developed into two R-software packages (driftsel and RAFM), intended for users without extensive skills in Bayesian statistics. This thesis also investigates the performance of the methods and the interpretation of the model estimates obtained as a by-product of the neutrality tests. As it turns out, the method developed here can detect signals of natural selection even in cases where usual QST-FST comparisons fail. In addition to the methodological contributions, the work included into this thesis has also an empirical dimension demonstrating occurrence of adaptive differentiation among three-spined (Gasterosteus aculeatus) and environmental adaptation in nine-spined stickleback (Pungitius pungitius) populations.
Source: Työväentutkimus Vuosikirja
Author: Markku Karhunen
Language: English
Subfield: Genetics
Domain: Life Sciences
Open Access status: green
OpenAlex ID: W1496096547