The effect of selection intensity on genetic drift variance

The theory of genetic drift has been well established in random mating populations and applied frequently to selected populations as an approximation. However, the magnitude of drift in selected populations may be different from that of random mating populations because of modification of the genetic architectures due to selection in the populations. The relationship between selection and genetic drift has not been properly developed in the literature of quantitative genetics. The objective of this research was to clarify the relationship between selection and drift for quantitative traits by considering both the effect of selection on the additive genetic variance and on effective population size. The new model shows that, for quantitative traits controlled by many genes, each with infinitesimal additive effects on the trait, directional truncation selection leads to an increase in genetic drift variance. This is contrary to the commonly accepted idea that selection is a countering force to genetic drift. The model was also extended to artificial stabilizing selection, natural stabilizing selection (also called optimal selection) and disruptive selection. The general conclusions were: (1) with stabilizing selection, drift variance will decrease as selection intensity increases if selection is symmetric with respect to the overall population mean and (2) with disruptive selection, the reverse will occur. Data from a directional truncation selection experiment with Tribolium castaneum were analyzed. The results showed that, as selection intensity increased from 48% to 40% and 24%, drift variance of pupal weight (estimated by the regression approach) increased from 5.32 to 10.45 and 20.74, respectively. These results qualitatively agreed with predictions of the new theory