Evolve and resequence provides a granular view of micro-evolution under different mating systems in Mimulus guttatus

Abstract

We performed 10 generations of experimental evolution in Mimulus guttatus and measured genome-wide change in replicated populations that were compelled to reproduce entirely by self-fertilization, entirely by outcrossing, or by a mixture of the two. We developed a novel testing framework based on ancestral haplotype inference to locate mating system loci. Our results confirm several outstanding theoretical predictions: Selfing populations showed increased homozygosity, widespread hitch-hiking, and higher stochastic changes in allele frequencies compared to outcrossing populations. Despite this variability, approximately 20 genomic regions (QTLs) demonstrated parallel evolution across treatments. We identified candidate genes within QTLs using RNA sequencing data from the ancestral lines. In several instances, we found closely linked candidate genes, suggesting that by inhibiting recombination inbreeding can allow for selection on favorable gene combinations. We observed a general down-regulation of candidate genes in selfing populations, mirroring known transcriptome differences between established selfing and outcrossing sister species. This suggests that gene expression is a significant component of the "selfing syndrome."