Evolutionary Genomics of the Genes Underlying Phenotypic Associations for Forest Trees

Natural populations of forest trees exhibit striking phenotypic adaptations along diverse environmental gradients, thereby making them appealing subjects for the study of genes underlying ecologically relevant phenotypes. A multitude of studies have linked phenotypic variation with genotypes at single-nucleotide polymorphisms (SNPs). The way in which these polymorphisms have typically been discovered enables a two-fold analysis with respect to patterns of diversity and divergence – one at the level of SNPs and one at the level of genes containing these SNPs. I will use this two-fold approach to address the presence of non-neutral patterns of evolution at the genes containing SNPs associated with phenotypes in loblolly pine (Pinus taeda L.). Specifically, I will test the null hypothesis that the more than 500 significant genetic associations discovered for loblolly pine phenotypes are located in genes with patterns of diversity and divergence predicted by the neutral theory of molecular evolution. A suite of analyses based on both polymorphism and divergence data were used to test this null hypothesis. In all analyses, the genes containing SNPs associated with phenotypes appeared non-neutral. The kind of phenotype under consideration, moreover, affected the prevalence and magnitude of deviations from neutral predictions, so that cellular phenotypes (gene expression levels and primary metabolite concentrations) relative to whole plant phenotypes (disease resistance and drought tolerance) were associated with SNPs located in genes with weaker signals of positive selection. The ramifications for the interpretations and design of future genome-wide association studies in conifers will be discussed.