ABSTRACT
Experimental work has revealed many genes and pathways affecting flowering time in Arabidopsis thaliana. Despite this, it has been difficult to reveal whether these genes also contribute to the natural variation in flowering-time in the worldwide population. Using a new polygenic association mapping method, we earlier revealed 33 loci associated with flowering time variation in the wild collected A. thaliana accessions from the 1,001-genomes project. Together they explained 66% of the additive genetic variation in time to flowering at 10°C in the greenhouse. Here, we explore how these loci together contribute to the flowering time variation between genetically defined subpopulations collected from the native range of the species. For some loci both alleles were present in all populations, whereas for others one of the alleles was missing in some subpopulations. The globally segregating alleles contributed mostly small individual effects, but together captured the overall pattern of early to late flowering between the subpopulations, illustrating the valuable contribution by polygenic adaptation for flowering time. Several of the loci with local alleles had relatively large effects on flowering and in this way contributed to the more extreme adaptations of some local populations. Several cases of long-range LD between genes in biological pathway were found, indicating a possible role of local coevolution of functionally connected polymorphisms in local adaptation. Overall, this study provides new insights to the polygenic architecture of flowering time adaptation that has facilitated the colonisation of a wide range of ecological habitats by this self-pollinating plant.
