The role of historic, species-scale versus recent, local-scale demographic processes in explaining population genomic diversity in Arabidopsis lyrata

Much research has been conducted to understand how genetic diversity is shaped by local demographic factors such as census size or mating system. However it remains often unclear to what extent within-population genetic diversity also bears the imprint of historic, range-dynamics-related demographic processes. Furthermore little is known how strongly such putatively neutral demographic processes affect intergenic and coding regions of the genome. In this study we estimated genomic diversity of 52 populations of North American Arabidopsis lyrata across its entire distribution at different functional regions of the genome. Population-phylogeographic analysis revealed the splitting of three old clades – southern Missouri, the Midwest and the eastern US along the Appalachians, as well as more recent splitting along the expansion routes since the end of the last glaciation cycle, from out of Wisconsin to the north and east to Lake Erie, and from the central Appalachians to the north and south. Best predictors of genomic diversity were the local mating system (selfing compared to outcrossing) and historic range dynamics since the last glaciation cycle. Historic demographic processes pre-dating the last glaciation cycle and admixture between clusters had a much smaller impact and were more visible in intergenic regions. Census size was only positively linked with nucleotide diversity in parts of the distribution area. The study highlights that for a species with a relatively recent expansion history, this history is one of the most important factors explaining diversity in coding and non-coding DNA, with the highest diversity in the area from which expansion happened.