The timing and the mode of evolution of wild grapes (Vitis).

Wild grapes are woody climbers, found mostly in temperate regions of the northern hemisphere, comprising the genus Vitis. Despite its importance, the evolutionary history of Vitis is still contentious. Past studies have led to conflicting hypotheses about the phylogeny, speciation events, and biogeographic history of the genus. Here we investigate the evolutionary history of Vitis using data from four chloroplast spacers (trnH-psbA, trnK-rps16, trnF-nahJ, and rpl32-trnL) and the nuclear gene RPB2-I, and we explore mechanisms that could have shaped the observed distribution of current species. Maximum likelihood and Bayesian analyses provided similar results, strongly supporting the presence of two subgenera and suggesting a species clustering within subgenus Vitis that mainly mirrors the disjunction between the Old and New World. Vitis vinifera subsp. sylvestris was found to be sister to the Asian species while three major clades were found in the American species. A network approach confirmed the main geographic groups and highlighted different chloroplast haplotype patterns between Asian and American species. Molecular dating analysis provided the time boundaries to discuss our results. Our study shows wild grape diversification to be a continuous and complex process that concerned the Tertiary as well as the Quaternary, most likely involving both geographical and climatic forces. Local variations in extent and timing of these forces were discussed based on observed differences between groups. In the context of the Tertiary-Quaternary debate, we provide evidence in favor of the "continuous hypothesis" to explain present diversity. Finally, two directions for future research are highlighted: (i) was the earliest grape American or Asian? and (ii) are all modern grape species real?