Abstract
Orobanchaceae is the largest family among the parasitic angiosperms. It comprises non-parasites, hemi- and holoparasites, making this family an ideal test case for studying the evolution of parasitism. Previous phylogenetic analyses showed that holoparasitism had arisen at least three times from the hemiparasitic taxa in Orobanchaceae. Until now, however, not all known genera of Orobanchaceae were investigated in detail. Among them, the unknown phylogenetic positions of the holoparasites Gleadovia and Phacellanthus are the key to testing how many times holoparasitism evolved. Here, we provide clear evidence for the first time that they are members of the tribe Orobancheae, using sequence data from multiple loci (nuclear genes ITS, PHYA, PHYB, and plastid genes rps2, matK). Gleadovia is an independent lineage whereas Phacellanthus should be merged into genus Orobanche section Orobanche. Our results unambiguously support the hypothesis that there are only three origins of holoparasitism in Orobanchaceae. Divergence dating reveals for the first time that the three origins of holoparasitism were not synchronous. Our findings suggest that holoparasitism can persist in specific clades for a long time and holoparasitism may evolve independently as an adaptation to certain hosts.