Format: ORAL

Brock Mashburn1,2, Margaret M. Hanes3, Lex A. J. Thomson4, Kenneth M. Olsen1, Christine E. Edwards2

1 Washington University in St. Louis, St. Louis, MO, USA 2 Missouri Botanical Garden, St. Louis, MO, USA 3 Eastern Michigan University, Ypsilanti, MI, USA 4 Australian Centre for Pacific Islands Research, Queensland, Australia

The last three decades have seen an explosion of studies showing long-distance dispersal to be a major force in determining the patterns of species distributions around the globe. The increasing prevalence of molecular phylogenies with widespread species sampling, coupled with improved dating methods, has revealed that dispersal and subsequent adaptive radiations are common and occur more rapidly than previously thought. Recent studies in the genus Hibiscus (Malvaceae) have demonstrated repeated dispersal events out of Madagascar alone, indicating that the island is not only a hotspot of biodiversity but also a source of colonizers to numerous continents. Perhaps most striking are the species in Hibiscus section Lilibiscus, a group that arose in Madagascar within the last ~2 million years. Members of Lilibiscus are largely endemic to oceanic islands, with 19 of the 30 species (63%) occurring on three volcanic archipelagos (Mascarene, Fijian, Hawaiian) spread across the globe. A parsimonious explanation for this distribution might be a single dispersal event out of Madagascar to the Mascarenes, from there to Fiji, and finally to Hawaii. We tested this hypothesis using a 2bRAD next-generation sequencing approach with over 700 samples and produced a phylogeny that includes every species in sect. Lilibiscus and closely related outgroups. The resulting species relationships revealed an unexpected colonization pattern, with at least three independent dispersal events out of Madagascar and two-way dispersal between islands in the Pacific and Indian Oceans. This suggests that we may still be underestimating the ability of species to regularly disperse across incredibly long distances and that species disjunctions could be much younger than expected. In addition, we uncovered a surprising biogeographical connection between volcanic archipelagos in the Pacific and Indian Oceans that requires further investigation.