Abstract
The Hawaiian silversword alliance (Asteraceae) is an iconic adaptive radiation of 33 species. However, like many island plant lineages, no fossils have been assigned to the clade. As a result, the clade's age and diversification rate are not known precisely, making it difficult to test biogeographic hypotheses about the radiation. In lieu of fossils, paleogeographically structured biogeographic processes may inform species divergence times; for example, an island must first exist for a clade to radiate upon it. We date the silversword clade and test biogeographic hypotheses about its radiation across the Hawaiian Archipelago by modeling interactions between species relationships, molecular evolution, biogeographic scenarios, divergence times, and island origination times using the Bayesian phylogenetic framework, RevBayes. The ancestor of living silverswords most likely colonized the modern Hawaiian Islands once from the mainland approximately 5.1 Ma, with early surviving silversword lineages first appearing approximately 3.5 Ma. In testing the progression rule of island biogeography, we found strong positive evidence of the dispersal process preferring old-to-young directionality, but strong negative evidence for speciation occurring on islands during their young growth phase. This work serves as a general example for how diversification studies benefit from incorporating biogeographic and paleogeographic components.
