Digging into the general dynamic model: Island ontogeny predicts shifting micro-evolutionary processes in a Mascarene flowering plant radiation

Island studies have been integral in the development of process-oriented biodiversity models such as the general dynamic model (GDM) of oceanic island biogeography. While empirical tests of the GDM primarily come from phylogeographic studies, tests incorporating comprehensive population-level sampling of island radiations are rare. In this study, we elucidate the evolutionary processes driving the diversification of Hibiscus section Lilibiscus in the Mascarene archipelago using population-level sampling and 2bRAD sequencing. Our goals were to: 1) assess species relationships and resolve taxonomic issues; 2) test patterns of intra- and interspecific genetic differentiation under a model of shifting speciation processes as islands age; and 3) utilize demographic modeling to infer the relative divergence times of populations in the archipelago. We found the Mascarene radiation of sect. Lilibiscus to be monophyletic and confirmed the presence of six morphologically distinct species. Species richness and phylogenetic relationships supported the expectations of the GDM. Namely, we found that morphologically similar populations of the same species on the intermediate-aged island, Mauritius, diverged earlier and showed greater divergence than morphologically disparate species on the youngest Mascarene island, Réunion. These patterns are consistent with the hypothesis that ecological selection may be affecting speciation on the youngest island, drift may be the most important force on the middle-aged island. Although not tested here, given that the oldest island is in a late stage of subsidence, extinction may be the most important evolution force occurring on the oldest island, consistent with the expectation that evolutionary processes occurring across each of these islands may be influenced by ontogenetic stage. We also found evidence that some species in the group may be naturally rare, suggesting that the rarity of species may not only be from recent habitat loss. Our study is the first to use population-level data to test complex predictions of the GDM and demonstrate that the most important evolutionary processes shift depending on island ontogeny.