Plastome phylogenomics reveals an early Pliocene North- and Central America colonization by long-distance dispersal from South America of a highly diverse bromeliad lineage.

Understanding the spatial and temporal frameworks of species diversification is fundamental in evolutionary biology. Assessing the geographic origin and dispersal history of highly diverse lineages of rapid diversification can be hindered by the lack of appropriately sampled, resolved, and strongly supported phylogenetic contexts. The use of currently available cost-efficient sequencing strategies allows for the generation of a substantial amount of sequence data for dense taxonomic samplings, which together with well-curated geographic information and biogeographic models allow us to formally test the mode and tempo of dispersal events occurring in quick succession. Here, we assess the spatial and temporal frameworks for the origin and dispersal history of the expanded clade K, a highly diverse Tillandsia subgenus Tillandsia (Bromeliaceae, Poales) lineage hypothesized to have undergone a rapid radiation across the Neotropics. We assembled full plastomes from Hyb-Seq data for a dense taxon sampling of the expanded clade K plus a careful selection of outgroup species and used them to estimate a time- calibrated phylogenetic framework. This dated phylogenetic hypothesis was then used to perform biogeographic model tests and ancestral area reconstructions based on a comprehensive compilation of geographic information. The expanded clade K colonized North and Central America, specifically the Mexican transition zone and the Mesoamerican dominion, by long-distance dispersal from South America at least 4.86 Mya, when most of the Mexican highlands were already formed. Several dispersal events occurred subsequently northward to the southern Nearctic region, eastward to the Caribbean, and southward to the Pacific dominion during the last 2.8 Mya, a period characterized by pronounced climate fluctuations, derived from glacial-interglacial climate oscillations, and substantial volcanic activity, mainly in the Trans-Mexican Volcanic Belt. Our taxon sampling design allowed us to calibrate for the first time several nodes, not only within the expanded clade K focal group but also in other Tillandsioideae lineages. We expect that this dated phylogenetic framework will facilitate future macroevolutionary studies and provide reference age estimates to perform secondary calibrations for other Tillandsioideae lineages.