The evolutionary history of vines in a neotropical biodiversity hotspot: Phylogenomics and biogeography of a large passion flower clade (Passiflora section Decaloba)

Because of their extraordinary flower and leaf morphology, passion flowers (Passifloraceae) have fascinated naturalists since their discovery. Within the large, diverse (600 species) genus Passiflora is an especially enigmatic and species-rich (120 spp.) subclade, Section Decaloba, which occurs in the Neotropics and has its center of diversity in Andean montane forests. A recent phylogenetic study of Passifloraceae showed that Section Decaloba was monophyletic, but was unable to resolve relationships within the clade, thus preventing inferences of evolutionary history and biogeography. The goal of this study was to elucidate the phylogeny and biogeography of Section Decaloba. We sampled 206 accessions representing 91 of the ~ 120 known species in section Decaloba and four outgroups, with samples derived predominantly from herbarium specimens. We generated DNA sequences using a high-throughput DNA sequencing technique called 2b-RAD, reconstructed the phylogeny, and conducted ancestral area reconstructions to infer the biogeographic history of the group. We recovered predominantly well-supported trees in which species were grouped into two main clades: 1) the Central American clade, within which the majority of nodes well supported and species were monophyletic and 2) the South American clade, a large clade that showed overall lower resolution and included several polyphyletic species and species complexes that need additional research. RASP analysis showed that section Decaloba originated in Central America around 10.4 Ma, and then dispersed to South America, the Greater Antilles, and the Bahamas. The South American clade diversified in the Northern Andes and then dispersed to the rest of South America, and Lesser Antilles. Results suggest that both long-distance dispersal and colonization of newly available habitats (i.e., in the Andes) likely promoted diversification of this clade. This study also illustrates how using herbarium specimens and a RAD-seq approach can produce phylogenies for broadly distributed, highly diverse, and poorly accessible groups of plants where field collections would be unfeasible.