Search

The monocot family Cyclanthaceae (Pandanales) comprises ca. 230 known species in 12 genera restricted to the Neotropics. The family has not been the subject of a species‐level molecular phylogenetic study to date, with multiple evolutionary and biogeographic questions about Cyclanthaceae still unanswered. In this study, we address genus‐ and species‐level relationships and the historical biogeography of Cyclanthaceae based on a broadly sampled molecular phylogeny. Two low‐copy nuclear genes (phyC, rpb2) and five plastid regions (atpB‐rbcL, psbA‐trnH, trnL‐trnF, trnQ‐rps16, matK) representing 99 species and all genera of Cyclanthaceae were sampled, spanning the taxonomic and biogeographic diversity of the family. Our results strongly support the monophyly of all Cyclanthaceae genera and confirm previously proposed hypotheses of intergeneric relationships. Infrageneric relationships are generally well supported, with some exceptions in the genus Asplundia that may require a broader sampling to clarify. The early diversification of Cyclanthaceae is estimated to date back to the Paleocene‐Eocene period in South America, where the family possibly arrived through a boreotropical route. The origin of most genera is inferred from the Paleocene of the Tumbes‐Chocó‐Magdalena region, possibly indicating an earlier origin for this biodiversity hotspot. The current distribution of Cyclanthaceae is likely to have been strongly influenced by major biogeographical events in the Neotropics, such as the uplift of the Andes and the opening of the South America dry diagonal. Further studies that include a broader sample of the large Asplundia‐Dicranopygium clade and Sphaeradenia group are required. [ABSTRACT FROM AUTHOR]

• A novel multilocus and highly supported phylogeny clarify several previously uncertain relationships in Myrcia. • Our results provide a solid backbone for further evolutionary and taxonomic work. • Previously unplaced Myrcia species formed a new circumscribed clade. • The first phylogenetic relationships for Myrcia sect. Gomidesia are shown with a morphological trait evolution and geographic range distribution approaches. • The inclusion of previously available molecular data with novel sequences of Myrcia and relatives significantly improved our phylogenetic estimates. Myrcia is the largest exclusively Neotropical genus of the plant family Myrtaceae with c. 770 species. Although several studies have elucidated the relationships within particular sections of the genus, to date no phylogeny has been produced that includes a broad taxonomic and geographic representation. Here we present a phylogenetic hypothesis of Myrcia and close relatives comprising 253 species and based on two nuclear and seven plastid markers. We combine previously available sequence data with 234 new sequences of the genus Myrcia for this study. We use this phylogeny to investigate the evolution of selected morphological traits and to infer the biogeographic history of the genus. Our results yield a highly supported phylogenetic tree where the Myrceugenia group is sister to the Myrcia and Plinia groups. Five Myrcia species previously considered unplaced emerge in a newly circumscribed clade. The monophyly of two Myrcia sections previously considered uncertain, Aulomyrcia and Gomidesia, are confirmed with strong support. Flowers with free calyx lobes, 2-locular ovaries, and anthers with symmetrical thecae are ancestral features of Myrcia. The Myrcia sect. Gomidesia is highly supported and recovered as monophyletic, with asymmetric anthers that retain their curvature after dehiscence as a morphological synapomorphy. The Atlantic Forest is the most likely ancestral area of the genus and most of its internal clades, from where multiple lineages colonized different regions of South and Central America, in particular the Brazilian Cerrado through multiple unidirectional range expansions. The southern Atlantic Forest is the ancestral area for Myrcia sect. Gomidesia , with lineages reaching the northern Atlantic Forest, Cerrado, Yungas, and other savanna vegetation of South America. Our results provide a solid backbone for further evolutionary and taxonomic work and clarify several previously uncertain relationships in this mega-diverse plant group, and shed light on its geographical range evolution. [ABSTRACT FROM AUTHOR]

• Tapinotaspidini broad phylogeny indicates monophyly of most genera. • Tapinotaspidini bees originated in Cerrado-like savannas in the Paleocene. • Transitions to forested areas slightly increase speciation rates. • Malpighiaceae is the ancestral host for floral oils. • Shifts to other oil plant families are possibly linked to habitat shifts. Worldwide distributed tropical savannas were established only in the Miocene, with climatic cooling and rise of C4 grasses. However, there is evidence for an earlier presence of savanna-like vegetation in southern parts of South America. Here we investigated the biogeographic history of a clade of solitary bees which have endemic groups in areas covered by savannas and other types of open vegetation as well as forested areas. We hypothesized that these bees originated in savanna-like biomes and that shifts to forested areas and floral host shifts increased species diversification along their evolutionary history. We reconstructed a comprehensive phylogeny for Tapinotaspidini bees based majorly on original DNA sequences. We then used macroevolutionary tools to estimate ancestral range area and reconstructed ancestral habitat (open versus forested) and host plant association to analyze the effects of shifts in vegetation type and flower hosts on their diversification. Tapinotaspidini bees originated in the Paleocene and in a savanna-type, Cerrado-like, which is reinforced by reconstruction of open vegetation as the most probable ancestral area, thus bringing additional evidence to a much earlier origin of this vegetation type in South America. Shifts to forested areas occurred at least three times in a period of 30 Ma and were responsible for slight increases in diversification rates. Malpighiaceae is the ancestral floral host; host broadening occurred only in the Miocene and at least in three occasions. Host shifts, i.e. from Malpighiaceae to other oil families, occurred in the Eocene and Miocene. Both host broadening and host shifts did not significantly alter diversification rates, however exploitation of other oil sources were important in occupying new habitats. The link between biomes and host plant shifts and changes in diversification rate brings us additional insights into the evolution of bees and associated flora in South America. [ABSTRACT FROM AUTHOR]