Multilocus phylogeny and phylogenomics of Eriochrysis P. Beauv. (Poaceae: Andropogoneae): Taxonomic implications and evidence of interspecific hybridization

WELKER, C.A.D.; SOUZA-CHIES T.T.; LONGHI-WAGNER, H.M.; PEICHOTO M.C.; MCKEIN M.R.; KELLOG E.A.

MOLECULAR PHYLOGENETICS AND EVOLUTION

ACADEMIC PRESS INC ELSEVIER SCIENCE

Lugar: Amsterdam; Año: 2016 vol. 99 p. 155 - 167

1055-7903

Species delimitation is a vital issue concerning evolutionary biology and conservation of biodiversity. However, it is a challenging task for several reasons, including the low interspecies variability of markers currently used in phylogenetic reconstructions and the occurrence of reticulate evolution and polyploidy in many lineages of flowering plants. The first phylogeny of the grass genus Eriochrysis is presented here, focusing on the New World species, in order to examine its relationships to other genera of the subtribe Saccharinae/tribe Andropogoneae and to define the circumscriptions of its taxonomically complicated species. Molecular cloning and sequencing of five regions of four low-copy nuclear genes (apo1, d8, ep2?ex7 and ep2?ex8, kn1) were performed, as well as complete plastome sequencing. Trees were reconstructedusing maximum parsimony, maximum likelihood, and Bayesian inference analyses. The present phylogenetic analyses indicate that Eriochrysis is monophyletic and the Old World E. pallida is sister to the New World species. Subtribe Saccharinae is polyphyletic, as is the genus Eulalia. Based on nuclear andplastome sequences plus morphology, we define the circumscriptions of the New World species of Eriochrysis: E. laxa is distinct from E. warmingiana, and E. villosa is distinct from E. cayennensis. Natural hybrids occur between E. laxa and E. villosa. The hybrids are probably tetraploids, based on the numberof paralogues in the nuclear gene trees. This is the first record of a polyploid taxon in the genus Eriochrysis. Some incongruities between nuclear genes and plastome analyses were detected and are potentially caused by incomplete lineage sorting and/or ancient hybridization. The set of low-copy nuclear genes used in this study seems to be sufficient to resolve phylogenetic relationships and define the circumscriptions of other species complexes in the grass family and relatives, even in the presence of polyploidy and reticulate evolution. Complete plastome sequencing is also a promising tool for phylogenetic inference.