Abstract Background and Aims Diospyros (∼700 species) is a pan-tropical genus with centres of diversity in poorly studied areas. Fragmented regional classifications have led to confusion in the understanding of this economically important tropical genus. With 20% of species and only six plastid loci previously sampled (atpB, ndhF, rbcL, trnK-matK, trnL-trnF and trnS-trnG), existing phylogenies have been poorly resolved and supported. Next-generation sequencing (NGS) provides the possibility of broad sampling of herbarium specimens. In this study, the three main aims are to: explore methods for integrating legacy Sanger and new NGS sequence data to make more comprehensive datasets; to build the most complete and well-supported phylogenies of Diospyros to date, and to identify the most informative genes for Sanger-based phylogeny building. Methods Approximately 200 Diospyros and outgroup species, totalling 324 samples were de-novo sequenced with NGS technology. One hundred and eleven newly collected samples resulted in 62 whole plastomes and 49 additional partial sequences while 213 herbarium samples generated 48 whole plastomes and 165 additional partial sequences of the six loci. DNA sequences from further 207 samples from 138 taxa were added from public databases. Single-locus, multi-locus and whole plastome phylogenies were reconstructed using a Maximum Likelihood approximation method.. Key results The six-locus tree elucidates the genus-wide relationships and the whole plastome phylogeny resolves multiple previously ambiguous relationships with robust support. There remain only two unresolved clades, possibly results of rapid radiation and/or hybridisation. Whole plastome analysis shows ycf1 to be the best resolving gene when combined with the six legacy loci. Conclusions We provide a well-resolved phylogeny for Diospyros offering a major advance in understanding evolution of the genus. The integration of whole plastome NGS data and legacy chloroplast sequence data enabled us to generate the most comprehensive view of evolutionary relationships for the genus to date (∼250 Diospyros taxa).
Meeprom et al. (Wed,) studied this question.