The pantropical genus Begonia is the sixth-largest genus of flowering plants, including 1870 species. The sections of Begonia are used frequently as analogues to genera in other families but, despite their taxonomic utility, few of the current sections have been examined in the light of molecular phylogenetic analyses. We present herein the largest, most representative phylogeny of Begonia published to date and a subsequent provisional sectional classification of the genus. We utilised three plastid markers for 574 species and 809 accessions of Begonia and used Hillebrandia as an outgroup to produce a dated phylogeny. The relationships between some species and sections are poorly resolved, but many sections and deeper nodes receive strong support. We recognise 70 sections of Begonia including 5 new sections: Astrothrix, Ephemera, Jackia, Kollmannia, and Stellandrae; 4 sections are reinstated from synonymy: Australes, Exalabegonia, Latistigma and Pereira; and 5 sections are newly synonymised. The new sectional classification is discussed with reference to identifying characters and previous classifications.

No abstract

Over the past 300 years of plant collecting for herbaria, the basic method of preservation has remained remarkably consistent-a plant press with absorbent paper. However, the difficulty of drying plant specimens in the humid tropics has led to a variety of additions to this basic technique, primarily to prevent the specimens succumbing to fungal and bacterial breakdown. These additions include drying gently in tents over low fires, soaking the specimens in alcohol before pressing and, more recently, drying the specimens with forced heat from hair dryers. The process of drying is, however, known to cause breakage and damage to the plant's DNA, as well as providing time for non-plant organisms (bacteria and fungi) to multiply in the tissue, potentially "swamping" the plant specimen's own DNA. Contemporary plant collectors therefore usually collect a separate sample for DNA work, usually rapidly dried in silica gel desiccant; historical collections, however, may have been treated with alcohol and/or heat. We have recently shown that Hyb-Seq provides a reliable method for retrieving robust sequence data from even very old and degraded herbarium specimens. In this study we have used a panel of specimens preserved using a variety of methods to assess whether Hyb-Seq is capable of retrieving informative amounts of sequence data from duplicate specimens preserved using a range of specimen preservation methods. We present data on the amount and quality of DNA and of sequence data retrieved, the variation in error types between preservation techniques and the utility of the data for phylogenetic analysis.