Abstract/Summary

The genus Urophyllum Wall. is a taxonomically problematic genus of Rubiaceae in Thailand due to the lack of a recent taxonomic revision and identification key to the species. This has led to confusion in the identification of species and no conservation status assessments for the genus. The aim of this thesis, therefore, is to produce a taxonomic revision of Urophyllum with an identification key to the species focused on Thailand that will contribute towards the Rubiaceae account in the Flora of Thailand which remains unwritten. An integrative taxonomic approach is applied in this thesis, using both morphological and molecular data for species delimitation. A rigorous approach of combining linear and geometric morphometric data with machine learning was applied to 130 specimens of 13 Urophyllum taxa to test groups (taxa). The results based upon linear morphometric data revealed that 86% of specimens were accurately classified based on the data supplied, with seven taxa classified with 100% accuracy. Geometric morphometrics performed worse with only 60% of specimens being accurately classified. However, a combined data approach improved the accuracy, with 91% of specimens being successfully classified, including 10 taxa with a perfect accuracy. Phylogenetic analyses were also performed on 18 species (39 samples) of Urophyllum to provide support for the morphological classification. Whole plastid genomes and nrDNA cistron sequences were mined from genome skimming data. The results of the phylogenetic analyses reveal an incongruent relationship between plastid and nrDNA. Combining morphometric data with molecular results reveal that nrDNA largely supports the classification of taxa using morphometrics, whilst plastid DNA reveals a geographic pattern. Moreover, the three misclassified taxa from the combined morphometric data were resolved to their own clade using nrDNA. The results from the morphological and molecular investigation were used to inform a taxonomic revision of Urophyllum in Thailand and as the basis to publish new species in Cambodia and Vietnam. This thesis therefore directly contributes towards the Flora of Thailand, and the Flora of Cambodia, Laos and Vietnam. The application of methods in this study also provides a robust framework for testing the species delimitation in a wider taxonomic context.