Species delimitation of the liquorice tribe (Leguminosae: Glycyrrhizeae) based on phylogenomic and machine learning analyses

The liquorice tribe Glycyrrhizeae is a leguminous herbaceous group of plants comprised of the genera Glycyrrhiza and Glycyrrhizopsis. Some Glycyrrhiza taxa contain glycyrrhizin, a pharmacologically significant sweet substance that also has applications in crafting industrial materials. Here, we utilized an expanded taxon sampling of Glycyrrhizeae to reconstruct the phylogenetic relationships in the tribe based on genome skimming data, including whole chloroplast genomes, nuclear ribosomal DNA, and low-copy nuclear DNA. We also launched machine learning analysis (MLA) for one species pair with controversial taxonomic boundary. The integrated results indicated Glycyrrhizopsis should be split from Glycyrrhiza, while the former genus Meristotropis should be treated as part of Glycyrrhiza. Glycyrrhizopsis includes two species, Glycyrrhizopsis asymmetrica and Glycyrrhizopsis flavescens, and we recognize 13 species in Glycyrrhiza: Glycyrrhiza acanthocarpa, Glycyrrhiza astragalina, Glycyrrhiza bucharica, Glycyrrhiza echinata, Glycyrrhiza foetida, Glycyrrhiza glabra, Glycyrrhiza gontscharovii, Glycyrrhiza lepidota, Glycyrrhiza macedonica, Glycyrrhiza pallidiflora, Glycyrrhiza squamulosa, Glycyrrhiza triphylla, and Glycyrrhiza yunnanensis. We propose a broader G. glabra that includes former Glycyrrhiza aspera, G. glabra s.s., Glycyrrhiza inflata, and Glycyrrhiza uralensis, and represents the glycyrrhizin-contained medicinal group. Our ancestral state inferences show the ancestor of Glycyrrhiza lacked glycyrrhizin, and the presence of glycyrrhizin evolved twice within Glycyrrhiza during the last one million years. Our integrative phylogenomics-MLA study not only provides new insights into long-standing taxonomic controversies of Glycyrrhizeae, but also represents a useful approach for future taxonomic studies on other plant taxa.