LAUSR

Abstract Ligusticopsis (Apiaceae, Apioideae) is now considered to have an East-Asia and Sino–Himalaya distribution. The genus was not recognized as a natural and separate genus and was treated as a synonym of Ligusticum both in Flora Reipublicae Popularis Sinicae and Flora of China since first established, though Pimenov et al. have made many taxonomic revisions to Ligusticopsis, phylogenetic relationships between Ligusticopsis and Ligusticum have been in dispute. Thirty-four plastomes and 35 nrITS from Apioideae were analysed by RAxML and MrBayes to reconstruct the phylogenetic relationships, along with carpology of 10 species and comparative analyses of 17 plastomes to investigate the evidence supporting the independence of Ligusticopsis. As a result, nine species suggested to be Ligusticopsis formed a highly supported monophyletic branch (Subclade A) inside Selineae both in maximum likelihood and Bayesian inference; the results of the comparative analyses further supported the monophyly of Subclade A, mainly in the location of genes at the IRa/LSC boundary, the sequence diversity exhibited by various genes (e.g. trnH-GUG–psbA and ycf2) and same codon biases in terminator TAA (relative synonymous codon usage = 1.75). Species in Subclade A also had shared characters in mericarps, combined with other characters of the plant, ‘base clothed in fibrous remnant sheaths, pinnate bracts, pinnate bracteoles longer than rays of umbellule, mericarps strongly compressed dorsally, median and lateral ribs filiform or keeled, marginal ribs winged, and numerous vittae in commissure and each furrow’ should be the most important and diagnostic characters of Ligusticopsis. Our phylogenetic trees and other analyses supported the previous taxonomic treatments of Pimenov et al. that Ligusticopsis should be a natural and separate genus rather than a synonym of Ligusticum.