Objective To examine the phylogenetic relationships among Salvia species fromChina and adjacent areas; and investigate the monophyly of each subgenus and section ofChinese Salvia. Method We tested this classification using molecular phylogeneticanalysis of43species of Salvia from China, six from Japan, and four introduced species.The nuclear ribosomal ITS region and three chloroplast regions (rbcL, matK, trnH-psbA)were analyzed by maximum parsimony (MP), maximum likelihood (ML), and Bayesianinference (BI) methods. Results (1) Our results showed that the Chinese (except S.deserta) and Japanese Salvia species formed a well-supported clade; S. deserta fromXinjiang grouped with S. officinalis of Europe in MP, ML and BI tree. In addition, allintroduced Salvia species in China were relatively distantly related to the native ChineseSaliva.(2) In FRPS, both sect. Pleiosphace and sect. Eusphace contain a single species indifferent subgenera. However, in our results, S. deserta (subg. Sclarea, sect. Pleiosphace)grouped with the introduced S. officinalis (sect. Eusphace) in clade I sensu Walker&Sytsma; Salvia trijuga (sect. Drymosphace), S. pauciflora (sect. Eurysphace), and allspecimens of S. plebeia (sect. Notiosphace) formed a clade in our cpDNA and cpDNA+nrDNA trees; in our cpDNA+nrDNA trees; species from subg. Allagospadonopsis andsect. Drymosphace (subg. Sclarea) were all in the same subclade, in which subg.Allagospadonopsis formed a well supported group that included several sect.Drymosphace species (i.e., S. cavaleriei Lévl. var. cavaleriei, S. cavaleriei Lévl. var.simplicifolia Stib., S. plectranthoides Griff, and S. yunnanensis). Conclusions (1) TheChinese Salvia were not monophyletic; S. deserta (sect. Pleiosphace) from Xinjiang wasclosely related to the European S. officinalis Linn (clade I) and Chinese (except S. deserta)and Japanese Salvia might own shared origin.(2) Sects. Eusphace and Pleiosphaceshould be classified into a new subgenus; sect. Notiosphace (S. plebeian R.Br.) should bean independent subgenus rather than being included in subg. Sclarea; there was no clearboundary between subg. Allagospadonopsis and sect. Drymosphace (subg. Sclarea). |