J Syst Evol ›› 1983, Vol. 21 ›› Issue (1): 34-43.

• Research Articles •

### A Preliminary Research on the Numerical Classification of the Genus Panax from China

Xu Ke-Xue, Li De-Zhong

1.  (Laboratory fir the Control of Drugs in Jindongnan Locality Shanxi Province)
• Published:1983-02-18

Abstract: In this paper, 10 species and varieties of the genus Panax from China were studied by means of numerical taxonomic methods. At first, the geometric approaches and statistical treatments were used and some new characters induced by the fundamental characters were defined for the numerical representation of the morphological characters of plants. Consequently, forty-seven morphological, three chemical, one geographical and one cytological characters were adopted. The principal component analysis and the similarity coefficients were computed on the standardized data. Based upon the correlation matrix and the distance matrix respectively, the Q and R cluster analyses were carried out, and UPGMA was used in both Q and R cluster analyses. According to the R cluster analysis, all characters are mainly divided into 5 sets: A, B, C, D and E (Fig. 2). The tree-like diagram illustrates that chemical constituents of triterpenoids and the chromosome numbers are related to some morphological characters, such as the roots, the rhizomes, the seeds and the leaves. It is of interest to note that the thicker the fleshy roots, the larger the seeds and the wider the teeth of leaflet, the higher the content of the tetracyclic triterpenoids of dammarane type it contains. On the other hand, the Q cluster analysis showed that Panax as a whole may be divided into two groups (Fig. 3). The first group includes P. ginseng, P. quinquefolius, P. notoginseng and P. zingiberensis and the second group includes P. stipuleanatus, P. pseudo-ginseng, P. japonicus var. japonicus, P. japonicus var. angustifolius, P. japonicus var. major and P. japonicus var. bipinnatifidus. The results of the computation of principal component analysis indicate that the first principal component consistes of the characters occurring in the sets A and B. It shows that the variation in Panax has two opposite directions. One of them, corresponding to the set A, is represented by the first group, and the other, corresponding to the set B, is by the second group. Finally, some questions about the use of the genus in medicine were discussed.