A NUMERICAL TAXONOMIC STUDY ON TWO GENERA YINSHANIA AND HILLIELLA (CRUCIFERAE)

The present paper is concerned with numerical classification of two genera Yinshania Ma et Y. Z. Zhao and Hilliella Y. H. Zhang et H. W. Li. An original data matrix, in which there are 27 OTUs of two genera with 35 ecological and morphological characters, was prepared. After computing the standardized matrix and the correlation coefficient matrix, cluster analysis for 27 OTUs was carried out by using 9 cluster methods, among which the UPGMA was shown to be the most suitable one by means of the cophenetic correlation coefficients (Table 3)as the criteria for assessing the best result obtained through cluster analysis. The dendrogram of UPGMA is presented in Figure 1.1. The results show that all these taxa with minut pustulese on the testa surface of valves are divided into two groups, i. e. the genus Yinshania (OTUs 1—12) and the genus Hilliella (OTUs 13—27).2. According to the traditional morphological method, the genus Yinshania is divided into two sections, Sect. Yinshania (OTUs 5,7—12) and Sect. Microcarpae Y. H. Zhang(OTUs1—4,6). The silicle of the former is elongate, but the latter is round. In the dendrogram, Y. furcatopilosa (Kuan)Y. H. Zhang(OTU7) and Y. henryi(Oliv.) Y. H. Zhang(OTU5)are both separated from the other two groups(OTUs1—4, 6 and OTUs 8—12), but from biological point of view it should refer to Sect. Yinshania, because its silicle is elongate. In Sect. Yinshania, both OTU5 and OTU7 are treated as a "Series", i. e. Ser. Henryanae Y. H. Zhang, different fromSer. Yinshania(OTUs8—12).3. Y. albiflora Ma et Y. Z. Zhao was reduced to a variety, i. e. Y. acutangula(O. E. Schulz)Y. H. Zhang var. albiflora(Ma et Y. Z. Zhao)Y. H. Zhang (OTU 11), by the first author in June, 1987. In the dendrogram it is very close to the Y. acutangula, R=0.715. The first author examined the type specimen (G. Giraldi 2423) of Cochlearia acutangula O. E. Schulz (=Y.acutangula)at the Botanisches Museum, Berlin-Dahlem, in July, 1987, and found that there are not remarkable differences between Y. albiflora and C. acutangula. The result in dendrogram supports the point that Y. albiflora is a synonym of Y. acutangula (Y. H. Zhang 1990).4. The phenetic relationships among species in the two genera are basically consistent with that of the classical classification But, the numericaltaxonomy reminds us to pay more attention in dealing some taxa.