J Syst Evol ›› 1982, Vol. 20 ›› Issue (4): 410-420.

• Research Articles • Previous Articles     Next Articles

Three numerical approaches to the classification of Chinese species of genus Abelia

Hu Chia-Ch'i, Chu Ming-Yuan, Shu Tzu-Sheng   

  1. (Fudan University)
  • Published:1982-11-18

Abstract: In the last 10—20 years there has bee n increasing awareness of the problem concerning the aims and practices of taxonomy. In particular, there has been growing interest in the development of numerical methods in biological taxonomy as an aid to making systematics a quantitative science, a step which comes in time to almost every scientific discipline. Numerical taxonomy is the evaluation by numerical methods of the affinity or similarity between taxonomic units and the employment of these affinities in erecting a hierarchic order of taxa. The present rapid development of these ideas is presumably a result of the development of computer techniques. Numerical taxonomic approach has been applied to the studies of entomology and microbiology in China to some extent since 1975. But so far it hasn’t been commonly used in botany. The present report is a preliminary study on 9 spp. of the genus Abelia. A set of binary data with 54 characters is used for computing association coefficient; and a set of quantitative data with 47 characters for distance coefficient and correlation coefficient. For the mathematical models were chosen the non-metric simple matching association coefficient, the geometrical distance of Riemannian space and correlation coefficient. Computational procedures are stepwise presented in detail and computer programmes are written in the background of Algol-60 language. Cluster analysis is compared with simple linkage, average linkage and multi-correlation. The results of DC and CC for 9 spp. of Abelia agree closely with the traditional taxonomy, because the data we collected mainly come from morphological characters. It would seem that the results of quantita tive data are more appropriate for seed plants. It is, therefore, postulated that our programes are complementary and very useful to a wide range of classification entities, such as microbes, animals and plants in present situation in China. In conclusion, a comparison between the conventional taxonomy and numerical taxonomy has been made, and a brief discussion of three problems, i.e. the monothetic versus polythetic, divisive versus agglomerative, weighting versus unweighting.