| As the traditional famous precious flower of China, peony is renowned as Peerless Beauty and Queen of Flowers. For a long time, it is the symbol of peace, happiness and prosperity. With the improvement of people's living standards and the development of flower industry, people's demand on the high quality peony cultivars is increasing day by day.Numerical taxonomy is a classification which introduces the mathematics and computer technology into the study of plant taxonomy. Currently, it is widely used in the classification of many ornamental plants, and the results proved to be ideal.In this study, pollen morphologies of 22 different cultivated peony cultivars were observed under LM and SEM, and External morphologies of 96 cultivated peony cultivars were observed and recorded. In order to make known the genetic relationship among different peony cultivar groups, this paper studied the numerical taxonomy on peony using its quantitative indices and external morphologic indices. It is intent to probe the important pollen indices and external morphologic indices in the classification of peony through principal component analysis; provide basic evidence for the classification, identification, genetic breeding, evolution and development and use of peony by the way of numerical taxonomy; and complement the study content of peony's quantitative classification. The results of the experiment are as follows:1. Pollen morphological characters of different cultivar groups. The differences of pollen morphologies among different groups are obliviously. Between the central plains cultivar group and Japan cultivar group, the difference is small, so they have a high genetic similarity. The difference of pollen morphologies between central plains cultivar group and northwest cultivar group is obvious. Therefore, the genetic relationship of central plains cultivar group with Japan cultivar group is closer than that with northwest cultivar group. The above results showed that pollen morphologies could be taken as one of the reference basis of peony cultivar groups' classification.2. Q cluster analysis results. This paper made Q cluster analysis on the 82 peony cultivars. Results show that classifying peony cultivars simply by morphological characters is feasible. Different genetic distance coefficients divided the 82 cultivars into different clustered groups, and the genetic relationship was closer in the same clustered group. The clustering results in Figl shows:the genetic relationship between some central plains cultivars and exotic cultivars (mainly referred to Japan cultivars) is quite close; and some central plains cultivars are clustered to one group with northwest cultivars. The main feature of northwest cultivars is the plaque in the base of petals, and most central plains cultivars in the same clustered group also have plaque or tinged. Clustering Fig1 shows that the genetic relationship between exotic cultivar group (mainly referred to Japan cultivars) and northwest cultivar group is closer than and central plains cultivar group. Synthesizing the morphological characters and pollen characters, this study did Q cluster analysis on 22 peony cultivars. Cluster analysis results in Fig3 show:in different genetic distance coefficients, northwest cultivar group is separated from central plains cultivar group and exotic cultivar group firstly, and then central plains cultivar group is separated from Japan cultivar group. Therefore, northwest cultivar group and central plains cultivar group have a far genetic relationship with exotic cultivar group.To sum up, numerical taxonomy is a feasible supplementary method which overcomes the insufficient of quantitatively determining the group's genetic relationship. Through numerical taxonomy, different cultivar groups could be classified ideally, and the genetic relationship is clearly revealed. This study shows that the genetic relationship of central plains cultivar group with exotic cultivar group (mainly referred to Japan cultivars) is closer than that with northwest cultivar group.3. R cluster analysis results. Characters selecting in the quantitative classification could determine the reliability of the cluster analysis results, therefore, to probe the relationships among different characters and the rationality of the characters' selection, it's essential to do R cluster analysis on the selected characters. R cluster analysis results show:the earlier the characters come together, the relevance is stronger. From Fig2, it could be known clearly that there are quite strong relevance between the length and the type of compound leaf, and between the flower-arch tall and the flower type. So, when selecting characters in cluster analysis, it's feasible to choose one character from each group; however, the other characters were not clustered together, and most of them were independent. Although a few characters were clustered together in the Clustering figure, yet they represent the morphological differences among different organs but no obvious Logical relations, therefore, these characters should be retained. The characters sample size is small, so it needs to be enlarged in the future study. The clustered result shows that the selection of the characters is reasonable basically, which provides quantitative basis for the characters' selection in future.4. Principal component analysis results. Principal component analysis was carried out on pollen and external morphologic characters of peony cultivar in this paper. Results shows that some high contribution rate synthetical characters are:style of exine sculpture, size of aperture, length of polar axis, length of equator axis, P/E, color of petals base, flower diameter, flower-arch tall, florescence, flower color, branches length of current year, leaflets type, crazing mode of terminal leaflet, petiole length, color of petiole socket, leaf width, tinged on the leaves, growth potential, plant traits and so on. Therefore, pollen morphologic characters is more important than external morphologic characters in classification of peony cultivars, which is in line with the conclusion that:pollen morphology is controlled by genes, and influenced less by the environment, so it takes a great deal of information about evolvement and possesses strong genetic stability. So it's widely used in plant classification, phylogenetic development, origin and evolvement, etc. The result also proofs that reproductive organs characters are more important than nutritious organs characters. |
PDF Full Text Request