基于统计的方法对苹果属山荆子组的分类研究-植物学专业论文.docxVIP

聊城大学硕士学位论文 ABSTRACT The quantitative character classification is an important method in the study of plant system. More than 75% of the Malus Mill. is an special species of China, which has great significant to Malus Mill.taxonomy system. In the paper forty-five characters of leaves, flowers and fruits were observed from thirty-seven wild populations of Malus Mill. (involved thirteen taxa). The descriptive analysis of characteristics, the dendrogram of cluster analysis and the principal component were performed for data processing of Sect. Gymnomeles Koehne, to assess the importance of character, and pick up the proper characters to determine the relationship between different taxa. Different population characters has slightly difference in Sect.Gymnomeles Koehne. But the characters which from different population are continuous transition, can not be as taxonomic characters. Some quantitative characters which show strong positive linear correlation and explain the relationship between these characters is stable, can provide convenience for the overall variability research. According to clustering analysis on several characters, different taxa are cross distribution. In the same species, the same poupulation is the same position. The result show that Sect.Gymnomeles Koehne shouldn’t divided and downgrade the level of Malus hupuensis. In a same species, the character diversity inner population is smaller than that of inter populations. The principal component analysis and clustering figure show that principal component analysis based on the large coefficient of characters as classification properties are not suitable. Considering the classification should be stable variation characters, the middle coefficient of characters are the best classification properties of the classification. Nine characters were important for the classification of the genus Malus in autumn and eight characters were used to identity in spring. Key words: Malus Mill.Sect.Gymnomeles Koehne; qua