作　　者： ; ; ; ;

机构地区： 广东海洋大学水产学院

出　　处： 《广东海洋大学学报》 2009年第6期7-11,共5页

摘　　要： 运用差异系数检验、主成分分析、判别分析和聚类分析等4种多元统计分析方法,采用5个比例性状,比较研究了广西北海、广东湛江和汕尾的3个翡翠贻贝野生种群的形态性状差异。结果表明:(1)3个种群间5个比例性状的差异系数在0.024～1.134之间,种群间尚未达到亚种差异水平;(2)主成分分析构建了3个主成分,PC1、PC2和PC3贡献率分别为40.134%、27.956%和20.614%,累计贡献率为88.704%;(3)利用逐步判别法选择3个贡献率较大的性状进行判别分析,建立了3种群判别函数,判别准确率为P1为73.0%～88.0%,P2为74.6%～86.8%,综合判别率为79.7%;(4)聚类分析显示,北海种群与汕尾种群形态较为接近,湛江种群趋异程度最大。各种分析方法从不同角度反映了群体间的形态学差异,这种差异是环境和遗传共同作用的结果。 Based on 5 morphological proportional characters, four multivariation analysis methods （difference coefficient test, principal component analysis, discriminant analysis and cluster analysis） were used to make comparative investigation of the morphological difference among three wild populations of Perna viridis in Bei Hai （BH）, Zhan Jiang （Z J） and SanWei （SW）. The results showed that： （1） The values of difference coefficient test were from 0.024 to 1.134, showing that the populations have not yet reached the discriminant level of different subspecies. （2） In the principal component analysis, three principal components were constructed. The contributory ratio of the PCI, PC2 and PC3 was 40.134%, 27.956% and 20.614%, respectively and the cumulative contributory ratio was 88.704 %. （3） The stepwise discrimination analysis revealed that the three populations differed significantly in morphology. Then the discrimination functions of the three populations were set up, the discrimination accuracy being 73.0%-88.0% for P1 and 74.6%-86.8% for P2, the integrative discrimination accuracy being 79.7%. （4） The results of cluster analysis indicated that the morphological characters were similar between the populations of BH and SW, but quite different from that of ZJ. In this study, all the results have revealed some morphological difference among the three populations, which resulted from the joint effect of genetic components and environment.

关 键 词： 翡翠贻贝 形态差异 主成分分析 判别分析 聚类分析

领　　域： [农业科学] [农业科学]