大黄鱼4个家系的形态差异分析  

Analysis on Morphological Variations with Genealogy of Pseudosciaena crocea

作　　者：丁文超[1] 李明云[1] 管丹冬[1] 冯威[1] 

机构地区：[1]宁波大学生命科学与生物工程学院,浙江宁波315211

出　　处：《宁波大学学报（理工版）》2009年第2期185-190,共6页Journal of Ningbo University：Natural Science and Engineering Edition

基　　金：国家高技术研究发展计划项目(2006AA10A405);长江学者和创新团队发展计划(IRT0734)

摘　　要：采用聚类分析、主成分分析和判别分析3种多元分析方法,结合传统形态学测定和框架测定,对大黄鱼(Pseudosciaena crocea)反交家系、岱衢洋家系、官井洋家系和正交家系进行形态差异比较.可数性状卡方分析表明,4个家系可数性状无显著差异;可量性状和框架数据的聚类分析结果表明,岱衢洋家系、官井洋家系和正交家系之间形态差异较小,而它们同反交家系差异显著;主成分分析提取了3个主成分,结果也表明反交家系与其他家系差异较大,与聚类分析结果相一致;通过对10个贡献率较大的形态变量进行逐步判别分析,建立了4个家系的判别公式,判别准确率P1为83.3%～100%,P2为87.5%～96.8%,综合判别率为92.5%.Traditional morphological data and truss network data are combined to generate three multivariate analytical methods, that is, cluster analysis, principal component analysis and multivariate discrimination analysis, to study the morphological variations with genealogy for four types ofPseudosciaena crocea. Analysis on meristic characters indicates that there are no significant differences among these genealogical findings. The results of cluster analysis of metric characters combined with truss network data suggest that the cross-back genealogy of Pseudosciaena crocea is significantly different from the other three. The result of principal component analysis reveals the same conclusion. Discriminant analysis demonstrates that there are significant differences among these four genealogies. The discriminate formula is established through abstracting ten morphological variables. The identification accuracy of genealogies reaches 83.3%-100% （P1） and 87.5%-96.8% （P2）, and the integrative identification accuracy is found to be 92.5%.

关 键 词：大黄鱼家系 形态差异 多元分析 

分 类 号：S917.4[农业科学—水产科学]