摘要
采用聚类分析、主成分分析等多元分析方法,以体质量和12个形态性状为指标,对分别采自中国大连(DL)、莱州湾(LZ)、海州湾(HZ)和象山湾(XS)海区的4个日本蟳(Charybdis japonica)地理群体进行比较分析。结果表明,日本蟳雄性个体比雌性个体大,雌、雄间的差异显著(P<0.05);聚类分析结果表明,莱州湾群体(LZ)和大连群体(DL)间的遗传距离值最小,象山群体(XS)与莱州湾群体(LZ)间的遗传距离值最大。主成分分析显示,4个群体雌蟹的前5个主成分差异贡献主要集中在大螯相关指标和甲长,累积贡献率为74.021%,群体间的差异受大螯和甲长的相关参数影响最大,而雄蟹前5个主成分的累积贡献率为73.212%,群体间的差异主要表现在第一侧齿和第一步足方面,主成分散点图的结果与聚类分析结果一致。逐步判别分析结果显示,4个群体雌蟹的综合判别率为84%,其中以大连群体(DL)最低(64%),象山湾群体(XS)最高(100%);4个群体雄蟹的综合判别率为70%,其中莱州湾群体(LZ)和大连群体(DL)的判别率最低(56%),象山湾群体(XS)的最高(100%)。研究结果表明,不同地区日本蟳群体间已经产生了一定程度的形态差异,但这些差异尚未达到亚种水平。本研究旨在为日本蟳不同地理种群的识别、亲缘关系的比较、种质资源的保护和利用等提供基础资料。
Geographic variation of twelve morphological traits and the body weights of Charybdis japonica was analyzed using multivariate methods, including cluster analysis, principal component analysis, and discriminant analysis. Four wild populations, located in Dalian, Laizhou Bay, Haizhou Bay, and Xiangshan Bay, were surveyed There were significant differences between male and female populations （P〈0.05）. Cluster analysis revealed that the Dalian stock was most closely related to the Laizhou Bay stock, and that the Xiangshan Bay stock was most distantly related to the Laizhou Bay stock. The cumulative variance contribution ratio, in terms of the five princi- pal components, was 74.021% for female populations and 73.212% for male populations. In discriminant analyses, the total accuracy rate for female population was 84%. The accuracy rate was lowest for the Dalian stock （64%） and highest for the Xiangshan stock （100%）. The total accuracy rate of discriminant analyses on male populations was 70%. It was lowest for the Laizhou Bay and Dalian stocks （56%） and highest for Xiangshan Bay stock （100%） All of the analyses indicated that there were morphological differences among the four populations of C. japonica, which were therefore characterized as subspecies.
出处
《中国水产科学》
CAS
CSCD
北大核心
2012年第4期604-610,共7页
Journal of Fishery Sciences of China
基金
国家虾产业技术体系资助项目(CARS-47)
农业科技成果转化资金项目(2010GB23260589)
山东省科技发展计划项目(011GGF01005)
关键词
日本蟳
地理群体
形态差异
多元分析
Charybdis japonica
wild population
morphological traits
multivariate analysis
