In this paper, 16 morphometric parameters were measured in five populations of soft shell turtle Trionyx sinensis, including Dongting (DT), Yellow River (HH) , Huangsha (HS), Japan (RB) and Luka (LK, the hybrid of DT ? and HH 6 ) ,and analyzed using multiple regression. Multivariate analysis revealed that the DT, HH, and LK populations were grouped in one cluster, while the HS and RB population were grouped in another cluster. According to principal component analysis, the extracted two components with contribution of 51. 93% and 11.02% interpreted 62.95% of the total variation, and the HS and RB populations were separated from the other three populations. In discriminant analysis, 9 out of 16 parameters were selected to establish discriminant formulae for the five populations, with the discriminant accuracy of 61. 7% -89.5%. The discriminant analysis also revealed two major clusters in the five populations of soft shell turtle.%通过对中华鳖Trionyx sinensis洞庭(DT)、黄河(HH)、黄沙(HS)、日本(RB)以及绿卡(LK)(洞庭与黄河杂交子代,DT♀×HH ♂)5个不同群体16项生物学外部形态性状的测量,采用多元分析方法对其进行比较分析.聚类分析结果表明,绿卡与其母本洞庭群体先聚合,然后再与其父本黄河群体聚为一支,而黄沙和日本群体另外聚为一支,然后两支再汇聚；主成分分析得到了2个主成分,其贡献率分别为51.93％和11.02％,累计贡献率达62.95％;在16个测量参数中挑选9个对判别贡献较大的参数建立5个群体的判别函数,判别准确率为61.7％～89.5％,判别分析结果也显示,中华鳖5个不同群体分为两大支.
展开▼
