中国南部海域不同地理群体条纹斑竹鲨的形态差异分析

Morphological Variations of Chiloscyllium plagiosum in Different Geographical Populations Along Southern China Coast

运用4种多元分析方法对海南琼海、广东雷州、广西北海、福建平潭以及台湾屏东等5个不同地理群体条纹斑竹鲨(Chiloscyllium plagiosum)的19项特征变量进行了群体间差异分析和判别分析.分析结果显示,5个群体间有16项特征变量差异极显著(p<0.01);主成分分析得到的6个主成分对不同群体间总变异方差的贡献率分别为31.644%,14.417%,13.078%,11.499%,8.673%,6.818%,累计贡献率为86.130%;主成分分析和聚类分析结果均显示,福建平潭群体和台湾屏东群体形态最为相似,此外,广东雷州、海南琼海和广西北海3个群体间形态也较相似,但与福建平潭、台湾屏东两群体差异较大.以逐步判别法建立了条纹斑竹鲨5个群体的判别函数,判别准确率分别为:80.0%～100%(P1),85.3%～100%(P2),综合判别率达到93.4%.

The white-spotted bamboo shark Ckioscyllium plagiosum is one of Chinese important economic fishes for its high yields. In recent years, biologists at home and abroad have done many researchs on the white-spotted bamboo shark, but there is still a lack of tile researchs on morpbological data. Studies of the genetic diversity and genetic structure of white spotted bamboo shark in different geographical populations along southern China coas＊ from a morphological point of view may provide improtant basic data for our white-spotted bamboo shark germplasm resource library. Four multivariation analytic methods were used to investigate the morpho- logical variations among 5 stocks of white-spotted bamboo shark C. plagiosum collected from Qionghai in Hainan province, Leizhou in Guangdong province,Beihai in Guangxi province, Pingtan in Fujian province and Pingdong in Taiwan, respectively. AN（）VA indi- cated that 16 morphometric proportional parameters showed significant morphological variations （p^0.01） among the 5 stocks. Through principal component analysis, 6 principal components were established, with their contributory ratio being 31. 644%, 14.417%,13.078%,11.499%,8.673%,6.818% respectively,and the cumulative contributory ratio being 86.130%. The results of clusters analysis and principal component analysis revealed that Pingtan stock and Pingdong stock had similar morphological charac- ters,on thc other hand, those from Leizhou, Qionghai and Beihai seemed each other. Discriminant function of 5 populations were es tablished. I）iscriminant accuracy ratio being 80.0%-100% for P, and 85.3%-100% for Pc,the integrative discriminant accuracy ratio being 93.4%.