长江口及邻近海域凤鲚4个不同地理群体的形态差异  

Analysis of morphological variations among four different geographic populations of Coilia mystus in the Yangtze River Estuary and its adjacent waters

作　　者：杨琴 赵峰[1,2,3] 宋超 张涛[1,3] 苗中博 庄平[1,3] YANG Qin;ZHAO Feng;SONG Chao;ZHANG Tao;MIAO Zhong-bo;ZHUANG Ping(East China Sea Fisheries Research Institute, Chinese Academy of Fishery Sciences, Shanghai 200090, China;College of Fishery and Life Sciences, Shanghai Ocean University, Shanghai 201306, China;Key Science Observing and Experimental Station of Fisheries Resource and Environment in the Yangtze Estuary and East China Sea, Ministry of Agriculture, Shanghai 200090, China)

机构地区：[1]中国水产科学研究院东海水产研究所,上海200090 [2]上海海洋大学水产与生命学院,上海201306 [3]农业农村部东海与长江口渔业资源环境科学观测实验站,上海200090

出　　处：《海洋渔业》2019年第3期294-303,共10页Marine Fisheries

基　　金：上海市科技兴农重点攻关项目(2017-02-08-00-07-F00075);农业农村部农业财政专项

摘　　要：为研究长江口及其邻近海域凤鲚(Coilia mystus)群体的形态差异,分别采集江苏吕四(吕四群体)、上海崇明(崇明群体)、浙江舟山(舟山群体)和浙江温州(温州群体)近岸海域的凤鲚共160尾,测量获得了11个传统形态可量性状和19个框架结构数据,将其转化为29个形态比例指标,并进行了多元统计分析。非参数检验分析表明:除吻长/体长以外,凤鲚4个群体其它性状比例均存在显著性差异(P<0.05);主成分分析构建了7个反映形态特征信息的主成分,其中主成分1、主成分2和主成分3的贡献率分别为44.27%、16.31%和8.36%,累计贡献率为68.94%,结果显示,影响4个群体形态差异的主要因素是凤鲚躯干部和尾部。判别分析利用贡献率最大的8个参数构建了4个群体的判别方程,温州群体的判别准确率最高,达90.0%,其次是吕四群体(87.5%)、崇明群体(67.5%)和舟山群体(60.0%),综合判别准确率为76.3%。聚类分析发现,崇明群体和舟山群体距离最近,最先聚为一组,后与吕四群体汇为一大组,而温州群体单独聚成一组。研究认为:崇明群体形态与舟山群体最为接近,与吕四和温州群体有显著差异(P<0.05)且差异性逐渐增大。In order to study morphological differences of C. mystus distributed in the Yangtze River Estuary and its adjacent waters, a total of 160 samples of C. mystus were collected from Lvsi (Ls population), Chongming Island of Shanghai (Cm population), Zhoushan (Zs population) and Wenzhou (Wz population).Eleven traditional morphological morphometric characters and nineteen frame structures were measured, and analyzed by multivariate statistical analysis. Non-parametric test analysis showed that there were significant differences( P <0.05) in the proportion and other characters of four populations except SL/BL . Principal component analysis constructed seven principal components reflecting morphological characteristics information, among which principal component 1, principal component 2 and principal component 3 played the main roles, and their contribution rates were 44.27%, 16.31% and 8.36% respectively and the cumulative contribution rate was 68.94%. The results showed that STL and TL were the main factors affecting the morphological differences of four populations. The discriminant equations of four populations were constructed by using eight parameters that had the largest contribution rate. The discriminant accuracy rates were Ls population (87.5%), Cm population (67.5%), Zs population (60.0%) and Wz population ( 90.0 %) respectively. The comprehensive discriminant accuracy rate was 76.3%. Cluster analysis showed that Cm and Zs populations were the closest to each other, which first clustered into a group, then clustered into a large group with Ls population, while Wz population clustered into a single group. It showed that the morphology of Cm population was the closest to that of Zhoushan population, which was significantly different ( P <0.05) from Ls population and Wz population, and the difference increased gradually.

关 键 词：凤鲚 形态 可量性状 框架结构 多元统计分析 

分 类 号：S931[农业科学—渔业资源]