作 者：
李军;骆小年;金广海;刘义新;李敬伟;杨培民
单 位：
辽宁省淡水水产科学研究院;辽宁省水生动物病害防治重点实验室
关键词：
唇;形态性状;相关分析;通径分析;多元回归方程
摘 要：
测量234尾3月龄唇的全长(x1)、体长(x2)、体高(x3)、头长(x4)、吻长(x5)、眼径(x6)、眼后头长(x7)、尾柄长(x8)、尾柄高(x9)、头高(x10)及体质量(y)11个形态学指标。运用相关分析、通径分析和回归分析,分别计算了形态性状的相关系数、影响体质量的主要形态性状的通径系数和决定系数,建立了主要形态性状与体质量的回归方程。试验结果表明,唇的体质量变异系数最大(21.5806%)。各形态性状间的表型相关系数均达极显著水平(P<0.01)。通径分析结果表明,体长、体高、眼径、尾柄高与头高达显著水平,是影响体质量的主要形态性状,其通径系数分别为0.4316、0.3487、-0.0621、0.2071和0.1101。主要形态性状对唇体质量的决定系数中,体长最大,为0.1863,其他性状由大到小依次为体高(0.1216)、尾柄高(0.0429)、头高(0.0121)和眼径(0.0039);两两性状对体质量的决定程度中,体长和体高对体质量的影响最大。复相关分析和回归分析结果表明,体长、体高、眼径、尾柄高和头高5个性状对体质量的相关指数达到0.892,回归关系达到显著水平(P<0.05)。应用逐步多元回归分析建立了以体质量为依变量,体长、体高、眼径、尾柄高和头高为自变量的回归方程:y=-3.778+0.550x2+1.417x3-0.839x6+1.915x9+0.715x10,方程回归关系达到极显著水平(P<0.01)。唇幼鱼体长、体高、眼径、尾柄高、头高和体质量分别为(5.30±0.34)cm,(1.15±0.11)cm,(0.51±0.03)cm,(0.49±0.05)cm,(0.99±0.07)cm和(1.98±0.43)g,在网箱内饲养26个月,体长、体高、眼径、尾柄高、头高和体质量分别为(32.40±2.87)cm,(7.22±0.84)cm,(1.18±0.09)cm,(2.78±0.32)cm,(5.33±0.64)cm和(576.94±143.82)g。
译 名：
Mathematical Analysis of Effects of Morphometric Attributes on Body Weight and Growth of Hemibarbus labeo in Cage Culture
作 者：
LI Jun;LUO Xiao-nian;JIN Guang-hai;LIU Yi-xin;LI Jing-wei;YANG Pei-min;Liaoning Province Key Laboratory of Aquatic Animal Disease Control and Prevention,Freshwater Fisheries Sciences Institute of Liaoning Province;
关键词：
Hemibarbus labeo;;morphometric trait;;correlation analysis;;path analysis;;multiple regression equation
摘 要：
Eleven morphometric attributes including total length(x1),body length(x2),body depth(x3),head length(x4),snout length(x5),interorbital(x6),postorbital head length(x7),caudal peduncle length(x8),caudal peduncle depth(x9),head depth(x10)and body weight(y)were measured in 234threemonth-old Hemibarbus labeo samples.The correlation index,path coefficients and determination coefficient were calculated based on morphometric attributes by correlation analysis,path analysis and multivariate analysis,and the significan morphometric attributes(P<0.05)were used to establish the multiple regression equations.The results showed that the maximal coefficient of variation was found in y(21.5806%),higher than that in variation of the other morphological attributes,with a normal distribution in morphometric attributes.The correlation coefficients of morphological attributes to body weight were all at very significant level(P<0.01).The path analysis revealed that partial regression coefficients of x2(0.4316),x3(0.3487),x6(-0.0621),x9(0.2071)and x1 0(0.1101)were found at significant level(P<0.05).The determination coefficients was larger in x2(0.1863)than that in x3(0.1216),x9(0.0429),x10(0.0121)and x6(0.0039).The determination coefficient of x2and x3to y was higher than that of others.The high value of correlation index between morphometric attributes and body weight was 0.892(r2)with significant level(P<0.05),indicating that the five attributes(x2,x3,x6,x9and x10)were key morphometric attributes to the body weight and practical in the selective breeding of H.labeo.The significant morphometric attributes(P<0.05)were used to establish the multiple regression equation as y =-3.778+0.550x2+1.417x3-0.839x6+1.915x9+0.715x10,in juvenile H.labeo[x2=(5.30±0.34)cm,x3=(1.15±0.11)cm,x6=(0.51±0.03)cm,x9=(0.49±0.05)cm,and x1 0=(0.99±0.07)cm,Y =(1.98±0.43)g].The fish cultured in the cage for 26months had x2=(32.40±2.87)cm,x3=(7.22±0.84)cm,x6(1.18±0.09)cm,x9(2.78±0.32)cm,x10(5.33±0.64)cm and y=(576.94±143.82)g.
