3月龄卵形鲳鲹形态性状对体质量的影响分析

Mathematical analysis of morphometric attribute effects on body weight for three-month-old Trachinotus ovatus

为了探明卵形鲳鲹(Trachinotus ovatus)幼鱼形态性状对体质量的影响程度,分别测定其全长(TL)、体长(SL)、头长(HL)等11个形态学指标和体质量(BW)。分别计算形态性状测量值和体质量数值之间的通径系数和决定系数,分析各形态性状对体质量性状的关联程度。分析结果显示,卵形鲳鲹被测量形态性状对体重的相关系数皆达极显著水平(P<0.01);在通径分析中,3月龄卵形鲳鲹仅全长、体宽、体高和尾柄高的形态性状测量值对幼鱼体质量数据的通径系数达到极显著水平(P<0.01),可将其确定为影响体质量的主要性状,其中全长对体质量的直接影响最大(P_2=0.493 6);决定系数分析显示,全长对体质量的决定系数最大,达到24.39%。卵形鲳鲹的4个形态性状与体质量性状的相关指数大小为R^2=0.888 1,说明该被选性状是与体质量关联的主要性状。设全长(X_2)、体高(X_3)、体宽(X_4)和尾柄高(X_(10))为自变量,设体质量(Y)为因变量,建立多元回归方程:Y=-32.217 8+0.252 4X_2+0.378 2X_3+0.440 8X_4+0.634 1X_(10),回归预测的估计值与实际值间差异不显著(P>0.05),故该方程可被用于卵形鲳鲹的实际生产。本研究可为卵形鲳鲹选育工作提供理论依据和测量指标。

The relationship between morphometric traits and body weight of Trachinotus ovatus was analyzed using body weight and 12 morphometric traits from three-month-old T. ovatus. In this study, I I body characters such as total length （ TL）, standard length （ SL）, head length （HL） were measured. Results showed that the correlation coefficients between independent variables（ morphometric traits）and dependent variable （ body weight） all reached extremely significant level （ P 〈 0. 01 ）. After progressive multiple regression analysis and partial regression coefficient test, standard length was eliminated from the variable data, because it was co-linear with total length, while head length, snout length, eye cross, snout tip to frontal of gill cleft, caudal peduncle length, and tail length were also kicked out from the variable data as they were not extremely significant in regression equation by multiple regression analysis. The path coefficients of total length, body width, body thickness, caudal peduncle length and caudal peduncle depth reached extremely significant level （P 〈 0. 01 ）. Among them, total length was the most predominant variable to affect body weight （P2 = 0. 493 6）. The results of determinant coefficients analysis revealed that total length had the predominant determinative effect （24.39%）. Whereas the other three morphometric traits exhibited a slight direct effect and significant indirect effect on body weight via total length. The result of determinant coefficients analysis was the same as path analysis. The high value of multiple correlation index R2 at 0. 888 1 between morphometric attributes and body weight suggested that the four selected attributes were practical. The morphometric attributes total length （ X2 ）, body width （ X3 ）, body sickness （ X4 ） and caudal peduncle depth （X10） were used to establish the multiple regression equations as Y= -32. 217 8 ＋0. 252 4X2 ＋0. 378 2X3 ＋ 0. 440 8X4 ＋ 0. 634 lX10. This study provides theoretical evidence and perfect measure targets for the breeding of T. ovatus.