三疣梭子蟹不同日龄生长性状相关性及其对体重的影响

利用定向交尾技术构建三疣梭子蟹15个半同胞家系和38个全同胞家系。测量每个家系80 d、100 d和120 d的全甲宽（X₁）、甲宽（X₂）、甲长（X₃）、体高（X₄）和体重（Y）。采用相关分析和通径分析方法计算不同日龄各性状的遗传相关、表型相关、以形态性状为自变量对体重作依变量的相关系数、通径系数、决定系数。结果表明，5个生长性状的表型相关系数在0.759～0.972之间，遗传相关系数在0.170～0.975之间。经显著性检验，各性状间的表型相关和遗传相关都呈极显著水平（P<0.01），各性状不同日龄对体重的直接影响不同，80 d：X₁> X₂> X₃> X₄，100 d：X₁> X₃> X₂> X₄，120 d：X₃> X₂> X₁> X₄， 80 d和100 d时主要通过全甲宽间接影响体重，120 d主要通过甲长和甲宽间接影响体重。建立全甲宽（X₁）、甲宽（X₂）、甲长（X₃）和体高（X₄）对体重（Y）的多元回归方程，不同日龄各生长性状与体重的复相关指数R²分别为0.960，0.971，0.892，说明它们都是直接影响体重的重要指标。

Correlation of growth traits of Portunus trituberculatus at the different ages and its impact on body weight

In order to find out the main growth traits of Portunus trituberculatus which have influence on body weight, and define degree of correlation among characters, thirty eight full-sib groups and fifteen half-sib groups of Portunus trituberculatus were obtained by directional mating techniques. Full carapace width (X_1), carapace width(X_2) ,carapace length(X_3) ,body height(X_4) ,and body weight (Y) of each family on 80 d, 100 d and 120 d were measured. The phenotype correlation coefficients and the genetic correlation coefficients were calculated by correlation analysis at the different ages. The first 4 morphometric attributes (X_1-X_4) were used as independent variables, and body weight (Y) was used as a dependent variable for path analysis. Path coefficients, and determination coefficients were calculated in path analysis. The results showed that the phenotypic correlation of the five traits was between 0. 759 and 0. 972, and the genetic correlation was between 0. 170 and 0. 975. All five correlation coefficients achieved very significant difference(P < 0. 01) levels. The direct effects by the characters have an order at the three stages, 80 d: X_1 >X_2 >X_3 >X_4,100 d: X_1 > X_3 > X_2 >X_4,120 d: X_3 > X_2 > X_1 > X_4, among them X_1 weighted the most to the body weight on 80 d and 100 d, X_1 and X_3 weighted the most to the body weight on 120 d. Full carapace width which have the largest direct impact on body weight (0.471 9) is the most important indicators for indirect selection of body weight. Multiple regression equation of (X_1,X_2,X_3,X_4) to body weight at different developmental stages was obtained as: 80 d Y = -54. 113 +0. 491X_1 +0. 632 X_2 + 0.325 X_3+0.451 X_4;100d Y= -123.484+0.627 X_1 +0.928X_2 +1.433 X_3 +0.152 X_4; 120 d Y= -158.239 +0.137 X_1 +1.410 X_2 +2.370 X_3 +0.800 X_4. The high compound correlation index(R~2 =0.960, 0.971,0.892) indicated that these growth traits had direct influence on body weight. Efficient scheme of multiple trait selection will be made if we obtain definite quantitative relationship between each two characters. This study provides theoretical evidence and perfect measure target for breeding of Portunus trituberculatus.