黑鲷形态性状对体质量影响的相关性分析

Analysis of Effects of Morphometric Traits on Body Weight of Acanthopagrus schlegelii

摘要: 随机选取12月龄大小两种不同规格黑鲷各69尾(大规格平均体长16.94 cm、平均体重70.98 g；小规格平均体长13.14 cm、平均体重39.27 g)，分别测定其体质量(Y)和全长(X₁)、体长(X₂)、头长(X₃)、吻长(X₄)、眼径(X₅)、眼间距(X₆)、眼后头长(X₇)、体高(X₈)、体厚(X₉)、尾柄长(X₁₀)、尾柄高(X₁₁)共11个形态性指标。采用相关分析、通径分析和多元回归分析方法，分别计算了黑鲷形态性状对体质量的相关系数、通径系数和决定系数。结果表明：大小不同规格黑鲷体质量和全长、体长、头长、吻长、眼径、眼间距、眼后头长、体高、体厚、尾柄长、尾柄高11个形态性状之间的相关性系数分别为0.292~0.836，0.132~0.464，都达到极显著水平(P < 0.01)；通径分析表明，大规格黑鲷有2个形态性状(全长和体高)对体质量的通径系数达到极显著水平(P < 0.01)，小规格黑鲷仅有眼间距一个形态性状对体质量的通径系数达到极显著水平(P < 0.01)；决定系数分析结果表明，大规格黑鲷全长和体高的决定系数较大，其中全长对体质量的决定系数最大(0.476)，体高主要通过全长影响体质量，小规格黑鲷眼间距的决定系数较大，其中眼间距对体质量的决定系数最大(0.471)。决定系数分析结果与通径分析结果的变化趋势相一致；大小不同规格所选形态性状与体质量的相关指数分别为R² = 0.222、R² = 0.736，说明所选性状是影响体质量的主要性状。应用逐步多元回归分析，经偏回归系数的显著性检验，对大规格黑鲷建立以体质量为因变量(Y)，全长(X₁)和体高(X₈)为自变量的多元回归方程：lgY = 156.340 + 7.968lgX1 + 1.738X8，对小规格黑鲷建立以体质量为因变量(Y)，眼间距(X₆)为自变量的多元回归方程：lgY = 14.041 + 2.343lgX₆。经回归预测大小规格估计值与实际值间的差异不显著(P > 0.05)，该方程可用于黑鲷实际生产，为黑鲷良种选育提供理论依据和测量指标。

Abstract: Two groups of 69 black porgies (large group with average body length 16.94 cm and body weigh 70.98 g; small group with average body length 13.14 cm and body weigh 39.27 g) were selected randomly and measured, which were at the same age (12-month old) but in different sizes. Eleven morphological characteristics were measured, including body weight (Y) and full-length (X₁), body length (X₂), head length (X₃), snout length (X₄), eye diameter (X₅), interorbital distance (X₆), the head length behind two eyes (X₇), body height (X₈), body thickness (X₉), the length of caudal peduncle (X₁₀) and the height of caudal peduncle (X₁₁). By ways of correlation analysis, path analysis and multivariate regression analysis, those parameters were calculated including the correlation coefficient, path coefficient, and coefficient of determination of body weight (Y) and morphological characters of these black porgies respectively. The results show that the correlation coefficients between body weight and the eleven morphological characters above are 0.292~0.836 and 0.132~0.464 respectively among these two different size black porgies, all of which show significant difference (P < 0.01). According to the results of the path analysis, as for these bigger-size black porgies, the path coefficient of body weight (Y) and two morphological indicators, full-length (X₁) and body height (X₈), reaches the most significant point (P < 0.01). As for these smaller-size group, only the path coefficient of body weight (Y) and interorbital distance (X₆) reaches the most significant point (P < 0.01). According to the results of the correlation analysis, the coefficient of determination of body weight (Y) and full-length (X₁) and the coefficient of determination of body weight (Y) and body height (X8) are higher, of which the coefficient of body weight (Y) and full-length (X₁) is the highest (0.476). In addition, the body weight (Y) is mainly influenced by body height (X₈) through full-length (X₁). The coefficient of determination of body weight (Y) and interorbital distance (X₆) among smaller ones is higher, and that of body weight (Y) and interorbital distance (X₆) is the highest (0.471). The analysis of coefficient of determination is consistent with that of the path coefficient. For these different-size black porgies, the coefficient of determination of body weight (Y) and morphological indicators are R² = 0.222 and R² = 0.736 respectively, which indicates that the morphological characteristics we choose are the main characters that affect the body weight (Y) of black porgies. By step-by-step multiple regression analysis, partial regression coefficient and the establishment of main multivariate regression equation are lgY = 156.340 + 7.968lgX₁ + 1.738X₈ (the dependent variable is body weight (Y), the independent variables are full-length (X₁) and body height (X₈)), as well as the establishment of subordinate multivariate regression equation is lgY = 14.041 + 2.343lgX₆ (the dependent variable is body weight (Y), the independent variable is interorbital distance (X₆)). There are a few differences between the estimate of the regression prediction and the experimental data (P > 0.05). We can draw a conclusion that the equation can be used to provide theoretical basis and measurement index for the selection and production of black porgies.