大菱鲆生长和耐高温性状的遗传参数估计

Estimation of genetic parameters for growth and upper thermal tolerance traits in turbot Scophthalmus maximus

对来自40个家系的753尾大菱鲆个体进行耐高温实验,测定其生长和耐高温性状。基于是否考虑全同胞家系效应,建立了两种动物模型,用于估算大菱鲆生长和耐高温性状的遗传参数。利用WOMBAT程序采用平均信息约束极大似然法估计各模型中性状的方差组分,用似然比检验法进行不同模型的差异检验。方差组分分析结果显示,估计体质量遗传参数时,采用模型Ⅰ效果较好,相应遗传力为(0.22±0.09);估计肥满度时,两模型间差异不显著,其遗传力范围为(0.21±0.18)～(0.39±0.11);估计耐热性时,采用模型Ⅰ比较理想,遗传力为(0.026±0.034)。表型和遗传相关分析结果表明,体质量与肥满度、耐热性的表型相关系数分别为0.13和0.04,肥满度与耐热性的表型相关系数为0.13;体质量与肥满度、耐热性遗传相关系数分别为0.01和-1.00,肥满度与耐热性的遗传相关系数为0.05。研究结果表明,没有考虑全同胞家系效应的模型Ⅰ是估计大菱鲆生长和耐高温性状遗传参数的较好模型。

Genetic parameters of growth and upper thermal tolerance traits of turbot Scophthalmus maximus were analyzed based on the data collected during experiments with 753 individuals from 40 families of turbot that were exposed to a thermal tolerance challenge,and two animal models were established by including a common full-sib effect or inverse.The variance components in each model were estimated with the average information restricted maximum likelihood method using WOMBAT software,and the differences between the two models were tested by likelihood ratio test.The results showed that better estimation was achieved with model Ⅰ for estimating genetic parameters of body weight and upper thermal tolerance,whose heritability was（0.22±0.09）and（0.026±0.034）,respectively.The differences between genetic parameters of relative fatness by the two animal models were not significant,and the range of heritability was（0.21±0.18）-（0.39±0.11）.The correlation analysis showed that the phenotypic correlation coefficients between body weight and relative fatness,body weight and upper thermal tolerance were 0.13 and 0.04,respectively,while the phenotypic correlation coefficient between relative fatness and upper thermal tolerance was 0.13.The genetic correlation coefficients between body weight and relative fatness,body weight and upper thermal tolerance were 0.01 and-1.00,respectively.The genetic correlation coefficient between relative fatness and upper thermal tolerance was 0.05.The results indicated that model Ⅰ without including a common full-sib effect was better than model Ⅱ to be used to estimate the genetic parameters in this study.