【作者】 顾宪明；

【导师】 刘海金；

【作者基本信息】 东北农业大学 ， 动物遗传育种与繁殖， 2008， 硕士

【摘要】 本研究以中国水产科学研究院北戴河中心试验站的野生牙鲆为亲鱼建立家系。在养殖到185日龄、235日龄、345日龄、455日龄、505日龄时对试验鱼的全长、体长、头长、尾柄长、眼后头长、体高、尾柄高、体重和肥满度进行了测定和计算;对影响牙鲆生长性状的主要因素进行了分析,应用育种界首推的非求导约束最大似然法(DFREML)和动物模型最优线性无偏预测(BLUP)方法进行了不同模型的比较研究,估计出生长性状的方差组分和遗传参数,并计算了固定效应值、永久环境效应值和动物个体的加性效应值(估计育种值)。本研究根据估计育种值对所有个体进行遗传评定,结果如下:(1)池、年、季、家系和测定日龄效应对所有生长性状均有显著影响。(2)应用单性状无重复观测值模型,两性状模型和单性状有重复观测值模型(重复率模型)对其进行遗传参数的估计。应用单性状无重复模型估计的遗传力范围:全长0.22～0.30、体长0.23～0.30、头长0.23～0.29、尾柄长0.23～0.29、眼后头长0.26～0.38、体高0.22～0.34、尾柄高0.20～0.29、体重0.19～0.31、肥满度0.22～0.43;应用两性状模型估计的遗传力范围:全长0.26～0.31、体长0.23～0.35、头长0.29～0.38、尾柄长0.26～0.38、眼后头长0.26、体高0.19～0.28、尾柄高0.14～0.29、体重0.20～0.35、肥满度0.11～0.36。应用重复率模型分析的结果为:全长0.3、体长0.29、头长0.32、尾柄长0.28、眼后头长0.35、体高0.32、尾柄高0.25、体重0.34、肥满度0.20。全长、体长、头长、尾柄长、眼后头长、体高、尾柄高、体重,8个生长性状间为显著表型正相关;肥满度与体重在185日龄、235日龄、345日龄时表现着显著负相关,相关系数分别为-0.349、-0.123、-0.282,455日龄表现着不显著的正相关,相关系数为0.074,505日龄表现着极显著的正相关,相关系数为0.167。性状遗传相关表现趋势与表型相关基本相同。(3)三种模型比较的结果表明,单性状有重复有观测值模型为最优模型。(4)各性状的育种值对牙鲆进行遗传评定的名次与综合育种值的评定的名次之间差别不大;各种遗传评定方法之间都存在着显著的秩相关。更多还原

【Abstract】 Samples derived from the research center in BeiDaiHe based in Chinese academy of fishery sciences, based on wild Japanese Flounder to build family. The fish were recorded when they were 185 day, 235 day, 345day, 455 day and 505 day. The study analyzed the major factors influencing the growth trait of Japanese Flounder, compared different models and estimated variance components and genetics parameters of total length, body length, head length, caudal peduncle length, postortital length, body depth, caudal peduncle depth, body weight and condition factor utilizing Derivative-Free Restricted Maximum Likelihood(DFREML) and animal model Best Linear Unbiased Prediction (BLUP).The fixed effects, the permanent environment effects and the fixed effects, the permanent environment effects and the animal genetic effects were calculated. Animal were ranked according to their estimated breeding value (EBV)to provide information for selection. The results showed as following:Pool, year, season, family and test-day significantly affected almost all growth traits.Genetic parameters were estimated with single-trait, multiple-trait and repeatability animal models. Total length, body length, head length, caudal peduncle length, postorbital length, body depth, caudal peduncle depth, body weight and condition factor showed low to moderate heritability estimates ranging from 0.22 to 0.30, from 0.23 to 0.30, from 0.23 to 0.29, from 0.23 to 0.29, from 0.26 to 0.38, from 0.22 to 0.34, from 0.20 to 0.29, from 0.19 to 0.31 and from 0.22 to 0.43 by single-trait animal model, respectively. Total length, body length, head length, caudal peduncle length, postortital length, body depth, caudal peduncle depth, body weight and condition factor exhibited low to moderate heritability by multiple-trait animal model, the heritability ranging between 0.26 to 0.31, between 0.23 to 0.35, between 0.29 to 0.38, between 0.26 to 0.38, between 0.26 to 0.38, between 0.19 to 0.28, between 0.14 to 0.29, between 0.20 to 0.35 and between 0.11 to 0.36, respectively. Heritability was estimated with repeatability animal model for total length, body length, head length, caudal peduncle length, postortital length, body depth, caudal peduncle depth, body weight and condition factor were 0.3, 0.29, 0.32, 0.28, 0.35, 0.32, 0.25, 0.34 and 0.20, respectively. Significant phenotypic positive correlations were found among eight growth traits including total length, body length, head length, caudal peduncle length, postortital length, body depth, caudal peduncle depth and body weight. Significant negative correlations were found at 185, 235 and 345 test-day between condition factor to body weight, the correlations coefficients is -0.349, -0.123 and -0.282, respectively. Insignificant positive correlations and significant positive correlations were found at 455 and 505 test-day whose correlations coefficients are 0.074 and 0.167. The genetic correlations trend of the traits was the same to phenotypic correlations.In comparison of different model, repeatability was the optimal model.The ranks of all kinds of one-trait estimated breeding values were basely same to comprehensive estimated breeding value. And all the genetic evaluation methods had significant correlations.更多还原