赤拟谷盗模拟种公畜选择──常规与改进方法的比较  

The Simulation of Sire Evaluation Using Tribolium castaneum for Comparison of Two Statistical Methods

作　　者：张劳[1] 王希斌[1] 张启能 王雅春 刘巍 

机构地区：[1]北京农业大学,北京华牧农科院畜牧所,天津华牧家禽育种中心,华牧家禽育种中心

出　　处：《Acta Genetica Sinica》1994年第6期441-446,共6页

基　　金：国家自然科学基金

摘　　要：本实验用实验昆虫赤拟谷盗两个世代各３０个家系的８５９和７２６只蛹的２３日龄蛹长和蛹宽作为选择性状模拟种公畜选择，对常规方法和改进方法进行比较。以Ｈｅｎｄｅｒｓｏｎ方法Ⅲ和综合指数作为常规方法，用约束最大似然法（ＲＥＭＬ）和ＢＬＵＰ估计方差组分和育种值为改进方法。结果表明，常规法估计遗传参数在两世代中不稳定，而改进方法弥补此缺点。对两世代个体和家系育种值排队，结果表明两种方法在前几名无差别，说明在选择强度大（留种率低）时，两种方法效果基本一致。The simulation of sire genetic evaluation using Tribolium castaneum for the comparison of two statistical methodes: Handerson'S method 3 and REML for the variance components, and the comparison of two methods:selection index and BLUP.The same data set of pupal length and pupal width in two generations for heritabilities and breeding values was analysed by the different methods. Results showed that heritabilities of two traits estimated by Handerson's methods 3 were lower than that estimated by REML and did not remain stable in two generations. The rankinq of each family average breeding value showed that there was no differece among the first five selected families by the two methods. It suggested that the usual meshod may be used for sire evaluation when selection intensities of traits are high.This project supported by National Natural Science Foundation of China ②Present adress: Huamn Toultry Dreeding Center,Tianjin 301700.③Present adress:Institute of Amimal Sciences, CAAS. Beijing 100094

关 键 词：赤拟谷盗 公畜 家畜 育种 实验昆虫 选择 

分 类 号：S813.2[农业科学—畜牧学]