IMPACT OF THE UNEVEN DISTRIBUTION OF THE NUMBER OF PROGENIES BY SIRES IN THE GENETIC EVALUATION OF ANIMALS IN ENVIRONMENTS WITH PRESENCE OF HETEROGENEITY OF ENVIRONMENTAL VARIANCE

Abstract A progeny test representing the milk yield was simulated to evaluate the effect of heterogeneity of environmental variance on genetic evaluation of sires in situations of unequal number of progeny per sire between herds, eliminating, at random, daughters per sires in environments of low and high environmental variability. In all situations involving loss of information in the environment of lower environmental variability, there was larger overestimation of the residual variance component, resulting in lower estimates of heritability. On the other hand, the decrease in the amount of information of herds from high environmental variability led to greater overestimation of additive genetic variance and lower overestimation of environmental variance. Although there were alterations in breeding values regarding sires, Spearman and Pearson correlations between breeding values, in all situations, were higher than 0.90. The uneven number of progeny per sire in the presence of heterogeneity of environmental variance led to the overestimation of variance components and changes in absolute breeding values of the sires, without, however, changing the classification values.

摘要 本研究模拟了产奶量后裔测定试验，以评估环境方差异质性对不同牧场种公牛后裔数不均情况下种公牛遗传评估的影响，并在低、高环境变异水平的环境中随机剔除各公牛的部分女儿。在所有涉及低环境变异性环境下信息丢失的场景中，残差方差组分的高估程度均更高，进而导致遗传力估计值偏低。反之，当高环境变异性牧场的信息体量减少时，会引发加性遗传方差的高估程度加剧，而环境方差的高估程度则有所降低。尽管种公牛的育种值存在一定变化，但所有场景下育种值间的斯皮尔曼（Spearman）相关系数与皮尔逊（Pearson）相关系数均高于0.90。当存在环境方差异质性且各公牛后裔数不均时，会导致方差组分被高估以及种公牛绝对育种值发生变化，但并不会改变其排名分类结果。