不同时间跨度表型资料对种猪遗传评估的影响

    Effects of phenotypic data from different time span on pig genetic evaluation

    • 摘要: 以杜洛克猪主要生长性状为研究对象,评估不同时间跨度、相同个体估计育种值(EBV)的准确性和排名的差异。利用DMU软件和单性状动物模型估计了主要生长性状的方差组分,并计算性状的遗传力;划分不同时间跨度,包括2年、3年、4年、5年和全部数据,评估验证群体个体的EBV准确性及其与利用全部数据时评估育种值的秩相关。结果表明,115 kg体重日龄(AGE)、30~115 kg日增重(ADG)、115 kg体重背膘厚(BF)、115 kg体重眼肌面积(LEA)和综合体型评分(BCS)的遗传力分别为0.22、0.16、0.38、0.30和0.09,除BCS外,均为中高遗传力性状。固定方差组分情况下,AGE、ADG、BF、LEA和BCS的EBV准确性变化范围分别为0.62~0.64、0.56~0.59、0.72~0.73、0.67~0.69和0.49~0.53;不固定方差组分情况下,各性状EBV准确性变化范围分别为0.64~0.65、0.51~0.59、0.61~0.73、0.67~0.69和0.42~0.53;不同条件下,秩相关均较为接近。利用群体全部数据计算方差组分作为先验值,优于利用阶段数据方差组分估计值时EBV准确性;不同时间跨度下,EBV排名差异较小,可适当缩小数据取值时间范围。

       

      Abstract: The accuracy and ranking difference of estimated breeding values(EBV) of the same individuals in different time span were evaluated for the major production traits of Duroc pig. Univariate animal model and DMU software were used to estimate the variances and heritabilities of major production traits. The accuracy of EBV and spearman correlation were compared between data from different periods, including two years, three years, four years, five years and all data. The results also showed that the heritabilities of age at 115 kg live weight (AGE), average daily gain between 30-115 kg (ADG), backfat thickness at 115 kg live weight (BF), loin eye area at 115 kg live weight (LEA) and body conformation score (BCS) were 0.22, 0.16, 0.38, 0.30 and 0.08 respectively. Except BCS, the heritabilities of other traits were medium and high. Furthermore, if the variance components were fixed, the accuracy ranges of AGE, ADG, BF, LEA and BCS were 0.62-0.64, 0.56-0.59, 0.72-0.73, 0.67-0.69 and 0.49-0.53 respectively. If the variance components were not fixed and estimated from the specific time span, the accuracy ranges of AGE, ADG, BF, LEA and BCS were 0.64-0.65, 0.51-0.59, 0.61-0.73, 0.67-0.69 and 0.42-0.53 respectively. For another, the spearman correlations were close for conditions of different time span. The results also showed that the prior variance components calculated from the whole data were better than the phase data, while the rank of EBV were close for different periods, so the time span of data could be reduced appropriately.

       

    /

    返回文章
    返回