BACKGROUND Control of the environmental variability by genetic selection offers possibilities for new selection objectives for productive traits. This methodology aims at reducing heterogeneity in productive traits and has been applied to several traits and species for which animal homogeneity is profitable. In horse breeding programmes, rank in competitions is a common selection objective but has been challenging to model. In this study, the parameters of environmental variability for the rank of a horse were computed to analyse the capability of a horse to maintain the best ranking across competitions that consist of long-distance races in which the adapted physical condition of the horse is essential. The genetic component of the environmental variance for the rank in endurance competitions was evaluated, which resulted in proposing a new transformation of horse scores in competitions. RESULTS Homogeneous and heterogeneous variance models were compared by assaying three random effects that affect both the rank and its variability, using endurance ride data consisting of 2863 records. The pedigree relationship matrix contained 5931 animals. The rank trait was transformed into a normalized variable to prevent false estimates of the genetic correlation by inappropriate artificial skewness. The models included the number of participants in the race, sex, and age as systematic effects. The rider, the rider-horse interaction, or an environmental permanent effect were tested as random effects, in addition to additive genetic and residual effects. The models were analysed using the GSEVM program. Estimates of heritability for rank ranged from 0.12 to 0.15. The heterogeneous variance model that fitted the rider was assessed as the best model based on the deviance information criterion. Estimates of genetic variance for rank variability ranged from 0.12 to 0.13. The genetic correlation between the rank and its environmental variability was low and did not differ from 0. CONCLUSIONS These results offer an opportunity to select animals for canalization by reducing the variability of race results and achieving the best positions, which could be a new selection objective by weighting estimated breeding values for rank and its variability in a selection index.

中文翻译：

---

挑战耐力比赛等级一致性的选择。

背景技术通过遗传选择控制环境变异性为生产性状的新选择目标提供了可能性。这种方法旨在减少生产性状的异质性，并已应用于动物同质性有利的几种性状和物种。在马匹育种计划中，比赛排名是一个常见的选择目标，但要建模很难。在这项研究中，计算了马匹等级的环境变异性参数，以分析马匹在包括长距离比赛在内的各项比赛中保持最佳排名的能力，在长距离比赛中，适应马匹的身体状况至关重要。评估了耐力比赛中排名的环境差异的遗传成分，从而提出了比赛中赛马得分的新转变。结果使用由2863条记录组成的耐力骑乘数据，通过分析影响等级及其变异性的三个随机效应，比较了均质和异质方差模型。系谱关系矩阵包含5931只动物。将等级特征转换为归一化变量，以防止由于不适当的人为偏斜而对遗传相关性做出错误估计。该模型将种族，性别和年龄的参与者数量作为系统性影响。除附加遗传和残留效应外，还测试了骑手，骑手与马的互动或环境永久性效应作为随机效应。使用GSEVM程序分析了模型。等级的遗传力估计范围为0.12至0。15.根据偏差信息标准，适合骑手的异质方差模型被评估为最佳模型。等级变异的遗传变异估计范围为0.12至0.13。等级与其环境变异性之间的遗传相关性很低，且与0并无差异。结论这些结果通过减少种族结果的变异性和获得最佳位置，为选择进行渠化的动物提供了机会，这可能是一个新的选择目标通过加权选择等级的估计育种值及其变异性。