Campos I L, Jahnel R E, Miglior F, Feng Z, Canovas A, Baes C F, Schenkel F S
Centre for Genetic Improvement of Livestock, Department of Animal Biosciences, University of Guelph, Guelph, Ontario, Canada.
Centre for Genetic Improvement of Livestock, Department of Animal Biosciences, University of Guelph, Guelph, Ontario, Canada.
J Dairy Sci. 2025 Jul 18. doi: 10.3168/jds.2025-26433.
Many studies have assessed the effect of heat stress on dairy cattle by integrating an environmental descriptor with test-day records. The genetic parameter of heat tolerance can be estimated using reaction norm models fitting a linear heat stress function (HSf). The HSf commonly fits the temperature-humidity index (THI) as a covariable and assumes a common THI threshold value for the onset of heat stress for all animals. Thus, it does not fully account for individual variation in the THI threshold in the response to heat stress and relies on the determination of a common threshold. The objective of this study was to estimate the genetic parameters and breeding values for heat tolerance by a model fitting Legendre polynomials (LP), which allows for individual variation of THI threshold value in response to heat stress, and to compare the estimated breeding values and genetic parameters to the estimates from a model fitting an HSf with a common THI threshold. Meteorological data collected from the closest weather station to each farm was combined with test-day records from 300,791 first-parity Holstein cows in 4,470 herds across Canada. The LP models from linear to cubic were compared by the mean squared error (MSE) and by inspecting the parameter estimates over the THI gradient. The ranking of bulls' EBV for heat tolerance was compared using the Spearman rank correlation between the LP and HSf models. The quadratic LP model was chosen for the comparison with HSf estimates. The EBV rank correlation for milk production traits under heat stress from the 2 alternate models was greater than 0.97 for all bulls. When the top 10% were compared, the rank correlation between the EBVs from both models was 0.97, 0.95, and 0.86 for fat, protein, and milk yields under heat stress. The results indicate no substantial change in the ranking of bull's EBV. Therefore, both models can identify bulls with high genetic merit for heat tolerance and can be used in genetic evaluations in Canada.
许多研究通过将环境描述符与测定日记录相结合，评估了热应激对奶牛的影响。可以使用拟合线性热应激函数（HSf）的反应规范模型来估计耐热性的遗传参数。HSf通常将温度湿度指数（THI）作为协变量进行拟合，并假定所有动物热应激开始时的THI阈值相同。因此，它没有充分考虑热应激反应中THI阈值的个体差异，而是依赖于一个共同阈值的确定。本研究的目的是通过拟合勒让德多项式（LP）的模型来估计耐热性的遗传参数和育种值，该模型允许THI阈值在热应激反应中存在个体差异，并将估计的育种值和遗传参数与拟合具有共同THI阈值的HSf模型的估计值进行比较。从每个农场最近的气象站收集的气象数据与加拿大4470个牛群中300791头初产荷斯坦奶牛的测定日记录相结合。通过均方误差（MSE）以及检查THI梯度上的参数估计值，对线性到三次的LP模型进行了比较。使用LP和HSf模型之间的斯皮尔曼等级相关性比较了公牛耐热性的估计育种值排名。选择二次LP模型与HSf估计值进行比较。对于所有公牛，两种替代模型在热应激下牛奶生产性状的估计育种值等级相关性均大于0.97。当比较前10%时，两种模型的估计育种值在热应激下脂肪产量、蛋白质产量和牛奶产量的等级相关性分别为0.97、0.95和0.86。结果表明公牛估计育种值的排名没有实质性变化。因此，两种模型都可以识别出耐热性遗传价值高的公牛，并可用于加拿大的遗传评估。
