The quality of colostrum administered to calves is based on its concentration in immunoglobulins G (IgG, g/L). Immunoglobulins A (IgA) and M (IgM) are also present but at a lower level. The gold standard reference analysis for these traits, radial immunodiffusion, is time-consuming and expensive. In order to define breeding strategies that are aimed at improving colostrum quality in dairy cattle, a large amount of data is needed, and the use of indicator traits would be beneficial. In the study presented here, we explored the heritabilities of reference (radial immunodiffusion) and near infrared-predicted IgG, IgA, and IgM concentrations and estimated their genetic correlations. First, the colostrum of 765 Holstein cows from nine herds was sampled to perform a reference analysis and the near-infrared spectra (400–2500 nm) were stored. We used a calibration set (28% of the initial samples) that was representative of the herds and cow parity orders to develop prediction equations for IgG, IgA, and IgM concentrations. Finally, these traits were predicted in the validation set (72% of the initial samples) to estimate genetic parameters for the predictions. Genetic correlations between reference and predicted values of each trait were estimated through bivariate linear animal models. The three near-infrared-predicted immunoglobulin fractions were genetically correlated with their reference value. In particular, the reference and predicted IgG concentrations were strongly correlated at both the genetic (0.854 ± 0.314) and phenotypic level (0.767 ± 0.019). Weaker associations were observed for IgA and IgM concentrations, which were predicted with lower accuracy compared to IgG. Simulation analyses suggested that improving colostrum quality by selective breeding in Holstein cattle based on near-infrared predicted colostrum immunoglobulins concentrations is feasible. In addition, less than 10 mL of colostrum are needed for spectra acquisition and thus implementation of such analyses is possible in the near future. The concentrations of colostrum immunoglobulins can be predicted from near-infrared spectra and the genetic correlation between the reference and the predicted traits is positive and favourable, in spite of the large standard errors of the estimates. Near-infrared spectroscopy can be exploited in selective breeding of dairy cattle to improve colostral immunoglobulins concentration.

中文翻译：

---

金标准法测定的牛初乳中免疫球蛋白浓度与其近红外预测具有遗传相关性

小牛初乳的质量取决于其在免疫球蛋白 G (IgG, g/L) 中的浓度。免疫球蛋白 A (IgA) 和 M (IgM) 也存在，但水平较低。这些特征的金标准参考分析，即径向免疫扩散，既耗时又昂贵。为了确定旨在提高奶牛初乳质量的育种策略，需要大量数据，使用指标性状将是有益的。在此处介绍的研究中，我们探讨了参考（径向免疫扩散）和近红外预测的 IgG、IgA 和 IgM 浓度的遗传力，并估计了它们的遗传相关性。首先，对来自 9 个牛群的 765 头荷斯坦奶牛的初乳进行采样以进行参考分析，并存储近红外光谱（400-2500 nm）。我们使用代表牛群和奶牛胎次顺序的校准集（初始样本的 28%）来开发 IgG、IgA 和 IgM 浓度的预测方程。最后，在验证集（初始样本的 72%）中预测这些性状以估计预测的遗传参数。通过双变量线性动物模型估计每个性状的参考值和预测值之间的遗传相关性。三个近红外预测的免疫球蛋白组分与它们的参考值在遗传上相关。特别是，参考和预测的 IgG 浓度在遗传 (0.854 ± 0.314) 和表型水平 (0.767 ± 0.019) 上都具有很强的相关性。观察到 IgA 和 IgM 浓度的关联较弱，与 IgG 相比，预测准确度较低。模拟分析表明，通过基于近红外预测的初乳免疫球蛋白浓度对荷斯坦牛进行选择性育种来提高初乳质量是可行的。此外，光谱采集所需的初乳少于 10 mL，因此在不久的将来可以实施此类分析。初乳免疫球蛋白的浓度可以从近红外光谱预测，尽管估计的标准误差很大，但参考和预测性状之间的遗传相关性是积极的和有利的。近红外光谱可用于奶牛的选择性育种，以提高初乳免疫球蛋白浓度。