McParland S, Berry D P
Animal and Grassland Research and Innovation Center, Teagasc, Moorepark, Fermoy, Co. Cork, Ireland.
Animal and Grassland Research and Innovation Center, Teagasc, Moorepark, Fermoy, Co. Cork, Ireland.
J Dairy Sci. 2016 May;99(5):4056-4070. doi: 10.3168/jds.2015-10051.
Knowledge of animal-level and herd-level energy intake, energy balance, and feed efficiency affect day-to-day herd management strategies; information on these traits at an individual animal level is also useful in animal breeding programs. A paucity of data (especially at the individual cow level), of feed intake in particular, hinders the inclusion of such attributes in herd management decision-support tools and breeding programs. Dairy producers have access to an individual cow milk sample at least once daily during lactation, and consequently any low-cost phenotyping strategy should consider exploiting measureable properties in this biological sample, reflecting the physiological status and performance of the cow. Infrared spectroscopy is the study of the interaction of an electromagnetic wave with matter and it is used globally to predict milk quality parameters on routinely acquired individual cow milk samples and bulk tank samples. Thus, exploiting infrared spectroscopy in next-generation phenotyping will ensure potentially rapid application globally with a negligible additional implementation cost as the infrastructure already exists. Fourier-transform infrared spectroscopy (FTIRS) analysis is already used to predict milk fat and protein concentrations, the ratio of which has been proposed as an indicator of energy balance. Milk FTIRS is also able to predict the concentration of various fatty acids in milk, the composition of which is known to change when body tissue is mobilized; that is, when the cow is in negative energy balance. Energy balance is mathematically very similar to residual energy intake (REI), a suggested measure of feed efficiency. Therefore, the prediction of energy intake, energy balance, and feed efficiency (i.e., REI) from milk FTIRS seems logical. In fact, the accuracy of predicting (i.e., correlation between predicted and actual values; root mean square error in parentheses) energy intake, energy balance, and REI from milk FTIRS in dairy cows was 0.88 (20.0MJ), 0.78 (18.6MJ), and 0.63 (22.0MJ), respectively, based on cross-validation. These studies, however, are limited to results from one research group based on data from 2 contrasting production systems in the United Kingdom and Ireland and would need to be replicated, especially in a range of production systems because the prediction equations are not accurate when the variability used in validation is not represented in the calibration data set. Heritable genetic variation exists for all predicted traits. Phenotypic differences in energy intake also exists among animals stratified based on genetic merit for energy intake predicted from milk FTIRS, substantiating the usefulness of such FTIR-predicted phenotypes not only for day-to-day herd management, but also as part of a breeding strategy to improve cow performance.
了解动物个体和群体水平的能量摄入、能量平衡及饲料效率会影响日常的畜群管理策略；个体动物水平上这些性状的信息在动物育种计划中也很有用。数据匮乏（尤其是在个体奶牛水平上），特别是饲料摄入量数据的匮乏，阻碍了在畜群管理决策支持工具和育种计划中纳入这些属性。奶牛养殖户在泌乳期间每天至少能获取一次个体奶牛的牛奶样本，因此任何低成本的表型分析策略都应考虑利用该生物样本中可测量的特性，这些特性反映了奶牛的生理状态和生产性能。红外光谱学研究电磁波与物质的相互作用，在全球范围内用于预测常规采集的个体奶牛牛奶样本和奶罐样本的牛奶质量参数。因此，在下一代表型分析中利用红外光谱学将确保在全球范围内可能迅速应用，而且由于基础设施已经存在，额外的实施成本可忽略不计。傅里叶变换红外光谱（FTIRS）分析已用于预测牛奶中的脂肪和蛋白质浓度，其比例已被提议作为能量平衡的指标。牛奶FTIRS还能够预测牛奶中各种脂肪酸的浓度，已知当身体组织被动员时，即奶牛处于负能量平衡时，其组成会发生变化。能量平衡在数学上与残余能量摄入量（REI）非常相似，REI是一种建议的饲料效率衡量指标。因此，从牛奶FTIRS预测能量摄入、能量平衡和饲料效率（即REI）似乎是合理的。事实上，基于交叉验证，奶牛从牛奶FTIRS预测能量摄入、能量平衡和REI的准确性（即预测值与实际值之间的相关性；括号内为均方根误差）分别为0.88（20.0兆焦）、0.78（18.6兆焦）和0.63（22.0兆焦）。然而，这些研究仅限于一个研究小组基于英国和爱尔兰两个对比生产系统的数据得出的结果，需要进行重复验证，特别是在一系列生产系统中，因为当验证中使用的变异性在校准数据集中未体现时，预测方程并不准确。所有预测性状都存在可遗传的基因变异。根据从牛奶FTIRS预测的能量摄入的遗传价值对动物进行分层后，动物之间在能量摄入方面也存在表型差异，这证实了这种FTIR预测的表型不仅对日常畜群管理有用，而且作为提高奶牛生产性能的育种策略的一部分也很有用。
