Bittante Giovanni
Department of Agronomy, Food, Natural resources, Animals and Environment (DAFNAE), University of Padova (Padua), 35020 Legnaro (PD), Italy.
J Dairy Sci. 2022 Jun;105(6):5084-5096. doi: 10.3168/jds.2021-21105. Epub 2022 Apr 22.
Milk urea content is receiving growing interest from science and industry as a tool to infer the protein adequacy of dairy cows' diets, nitrogen excretion and its environmental impact, and efficiency of animals' protein metabolism. Fourier-transform infrared (FTIR) prediction is a high-throughput method for rapidly and cheaply evaluating milk urea content at the population level. Existing knowledge of the major sources of variation (e.g., year, season, farming system, individual herd, and the cow's breed, parity, stage of lactation, and productive potential) is fragmentary. The objective of this work was to study at the population level the simultaneous effects of all the major sources of variation and their most important interactions. Milk urea content in 1,759,706 test day milk samples collected from 291,129 lactations of 115,819 cows from 6,430 herds over 8 yr was predicted by FTIR. The milk urea content data (and also milk protein percentage, for comparison) were analyzed using a linear model that included the effects of parity, days in milk (DIM) class, year, month, herd intensiveness level, cow productivity level, breed, and herd intensiveness and cow productivity levels within breed. All sources of variation of milk urea content proved highly significant, the most important in terms of F-value being breed > year > herd intensiveness level > parity. The ranking for milk protein was very different (DIM class > herd intensiveness level > parity > breed). The patterns of the least squares means for urea and protein contents of milk were also very different and sometimes contrasting. The seasonal variation in urea was sinusoidal. Urea content increased during the first 4 mo of lactation and then remained almost stable before decreasing after 11 mo. Specialized dairy breeds had lower average milk urea content than dual-purpose breeds; in the former case it was lower in Holsteins than in Brown Swiss, and in the latter it was lower in Simmentals than in Alpine Greys. The effect of herd intensiveness level was much stronger than the effect of cow productivity level; the increase in milk urea with increasing herd average daily milk yield was almost linear in the case of dairy breeds but curvilinear in dual-purpose breeds. The large differences in breed and the modest relationships with the cow's productive potential require further analysis at the genetic level to obtain information of potential use in genetic improvement of the dairy cow populations.
作为推断奶牛日粮蛋白质充足性、氮排泄及其环境影响以及动物蛋白质代谢效率的一种工具，乳尿素含量正日益受到科学界和产业界的关注。傅里叶变换红外光谱（FTIR）预测是一种在群体水平上快速且低成本地评估乳尿素含量的高通量方法。关于变异的主要来源（如年份、季节、养殖系统、个体牛群以及奶牛的品种、胎次、泌乳阶段和生产潜力）的现有知识是零散的。这项工作的目的是在群体水平上研究所有主要变异来源及其最重要的相互作用的同时影响。通过FTIR预测了从8年期间6430个牛群的115819头奶牛的291129次泌乳中采集的1759706个测定日乳样中的乳尿素含量。使用线性模型分析了乳尿素含量数据（以及用于比较的乳蛋白百分比），该模型包括胎次、泌乳天数（DIM）类别、年份、月份、牛群集约化水平、奶牛生产水平、品种以及品种内牛群集约化水平和奶牛生产水平的影响。乳尿素含量的所有变异来源都被证明具有高度显著性，就F值而言，最重要的是品种＞年份＞牛群集约化水平＞胎次。乳蛋白的排名则大不相同（DIM类别＞牛群集约化水平＞胎次＞品种）。乳中尿素和蛋白质含量的最小二乘均值模式也非常不同，有时甚至相反。尿素的季节性变化呈正弦曲线。尿素含量在泌乳的前4个月增加，然后在11个月后下降之前几乎保持稳定。专门化奶牛品种的平均乳尿素含量低于兼用品种；在前一种情况下，荷斯坦奶牛的乳尿素含量低于瑞士褐牛，在后一种情况下，西门塔尔牛的乳尿素含量低于阿尔卑斯灰牛。牛群集约化水平的影响比奶牛生产水平的影响要强得多；对于奶牛品种，随着牛群平均日产奶量的增加，乳尿素的增加几乎呈线性，但对于兼用品种则呈曲线关系。品种间的巨大差异以及与奶牛生产潜力的适度关系需要在基因水平上进行进一步分析，以获取可能用于奶牛群体遗传改良的信息。
