[目的] 探索和量化在爱尔兰商业奶牛场测定日、与奶罐样品的乳脂率和乳蛋白率的差异。[方法] 数据来自2021年2 841 个畜群的5 742 次产奶量记录,奶罐样品与测定日记录在同一天进行。检测日奶样乳脂率和乳蛋白率使用ICAR批准的上午或下午交替采集奶样的预测方程来评估。[结果] 测定日与奶罐样品的乳脂率和乳蛋白率之间的Pearson相关系数分别为0.82和0.96。这些相关性对应于乳脂率和乳蛋白率的平均差异为0.1%和0.01%,测定日产奶量记录显示两种测量值均较低。使用EDIY（自动采集测试样本的电子自制仪表）和Non-EDIY（由牛奶记录员手动采集测试样品）两种设备测定的牛群乳脂率相关性分别为0.87和0.72。测定日产奶量较高的畜群表现出较低的乳脂率相关性。同样,在产奶高峰季节采集的奶样与非高峰季节采集的奶样相比,乳脂率差异更大。[结论] 与奶罐奶样相比,在测定日产奶量记录报告中,乳脂率变化较大。这种变化的程度受产奶量、季节、记录设备或者这3 个因素的综合影响。

[Objective] This study was to explore and quantify the differences in fat and protein percentages between milk samples collected on test day milk recordings and those obtained from daily bulk collections in commercial dairy herds in Ireland. [Method] The dataset comprised 5 742 milk recordings from 2 841 herds in 2021,where daily bulk collections were taken on the same day as the milk recording. The predicted fat and protein percentages for test day milk were estimated using either the AM or PM milk samples and were incorporated into prediction equations approved by ICAR. [Result] The Pearson correlation between fat and protein percentages from test day milk recording and daily bulk samples was 0.82 and 0.96,respectively. These correlations correspond to mean differences of 0.1% and 0.01% for fat and protein percentages,with the average test day milk recording showing lower values for both measurements. Fat percentage correlations for herds using EDIY （electronic do it yourself meters which take a test sample automatically） and Non-EDIY （manual recordings where a milk recorder visits and takes a test sample） recording devices were 0.87 and 0.72, respectively. Herds with a higher average test day cow yield demonstrated a lower fat percentage correlation. Similarly,samples taken during peak milk production season exhibited larger differences in fat percentages compared to those taken during the off-peak milk production season. [Conclusion] The findings of this study clearly indicate that the largest variation is observed in the fat percentage reported in test day milk recordings when compared to the corresponding daily bulk samples. The extent of this variation is affected by yield,season,recording type,or a combination of all three factors.