Multivariate analysis relating milk production, milk composition, and seasons of the year

Abstract Our objective was to quantify the relationship between seasons of the year, milk production, and milk composition of a dairy farm based on data for 48 consecutive months, using multivariate statistical analyses. The dataset contained information on productive indexes and milk composition from the bulk tank milk, which was measured from milk samples, collected monthly and used to determine the total dry extract and defatted dry extract, lactose, fat, protein, somatic cell count, and total bacterial count. Seasons of the year and milk production/hectare were also considered. Factor, cluster, and discriminant analysis were used to study the relationships between the above-mentioned variables. A positive relationship was noted between season and total dry extract, defatted dry extract, milk fat, and protein, with higher values being observed in winter and spring. Similarly, a positive relationship was noted between season and milk production/hectare, lactose content, with an increase in milk production and lactose content in winter and spring, which was negatively related to the somatic cell count and total bacterial count. Milk production and composition varied mainly with seasons. Multivariate analyses facilitated a better understanding of the relationship between milk production and composition on this dairy farm.摘要 本研究旨在基于连续48个月的监测数据，结合多元统计分析方法，量化某奶牛场的季节、产奶量与乳成分三者间的关联关系。本数据集涵盖了每月采集的集散罐牛乳（bulk tank milk）样本的生产指标与乳成分信息，检测指标包括总干物质、脱脂干物质、乳糖、乳脂肪、乳蛋白、体细胞计数及总细菌数。此外，本研究还纳入了季节与单位公顷产奶量作为分析变量。本研究采用因子分析、聚类分析与判别分析对上述变量间的关联展开研究。研究发现，季节与总干物质、脱脂干物质、乳脂肪及乳蛋白含量呈正相关，上述指标在冬、春两季数值更高。同理，季节与单位公顷产奶量、乳糖含量亦呈正相关，冬、春两季的产奶量与乳糖含量均有所提升，而体细胞计数与总细菌数则与该两项指标呈负相关关系。产奶量与乳成分主要随季节发生变化。多元统计分析方法有助于更深入地理解该奶牛场产奶量与乳成分间的关联。