Seasonal variation of milk composition poses a challenge for maintaining dairy product quality. Multivariate statistical techniques were used here to identify seasonal patterns in mineral levels, fatty acid profiles and selected metabolites for New Zealand raw milk collected from 24 individual cows in early-season (November), mid-season (January/February) and late-season (March). Partial least-squares discriminant analysis (PLS-DA) showed clear seasonal differences in mineral and fatty acid profiles that may have been related to supplementary feeding, dilution effects (higher total solids in late-season milk) and the lactation cycle. Fatty acid profiles showed an inverse relationship between the C6–C12 fatty acids and the branched- or long-chain fatty acids, which was thought to be related to feeding of maize in mid- and late-season. Targeted metabolite analysis suggested elevated phosphatase activity in March, and this warrants further investigation as an indicator of metabolic changes related to declining milk quality at the end of the milking season.

牛奶成分的季节性变化对保持乳制品质量提出了挑战。这里使用多元统计技术来确定从季节早期（11月），季节中期（1月/ 2月）和季节后期从24头母牛身上采集的新西兰原奶的矿物质水平，脂肪酸谱和所选代谢物的季节性模式（游行）。偏最小二乘判别分析（PLS-DA）显示矿物质和脂肪酸谱的明显季节性差异，这可能与补充喂养，稀释效应（后期乳中总固体含量较高）和泌乳周期有关。脂肪酸谱显示C6–C12脂肪酸与支链或长链脂肪酸之间呈反比关系，这被认为与季节中期和后期的玉米饲喂有关。