Variation in macronutrients and minerals in bovine milk and their relationships with milk phosphorus content

Abstract

Under current Dutch regulations, accurate assessment of the amount of P secreted in milk is essential, as it determines manure P output. The two main aims were: 1) to predict P content in bovine milk using a broad range of predictor variables, and 2) to obtain predicted milk P contents representative of the Dutch dairy cow population. A secondary objective was to evaluate seasonal variation in milk P content. Weekly bulk milk samples (week 14 in 2017 up until week 13 in 2018) were collected from 14 dairy plants located across the Netherlands and pooled per week as representative samples of Dutch bovine milk. Milk samples were analysed for macronutrients and mineral contents. The mean P content of milk was 101.2 mg/100 g, and significant seasonal variation was observed, with the highest values found during winter and the lowest during summer. The contents of fat, protein, casein, Ca, Mg and Mn in milk were found to be highly correlated with the milk P content. The preferred multiple regression equation to predict the milk P content (mg/100 g) included the predictor variables milk fat (g/100 g), Ca (mg/100 g) and K (mg/100 g), viz. milk P content =-58.6 (± 14.09) + 0.28 (± 0.104) × Ca + 11.46 (± 2.559) × fat + 0.48 (± 0.094) × K, and explained 80% of the variation (R2adj) in milk P content. The contribution of milk K content to explain variation in milk P content cannot be physiologically explained.