Mineral concentrations in diets, water, and milk and their value in estimating on-farm excretion of manure minerals in lactating dairy cows

Thirty-nine commercial dairies in Merced County, California were enrolled in the present study to (1) compare lactating cow mineral intakes (via drinking water and total mixed ration) to the National Research Council (NRC) requirements, (2) evaluate the association between dietary concentrations of minerals with and without drinking water and adjusted for mineral concentrations in milk, and (3) compare 4 different methods to estimate excretion of minerals using either assays or estimations of milk mineral outputs and total daily mineral intake per cow with or without minerals coming from drinking water. Dairies were selected to represent a range of herd milk yields and a range of water mineral contents. Samples of total mixed ration, drinking water, and bulk tank milk were taken on 2 different days, 3 to 7 d apart in each farm. Across-farm medians and percentile distributions were used to analyze results. The herd median milk yield interquartile ranged (10th to 90th percentile) from less than 25 to more than 39 kg/d and the concentration of total solids in water interquartile ranged from less than 200 to more than 1,490 mg/L. Including drinking water minerals in the diets increased dietary concentrations by <4% for all minerals except for Na and Cl, which increased by 9.3 and 6.5%, respectively. Concentrations of P and K in milk were essentially the same as the NRC value to estimate lactation requirements. However, NRC milk values of Ca, Cl, and Zn were 10 to 20% greater than dairy farm values; and Na, Cu, Fe, and Mn were no less than 36% below NRC values. Estimated excretion of minerals via manure varied substantially across farms. Farms in the 10th percentile did have 2 to 3 times less estimated mineral excretions than those in the 90th percentile (depending on the mineral). Although including water minerals increased excretion of most minerals, the actual median effect of Ca, Mg, S, Cu, Fe, and Mn was less than 5%, and about 8% for Na and Cl. Replacing assayed concentrations of minerals in milk with NRC constants resulted in reduced estimated excretion of Ca, Na, Cu, Fe, and Zn, but median differences were <5% except for Na which was 7.5%. Results indicate that not including mineral intake via drinking water and not using assayed concentrations of milk minerals lead to errors in estimation manure excretion of minerals (e.g., Ca, Na, Cl, and S).