Influence of varying milking intervals on milk composition - a physiological approach on secretion of “normal milk”

摘要

A group of approx. 40 high-yielding German Holstein Frisian cows were milked robotically (VMS®, DeLaval) throughout 400 days. During 20 session days (20-day intervals), samples (quarter foremilk [QFM], quarter composite milk [QCM] and cow composite milk [CCM]) were drawn continuously for 24 hours. Cell count (SCC) determination (Fossomatic) was done in all milk fractions, culturing in QFM. The milk constituents NAGase activity, lactate, fat, protein, and lactose were analysed in QCM. Data analyses (SAS, PROC GLM) were directed to evaluate milk secretion, SCC and milk composition in relation to the milking interval (MI). The overall SCC in QCM was 4.75 lg (= 56,234 cells/ml). The mean milking frequency (MF) over 24 h was 2.72 with an average MI of 9:01 ± 2:32 hours and a mean milk yield of 26.87 ± 10.59 kg/cow/day. Distribution of all 1,802 milkings was as follows: MF1: = 0.93%; MF2: 38.04%; MF3: 49.44%; MF4: 11.04% and MF5: 0.47%. The lactation stage (days in milk, DIM) had a marked influence on the udder secretion rate, being independent from MI (4 - 6 h, 6 - 8 h; 8 - 10 h; 10 - 12 h, 12 -14 h). Milk secretion fell to approx. 50% between DIM < 100 and DIM > 299. The quarter secretion rate (MI: 6 h = 356 g/h; MI: > 12 h = 232 g/h) showed highly significant differences between all MI and a non-linear pattern in relation to the MI applied. All milk constituents but lactate showed non-linear, MI-dependent secretion patterns. Values for NAGase and lactose for MI < 10 hours suggested inflammation, but mean SCC amounted to 4.75 lg. Due to its insensitivity to MI, lactate could be useful for automatic mastitis detection, also when the cell count level is < 100,000 cells/ml.