Abstract:

To meet the seasonal regulation of production management and to explore the variation features of milk compositions and some related traits, based on the DHI data and derived data of Chinese Holstein lactating cows in the north of China, the data were grouped by parities and the natural months and the abnormal ones are deleted to obtain complete data 6 520 sets which contain 9 traits-daily milk yield(DMP), milk solid percentage(MSP), milk fat percentage(MFP), milk protein percentage(MPP), milk lactose percentage(MLP), somatic cell count(SCC), body condition score(BCS), lossed milk production(LMP) and fat-to-protein rate(FPR). The natural month, parity and their interaction were assumed as factors affecting the 9 traits. A GLM variance analysis method of SAS software showed that both of the natural months and lactation parities affected the above 9 traits significantly (P<0.000 1, or P<0.001, or P<0.05). However, the interaction between parities and the natural months showed no significant effect on MFP, MPP, MSP, SCC and FPR(P>0.05)while their interaction showed significant effect on DMP (P<0.000 1), MLP, BCS and LMP(P<0.05). The Duncan multiple comparison of natural months showed that DMP(23.58 kg·d^-1), MSP(12.35%), MPP(3.02%) and MFP(3.81%) all were the lowest in June and SCC was the highest (385.2×1 000·mL^-1) in August and LMP was the highest (1.12 kg·d^-1) in November and FPR was the highest (1.32) in February which means the cows consumed the body fat quickly. The Duncan multiple comparison of parity showed that MSP, MPP and MLP dropped rapidly in parity 4 (P<0.05). The SCC went up significantly in parity 4 and LMP increased significantly following the parity (P<0.05), and this means that the quality of the milk and sanitation indexes of milk production deteriorated and the milk loss production increased following the parity. Secondly, the canonical correlation analysis of the 9 traits showed that BCS were high positive correlation with SCC (0.830 6) and LMP (0.836 0) while SCC and LMP were high positive correlation (0.786 1). Thirdly, the relation equations between MSP(%) or MLP(%) and the natural months in mixed cow herd were built using Model Wood (MSP=12.862x^{－0.031 7}*e^{0.004 59x}, MLP=4.982 4x^{－0.019 6}*e^{0.001 06x}, x as month). The change patterns of milk composition and related traits in different natural months and parities obtained from the data provide scientific evidences for regulating feeding management and nutrient supply for cows accurately.