Natural variations of milk somatic cell count in dairy cows :

Abstract: The main objectives of this thesis were to find out which factors affect the day-to-day variations of somatic cell count (SCC) in bovine milk and how the results of on-line measurements of SCC can be applied in automatic milking systems. The effect of milking interval and milk fraction during sampling on SCC was investigated. Mastitis is regarded as one of the most costly diseases in the world in dairy production. Most of the mastitis cases are without visible symptoms (subclinical) and pass unnoticed. Cows having subclinical mastitis contribute the most to the total production loss for the farmer. With the new technology of on-line measuring of milk SCC comes the possibility to measure the SCC at every milking at a very low cost, getting an good overview of the variations in SCC. But it is of great importance to be able to interpret these consecutive SCC values and to know how and why the variations of milk SCC fluctuate. After investigating what factors that have an impact on an uninfected gland two experiments were performed, the first regarding the effect of milking interval on SCC and the second, the importance of the selected milk fraction. Both factors are assumed having an effect on the SCC. In the first experiment data from 83 different cows during a four-month period were examined for the influence of milking interval on the SCC. The cows were housed in a barn with the automatic milking system VMS, provided by DeLaval. The VMS was equipped with an on-line cell counter, OCC. The data was analysed using SAS REG procedure. Even though a trend between lower SCC and longer milking interval could be seen on all milking data, the correlation was not significant on total milkings level. However, 82% of the single cows had a negative correlation between SCC and milking interval, of which 34% were significant (P < 0.05). Although the results of this study showed that milking interval affects an uninfected udder, there was no possibility of creating a compensating algorithm applicable on all cows. The second experiment was performed in Switzerland during 40 days. A special milking machine with four separate milk containers was used enabling the measuring of quarter SCC. In total, 3420 milk samples from 19 cows were tested for SCC. All analyses were made by the Institution of Veterinary Physiology, University of Bern, using the Wilcoxon Signed Rank Test in SigmaStat. The result of this study showed that foremilk SCC did not very well represent the SCC of total milk of single quarters or whole udder milk. Foremilk SCC was significantly higher (P < 0.001). The study also showed that single quarter milk was not representative when estimating whole udder milk SCC. The mean of measured SCC of whole udder milk was significantly higher (P < 0.001) than the volume corrected SCC calculated from total quarter milk.