Benefits from daily milk recording data

Daily milk recording data was collected from 31 farms located in different regions of Germany. Varying with an interval of two or one week content samples were taken for the a.m., p.m. milking and for the mixed sample of the a.m. and p.m. milking. Correlations between milk yields of consecutive testdays are on average 0.96 though there are differences between farms which indicates different quality of data recording on the farm level. Milk, fat and protein yields of randomly chosen testdays are representative as long as the corresponding testing interval does not exceed six weeks. From a validation of the actual method for estimation daily yields from single a.m. or p.m. milkings it can be concluded that use of evening milkings is less accurate than morning milkings.