Non-destructive Determination of Bovine Milk Progesterone Concentration during Milking Using Near-infrared Spectroscopy

摘要

In the current dairy industry, an intensive demand for estrus detection and early diagnosis of pregnancy has been increasing. Progesterone is a steroid hormone that is secreted from corpus luteum into bovine blood and milk, and has a role of maintenance of estrus cycle and pregnancy. Therefore, progesterone concentration in bovine milk is used as an important indicator of estrus detection and early diagnosis of pregnancy. Current method for milk progesterone determination requires a hormone extraction procedure that is time consuming, various types of instruments, reagents management, and various assay methods that are destructive in nature. In contrast, near-infrared spectroscopy (NIRS) is a time saving and non-destructive analytical method that can be used for online real-time determination of milk constituents content such as milk fat, protein, lactose, milk urea nitrogen and somatic cell count. However, there has been limited study on using NIRS for online real-time determination of progesterone concentration in milk during milking. Thus, the objective of this study was to develop an online real-time NIR spectroscopic sensing system for milk progesterone determination during milking by using a specific enzyme immunosorbent assay as a reference (chemical) method. Milk spectra with a wavelength range of 700 to 1050 nm and milk samples were collected every 20 s during milking from four lactating Holstein cows for 28 days using the NIR spectroscopic sensing system. Calibration models were developed using partial least squares analytical method and the precision and accuracy of the models was validated. Milk progesterone concentration for each milking was calculated by taking the progesterone concentration of the milk predicted values and milk yield obtained every 20 s, and was compared with the milk progesterone concentration chemical analysis value for one milking (bucket milk). The results obtained show that the measurement accuracy for one milking of milk progesterone concentrations was reasonably good. By installing the NIR spectroscopic sensing system developed in this study into a milking robot, it could predict milk progesterone concentration for one milking with almost the same accuracy as chemical analysis. Thus, recording this predicted value every milking and monitoring the continuous transition of the milk progesterone concentrations, it becomes possible to use it for the detection of estrus status and for the diagnosis of pregnancy of each cow.