Improving udder health management in dairy herds with automatic milking systems

The increased use of automatic milking systems (AMS) in combination with modern sensor and big data technologies enables a transformation of mastitis management in dairy herds worldwide. The overall goal of this thesis therefore was to explore the potential use and benefit of using frequently measured data to optimize on-farm decision making in udder health management in herds using AMS.

In this thesis, we found that most risk factors in dairy herds using AMS are the same as in herds using conventional milking systems (CMS). However, the use of an AMS asks for specific attention for environmental mastitis. Hygiene of dairy cows and milking machine, which is related to the infectious pressure of environmental mastitis pathogens, is even more important in AMS than in CMS herds. On the other hand, epidemiological modelling revealed that transmission parameters of Staphylococcus aureus and Streptococcus agalactiae seem lower in AMS than in CMS dairy herds. Segregation of chronically infected cows would largely reduce the transmission between cows and environmental transmission route might be underestimated in AMS herds. The use of AMS does not result in impaired udder health status or more antimicrobial usage in AMS herds than in CMS herds. The distribution of mastitis pathogens is comparable between AMS and CMS herds. In both types of herds, antimicrobial use is associated with the attitude of the farmer to udder health.

The potential use and benefit of online sensor, based on an automated California Mastitis Test, to optimize on-farm mastitis decision making was evaluated by using a number of state-of-the-art statistical and epidemiological modeling methods. It was found that the measurements of an online somatic cell count (SCC) sensor moderately agree with the SCC measured in the Dairy Herd Improvement laboratories, and the agreement between these two measurements is positively associated with herd SCC. Online SCC sensor is valuable for different aspects of preventive mastitis management as well as for decisions on the use of antimicrobials to treat mastitis. Modern sensor technology holds great potential to support mastitis management and gives new insights in the infection dynamics of mastitis in dairy cows, of which the regularly fluctuating SCC described in this thesis is a fascinating example.

Based on the findings of this thesis, we conclude that the frequently measured data in AMS herds hold great potential to support data-driven mastitis management decision making. This can be done for instance by monitoring individual cow udder health, identifying herd-specific risk factors automatically, capturing patterns of intramammary infection dynamics and quantifying the transmission process of infectious mastitis pathogens. Concerns that now limit the implementation of sensor technologies must be addressed. Further research to integrate the data from different sources, and algorithms to turn the data into interpretable information that can be used by the farmer and his advisor are needed.

Zhaoju Deng obtained his DVM at Huazhong Agricultural University in China after 5 years of study and then continued master study at China Agricultural University, in which he started research on epidemiology of bovine mastitis and the innate immune response of bovine mammary gland to Prototheca zopfii. In 2015, he was granted a scholarship from the China Scholarship Council to pursue a PhD degree in mastitis epidemiology on Dutch dairy herds using AMS. In August 2020, he started to work on population health in Chinese dairy herds informally at the Department of clinical veterinary medicine of the China Agricultural University. His main interest lies in applying statistical methods on on-farm data (sensor data and manually collected data) to improve herd health management.

Text and picture: Zhaoju Deng

Leave a Reply

Your email address will not be published. Required fields are marked *