PRECISION DAIRY TECHNOLOGIES AND BEHAVIORAL INDICATORS FOR IMPROVING UDDER HEALTH AND MASTITIS MANAGEMENT IN AUTOMATIC AND CONVENTIONAL MILKING SYSTEMS

Abstract

This dissertation integrates four complementary studies that collectively investigate how Precision Dairy Technologies (PDT) can be used to enhance the monitoring, understanding, and management of clinical diseases and cow behavior in dairy production systems. This manuscript is influenced by the US dairy industry's rapid technological transformation over many years, characterized by increasing herd sizes, reduced availability of skilled labor, and growing availability of automated systems for decision-making. In this context, PDT, such as automatic milking systems (AMS) and activity sensors, provide daily, continuous, individual, and quarter-level data that can be utilized for early detection of disease, behavioral changes prior to and after health events, and the implementation of prophylactic management strategies. Mastitis, as one of the most prevalent and costly diseases in dairy cattle, is the main focus across this manuscript, particularly because of its relevance to both milking performance and animal welfare, and its importance in systems with restricted antimicrobial availability, such as organic dairies.The first chapter of this dissertation touches on the conceptual and biological foundation for the subsequent studies by synthesizing the relevance of PDT in dairy systems, with emphasis on automated monitoring and udder health. Furthermore, it highlights how AMS and sensor technologies have transformed dairy management from periodic observation to continuous, data-driven decision-making. Milking behavior, activity, and lying behavior parameters emerge as key indicators of health status, adaptation to a new system, and the level of stress cows experience in their respective systems. This chapter also identifies important knowledge gaps, particularly the limited integration of behavioral indicators and the lack of available treatments for cows at dry-off and for those managed under an organic farming, and lays the foundation for the observational and experimental studies developed in the following chapters. The second chapter examines differences in milking behavior and performance during early lactation among primiparous and multiparous cows of different breeds managed under an AMS. The findings show that parity and breed significantly influence both milking behavior and performance in an AMS setting. As expected, Holstein cows produced the highest milk yield, followed by Holstein x Jersey crosses and Jersey cows. The frequency of most undesirable outcomes decreased as lactation advanced, indicating habituation, particularly in primiparous cows, regardless of breed. These results highlight the importance of parity-specific management in AMS to improve dairy management and efficiency. The third chapter extends the focus to the relationship between pathogen-specific infections causing clinical mastitis and milking behavior and performance parameters throughout lactation in Holstein cows managed under an AMS. The results show that different pathogen categories are associated with distinct patterns in specific milking variables recorded at each milking in cows with clinical mastitis. In particular, cows affected by Gram-negative and Gram-positive pathogens had a higher probability of alterations in milking behavior and performance variables than healthy cows. Overall, this chapter demonstrates that AMS data can provide valuable insights into udder health status and combining them with other PDT may help detect clinical mastitis cases earlier, improving diagnostic accuracy and enabling prompt treatment. The fourth chapter focuses on behavioral responses, specifically lying behavior and walking dynamics, as a proxy for health and welfare status in dairy cows. This study examines changes in lying time and activity patterns in relation to health events, such as mastitis and lameness, throughout lactation. The results indicate that cows experiencing both mastitis and lameness exhibit alterations in these parameters around the time of diagnosis. These behavioral changes are seen as elements of a sickness behavior response, often indicating inflammation and discomfort. This reinforces the value of combining behavioral monitoring with other tools to improve early detection of diseases in cows under AMS. The fifth chapter evaluates the effect of an intramammary infusion of a carvacrol-based botanical product administered at dry-off on udder health during the subsequent lactation in organic-certified dairy cows. Given the limited availability of antimicrobial treatments in organic dairies, this study seeks to address an important gap in alternative mastitis control strategies. The results indicate that the application of the plant-based botanical product did not improve udder health parameters, including somatic cell counts, bacteriological cure rates, or the incidence of new intramammary infections, compared with the control treatment. Overall, the findings do not support the use of this organic treatment for improving udder health, highlighting the need for further research into safe and effective non-antimicrobial options for the dry-off period in organic systems. Collectively, the results of this dissertation indicate that PDT provides important insights for monitoring dairy cow behavior and health, but their interpretation requires a deep understanding of biological variation across factors such as parity, breed, pathogens involved, and behavioral parameters. The data obtained from the AMS show us their diagnostic value, which is enhanced when combined with behavioral indicators, which in some cases reflect more systemic physiological responses to disease. These studies also emphasize that clinical diseases such as mastitis are complex, and pathogen-specific dynamics significantly influence both sensor-based indicators and milk production losses. Finally, the evaluation of an alternative dry-off treatment highlights the complexity of developing non-antibiotic strategies for mastitis control, particularly in organic systems where therapeutic options are restricted. Together, these chapters contribute to a more integrated understanding of dairy cow health in precision monitoring environments and provide evidence supporting the relevance of data-driven decision-making tools in modern dairy production systems.