Analysing the Movement and Behaviour of Housed Dairy Cows

Cows in modern dairy systems are at risk of comprised health and welfare, and monitoring changes in behaviour can help identify early-warning signs. This thesis uses a local positioning system to detect changes in group-level behaviour. The proximity interaction network structure and consistency of a herd housed in a closed barn on a commercial farm in Essex is explored. Next, the network structure, alongside group-level space-use patterns, on the commercial farm in Essex are compared to those of a second dairy cow herd housed in an open barn (RVC Research farm). In the subsequent chapters, the relationship between barn temperature and bunching behaviour, a potentially maladaptive response to warmer than average temperatures, was investigated in both herds, through various bunching metrics: range size, inter-cow distance and nearest neighbour distance. The herd on the commercial farm in Essex was highly connected and temporally unstable, with inter-individual variation in interactions in the non-feeding zone, and social differentiation across functional zones. No social assortment by parity, days in milk or lameness state was detected. The herd on the RVC Research farm were less connected than the herd on the commercial farm in Essex. Inter-individual variation in proximity interactions was found in the feeding zoneof the RVC Research farm, alongside social differentiation across functional zones. Cows showed preferences for specific areas of the non-feeding zones, more so on the commercial farm in Essex than on the RVC Research farm. Cows increased their bunching behaviour ≥ 20°C in terms of all bunching metrics on the commercial farm in Essex. This pattern was observed for nearest neighbour distance on the RVC Research farm ≥ 15.91°C. This thesis demonstrates the use of precision livestock farming to monitor changes in group-level behaviour to improve the health and welfare of livestock.