Calving time identified by the automatic detection of tail movements and rumination time, and observation of cow behavioural changes

Abstract The use of electronic devices to improve animal health, welfare and farm efficiency in precision livestock farming is a developing area of great scientific and commercial interest. In particular, the use of on-site dairy farm instruments to detect calving is a tool in reproduction livestock farming. The primary aim of this study was to validate the ability of the Moocall device (MD) to detect calving cows. In addition, behavioural changes in parturient dairy cows were evaluated using video-based observations. The MD was applied approximately 9 days before cow delivery. Observational sessions were conducted three times a day for each cow from the day before MD application to calving time. The sensitivity (Se) and specificity (Sp) at 3 and 24 h before calving were measured to test the effectiveness of the MD. In addition, behavioural changes were investigated before and after the MD application as well as before and during calving time. The 3 h Se and the 3 h Sp obtained were 95.2 and 71.4%, respectively. No false negatives were observed in the 24 h before delivery (24 h Se = 100%) while the 3 h Se was 95.2%. The MD was well tolerated by the dairy cows since no change in behaviours was observed in this study among the cows with or without the MD, except for an increase in eating behaviour in the animals with the MD. As regards, the behavioural pattern during calving time (8 h before calving) in comparison with the previous phases, a significant increase in tail contraction frequency and raised tail position, and a decrease in eating behaviour and rumination time were observed. The first principal component (PC) was primarily explained by these variables, and calving cows best contributed to this PC. According to the results of the present study, the use of the MD can be a useful tool in detecting the moment of calving.