Quantification of group and individual behavioural responses to a nutritional challenge in dairy goats

Monitoring individual animal behaviour using sensor technology provides valuable insights into adaptation to environmental challenges. This study investigates the effects of a controlled nutritional challenge on the behavioural patterns of dairy goats using accelerometer-derived behavioural data. Fourteen goats were equipped with collar-mounted accelerometers for 19 days, including two days of nutritional challenge in the middle. The challenge consisted of a 7-day control period with standard total mixed ration (TMR), 2 days of straw feeding only, and a 10-day recovery period on the TMR. The ACT4Behav pipeline was used to classify behaviour from accelerometer data, and multiple features were extracted to characterize activity profiles. Post-challenge behaviours were compared to pre-challenge ones, considered as baseline. At the group level, mixed linear model analysis showed that rumination time decreased during the nutritional challenge, while time spent with the head in the feeder increased, particularly during the day. Lying time remained unchanged during the day but increased at night. At the individual level, high variability in behavioural responses was observed. Two novel indices, the Challenge Effect Index (CEI) and the Recovery Index (RI), were introduced to characterise the diversity of behavioural responses and recovery strategies among goats. While some goats fully recovered their baseline behavioural patterns, others exhibited higher levels of certain behaviours after the challenge. These results demonstrate that group-level analyses are not sufficient to assess the impact of environmental challenges, as individuals may have different response strategies.