Understanding behavioral responses of dairy cows to thermal load under different milking-system environments is essential for welfare monitoring in subtropical production systems, yet field-based evidence remains limited. This study compared heat-related behavioral responses between an automated milking system (AMS) and a conventional parlor system under commercial subtropical conditions, with the hypothesis that cows in the AMS would exhibit more stable and less variable behavioral patterns under heat stress. Panting, feeding, and rumination were continuously monitored and analyzed at the barn level in relation to the temperature–humidity index (THI). The AMS barn showed lower panting duration and reduced day-to-day variability, whereas the traditional barn exhibited higher panting levels and greater short-term fluctuations. Panting increased consistently with THI, showing a clear temporal trend (β ≈ 0.65 min/day), and emerged as the most sensitive indicator of thermal load, while feeding and rumination showed weaker responses. During heatwaves, panting increased markedly in both systems, with limited changes in feeding and rumination. Multivariate analyses (PCA) showed more compact clustering in the AMS barn, with higher variance explained along the first principal component (PC1 ≈ 44%), compared to greater dispersion in the traditional barn. These findings demonstrate that cows managed in the AMS exhibited more stable and less variable behavioral patterns under heat stress, whereas cows in the conventional parlor system showed higher panting levels and greater short-term fluctuations. These system-level differences directly address the study objective and support panting as a practical indicator for heat-stress monitoring in subtropical dairy systems.
Chao et al. (Fri,) studied this question.