Behavioural and physiological non-invasive indicators of heat stress in dairy cows

Climate change is increasing the frequency, intensity, and duration of heat waves worldwide, raising major concerns for dairy cattle welfare, health, and productivity. Dairy cows are particularly vulnerable to heat because of their high metabolic heat production associated with milk yield. In response to hot conditions, cows activate a range of physiological and behavioural adaptations, including increased respiratory rate, changes in posture and activity, and modifications in shade-seeking behaviour. However, many currently used indicators of heat stress are invasive and difficult to apply continuously under field conditions. In addition, they do not allow to assess the stress (affective response) experienced by animals. This thesis investigates the behavioural and physiological responses of dairy cows exposed to heat under contrasting climatic contexts (Mediterranean and continental climates) and production systems (rustic versus high-producing breeds). The objective is to better understand how cows adapt to hot conditions, how these responses evolve during repeated heat waves, and whether thermal discomfort can be detected using non-invasive indicators, including facial-expression-based measures and udder infrared temperature. Overall, cows exposed to hot conditions showed clear thermoregulatory and behavioural adjustments. Behavioural indicators such as posture, activity, and shade use appeared closely linked to environmental conditions, supporting their relevance for assessing adaptive responses to heat. Changes in several facial features were observed under hot conditions, suggesting that cows express discomfort due to heat through multiple facial patterns. These findings support the potential of facial expressions as indicators of animals’ affective responses to heat. Finally, while infrared surface temperatures showed some sensitivity to heat, they did not reliably reflect core body temperature, although certain measures (e.g., udder temperature) may have potential as early indicators. This thesis advances the understanding of dairy cows’ responses to heat by integrating behavioural, physiological, and affective indicators across diverse contexts. It highlights the value of non-invasive approaches for early detection of thermal discomfort and opens new perspectives for improving animal welfare assessment under climate change.