Search for a command to run...
Heat stress (HS) negatively affects the health, reproduction, and milk production of dairy cows, indicating a major challenge to animal welfare and farm profitability. Although environmental and physiological indicators are commonly used to monitor HS, cellular biomarkers such as heat shock protein 70 (HSP70) offer a promising approach for detecting HS at the molecular level. This research aimed to evaluate and compare HSP70 levels in the milk, blood, and saliva of lactating dairy cows across seasons, assess their environmental responsiveness using the temperature-humidity index (THI) as the standard environmental indicator of HS under temperate Australian conditions, and investigate the potential of milk HSP70 as a noninvasive biomarker of HS at the cellular level. Twenty multiparous Holstein Friesian cows were monitored across 4 seasons (summer, autumn, winter, and spring) over 1 yr, and HSP70 concentrations were measured using a competitive ELISA system. Weather data, including THI, as well as milk yield and composition traits, were also recorded. Results showed significant seasonal variation in HSP70 levels, with the highest concentrations observed in summer across all sample types. Blood exhibited the highest (416.45 ng/mL) HSP70 concentrations overall, followed by saliva (304.89 ng/mL) and milk (279.02 ng/mL). Milk HSP70 showed the highest variability (CV = 53.76%), indicating that additional factors beyond HS may influence its expression, whereas blood and saliva appeared to be more consistent. The associations between HSP70 concentrations in biological fluids were initially evaluated using Pearson correlation analysis. To account for repeated measurements in cows and identify key predictors, linear mixed-effects models with cow identification as a random effect were then fitted. Moderate positive correlations were found between the concentration of milk HSP70 and both blood (r = 0.58) and saliva HSP70 (r = 0.53), indicating that milk HSP70 may reflect systemic stress responses in dairy cows. Multivariable linear mixed models identified average THI as a significant predictor of HSP70 concentrations across fluids. In the milk HSP70 model, DIM was also a significant predictor, and milk yield was not. Overall, this study provides preliminary evidence that milk HSP70 may have potential as a noninvasive indicator of HS in dairy cows. These findings also contribute to a better understanding of HS physiology and may inform future approaches for herd-level monitoring and management under climate change.