Effects of extreme heat on milk quantity and quality

CONTEXT Heat exposure, particularly when co-occurring with extreme humidity levels, is amongst the most pronounced and challenging extreme weather conditions in European dairy production. Since these conditions are expected to increase in magnitude and frequency due to climate change, it is crucial to better understand the actual impact under managed production within the European dairy farm population. OBJECTIVE In this paper, we estimate the impact of heat stress, i.e. hot and humid weather conditions on milk quantity and quality in Flanders, a high intensity dairy production region in Belgium. The novelty of our approach is twofold. First, we provide insights in the non-linear response of dairy production (quantity and quality) to extreme heat using farm-level production data. Second, using data from the entire population of farmers in Flanders, and covering 6 years (N = 178.843), our results provide a maximum of representativeness and allows to infer on economic relevance of the effects. METHODS We use monthly milk deliveries within the grazing season for the period 2009 to 2014, which we match with temperature and humidity conditions at the barn-level. Using fixed effects regression, we particularly focus on heat shocks, i.e. the deviations from the average climate at the barn location, on production shocks, i.e. deviations from the average production of the barn. We use non-linear restricted cubic splines regression to estimate the response of milk quality and quantity to hourly temperature-humidity (THI) exposure. RESULTS AND CONCLUSION We find reductions of milk quantity and protein contents under hot and humid conditions, while milk fat content appears unaffected. These results are in contrast to earlier findings of heat exposure on dairy production in two ways. First, we find the critical temperature-humidity threshold above which production is affected to be substantially higher than in earlier/experimental studies. Second, not all quality components are affected by these conditions. We conclude that ...