Reaction norm modelling of heat stress for milk production in three dairy cattle breeds

Due to climate change, extreme weather events are becoming more frequent, which include days with higher temperatures, especially in the summer period. These more frequent heat events can have physiological consequences on dairy cattle when the ambient temperature exceeds their thermoneutral zone. The present study analysed heat stress effects on milk yield in three German dairy cattle breeds with data from southwest Germany. Daily milking and climate data (temperature and humidity) were collected between June 2022 and December 2024 from 5 182 dairy cattle from German Holstein Friesian, German Fleckvieh, and German Brown Swiss, on 46 farms in the state Baden-Württemberg, Germany. The milk yield was modelled as a function of the temperature-humidity index (THI). The average daily THI was determined and transformed by a double hinge function. Random regression reaction norm models with higher order Legendre polynomials were used to describe the genetic variation in milk yield across different THIs. The estimated variances of the animal effect in relation to the environmental variances were comparable across the breeds. The pattern of additive genetic variance was nearly linear, while the permanent environmental variance followed a u-shaped curve along the THI trajectory for all three breeds. Genetic correlations of milk yields at different THIs were small and, in a few cases, also negative for larger differences between THI, which implies that animals vary genetically in their response to thermal conditions. These correlations must be validated using large data sets. Breeding for improved heat tolerance and overall resilience towards thermal stress is expected to become an important objective, particularly considering the rising frequency of heat extremes.