Poppe, M. (2022). Genetic improvement of resilience in dairy cattle using longitudinal data. PhD thesis, Wageningen University, the Netherlands Resilience is the ability of cows to be minimally affected by, and to quickly recover from environmental disturbances, such as pathogens, extreme weather, or changes in feed quality. Resilience is important for welfare of cows and work pleasure of farmers, especially given an expected increase of disturbances in the future. This thesis focuses on developing indicators of resilience from longitudinal data, based on the expectation that longitudinal data records show response patterns to disturbances, informative about resilience. Resilience indicators based on response patterns can potentially assist in selecting the genetically most resilient animals for breeding, and in identifying resilience at herd level. The main focus of this thesis was on the resilience indicators 'variance' and 'autocorrelation' of daily deviations from expected values of milk yield and step count. Resilience is indicated by low variance (small deviations from baseline) and low lag-1 autocorrelation (weak dependency between subsequent deviations -quick recovery). Variance and autocorrelation of daily deviations from expected milk yield were heritable (0.21 and 0.09, respectively) and both were genetically similar within and between lactations. Variance seems informative about vulnerability and strength of response to disturbances, based on genetic correlations with milk loss upon reallife disturbances and with health, longevity and fertility. Autocorrelation had a favorable genetic correlation with recovery rate upon disturbances. Furthermore, variance was associated with lifetime gross margin, which confirms its economic importance. Levels of variance and autocorrelation differed considerably between herds, suggesting herd management plays a role in resilience. When based on step count instead of milk yield, variance and autocorrelation were heritable (0.14 and 0.04), and autocorrelation was favorably genetically correlated with health traits and fertility. However, variance had unfavorable genetic correlations with health. Additional traits based on step count were also heritable (especially mean step count; 0.45), but more research is needed into their relations with resilience. Genetic selection on a resilience index based on variance and autocorrelation of milk yield deviations, and potentially autocorrelation based on step count deviations, is expected to result in more resilient cows, that are less affected by disturbances and that recover more quickly.
Contents