Dairy calf characteristics and the association with performance later in life

Early life characteristics of calves may have the potential to improve the welfare of dairy cattle when they are able to predict cow performance in terms of fertility, production, and health later in life (lifelong performance). This might give the opportunity to early select promising animals for longevity, which is a sign of good welfare but also reduces economic costs (less young stock needed) and therefore contributes to a sustainable dairy production. This study aimed to investigate the performance of cows at three life stages (calf, heifer, cow) and to associate life stage performance with lifelong performance. A dataset was used including measures on activity, behaviour and physiological responsiveness to environmental challenge of 374 calves, 336 heifers and 688 cows (all Holstein Friesian). In total, 842 unique animals were included in the study. From 92 animals data was available of all three life stages. Within each life stage a principle component analysis was performed to reduce the number of observed variables into a smaller set of principle components (PCs), explaining the largest percentage of variation of the original dataset. Spearman rank correlations were calculated between PCs and measures of performance (production, growth, health and fertility), between PCs obtained in different life stages, and between measures of performance within and between life stages. Several significant correlations between PCs per life stage and measures of production, growth, health and fertility were found, but all had low values (rs=0.18). Correlating the PCs between life stages, only resulted in a significant correlation between the PCs ‘Cow fear’ (based on milkability and fear score) and ‘Heifer regularity’ (consistent day to day activity patterns) (rs=- 0.23) and between the PCs ‘Cow regularity’ (consistent day to day activity patterns and number of times lying down) and ‘Heifer eating’ (latency to eat after transport)(rs=-0.22). Several significant correlations between production, growth, health and fertility variables were found, but again all correlations were low (rs