Evaluation of maternal protective behaviour in dairy cows housed in a cow–calf contact (CCC) system

There is an increased interest among dairy farmers in management systems allowing cow–calf contact (CCC). This study evaluated maternal protective behaviour in dairy cows using Qualitative Behaviour Assessment (QBA) in combination with categorical behavioural scoring systems, selected objective recordings from the automatic milking system (AMS) and computer-controlled smart gates (SG). Maternal protective behaviour was assessed during the first 48 hours after birth by observation of 23 cow-calf pairs in the calving pen (“undisturbed”) and when the cow was restrained in the headlock in the maternity pen for milking while the calf was handled (“restrained”). Data on individual cow teat cup kick-offs (KO), somatic cell counts (SCC) and frequency of cow visits to the calf through SG were registered when the animals were housed in the CCC system. QBA was analysed using principal component analysis (PCA), revealing two main components, labelled mood (PC1) and alertness (PC2), which can be recognised from QBA studies of cattle and other species. A significant, positive association between QBA PC1 “undisturbed” and “restrained” was found, indicating that the cows’ behaviour when undisturbed may partly predict their behaviour when being milked. There was little variation in the categorical maternal protective behaviour scores. All cows except one were scored as exhibiting maternal behaviours labelled as “interested” (licking, vocalizations, shifting of the body position to facilitate suckling, nose-calf-contact) when the cow and calf were “undisturbed”, while most cows (74%) were scored to stand quietly and being attentive to their calf when “restrained”. All cows used the SG to access their calf and the frequencies of the visits varied greatly among cow-calf pairs. We ran regression models to investigate the associations between QBA and SCC, KO and SG passages. No associations were found between QBA scores and these variables, and there were no interrelationships between the selected recordings. The lack of significant associations might be due to a limited sample size, as only 23 cow-calf pairs were included in the study. Nevertheless, QBA may give important, supplemental information about maternal protective behaviour in cattle, but the method needs further validation.