Bulk milk quality traits differ according to the ClassyFarm welfare assessment score

The Italian on-farm welfare assessment protocol, named ClassyFarm, evaluates the risk associated with farm housing and management conditions (resource-based indicators) united to the effects of the farming system on the animals (animal-based indicators). It returns an overall welfare score expressed into percentage from 0 (poor) to 100 (excellent). This study aimed to investigate the effect of the farm welfare score on bulk milk quality traits in dairy cow farms, differentiating between loose housing (LH) and tie-stall (TS) systems. Welfare protocol data of 194 farms (89 TS and 105 LH) located in Northern Italy were combined with their milk analyses retrieved from the milk payment systems archive of 2022 (n=6,971) for a retrospective study. A mixed linear model was used for variance analysis considering welfare score classes (insufficient, good, excellent), housing (LH and TS), month, and interaction between housing and welfare as fixed effects. Herd-season was the random effect. For TS, 90% of farms were classified as good, 4% as excellent, 6% as insufficient, while for LH, the corresponding percentages were 75%, 20%, and 5%, respectively. Welfare class and its interaction with housing type affected fat, protein, casein, lactose content as well as somatic cell score (SCS), total bacterial count (TBC) and fat-to-protein ratio. Farms in the worst welfare class exhibited the highest TBC and SCS and the lowest lactose content. The impact of rearing conditions on milk quality remains a complex matter because of the interplay of numerous factors and lack of knowledge regarding cause-effect relation.