„The paper reports a comparison of the breeding value estimation for clinical mastitis when single or multi-trait models were employed. The evaluation was carried out on a set of 13,775 Holstein cows, with 33,809 lactations. Monitoring of clinical mastitis occurred between 2000 and 2018, on seven farms with an average number of cows ranging from 100 to 800 heads, average milk production of 8,600 kg to 10,100 kg of milk per lactation. The incidence of clinical mastitis was expressed as the number of clinical mastitis cases within 30 days from calving and throughout lactation, i.e. from calving to 305 days of lactation. In addition to clinical mastitis, the multi-trait model included some udder type traits: fore udder attachment, udder cleft, udder depth and udder width. A linear type scoring was available for cows on the first lactation. Therefore, only cows with information about production and clinical mastitis in the first lactation available were included in the analysis. A linear animal model with repeatability effect was used to estimate genetic correlations and the breeding values.
Phenotypic correlations between the number of clinical mastitis and linear type traits were weak, mostly around 5%. Except for udder width and depth, they were stronger for the number of clinical mastitis cases for the entire lactation than for the first 30 days of lactation. Genotypic correlations were more substantial than phenotype correlations, ranging from 13 to 35% in absolute terms. For the multi-trait model, there was an increase in the reliability of breeding value estimates, but only for the occurrence of clinical mastitis up to 30 days of lactation.“