Relationships between linear type scores, objective type measures, and indicators of mastitis

Objectives were to examine the usefulness of objective measures of conformation and linear scores for type for predicting indicators of mastitis. Approximately 300 cows in five herds were scored in the experimental linear classification program of the Holstein-Friesian Association of America, measured objectively for numerous components of conformation, and evaluated for four indicators of mastitis. Indicators of mastitis were average natural logarithm of somatic cell counts during the lactation, percent of samples (taken approximately every 3 mo) infected during a lactation, percent of samples infected with Staphylococcus aureus, Streptococcus agalactiae, Escherichia coli, Klebsiella, and percent of samples with signs of mastitic milk. Relationships between conformation and mastitis often were inconsistent because of moderate or low correlations between indicators of mastitis. Indicators of mastitis were predicted from results of multiple regression analyses with subjective scores and objective measures of type as independent variables. Largest differences of predicted indicators of mastitis from variation in any single type trait were sufficient to be important economically, but average differences were much smaller. Selection to reduce frequencies of cows with deep udders, especially low rear udders, widely placed teats, rear teats too far back, and teats that are short and wide, may augment, modestly, efforts to reduce incidence of mastitis through improved sanitation, antibiotic treatment, and proper milking procedures.