Manageable risk factors associated with bacterial and coliform counts in unpasteurized bulk milk in Flemish dairy herds.

Associations between herd management practices and both bacterial counts (BC) and coliform counts (CC) from 254 and 242 dairy herds in Flanders (Belgium), respectively, were studied. Data were analyzed using multivariable, multilevel linear regression analysis, allowing variance components analyses. Both BC and CC fluctuated throughout the year, although the milk quality parameters followed an opposite pattern. Bacterial count values decreased with each increase of the cleaning frequency of the cubicles (once per week, once per day, twice per day, or more than twice per day) between January and March. Herds with a conventional milking parlor had substantially lower BC than herds where the cows were milked using an automatic milking system. Lower BC were observed when the milking parlor was equipped with an automatic cluster removal system, when premilking teat disinfection was applied, when the dry cows were supplemented with a mix of minerals and vitamins, and when the teats were prepared either first wet and dried or via an automatic milking system. Milking cows with a high-pipeline milking parlor setup or with an automatic milking system was associated with substantially higher CC values. Herds where prepartum heifers were often treated with antimicrobials before calving had a lower CC than farms where heifers were either not or only rarely treated. Most variation in BC and CC resided at the herd level rather than at the observation level, indicating that management is important in the control of both BC and CC. Still, only a small proportion of the total variance was explained by factors capturing information related to the milking, herd health, and dry cow management, which suggests that the bacteriological milk quality and, in particular, CC is primarily driven by other factors than the ones included in this study.