Abstract:
Fat is the most variable milk component, and maintaining milk fat continues to be a challenge on commercial dairy farms. Our objectives were to establish associations between herd-level risk factors for milk fat depression and bulk tank milk fat content in commercial dairy herds feeding monensin. Seventy-nine Holstein commercial dairy herds in the northeast and Upper Midwest United States were enrolled in an observational study. Data were collected on herd characteristics, total mixed ration (TMR) samples, all component silage samples, and bulk tank milk samples. The unconditional univariable association of each explanatory variable and bulk tank milk fat percentage was evaluated using simple linear regression and multivariable regression models. Milk fat content of trans-10 C18:1 had an exponentially negative relationship to herd milk fat percentage. In general, milk fat content of fatty acids synthesized de novo in the mammary gland were positively related to herd milk fat, and the content of several trans-C18:1 fatty acids, which would be products of alternate pathways of ruminal biohydrogenation, were negatively related to herd milk fat. Variables related to TMR composition did not have univariable relationships with herd milk fat percentage. Herds that had >49.8% of the TMR particles on the middle screen of the Penn State particle separator had higher milk fat percentage than those with ≤49.8%, and herds with >54.0% of TMR particles in the bottom pan had lower milk fat percentage than herds with ≤54.0%. Dietary content of monounsaturated fatty acids (C16:1 and C18:1) had negative relationships with herd milk fat percentage; however, no single diet component accounted for more than 11% of the variation in herd-level milk fat percentage. Univariable monensin dose was not associated with herd milk fat percentage. The relative lack of significant univariate relationships with herd-level milk fat suggests many factors contribute to milk fat content, and herds experiencing low milk fat will need to examine many potential risk factors when working to troubleshoot this challenge.