THE EFFECT OF MONENSIN ON LACTATION DAIRY COWS: A DOSE RESPONSE EVALUATION

Abstract

Monensin (Rumensin®) was fed at doses of 0, 8, 16, or 24 ppm to 966 dairy cows in nine different geographical locations in the USA and Canada. A dose response analysis was conducted on the primary variable, milk production efficiency, to determine the most appropriate dose response function, establish a minimum effective dose, and, when possible, determine a maximum effective dose. Linear mixed models (SAS® Proc Mixed v6.12) were fit to the data. Linear contrasts comparing the non-zero doses of monensin to the control were done to initially determine a minimum effective dose from the 3 non-zero design points. In addition, eight predefined linear contrasts were used to initially determine the general linear-plateau shape of a dose response function for each primary variable. A weighted regression analysis of the least squares means and corresponding standard errors was used when it was necessary to discriminate between the competing linear-plateau functions. A non-overlapping confidence interval process was followed, if it was deemed appropriate, to establish a minimum effective dose for a nondesign point. In cases where the dose response function had a plateau, the dose where the plateau began was classified as the “maximum effective dose” (minimum dose for maximum effect). In cases where the dose response function did not have a plateau, the maximum effective dose was the largest dose used in the study if the response rate was significant.