Abstract #M266

# M266
Estimation of dry matter intake of individual cows fed in a group setting using common on-farm measurements.
M. E. Iwaniuk*1, E. E. Connor2, R. A. Erdman1, 1University of Maryland, College Park, MD, 2USDA-ARS, Beltsville, MD.

Feed is the single largest expense for producing milk on dairy farms so producers are interested in exploring methods to improve dairy feed efficiency (FE). To calculate FE and to allow for genetic selection of the most feed-efficient cows within a herd, knowing the dry matter intake (DMI; kg/d) of individual cows is required. The vast majority of dairy cows are fed in large groups such that the DMI of an individual cow within a group is unknown. The objective of this study was to develop and validate a system to estimate DMI of individual cows using measurements that are already commonly recorded on dairy farms. The proposed approach is a modification of an original model developed by Jonker et al. (1998; J. Dairy Sci. 81:2681–2692). For this study, the data set included 2-wk averages for dietary nitrogen (DietN) concentrations (g/kg DM) as well as DMI and milk production (kg/d), milk protein (g/d), and milk urea N concentrations (MUN; mg/dL) from 167 individual cows (2,157 cow observations) across 52 2-week periods. The data were used to predict N losses in milk, urine, and feces to estimate DMI of individual cows fed the same diet by 2-wk periods using the following equation: DMI (kg/d) = [MilkN + A + (B × MUN)]/(0.8 × DietN + 5 – C) where: MilkN = milk N output (g/d), A = intercept for predicted urinary N losses (g/d), B = coefficient used to predict urinary N output based on changes in MUN, 0.8 = availability of DietN digested, DietN = dietary N concentration (g/kg diet DM), 5 = mean metabolic fecal N (MFN; g/kg diet DM), and C = adjustment in differences in diet N availability and MFN. Predicted DMI explained 59% of the total variation in measured DMI (R2 = 0.59; RMSE = 2.15; P < 0.0001). Residual plots of the deviations in actual vs. predicted DMI indicated no mean or linear biases. Therefore, it was concluded that DMI can be estimated on an individual cow basis using common on-farm measurements even if the cows are fed in a group setting. These results will enable producers to calculate individual cow DMI as well as FE to allow for the genetic selection of more feed-efficient cows.

Key Words: DMI, dairy cow, prediction model