ABSTRACT
The objective of the current study was to quantify the change in the prediction of dry matter intake (DMI) resulting from the inclusion of rumination time (RT) in the 2001 National Research Council (NRC) DMI prediction model. Forty-one Holstein cows fed the same total mixed ration were involved in a 10-wk study. Individual DMI were measured daily. The accuracy and precision of the original NRC prediction model, based on body weight, fat-corrected milk, and week of lactation as independent variables, was compared with the accuracy and precision of the same model with RT as an additional independent variable. The RT estimate was significant in the model developed but had a low value (0.031 kg/h). Root mean square prediction errors were very similar in the 2 models (1.70 and 1.68 kg/d) as were the other indicators (R(2), linear bias, random error, and concordance correlation coefficient) selected to compare the models in this study. These results indicate no gain in DMI prediction precision or accuracy when RT is included in the NRC model.