The objective of this analysis was to compare the rumen submodel predictions of four commonly used dairy ration programs to observed values of duodenal flows of crude protein (CP), protein fractions and essential AA.

The literature was searched and 40 studies — including 154 diets — were used to compare observed values with those predicted by AminoCow, Agricultural Modeling and Training Systems (AMTS), Cornell-Penn-Miner (CPM) and National Research Council 2001 (NRC) models. The models were evaluated based on their ability to predict the mean, their root mean square prediction error, error bias, and adequacy of regression equations for each protein fraction. The models predicted the mean duodenal CP flow within 5 percent, with more than 90 percent of the variation due to random disturbance. The models also predicted within 5 percent the mean microbial CP flow except CPM, which overestimated it by 27 percent. Only NRC, however, predicted mean rumen-undegraded protein (RUP) flows within 5 percent, whereas AC and AMTS underpredicted it by 8 to 9 percent and CPM by 24 percent.

Regarding duodenal flows of individual AA, across all diets, CPM predicted substantially greater (>10 percent) mean flows of Arg, His, Ile, Met, and Lys; AMTS predicted greater flow for Arg and Met, whereas AminoCow and NRC estimations were, on average, within 10 percent of observed values. Overpredictions by the CPM model were mainly related to mean bias, whereas the NRC model had the highest proportion of bias in random disturbance for flows of essential AA. Models tended to predict mean flows of essential AA more accurately on corn silage and alfalfa diets than on grass-based diets, more accurately on corn grain-based diets than on non-corn-based diets, and finally more accurately in the mid range of diet types. The models were accurate at predicting mean dry matter intake. The AminoCow, AMTS and NRC models were all sufficiently accurate to be used for balancing essential AA in dairy rations under field conditions.

Source: D. Pacheco, R.A. Patton, C. Parys, H. Lapierre/Journal of Dairy Science