Evaluation of equations predicting the net portal appearance of amino acid nitrogen in ruminants.

Martineau, R., Côrtes, C., Ortigues-Marty, I., Ouellet, D.R., and Lapierre, H. (2014). "Evaluation of equations predicting the net portal appearance of amino acid nitrogen in ruminants.", Journal of Dairy Science (JDS), 97(3), pp. 1670-1681. doi : 10.3168/jds.2013-7249  Access to full text


A better assessment of digestible protein and AA flows is required to improve the predictions of animal performance in ruminants (e.g., growth and yields of milk and milk protein). In that respect, 2 recent meta-analyses were conducted in our laboratory to establish the relationships between net portal appearance of AA nitrogen (NPA-AAN) and dietary characteristics either from the National Research Council (Washington, DC) or Institut National de la Recherche Agronomique (INRA; St Genès Champanelle, France). Three prediction equations were selected from these meta-analyses: one equation based only on N intake (NI) and 2 equations based on NI, the intake of neutral detergent fiber, plus the dietary concentration of either total digestible nutrients or digestible organic matter. In the current meta-analysis, 2 new equations were developed to predict NPA-AAN from the estimated supply of metabolizable protein (MP) and the protein truly digestible in the intestine (PDI). The reliability of these 5 equations to predict NPA-AAN was evaluated using an independent database. On average, NPA-AAN predictions based on the supply of MP or PDI had the highest coefficient of determination and the lowest root of mean square prediction error and mean and regression biases compared with predictions based on dietary characteristics, suggesting better reliability with the former. No major difference was detected between NPA-AAN predictions based on parameters from the National Research Council or INRA, except that predictions based on MP had the lowest mean and regression biases. In each equation, mean of residual NPA-AAN (observed NPA-AAN minus predicted values) was lowest and negative for sheep compared with dairy cows, suggesting that NPA-AAN were overpredicted in sheep. Many continuous variables biased NPA-AAN predictions based on NI only, but none of the tested variables biased the predictions based on the supply of MP or PDI, corroborating the better reliability for the prediction equations based on the supply of digestible protein. Of the tested continuous variables, only the dietary concentration of crude protein (CP) biased NPA-AAN predictions based on NI plus dietary characteristics. The NPA-AAN responses to dietary CP concentration were overpredicted as dietary CP concentration increased and underpredicted as CP decreased, suggesting that ruminants were more efficient at converting ingested N into digestible protein when fed low-CP diets compared with high-CP diets.

Date modified: