Cornell Net Carbohydrate and Protein System: A model for precision feeding of dairy cattle

The Cornell Net Carbohydrate and Protein System (CNCPS) predicts cattle requirements and nutrient supply for site-specific situations. This paper describes the CNCPS version 6 (CNCPSv6), which represents a re-engineering and updating of CNCPS version 5 with the following objectives: (1) improve the organization of the model and user interface to improve speed and accuracy in formulating diets for a herd of dairy cattle, (2) expand the carbohydrate pools to include sugars, soluble fibers, and organic and volatile fatty acids, (3) integrate a fat model to account for ruminal lypolization and biohydrogenation, and absorption of fatty acids in the small intestine, and (4) update the computational sub-models with new information. The CNCPSv6 model was re-designed using object-oriented programming in which physiological functions (e.g. growth, lactation, pregnancy) and anatomical compartments (e.g. rumen, intestines) were programmed as objects. The interface uses farm, location, and group flow, which decreases the number of inputs required per cattle group and allows for more rapid evaluation of diets, feed requirements, and nutrient excretion by location, group, and herd. The revised implementation of the body reserves sub-model allows accounting for fluxes in energy reserves when formulating diets. Updated equations and coefficients include the prediction of rumen ammonia balance and feed passage rates, indigestible DM, MP lactation efficiency, and DMI. The CNCPSv6 was evaluated with data from individually fed lactating dairy cows from three independent studies. As implemented, CNCPSv6 accounted for a similar proportion of the variation (86%) in first limiting (ME or MP) milk production as CNCPSv5 but with a lower bias (1% versus 4%, respectively). We concluded the re-designing and updating of the CNCPS improved its ability to formulate and evaluate a feeding program for a herd of dairy cattle with greater accuracy and efficiency.