A critique of the Cornell Net Carbohydrate and Protein System with emphasis on dairy cattle. 1. The rumen model

The Cornell Net Carbohydrate and Protein System (CNCPS) is primarily a nutr ient supply model, since it does not model the utilisation of the absorbed nutrients, other than by using the NRC (1988) net energy and NRC (1985) ava ilable protein systems to calculate the animal's energy, protein and amino acid (AA) requirements. The model incorporates details of the metabolism in the rumen not found in other published models of energy and protein requir ements, particularly rates of carbohydrate degradation, predicted rumen pH and rumen nitrogen and peptide balance. The model does not predict volatile fatty acid (VFA) proportions or methane production. Microbial growth is as sumed to be dependent upon the rate of carbohydrate degradation, whereas ot her mathematical models of the numen relate microbial growth to the concent ration of nutrients in the rumen, and also require estimates of rumen volum e and microbial mass. The microbial yield of structural carbohydrate (SC) b acteria is not limited or altered by the rumen ammonia supply, with the res ult that stoichiometrically unsound amounts of microbial protein and negati ve rumen ammonia levels can be predicted. Only estimated peptide supply mod ifies die growth of non structural carbohydrate (NSC) bacteria. Solid outfl ow rates adopted are low at higher levels of feeding compared to ARC (1984) and AFRC (1992), with the result that the proportion of OM digested in the rumen is predicted to be higher (c. 0.75) than the mean value 0.65 adopted by ARC (1980). The effects of liquid outflow rate upon the outflow of NPN, AA and soluble proteins, pectins, sugars, organic acids and VFA are ignore d. Only peptides are affected by the liquid outflow rate in the model. The degradation rates for carbohydrate fractions A and B1 proposed are very hig h, exceeding the possible rate of microbial growth so that microbial synthe sis does not respond to considerable variations in these high rates. The ad opted maximum microbial yields of SC and NSC rumen bacteria are lowered by 20% to allow for the effects of protozoal predation, which has the effect o f compensating for the high predicted microbial yields adopted, but the pro tozoa are not accounted as contributing to the microbial AAN output. Starch disappearance in the rumen is not corrected for protozoal ingestion and it s re-appearance in the intestine, nor is there is any accounting for the pr otozoal contribution to fat uptake in the rumen. Predicted TDN values of fo rages are therefore sensitive mainly to the rate of cell (eNDF%) degradatio n selected. Only one dietary parameter, effective neutral detergent fibre ( eNDF%) defines or modifies maximum microbial yield, microbial maintenance, realised microbial yield, and rates of SC and NSC degradation. The conseque nce is that both energy and protein supply to the cow are affected by the p arameter eNDF when values fall below 24.5% in diet DM.