Amino acid balancing using the factorial system in the CNCPS and CPM Dairy

Larry Chase and I frequently receive questions regarding amino acid balancing within the factorial system utilized by the Cornell Net Carbohydrate and Protein System (CNCPS) and CPM Dairy to evaluate amino acid nutrition of dairy cattle. The general form of these questions is: "I ran this ration through the Cornell model and it told me that (isoleucine, leucine, valine, arginine – take your pick) is limiting. Is it real??" The general answer that Larry and I give is an unsatisfying "We don’t think so, but we’re not sure."

It is important to understand how this system works when we are interpreting output from the amino acid (AA) submodel. The AA submodel combines a prediction of individual AA supply from microbial and feed protein sources with a prediction of the requirement for each AA. The calculation of AA supply relies on a prediction of microbial protein yield combined with the average AA composition of rumen microbes plus a prediction of the amount of feed protein that escapes the rumen combined with the average AA composition of the insoluble protein fraction of the feed. The model uses the AA composition of the insoluble protein in the feedstuff because, for some feeds, the AA composition of the soluble protein in the feedstuff is quite different than the insoluble protein in the feedstuff. The calculation of amino acid requirement uses the net amount of protein synthesized for each function in the cow (maintenance, growth, gestation, and lactation), its amino acid composition, and an efficiency factor for conversion of absorbed (metabolizable) amino acids to the net amino acid requirement.

If we break each calculation down into its components and evaluate each component, we can begin to determine which steps may have greater variation associated with them. The prediction of total metabolizable protein supply to the cow seems to be quite accurate, although there probably is a slight overestimation of microbial protein supply and underestimation of feed protein supply in the current version of the model. The greatest culprits for variation in the prediction of AA supply are the AA compositions used for microbial protein and insoluble feed protein. Table 1 depicts the mean values and variation associated with the AA composition of microbial protein. This variation occurs as a result of sampling technique in the various experiments, amino acid analytical variation, and variation due to changes in the microbial population when cattle are fed different diets. At this time, we do not have a satisfactory method to adjust microbial AA composition for differences in the intake and composition of diets fed to cattle.

There are fewer steps in the calculation of AA requirements, yet these probably also are a major source of variation. In a high-producing dairy cow, most of the AA requirement is driven by lactation; maintenance has a relatively small requirement for AA. The AA submodel uses an average AA composition of milk true protein. Although there certainly will be some variation associated with these values, it probably is relatively small because the AA composition of proteins in whey and casein fractions is genetically encoded, and the proportions of whey and casein, and the types of caseins present, will not vary to a large extent in milk protein. Most of the variation on the requirement side is present in the efficiency factors used for maintenance and lactation (Table 2). The efficiency factors used for maintenance are .85 for all essential AA except for the branched-chain AA (.66). The branched-chain AA are discounted because in nonruminants, branched-chain AA are catabolized by muscle tissues. More recent data would indicate that this may not be the case in ruminants and that a value equal to the other essential AA may be more appropriate; however, until we have data to substantiate a change, we will leave these coefficients as-is in the model. Of greater interest and importance are the efficiency factors used for lactation. The system is particularly sensitive to these factors. We have recently updated these efficiency factors (Overton, 1999), and the updated factors will be included in CNCPS version 4.0. The new factors will result in valine and arginine (and perhaps isoleucine) coming up limiting with some frequency. Again, is this real??

There are a small number of experiments published in the scientific literature in which cows were administered (usually by infusion into the abomasum) either branched-chain AA or arginine. Mackle et al. (1999. J. Dairy Sci. 82:161-171) infused either branched-chain AA (150 g/d), casein (600 g/d), or a combination of the two (44 g/d of branched-chain AA plus 600 g/d of casein) into the abomasum of lactating cows. Milk yield, milk protein percentage, and milk protein yield was not affected by infusion of branched-chain AA, but percentage and yield of milk protein were increased slightly by infusion of casein. Supplementation of the casein infusate with branched-chain AA had no effect beyond the casein infusion alone on performance in this experiment.

In another experiment published recently, Robinson et al. (1999. J. Anim. Sci. 77:2781-2792) formulated a basal diet using a previous version of the CNCPS to be co-limiting in intestinally absorbable supplied of methionine, lysine, and isoleucine and then infused isoleucine into the abomasum, fed a rumen-protected methionine and lysine product, or both infused isoleucine into the abomasum and fed a rumen-protected methionine and lysine product. Feeding the rumen-protected methionine and lysine supplement tended to increase percentage and yield of milk protein and infusing isoleucine into the abomasum tended to increase milk yield; however, the combination of isoleucine, methionine, and lysine did not affect yields of milk and milk components compared with the control.

Even fewer data are available on the effects of arginine infusion into the abomasum. Vicini et al. (1988. J. Dairy Sci. 71:658-665) infused 178 g/d of arginine into the abomasum and did not report any significant effects on production, although the experimental design that they utilized did not lend itself to detecting differences in production. Hopkins et al. (1994. J. Dairy Sci. 77:1084-1092) infused a mixture of isoleucine, leucine, valine, and arginine into the peritoneal cavity and did not affect yields of milk and milk protein, but had small effects on milk fat synthesis.

It’s time again for the annual Feed Dealer Seminars. Larry Chase and Tom Overton will be discussing what’s new in forage analyses, amino acid nutrition of dairy cows, ionophores in relation to transition dairy cows, and choline for transition cows. Locations, times, and contacts follow:

Please email Tom Overton at tro2@cornell.edu to be added to our list of subscribers for notification of new issues.