Estimates of Retained Nitrogen Used to Correct Estimates of Bioavailable Energy

Bioavailable energy (BE) values of poultry feedingstuffs are frequently corrected to zero nitrogen balance (RNb). An implicit assumption is that RNb is a valid estimate of retained nitrogen (RN); when RN is estimated by comparative slaughter (RNS), it is common to find RNb>RNS suggesting that one or both is an inaccurate estimate of RN. An experiment with broiler chicks provided estimates of RNb, RNS, and of the BE of a lipogenic diet. The RNb was generally larger than RNS and slightly more variable. The effects of the nitrogen (N)-corrections on apparent metabolizable energy (AME) resulted in AMEb (AME corrected to RNb = 0) differing significantly (P<.01) from AMES (AME corrected to RNS = 0), but in terms of precision, there was little difference between the two estimates. In a second experiment with adult Single Comb White Leghorn cockerels, RNb was, on average, greater than RNS, but the ability to predict the initial N content of the experimental birds was much inferior to that in the chick experiment. The RNb was less variable than RNS, and the greater precision was carried through into the N-corrected BE estimates. In practical bioassays, RNb is preferred to RNS for ease of measurement and conservation of birds; however, there is evidence that the estimate RNb may differ from RN, and this raises questions about the constants used in making nitrogen corrections.