Provision of Supplemental Feed and the Application of a Nitrogen Correction in Bioassays for True Metabolizable Energy

Abstract

Abstract Four experiments were made to study the feasibility of reducing the severity of the fast in the bioassay for true metabolizable energy (TME) and to measure the effect of correcting the fecal plus urinary energy (FE + UE) outputs to zero nitrogen balance. Each experiment followed the general procedures of a TME bioassay. Test materials included corn, oats, meat meal, fish meal, wheat, soybean meal, and mixtures of purified ingredients. In the first experiment, graded levels of a starch:oil:glucose mixture were given to fasted cockerels and the excretion of energy and nitrogen during the subsequent 48 hr was measured. The control birds in the second experiment were unfed, which is usual, or precision-fed a known quantity of one of three energy sources; each test material was fed at levels of 15, 30, and 45 g. In the third experiment, control birds were fed 0, 15, 30, or 45 g of an energy source while the test materials were fed alone (45 g) or at 30 g in combination with 15 g of the energy source. In the fourth experiment, a pelleted energy source, which contained 5% by weight of casein, was available free-choice to half of the birds during a 48-hr preliminary fast and the 48-hr collection period of a TME bioassay. Provision of supplemental energy reduced, but did not prevent, body weight and nitrogen loss. When control birds were precision-fed with an energy source, reasonable TME values were obtained, but bias was introduced if the source was not completely digestible; the latter problem can be overcome by giving the same amount of the energy source to all birds. Supplementing low inputs with an energy source can provide acceptable data but the procedure introduces assumptions that must be monitored. Free-choice feeding of energy sources was not satisfactory due to feed spillage, excreta contamination, and highly variable intakes; error variance estimates of FE + UE increased at least fivefold while those for FE + UE corrected to zero nitrogen balance (FEn + UEn) increased twelvefold. The preferred technique, if a 72-hr total fast is to be avoided, is to feed each test material at more than one level (e.g., 20 and 40 g). The application of a nitrogen correction had a profound effect, reducing error variance estimates by more than 40% in each experiment. The correction improves the fit of the linear model on which the TME bioassay is based. The provision of an energy source to reduce tissue protein degradation and thus obtain better estimates of metabolic and endogenous energy losses is unnecessary when the correction is applied. The correction is strongly recommended.