Influence of food input and procedure of determination on metabolisable energy and digestibility of a diet measured with young and adult birds.

Two series of balance trials were performed with adult cockerels and with broiler chickens during their 5th week of life, and one with adult colostomised hens. The latter served to determine the digestibility of protein of the diet under test. All birds were fed the same diet with one exception: in the case of colostomised hens limestone (100 g/kg) was added before feeding. The main purpose of the experiments was to compare the Sibbald procedure for determining the so-called true metabolisable energy (TME) with apparent metabolisable energy (AME) or TME values obtained by applying a conventional addition method (CAM). The results showed that the Sibbald procedure was less precise than CAM. The Sibbald procedure delivered incorrect TME and AME values, the reason for this being the use of starved birds. This led not only to wrong intercepts but also to misleading regression coefficients in the regression equations used to calculate energy excretion on food intake. In CAM there existed no differences between AME and TME. This was not true for the Sibbald procedure, in which AME values differed considerably from TME values. Because of the errors inherent in the Sibbald procedure it is suggested that it should be replaced by CAM, in which several feeding inputs should be used.