The chemical composition of wheat is believed to influence its apparen
t metabolisable energy (AME) value for young poultry. In a preliminary
study, 10 samples of wheat were assessed biochemically and AME was de
termined by chick bioassay. Diets were formulated to include wheat at
90%. Biochemical analyses of the wheat and excreta included starch, di
etary fibre components, nitrogen, lipid and gross energy. Although the
wheats tested did show a compositional variation, none were identifie
d as being low-AME types. Mean values ranged from 11.98 to 14.90 MJ/kg
of dry matter. Significant differences were found between the content
s of starch (g/kg DM; range of 608-744, p < 0.001) and protein (N x 5.
83, g/100 g DM, range of 9.57-14.69, p < 0.001), although these were n
ot correlated with AME. AME was correlated with apparent starch digest
ibility in each trial (r2 = 0.96 and 0.74) and apparent starch digesti
bility was found to be independent of the NSP level or composition. Th
e arabinoxylan content was not correlated with AME or energy metabolis
ability. No relationship was observed between the AME and starch or th
e non-starch polysaccharide (NSP) content. Recoveries of NSP from excr
eta were high (up to 86%) and AME was independent of NSP digestibility
. A strong negative correlation (r2 = -0.92) existed between the AME a
nd the ratio of arabinose-to-xylose in the NSP. Fractionation of the N
SP into soluble and insoluble components revealed a variable ratio of
between 0.28 and 1.00. No significant correlation existed between eith
er of these two variables and the AME.