Comparison of two direct bioassays using 3-week-old broilers to measure the metabolisable energy of diets containing cereals high in fibre: differences between true and apparent metabolisable energy values

1. Nitrogen-corrected apparent and true metabolisable energy contents (AME(N) and TMEN) of 12 diets, containing different amounts of maize, wheat, barley and rye, barley with beta-glucanase and maize with 0.05 g/kg guar gum were measured using 3-week-old male broilers. 2. The AME(N) method involved: 4 d of dietary adaptation, 24 h of starvation, 54 h of ad libitum feeding, 24 h of starvation and daily total droppings collection. TMEN was measured after a 48 h starvation period followed by 2 d of tube-feeding two doses of 15 g of the diets and quantitatively collecting all the droppings after 48 h of starvation. The endogenous energy losses (EELN) were measured after tube-feeding 8 chickens with 30 g of glucose, also over 2 d. 3. The results show that a significant interaction exists between the nature of the diets and the method used to measure metabolisable energy. Differences between TMEN and AME(N) values ranged from 1.03 to 1.98 MJ/ kg dry matter. These differences were positively correlated with the total pentosan contents of the diet (r = 0.94), the natural logarithm of the viscosities of the water extracts (r = 0.87), the sum of the contents of total pentosans and total beta-glucan (r = 0.85), the calculated total non-starch polysaccharide contents (r = 0.79), the neutral detergent fibre contents (r = 0.76) and the total beta-glucan contents (r = 0.70). Furthermore, a significant interaction was found between diets and the method used to derive lipid digestibility. The differences between apparent and true lipid digestibility ranged from 0.05 percentage points for the wheat diet to 0.20 percentage points for the rye diet. 4. From the results of this trial, it appears that TMEN overestimates the energy value of high fibre diets.