Tube-fed broiler cockerels were used in three experiments to measure the effect of different enzyme treatments on true metabolisable energy (TME) and true nitrogen digestibility (TND) of commercial solvent-extracted, heat-treated soybean meal (SBM). 2. In experiment 1, proteases P2 and P3 (from Aspergillus niger) and alpha-galactosidase (from A. oryzae) improved TME and TND while protease P1 (from Bacillus subtilis) had little effect. The effects of enzyme treatment were similar whether treatment was applied by pre-incubation of enzymes (proteases P1, P2 and P3, 1.0 g/kg; alpha-galactosidase, 0.1 g/kg) with SBM for 2 h at 50 degrees C or by simple mixing of enzymes (proteases P1, P2 and P3, 0.25 g/kg; alpha-galactosidase, 0.1 g/kg) with SBM prior to feeding. 3. In experiment 2, the effects of mixing SBM with each of protease P1 (0 or 0.25 g/kg), protease P3 (0 or 0.25 g/kg) or alpha-galactosidase (0 or 0.1 g/kg) alone or in all possible combinations were studied. Effects of protease P1 were limited, but protease P3 and alpha-galactosidase improved TME and TND. There were significant interactions between protease P3 and alpha-galactosidase for both TME and TND; the response when enzymes were applied together was less than that expected from application of the enzyme preparations individually. 4. In experiment 3, the effect of varying the concentration of protease P3 (0, 0.1, 0.25 and 1.0 g/kg SBM dry matter) and alpha-galactosidase (0, 0.025, 0.0625 and 0.25 g/kg SBM dry matter) mixed with SBM alone or in all possible combinations of protease P3 and alpha-galactosidase were assessed. Increases in TME and TND for both enzymes were non-linear with the greatest response occurring at the lowest concentration of each enzyme. There were also interactions between the two enzyme preparations. 5. Finally, either protease P3 (0, 0.1 and 1.0 g/kg SBM dry matter) or alpha-galactosidase (0, 0.025 and 0.25 g/kg SBM dry matter) were mixed with SBM alone or in all possible combinations and treated SBMs incorporated into semi-purified diets containing 450 g SBM/kg as the sole source of dietary N and offered to growing broiler chicks for 21 d. Similar patterns of increases in chick growth rate and diet digestibility to those recorded in experiment 3 were obtained when protease P3 and alpha-galactosidase where included in the diets. 6. It is concluded that responses measured by tube-feeding SBM treated with protease P3 and alpha-galactosidase were similar to responses obtained with growing broiler chicks. Further, these responses were obtained by simple addition of enzymes to diets and nutritive value of SBM was improved by both protease and alpha-galactosidase treatment.
