Predicting ruminal microbial efficiencies of dairy rations by in vitro techniques

The efficiency of microbial production (EMP) of nine mixed diets consisting of approximately even parts of roughage and concentrates was investigated in vivo and in vitro. EMP in vivo was estimated based on ruminal N-15 infusions Of dairy cows and varied from 27.8 to 34.6 g of nitrogen (N) per kg truly fermented OM (mean = 31.9 g). EMP in vitro was estimated by partitioning factor (PF ratio of mg of OM truly degraded to mi gas produced thereby), by two microbial N balances and by purine base (PB: adenine and guanine) analysis. Both microbial N balances were based on the determination of undegraded feed N by neutral detergent (ND) solution treatment and N analysis. Changes in the NH, concentrations in the incubation medium were taken into account in one of the microbial N balances. All in vitro incubations were conducted in N-low (-NH4) and N-rich ( + NH4) medium. In vitro digestibility measurements were consistently higher (P < 0.0001) in N-rich medium but the effect of N-level on in vitro microbial efficiency varied in dependence of how microbial production was analyzed, Of the estimates of EMP in vitro only PF values were significantly related to EMP in vivo, the relationship being slightly closer for the - NH4 (r = 0.77, P = 0.016) than for the + NH4 (r = 0.74, P = 0.022) medium. PB per unit ATP (<mu>mol/mmol) tended to be inversely related to EMP in vivo in both -NH4 (r= -0.52, P=0.148) and +NH4 (r = - 0.32, P = 0.403) medium. The amount of substrate truly degraded in vitro as determined by ND-solution treatment was in good agreement with the sum of the fermentation products recovered suggesting that this treatment can also be used for the estimation of in vitro true degradability of mixed diets. (C) 2001 Published by Elsevier Science B.V.