Comparison of microbial fermentation in the rumen of dairy cows and dual flow continuous culture

Four ruminal and duodenal cannulated lactating dairy cows and four dual flow continuous culture fermenters were used in a replicated 4 × 4 Latin square to compare fermentation and microbial ecology of in vivo and in vitro systems. Diets contained 32% maize silage, 19.8% alfalfa-grass hay, and 48.2% concentrate, and were arranged in a 2 × 2 factorial with two levels of non-fibrous carbohydrate (40%, 25%) and two levels of ruminally degradable intake protein (12%, 9%). Concentrations of viable bacteria were greater (P = 0.01) in vitro compared with in vivo, averaging 5.04 × 10⁹ and 2.75 × 10⁹ cells per ml, respectively. Cellulolytic bacterial concentrations (cells per ml) were lower (P = 0.04) in vitro (4.09 × 10⁷) compared with in vivo (6.15 × 10⁷), but concentrations of proteolytic and amylolytic bacteria were similar (P > 0.1). Protozoal concentrations (cells per ml) were greater (P = 0.0001) in vivo (3.72 × 10⁵) compared with in vitro (2.8). Digestion of neutral detergent fiber (NDF) and total non-structural carbohydrate (TNC) exhibited square × NFC interactions (P < 0.002), and comparison of these interaction means revealed that the difference in NDF and TNC digestion between ruminal and in vitro fermentations occurred only for the 40% NFC diets. Mean TNC digestion was 87.6, 61.7, 60.9, and 60.4% whereas NDF digestion was 33.8, 55.9, 57.6, and 58.4% for the in vitro-40% NFC, in vitro-25% NFC, in vivo-25% NFC, and in vivo-40% NFC treatments, respectively. These data suggest that the continuous culture system had difficulty simulating digestion of high TNC diets. Other differences between fermentation in continuous culture and in vivo can be attributed mainly to lack of absorption from the fermenters, defaunation of the in vitro system, and endogenous protein contamination in vivo.