Three ruminally cannulated mature steers were fed, in a crossover design, diets with concentrate to roughage (C: R) ratios of 80: 20, 55: 45 and 30: 70, twice a day (09: 00 and 15: 30) in a predetermined feed intake sequence of ad libitum, 80% ad libitum and 90% ad libitum. Following the faeces collection period of 12 days, ruminal contents were removed at 2 h post feeding and were weighed, mixed, sampled and returned. DM intake decreased and then increased (non-linear; P ≤ 0.01) as C: R ratio increased. DM digestibi1itiies of the three intake levels were respectively 72.9, 73.8 and 76.1% for the 80: 20 diet, 64.9, 65.0 and 65.1% for the 55: 45 diet and 55.3, 56.3 and 57.1% for the 30: 70 diet. DM digestibility was not influenced (P = 0.28) by feeding level, but was linearly reduced (P  ≤ 0.05) as C: R ratio decreased, which indicated that no significant associative effect had occurred. Both a decrease in feeding level and in C: R ratio increased rumen OM retention linearly (P ≤0.01). Volatile fatty acid concentration (VFA) 2h post-feeding was not influenced (P  ≤ 0.05) by feeding level, but was reduced linearly (P  ≤ 0.01) by a decrease in C: R ratio. A decrease in C: R ratio increased the molar percentage acetic acid and decreased propionic acid linearly, while feeding level had no significant influence. The lowest rumen pH (6.1) 2h post feeding occurred at 80% of ad libitum on the 80 : 20 diet, which indicates that the ruminal environment was relatively stable on all treatments.

Keywords: Dietary energy concentration, digestibility, feeding level, steers.