Two experiments were carried out with Holstein Friesian early lactation cows of high genetic merit to evaluate the effects on intake and milk production when the animals were offered a concentrate supplement either within a complete diet (CD) or separately from grass silage (separate feeding, SF) through a computerized out-of-parlour feeder system. This system offered the daily allowance of concentrates in four × 6h time windows and the cows could choose to have several accesses to concentrate within each time window. In experiment 1 22 first lactation cows were used in a two-treatment change-over design with 10-week periods, while in experiment 2 48 animals (28 in lactation 1 and 20 in lactation 2 or over) were used in a continuous design with an average of 80 days on experiment. In both experiments the diets on the CD treatment were offered ad libitum with a concentrate proportion of 0·61 in total dry-matter (DM) intake while on the SF treatment grass silage was offered ad libitum and the allowance of concentrate was made equal to the mean concentrate intake of the CD treatment during the previous week. All the animals in experiment 2 were turned out to pasture on 5 April for 120 days to examine the residual treatment effects. Twelve mid to late lactation cows and eight rumen-fistulated beef steers were also used to evaluate the effects of the two feeding methods on nutrient utilization and rumen fermentation, respectively. In the latter two trials, the concentrate on the SF treatment was also offered four times daily but in a conventional trough-feeding arrangement.

In experiment 1 both silage and total DM intakes were similar between the two treatments. However, in experiment 2 silage DM intake was significantly increased fP < 0·002) and total DM intake tended to be higher (P > 0·05) with the SF treatment. Feeding the complete diet tended to increase milk yield in both experiments (P = 0·057 in experiment 1), but significantly reduced both the concentration and yield of fat in experiment 2 (P < 0·05). There were no significant treatment effects on the concentration of protein in either experiment, nevertheless in experiment 1 the yield of protein was significantly increased (P < 0·05) with the CD treatment. When the animals in experiment 2 were turned out to pasture, there were no significant residual effects of the indoor feeding treatment on milk production, or yield of fat or protein. However, the concentrations of fat and protein in milk, obtained during the first 60 days on pasture, were significantly higher for the cows offered the concentrate separate from silage during the indoor feeding treatment. The nutrient utilization and rumen fermentation trials indicated no significant difference between the CD and SF treatments in either whole tract apparent digestibility of nutrients (DM, organic matter, energy or nitrogen (N)) or rumen fermentation characteristics (pH, ammonia-N concentration, total volatile fatty acids (VFAs), or molar proportion of each individual VFA in the rumen liquor). It is concluded that feeding the concentrate within a complete diet rather than separately from the silage had no significant effects on total DM intake, milk production, nutrient digestion or rumen fermentation.