Seventy-eight British Friesian type cows, mean calving date 13 January, were used in a 3 × 2 factorial design experiment to examine the direct and residual responses to replacing 0·8 kg/day of a concentrate containing 184 g/kg crude protein with an equal quantity of fish meal when using three levels of total supplement feeding (0·8, 4·0 and 7·2 kg/day). In addition, all animals had access ad libitum to a high-quality grass silage (in vivo digestible organic matter 750 g/kg dry matter) during the treatment period. Treatments were applied from day 8 post partum until 22 April, when all animals went to pasture, giving a mean treatment period of 91 days. At pasture the animals were rotationally grazed as three groups, based on the three levels of total supplement offered during the treatment period, at the same stocking rate. The effects of treatments in terms of direct effects during the treatment periods, residual effects at pasture and also total lactation were assessed. Also during the treatment period the effects on rumen volatile fatty acid contents and blood composition were monitored. In addition, total diet digestibility and food utilization studies were carried out on six animals per treatment.
Level of supplementation significantly influenced milk output during both the treatment and full lactation periods with the total lactation responses being 2·0 and 1·0 kg milk per kg additional supplement between the food levels of 0·8 to 4·0 and 4·0 to 7·2 kg/day respectively. Level of supplementation also significantly influenced milk fat concentration during the treatment, residual and full lactation periods and milk protein concentration during the treatment period only.
The replacement of 0·8 kg conventional concentrate by 0·8 kg fish meal significantly increased milk yield during the final 21 days on treatment (mean yield per day 20·6 and 21·9 (s.e. 0·44) kg for without and with fish meal treatments respectively) but there were no significant residual or total lactation effects. From the data it was calculated that at low levels of supplementation 0·8 kg fish meal could be used to replace 1·9 kg conventional concentrate but at more moderate levels of nutrition any substitution would be much lower and uneconomic.