1. An experiment was carried out over 2 years to study the effect of stocking rate on the efficiency of pasture utilization by dairy cows on irrigated perennial pasture.
2. Two matched groups of milking cows grazed throughout the experiment on two unequal areas of pasture at stocking rates of 1·0 and 1·7 cows/acre during the first season and 1·2 and 2·0 cows/acre during the second season (L and H groups, respectively). The same rotational grazing technique was used on both areas.
3. In each season the experimental period lasted 249 days. During the first 200 days and 215 days of this period in the first and second seasons, respectively, pasture, grazed or conserved (from the same proportion of both areas), comprised the entire ration.
4. The H group produced 57 and 59% more milk per acre than the L group during the first and second seasons, respectively. Milk production per cow in the H group was not significantly depressed. The solids-not-fat content of the milk in both seasons, and the butterfat content in the first season, were significantly depressed at the higher stocking rate.
5. The body-weight gain of the cowa in the H group, relative to those in the L group, was depressed by 68% (from 170 to 55 lb./cow) and 88% (from 172 to 20 lb./cow) in the first and second seasons, respectively.
6. The yield per acre of utilized pasture, based on the nutrient requirements of the cows, was 39 and 38% higher at the higher stocking rate in the first and second seasons, respectively. The estimated individual intake of these cows in the two seasons was depressed by 14 and 20%, respectively. The estimated yield per acre of utilized pasture ranged from 4092 lb. s.e. at 1·0 cows/acre to 6519 lb. s.e. at 2·0 cows/acre.
7. The mean daily herbage intake during the second season, estimated by a herbage clipping technique, was 29·5 and 22·9 lb. dry matter per cow for the L and H treatments, respectively. The cows in the H group ate, at each grazing, all the herbage that had grown since the previous grazing, whereas the estimated intake of the L group was equivalent to only 80% of this regrowth. However, the estimated rate of regrowth of the swards on the two treatments was the same.
8. Herbage intake estimates derived from a faecal-indicator technique, and observations of the grazing behaviour of the cows provided more detailed information on the pattern of herbage availability on successive days during the grazing of paddocks at each stocking rate.
9. The results of this experiment are discussed in relation to the design of grazing experiments and to farming practice.