Losses of nitrogen ( N ) through ammonia (NH,) volatilisation and denitrification were determined following the application of cattle slurry to grassland in autumn or spring. Denitrification was examined on two contrasting soils. A system of small wind tunnels was used to measure N H , loss and an acetylene inhibition technique for denitrification. Between 31 and 84 % of the ammonium N (NH:-N) applied in slurry was lost through N H , volatilisation. Acidifying the slurry to pH c 5.5 prior to application reduced these losses to between 14 and 57%. On a ji-eely drained loam soil, denitrification j?om unacidified slurry applied in the autumn at 80 m3 ha-' was continuous throughout the winter, with the maximum rate of 0-91 kg N ha-' day-' occurring a few weeks after slurry application. The total denitrification losses were equivalent to about 29 % of the NH:-N applied for this treatment and 41 % for the acidified slurry. The nitrification inhibitor dicyandiamide reduced the amount of N lost through denitrification by 70 % when applied with the slurry at 25 kg ha-', by 55% at 20 kg ha-' and by 30 % at 15 kg ha-'. The nitrification inhibitor nitrapyrin did not appreciably reduce denitrification. Denitrification losses were consistently small from slurry applied to the j-eely drained loam soil in spring, or to a poorly drained, silty clay in autumn or spring. Neither nitrification inhibitor was of benefit in these situations.

No abstract

Three experiments were conducted to examine the influence of slurry application rate, wind speed and applying slurry in narrow bands on ammonia (NH3) volatilization from cattle slurry surface-applied to grassland. The experiments were conducted in the field using a system of small wind tunnels to measure NH 3 loss. There was an inverse relationship between slurry application rate and the proportion of NH4+-N volatilized. From slurry applied at 20, 40, 60, 80, 100 and 120 m 3 ha -1, the respective proportions of NH4+-N lost by NH 3 volatization in 6 days were 60, 56, 49, 40, 44 and 44%. The negative relationship was most pronounced in the first 24 hours after application when 57-77% of the total loss for 6 days occurred. Wind speed had a positive effect on NH 3 volatilization, although the effect was small in relation to the total loss; increasing the wind speed from 0.5 to 3.0 m s -1 increased the total 5 day loss by a factor of 0.29. The effect of wind speed was also most pronounced in the first 24 hours when much of the NH 3 loss took place. The effect of reducing the surface area of the applied slurry was examined by comparing NH 3 volatilization from slurry broadcast across plots with that applied in narrow bands. Although the rate of NH 3 volatilization was considerably smaller from the banded application immediately after the slurry was applied, the difference between the treatments progressively narrowed until 2 days after application, after which a higher rate was maintained from the banded slurry. After 5 days the total loss from the banded application was 83% of that from broadcast slurry.

Cattle slurry was applied to grassland on two contrasting soils in autumn and spring between 1987 and 1990. Slurry was applied with or without the addition of acid, to lower the pH to 5-5, and, in autumn only, with or without a nitrification inhibitor. Ammonia volatilization, denitrification and apparent recovery of N by the cut herbage accounted for 61-86% of the ammonium nitrogen (NH4*-N) applied in slurry. Estimates from lysimeter experiments indicated that nitrate leaching from autumn application may have accounted for an additional 1-2% only. Acidifying slurry reduced volatilization losses to 1-12% of the NH4*-N applied, and the nitrification inhibitor halved denitrification losses from autumn applications. Reductions in nitrogen losses were reflected in significant increases in first-cut herbage yields which, for slurry applied in the autumn with acid and the nitrification inhibitor, were generally greater than those from 120 kg ha"'N as ammonium nitrate applied in the spring. There were no significant differences between treatments at subsequent cuts in each season but, owing to the large increases at first cut, total yields were significantly higher for autumn-applied slurry with acid and nitrification inhibitor.

Experiments were conducted on a grassland site at Wrest Park, Silsoe, Bedfordshire between 1987 and 1989, to compare herbage yields from slurry applied by deep and shallow injection, low trajectory and conventional vacuum tanker methods. Slurry application rates for all spreaders were calibrated at 86 ± 5 t ha~^, an equivalent of c, 200 kg NH4+-N ha-' applied in autumn or in spring.As expected, herbage yields following spring applications were higher than from autumn applications, with average mineral fertilizer equivalents of 122 and 89 kg N ha-i respectively. Yields from the conventional and low trajectory spreaders showed no consistent differences. However, in both years, first cut yields from plots were significantly lower (P < 0-05) where slurry had been injected than where surface applications had been used by an average of 0-7 t DM ha-i. Subsequent cuts in 1988 demonstrated higher residual effects from injection so that annual total yields were similar from all slurry applications irrespective of spreader type.Analysis of N content revealed high N levels in herbage from deep injection plots. Mean concentrations of N in the herbage, expressed as a percentage of the dry matter, were 1-43 for surface treatments and 1-79 for deep injection in I Present address and correspondence: Water Research Centre,