The model simulates the cycling of N in grassland systems grazed by beef cattle and predicts the annual amount of N in liveweight gain, and the amounts lost through ammonia volatilization, denitrification and leaching, on the basis of fertilizer application and soil and site characteristics. It aims to provide a better understanding of the way in which these various factors interact in their influence on N transformations. The model has been programmed to run on IBM-compatible personal computers and responds rapidly to changes in input parameters.The model has been constructed from the average annual amounts of N passing through various components of the N cycle in ten field systems grazed by beef cattle. The amounts were either measured directly or were calculated from empirical sub-models, assuming a balance between inputs to, and outputs from the soil inorganic N pool. The model is given wide applicability through the inclusion of a mineralization sub-model which is sensitive to soil texture, sward age, previous cropping history, and climatic zone. Another important sub-model determines the partitioning of soil inorganic N to either plant uptake or the processes of loss: the proportion partitioned to plant uptake decreases as the total amount of soil inorganic N increases.Outputs from the model indicate that fertilizer N has a strong influence on ammonia volatilization, denitrification and leaching at a given site but that, over a range of sites with a given rate of fertilizer N, total loss and the proportions lost by the three processes are greatly influenced by the amount of N mineralized by the soil. The model indicates how fertilizer N should be matched with mineralization to limit gaseous and leaching losses and to achieve optimum efficiency of N use in grazing systems.

A detailed description is given of the construction and operation of a system of small wind-tunnels used in the measurement of ammonia volatilisation from grass swards. In the design of the system, emphasis was placed on the requirements that (i) natural sward conditions should be influenced as little as possible, and (ii) airspeed through the tunnels should be controllable in the range normally encountered in the field. An analyser is described which incorporates an ammonia-sensing probe; it was used to monitor automatically and continuously the concentration of ammonia in air entering and leaving the tunnels. The tunnels were found to have only small effects on sward conditions. Differences between simultaneous measurements of temperature made inside and outside the tunnels were rarely more than 1°C for air temperature and 2°C for soil temperature at 3cm depth. The potential of the system was examined in a study comparing the loss of ammonia through volatilisation from swards treated with natural urine or an artificial urine. There were marked differences in the rates of volatilisation from the two urines although the total loss of nitrogen was similar from both, amounting to about 3% of the nitrogen applied.

Two field experiments commencing in winter (December) and spring (April) were conducted to determine the fate of nitrogen (N) in cattle slurry following application to grassland. In each experiment three methods of application were used: surface application, and injection f the nitrification inhibitor, nitrapyrin. Slurry was applied at 80 t ha-', (~2 4 8 kg total N ha-' in the winter experiment, and 262kgNha-' in the spring experiment). From slurry applied to the surface, total losses of N through NH, volatilization, measured using a system of wind tunnels, were 77 and 53 kg N ha-' respectively for the winter and spring experiments. Injection reduced the total NH, volatilization loss to -2 kg N ha-'. Following surface application, loss by denitrification, measured using an adaptation of the acetylene-inhibition technique, was 30 and 5 kg N ha-' for the two experiments. Larger denitrification losses were observed for the injected treatments; in the winter experiment the loss from the injected slurry without nitrapyrin was 53 kgN ha-', and with nitrapyrin 23 kg N ha-'. Total denitrification losses for the corresponding injected treatments in the spring experiment were 18 and 14 kg N ha-'. Apparent recoveries of N in grass herbage in both experiments broadly reflected the differences between treatments in total gaseous loss.

The amounts of organic matter in the stubble, litter, root and soil macro-organic matter fractions of two swards of perennial ryegrass that had received normal applications of either fertilizer or cattle urine were, on average for the four fractions, about 3000, 500, 11,500 and 8,800 kg ha -1. The swards had been established 8 or 15 years previously and each was sampled at intervals over a period of about one year. The amounts of N contained in the four fractions were, on average, 68, 12, 249 and 240 kg ha -1, a total of 569 kg N ha-~.With other swards, increasing rates of application of fertilizer N were found to have little effect on the amounts of organic matter in stubble and roots. Concentrations of N in the organic matter of the stubble and roots, however, increased significantly with increasing rate of fertilizer application, though, with stubble, moderate rates of application had little effect.Assessments based on these data, together with other published information, indicate that the amount of N mineralized from the combined stubble, litter, root and macro-organic matter fractions during the first year after ploughing may range from about 40 kg to at least 360 kg N ha -~ depending on the age of the sward and its recent management. The amount mineralized is likely to increase with age of sward, with increasing rate of fertilizer N and with utilisation by grazing rather than cutting.

S U M M A R YComponents of the N cycle were studied at Hurley, UK, in 1985-87. In grass-clover {Lolium perenne and Trifolium repens) swards, grazed at three intensities, low total inputs of N were associated with low outputs and losses of N. Nevertheless, the flows (intake and excretion) of N through animals were substantial and gave rise, at the higher intensities of grazing, to an acceptably high agricultural output per hectare. This was considered evidence of a fast and efficient recycling of N between plants, animals and soil. The release of N to the environment (as nitrogenous gases and nitrate) was substantially less from the grass-clover swards than from a grass sward fertilized with 420 kg N/ha, and this was at the expense of only 20% loss in production. The mechanisms which might account for the high efficiency of utilization and recycling of N in grass-clover swards are discussed in the context of the balance of the supply of C and N to plant and soil biomasses under grazing. The results confirm that optimizing agricultural output in grass-clover swards has little adverse effect on the environment, but the importance to this end of sustaining a large proportion of N-deficient grass in grass-clover swards is emphasized.