Dynamics of ammonia volatilization and nitrous oxide production from urine patches in grazed pastures

Abstract

A continuously aspirated enclosure method was used to measure ammonia (NH₃) volatilization from simulated sheep urine patches in a perennial ryegrass (Lolium perenne) / white clover (Trifolium repens) pasture in the field during summer, autumn and winter periods. Volatilization was essentially complete after 100-200 hours. Mean volatilization losses from urine treated plots were 22.2% of the applied nitrogen (N) in summer, 24.6% in autumn and 12.2% in winter. Corresponding losses from the urea treated plots were 17.9%, 28.9% and 8.5%. Seasonal differences were significant (P ≤ 0.05) for both N sources, out differences between N sources during any particular season were not significant. Repeated applications of urine or aqueous urea to the same area of pasture were made during summer to simulate the possible effects of high stocking rates and sheep camp areas. Significantly greater (P ≤ 0.05) subsequent volatilization losses were produced, averaging 29.6 and 37.5% from the second and third applications, respectively.

Theoretical considerations were presented for the development of a simplified NH₃(g) volatilization model appropriate to urine patches. Volatilization rate was calculated to be directly proportional to the amount of ammoniacal-N in the topsoil, and inversely proportional to soil moisture content and the extent of exchange reactions with the charged sites on the soil colloids. Temperature and pH also markedly affect the rate of ammonia volatilization but in a non-linear manner. An increase in either of these parameters was calculated to increase the rate of ammonia volatilization. It was shown that the dominant factor determining the rate of NH₃(g) volatilization is the soil surface pH. Input data for calculating NH₃(g) losses are: a knowledge of the disposition of the applied N within the soil profile; the rate of urea hydrolysis in the topsoil; and soil surface pH and temperature measurements throughout the duration of a volatilization event. The model was verified using field experimental data from the present study and also published data from independent sources. It was considered that the model offers the potential for determining NH₃(g) volatilization losses following urine or aqueous urea applications to short pasture in non-leaching, non-nitrifying environments.

Field, growth cabinet and laboratory measurements of nitrous oxide (N₂O) emissions from simulated urine patches were also conducted. A sensitive electron-capture gas chromatographic procedure was combined with a short duration enclosure method to monitor the build-up of N₂O in the enclosed headspace above the pasture surface. In a field experiment, plots received aqueous solutions containing 7.2 g N as either sheep urine, calcium nitrate or ammonium sulphate and after 10 days lost 6.4, 6.8 and 7.7 mg of the applied N respectively. A control plot treated with distilled water released 1.1 mg N₂O-N during the same period. Diurnal fluctuations in N₂O emission rates from both N treated and untreated control plots were significantly correlated (r ≥ 0.980) with soil temperature (10 cm depth) although the magnitude of the temperature fluctuations (± 2°C) were insufficient by themselves to produce the large (e.g. 10 fold) variations in daily N₂O emission rates observed. Fluxes of N₂O from untreated pasture soil ranged from 0 - 2.1 mg N₂O m-2 day-1.

In growth cabinet and laboratory experiments, N₂O emissions were measured from blocks of freshly cut pasture son (165 x 165 x 150 mm) treated with aqueous solutions containing 0.5 g N as either sheep urine, calcium nitrate, ammonium sulphate or urea. Pasture blocks watered to 27.5% average soil moisture content lost significantly more (P ≤ 0.05) N₂O than blocks maintained at 14.0% average soil moisture content but within each moisture regime, differences in total N₂O-N losses between treatments were not significant. Peak emissions occurred on the days following watering with similar patterns of release apparent from each N source.

Emission rates of N₂O immediately following sheep urine applications to blocks of fresh pasture soil were significantly greater (P ≤ 0.05) than initial rates of production from similar applications of aqueous calcium nitrate, ammonium sulphate or urea. The magnitude of the initial pulse of N₂O from the sheep urine was unrelated to soil moisture content and amounted to about 30% of the N₂O loss from each simulated urine patch (i.e. 0.1% of the applied urine-N).

Measured N₂O losses from sheep urine and inorganic N fertilizers (ammonium sulphate, calcium nitrate and urea) were small with maximum losses estimated at < 2% of the applied N after 3 months. It was concluded that direct gaseous N₂O emissions from typical silt-loam pasture soils in Canterbury are of little agronomical importance.

The practical implications of the above results are presented and discussed.