Initial emission of nitrous oxide from sheep urine applied to pasture soil

Abstract

Concern has been expressed (Rasmussen er nl., 1976; Crutzen and Ehhalt, 1977; Bremner and Blackmer, 1978; Cicerone et al., 1978) about possible detrimental effects to the earth’s ozonosphere due to a suggested increase in tropospheric N,O. This increase could result from the expanding use of artificially-manufactured nitrogenous fertilizers. In pastoral agriculture in a number of countries (e.g. New Zealand and Australia) high concentrations of N are constantly being voided as urine on pastures by grazing animals. This may present a situation for NzO production and release to the atmosphere. Whereas artificial fertilizers are normally applied by broadcasting or banding, sheep and cattle urine is voided in discrete patches in which the concentration of applied N frequently exceeds the equivalent of 1000 kg haa’ (Doak, 1952; Barlow, 1974). Under these conditions soil pH increases and high concentrations of NOT-N can accumulate. Such effects may give rise to chemodenitrification routes for subsequent NzO release (Nelson and Bremner, 1970). Nitrous oxide fluxes from urine patches in grazed pastures have not been studied extensively, mainly because of the difficulty in measuring the very small amounts of N20 released in air samples. Using the highly sensitive electroncapture gas chromatograph method (Rasmussen et al., 1976) we report long-term release rates of N,O from nitrogenous fertilizers and sheep urine applied to grazed pasture blocks and short-term release rates from laboratory incubation of pasture soils. Our results demonstrate a positive stimulation of initial N,O release when sheep urine was applied to the soil compared with similar applications of ammonium sulphate and urea. An established ryegrass (Lolium perenne L.) white clover (Trifolium repens L.) grazed pasture in the Lincoln College Sheep Stud Farm was used. The soil is a Templeton silt loam (Cox et ul., 1971) a Dystric Ustochrept. It has the following properties: pH (1: 2.5 water) = 6.2, total N =0.31x, NOT-N = 26 peg gg’, NH:-N = 45 fig g’, organic C = 4.2%, field capacity = 34%. Eight blocks (165 x I65 x 150 mm) of soil complete with undisturbed pasture herbage were cut from the site and fitted into 5 1 polypropylene containers. These were transferred to a laboratory window-box and maintained there under ambient lighting and temperature conditions (20°C) for 45 days. During this period watering (to 32% average moisture capacity-see Fig. 1) was carried out. All pasture blocks except the controls received on each of 3 occasions 0.5 g N as either ammonium sulphate, urea or sheep urine in lOOmI of water: equivalent approximately to 200kgNhaa’, giving a total application of about 600 kgN ha-‘. Control pasture blocks received 100 ml of distilled water. Duplicates of each treatment were used. The sheep urine was collected from cannulated Romney ewes, analysed for pH (7.1) total N (1.37 g lOOmI-‘) and urea-N (1.26 g 100 ml-‘) and stored frozen till required. Subsequent analysis upon thawing showed no chemical changes associated with these N fractions. Nitrous oxide fluxes were measured using an intermittent enclosure technique in which air-tight lids were fitted to the polypropylene containers and the released N,O allowed to accumulate in the enclosed headspace (approx. 1 1). The lid was closed for about 1 h, during which time at least 2 air samples (I ml) were removed using a gas-sampling valve (Carle 8 port) and analysed for N,O. The analytical system consisted of a Varian 2800 gas chromatograph fitted with a 3 m x 3 mm O.D. stainless steel column of Porapak Q maintained at 20°C. This was connected to a PyeUnicam linearized pulsed electron-capture detector maintained at 35OC. The carrier gas was O,-free Nz at a Bow rate of 50 ml min’. Adequate separation of NzO from the potentially interfering gas CO, was achieved (Cicerone et al., 1978): their retention times being 195 s and 242 s respectively. The primary standard calibration gas was a 104 ~1 NzO 1-l in N2 gas mix (Matheson, U.S.A.). Secondary NzO-air standards were prepared by diluting known volumes of the primary standard gas with known volumes of compressed air. N,O peak areas were calculated automatically on a Varian Aerograph Model 408 integrator. To study initial N,O release rates under non-saturated soil aerobic conditions, small scale incubations were performed. Samples of air-dry soil (10 g, <2 mm) were placed in 155 ml serum bottles fitted with gas tight neoprene septa. The headspace gas was flushed with compressed air for 5 min and the bottles incubated at 20°C. A solution containing 25 mg N in 2.5 ml of water was added by syringe to 3 of the bottles and the initial NzO release rates were monitored using the analytical system described earlier. This experiment was repeated 2 more times and the accumulated results shown in Fig. 2. For the long-term pasture block experiment the average N,O release during the period of lid closure was calculated using the equation: