Nitrification inhibitors and fertiliser nitrogen application timing strategies to reduce N2O. Urine site in Devon, 2012

t a grassland site near Rothamsted Research, North Wyke in south-west England (sandy clay loam topsoil texture) small field plots (10 x 3 m) were arranged in a randomised block design with three replicates per treatment. Cattle urine (at 5 L m-2) was applied to grassland in mid-March and in early September 2012. A control treatment was included where no urine was applied. In separate treatments, two commercially available nitrification inhibitors were tested; dicyandiamide (DCD) and an additive containing two pyrazole derivatives (1H-1,2,4-triazole and 3-methylpyrazole) were pre-mixed with the urine prior to application to give application rates of 15 kg/ha and 5 L/ha for the DCD and the pyrazole derivatives, respectively. The urine was collected from lactating dairy cows at Reading University, kept refrigerated at <4°C and applied within two days of collection. Following urine application to five 0.36 m2 areas of the plot, measurements of direct nitrous oxide (N2O) emissions were made over c.12 months, using 5 static chambers each 0.4 x 0.4 m (1 per 0.36 m2 area, giving a total measured surface area of 0.8 m2) and analysed by gas chromatography. In a separate area of the plot, ammonia (NH3) emissions were measured for 7 days after each urine application, using a wind tunnel technique (one per plot). Nitrate (NO3) leaching losses were measured following the September urine application using porous ceramic cups (6 per plot) installed to a depth of 90 cm during the period of over-winter drainage (Webster et al., 1993) with samples collected every 50 mm of drainage or every 2 weeks whichever occurred sooner. Drainage volumes were estimated using IRRIGUIDE (Bailey and Spackman, 1986) and were combined with NO3 concentrations to quantify the amounts of NO3-N leached. Indirect N2O-N emissions were estimated from the measured NO3-N and NH3-N losses and using the Intergovernmental Panel on Climate Change default emission factors. Grass yields and N offtakes were also measured following grass cuts in May and July 2012 for the March urine application, and in May 2013 for the September urine application. The Devon 2012 urine experiment contains data sets of: annual nitrous oxide emission, annual nitrous oxide emission factor, total ammonia loss, overwinter nitrate leaching loss, soil moisture, top soil mineral nitrogen, temperature, rainfall and associated crop (grass yield and nitrogen offtakes) and soil measurements.

Direct N2O emissions were measured with five static flux chambers (40 cm wide x 40 cm long x 25 cm high) per plot, covering a total surface area of 0.8 m2. The chambers were of white (i.e. reflective) PVC and un-vented with a weighted lid and neoprene seal allowing an air-tight seal to form following chamber enclosure. Chambers were pushed into the soil up to a depth of 5 cm and remained in place throughout the experiment, except during urine/dung application and grass cutting when chambers were removed, locations were marked, and chambers were re-instated to the same position as prior to removal. Chambers remained open except for a short time on each sampling day. On that day, ten samples of ambient air were taken to represent time zero (T0) N2O samples. From each chamber, after a 40-minute enclosure period (T40) a headspace sample was taken using a 50-ml syringe. Using a double needle system the sample was flushed though a pre-evacuated 20-22 ml glass vial fitted with a chloro-butyl rubber septum and held at atmospheric pressure. The N2O flux was calculated using an assumed linear increase in N2O concentration from the ambient N2O concentration (T0) to the N2O concentration inside the chamber after 40-minutes enclosure (T40) (Chadwick et al., 2014). Throughout each experiment, the linearity of emissions through time was checked routinely from three chambers located on the urine only treatment. A minimum of four samples were taken from each chamber at 20 min intervals commencing at closure i.e. T0 and spanning the T40 sampling time. In order to minimise the effect of diurnal variation, gas sampling was carried out between 10:00 am and 14:00 pm and where possible between 10:00 am and 12:00 pm as suggested by IAEA (1992) and referred to in the IPCC good practice guidance (IPCC, 2000). Gas samples were analysed as soon as possible after collection (to minimise potential leakage) using gas chromatographs fitted with an electron-capture detector and an automated sample injection system. Following receipt in the laboratory, three replicates of one standard N2O gas were kept with the samples and were used to verify sample integrity during storage. The gas chromatographs were calibrated on a daily basis using certified N2O standard gas mixtures. Following urine/dung application, N2O flux measurements were carried out for 5 days immediately following urine/dung application, daily for a further 5 days during the next week, twice weekly for the next two weeks, every other week over the next c. four months, decreasing in frequency to monthly until the end of the 12 month sampling period. Prior to the urine/dung application N2O measurements were taken to provide baseline information. This sampling schedule resulted in an annual total of c.35 sampling days starting from the day of each of the urine/dung application. Measurements were taken over 12 months to follow IPCC good practice guidance and so that the results were directly comparable to the IPCC 2006 methodology default emission factor.

Nitrous oxide fluxes from the five replicate chambers per plot were averaged. Cumulative fluxes were calculated using the trapezoidal rule to interpolate fluxes between sampling points.