Nitrous oxide emissions from fertilised UK arable soils: Fluxes, emission factors and mitigation

Cultivated agricultural soils are the largest anthropogenic source of nitrous oxide (N2O), a greenhouse gas approx. 298 times stronger than carbon dioxide. As agricultural land covers 40-50% of the earth's surface agricultural N2O emissions could significantly influence future climate. The timing, amount and form of manufactured nitrogen (N) fertiliser applied to soils are major controls on N2O emission magnitude, and various methods are being investigated to quantify and reduce these emissions. A lack of measured N2O emission factors (EFs) means that most countries report N2O emissions using the IPCC's Tier 1 methodology, where an EF of 1% is applied to mineral soils, regardless of soil type, climate, or location. The aim of this research was to generate evidence from experiments to contribute to improving the UK's N2O agricultural inventory, by determining whether N2O EFs should vary across soil types and agroclimatic zones. Mitigation methods were also investigated, including assessing the impact of the nitrification inhibitor (NI) dicyandiamide (DCD), the application of more frequent smaller doses of fertiliser, and the impact of different rates and forms of manufactured N fertiliser. Nitrous oxide emissions were measured at one cropland site in Scotland and two in England for 12 months in 2011/2012, along with soil and environmental variables. Crop yield was also measured, and emission intensities were calculated for the contrasting fertiliser treatments. The greatest mean annual cumulative emissions from a range of ammonium nitrate (AN) fertiliser rates were measured at the Scottish site (2301 g N2O-N ha−1), which experienced 822 mm rainfall compared to 418 mm and 472 mm at the English sites, where cumulative annual emissions were lower (929 and 1152 g N2O-N ha−1, respectively). Climate and soil mineral N influenced N2O emissions, with a combination of factors required to occur simultaneously to generate the greatest fluxes. Emissions were related to fertiliser N rate; however the trend was not linear. EFs for AN treatments varied between sites, but at both English sites were much lower than the 1% value used by the IPCC, and as low as 0.20%. DCD reduced AN- and urea-generated N2O emissions and yield-scaled emissions at all sites. AN application in more frequent smaller doses reduced emissions at all sites, however, the type of fertiliser (AN or urea) had no impact. A significant difference in mean annual cumulative emissions between sites reflected differences in rainfall, and suggests that location specific or rainfall driven emission estimates could be considered.