Statistical analysis of nitrous oxide emission factors from pastoral agriculture field trials conducted in New Zealand.

Between 11 May 2000 and 31 January 2013, 185 field trials were conducted across New Zealand to measure the direct nitrous oxide (N2O) emission factors (EF) from nitrogen (N) sources applied to pastoral soils. The log(EF) data were analysed statistically using a restricted maximum likelihood (REML) method. To estimate mean EF values for each N source, best linear unbiased predictors (BLUPs) were calculated. For lowland soils, mean EFs for dairy cattle urine and dung, sheep urine and dung and urea fertiliser were 1.16 ± 0.19% and 0.23 ± 0.05%, 0.55 ± 0.19% and 0.08 ± 0.02% and 0.48 ± 0.13%, respectively, each significantly different from one another (p < 0.05), except for sheep urine and urea fertiliser. For soils in terrain with slopes >12°, mean EFs were significantly lower. Thus, urine and dung EFs should be disaggregated for sheep and cattle as well as accounting for terrain.