Yield-scaled greenhouse gas emissions associated with the use of stabilized nitrogen fertilizers in subtropical paddy rice

In this work, the hypothesis that stabilized nitrogen fertilizers (SNF) might be useful to boost paddy rice production with a reduced impact on nitrous oxide (N2O) and methane (CH4) emissions from soil was tested by conducting a field experiment on a subtropical typic Albaqualf in Southern Brazil over two rice growing seasons (GS). Treatments comprised the use of no nitrogen fertilizer (control) and application of common urea, urea + NBPT [urea plus the urease inhibitor N-(n-butyl) thiophosphoric triamide], urea + NBPT + DCD (urea plus NBPT and the nitrification inhibitor dicyandiamide), urea + Cu B (urea plus copper and boron), urea + S (urea plus sulfur) and urea + zeolite (urea plus zeolite). Field plots were arranged according to completely randomized block design. Methane and N2O fluxes were measured by using the static chamber technique and gases analyzed by gas chromatography. Seasonal partial global warming potential (pGWP) values were calculated as the combination of the seasonal levels for the individual gases multiplied by the respective radiative forcing potentials (viz., 28 for CH4 and 265 for N2O), and yield-scaled greenhouse gas (GHG) emissions were taken to be the ratios of pGWP to rice grain yields. N2O emissions were 0.9 kg ha-1 on average in GS1; also, they were similar irrespective of treatment and whether or not N fertilization was applied. In GS2, N2O emissions were lowest with the control treatment and all SNF led to emission levels comparable to those of common urea (average 4.5 kg ha-1). CH4 emissions in GS1 peaked at 474 kg ha-1 and exhibited no significant differences among fertilizers. On the other hand, urea + NBPT reduced CH4 emissions relative to urea + S and urea + zeolite in GS2. pGWP averaged at 11 216 kg eq.CO2 ha-1 across the two GS without N fertilization (control treatment) and at 11 803 kg eq.CO2 ha-1 with fertilization. The increase in grain yield resulting from nitrogen fertilization was similar irrespective of nitrogen source, and averaged at 9300 kg ha-1 in GS1 and 8200 kg ha-1 in GS2. Yield-scaled pGWP was only influenced by fertilization in GS1, where urea + NBPT and urea + zeolite reduced it relative to the control treatment. The starting hypothesis was thus rejected since SNF affected rice grain yield and soil GHG emissions similarly to common urea over two growing seasons in subtropical paddy rice ecosystems of southern Brazil.