Spatio-temporal variation in soil derived nitrous oxide emissions under sugarcane

Nitrous oxide (N2O) is a significant greenhouse gas with a global warming potential that is 300 times than that of carbon dioxide. Soil derived N2O emissions usually display a high degree of spatial and temporal variability because of their dependence on soil chemical and physical properties, and climate dependent environmental factors. However, there is little research that incorporates spatial dependence in the estimation of N2O emissions allowing for environmental factors in the same model. This study aims to examine the impact of two environmental factors (soil temperature and soil moisture) on N2O emissions and explore the spatial structure of N2O in the sub-tropical South East Queensland region of Australia. The replicated data on N2O emissions and soil properties were collected at a typical sugarcane land site covering 25 uniform grid points across 3600 m2 between October 2007 and September 2008. A Bayesian conditional autoregressive (CAR) model was used to model spatial dependence. Results showed that soil moisture and soil temperature appeared to have substantially different effects on N2O emissions after taking spatial dependence into account in the four seasons. There was a substantial variation in the spatial distribution of N2O emission in the different seasons. The high N2O emission regions were accompanied by high uncertainty and changed in varying seasons in this study site. Spatial CAR models might be more plausible to elucidate and account for the uncertainty arising from unclear variables and spatial variability in the assessment of N2O emissions in soils, and more accurately identify relationships with key environmental factors and help to reduce the uncertainty of the soil parameters.

► Soil moisture and soil temperature have substantially different effects on N2O emissions across four seasons of emissions.
► High N2O emissions were associated with high uncertainty and varied between seasons.
► Inadequate account of spatial correlation structure might lead to misinterpretation of N2O emission data.
► Spatial CAR model may reduce the uncertainty associated with the soil parameters influencing N2O emissions.