Incorporating rhizosphere processes into field-scale (agro)ecosystem models

Three different strategies for incorporating rhizosphere processes within field-scale models are compared, taking triple-cropped irrigated rice production as a common system and CH4 emission as a common focus of interest. The strategies may be characterised as homogeneous (model I; root C deposition is added to the bulk soil compartment), areal (model II; roots contribute via aerenchymatous exchange to an increased soil-atmosphere interfacial surface area), and volumetric (model III; roots create around themselves a specific rhizosphere compartment). Model I is simpler than model II, which is simpler than model III. With identical parameters all models lead to similar seasonally integrated CH4 emissions, but when the pattern of emission and the simulated CH4 concentration in the soil is brought into the reckoning, the following order of precedence (greater is better) becomes clear: model III greater than or equal to model II > model I. Current field-scale models of soil organic matter (SOM) transformation, especially in rice soils, could be improved by taking explicit account of the rhizosphere and the processes which occur within it.