Xiaoxian Zhang

Sustainable soils and crops, Rothamsted Research, Harpenden, AL5 2JQ United Kingdom

One of the biggest uncertainties in predicting the response of atmospheric CO₂ to global warming is heterotrophic soil respiration. While it has been acknowledged that the emission of CO₂ from soils is modulated by a myriad of abiotic and biotic processes, most studies have largely, in some cases exclusively, focused on the role of microorganisms. They overlook the significance of physical and hydrological processes, thereby interpreting changes in moisture and temperature sensitivity of soil respiration as microbial mechanisms. As a result, our current understandings cannot reconcile contrasting phenomena such as why, with the increase in temperature, soil microorganisms acclimate to reduce their respiration in some biomes but not in others. We show that these inconsistencies are induced by pore-scale physical processes that mediate bioavailability of oxygen (O₂) to microorganisms residing in the hydrated pore space. We first show that, due to the hierarchical structure of soil, water in soil does not exists in “thin films” as the literature often call it; instead, it agglomerates spatially and atmospheric O₂ thus needs to overcome significant resistances in its diffusion and dissolution in water before reaching the microprograms. As temperature rise reduces O₂ dissolution while microbial respiration at microsites alters the local concentration of O₂ and its diffusion, all varying with soil water content, the impact of these biotic and abiotic factors on soil respiration is far more complicated than existing soil carbon models could describe. We develop a multiscale model to simulate emission of CO₂ from whole soil profile. Vertically, it includes the macroscopic transport of gaseous O₂ from the atmosphere to the soil, and horizontally, it includes processes involved in gaseous O₂ diffusion at the water-air interface and its subsequent diffusion and reduction by microorganisms at reactive sites in hydrated pore space. The macroscopic process is described by the diffusion coefficient of gaseous O₂, and the microscopic processes and soil structure are characterized by a memory function, a pore-scale substrate accessibility function, and a microbial O₂ reduction rate function at microsites which depends on substrates, microbial biomass, and temperature. We prove upscaling soil structure and these processes occurring in the pore space over each horizontal soil layer and then to soil profile makes soil depth, soil temperature and water content, microbial biomass and substrate nonlinearly integrated in their influence on CO₂ emission in the soil profile. We demonstrate that the model can explain the paradox in some benchmark field warming experiments and show that moisture and temperature sensitivity data measured from incubation experiment differ significantly from emissions of CO₂ from soil profile. While there is no dispute over the importance of soil microorganisms, we show that these physical processes can produce the temperature sensitivity of soil respiration measured from all field and incubation experimentally. Inadvertently interpreting them as microbial processes would give rise to substantial errors in predicting interaction between soil carbon and global warming and the future development of soil carbon model should consider physical processes.