X-ray CT-derived pore structure determines soil gas transport under high soil water content in clayey but not sandy soils

• X-ray CT-derived pore metrics predict soil gas transport under wet conditions. • In clay soil, connected macropores (>100 µm) govern D s / D 0 and K a across potentials. • In sandy soil, transport tracks air-filled porosity more than CT topology. • Gas transport is structure-limited in clay and water-limited in sand. • RTS is recommended for clay soils, whereas NTS is better suited for sandy soils. Soil gas transport is fundamentally governed by pore structure characteristics and water content, yet the predictive value of X-ray computed tomography (CT)–derived pore metrics across soil textures under high water content remains unclear. We conducted two 8-year field experiments on contrasting soils: clay soil (Shajiang black soil) and sandy soil (Fluvo-aquic soil), each comprising tillage systems (no-tillage, NT; rotary tillage, RT) and straw management (straw return, S; straw removal, RS). We measured relative gas diffusivity ( D s /D 0 ), air permeability ( K a ) and computed effective oxygen diffusion coefficient ( D ) across different matric potential (-1∼-330 hPa) for soil depths of 0-10 cm and 10-20 cm. CT was used to quantify pore structure characteristics, including image-based porosity, pore size distributions, and pore skeletonization parameters (branch, junction, tortuosity). In clay soil, RTS significantly enhanced D s /D 0 and K a by 32.3% and 34.3% ( P <0.05), respectively, at -330 hPa in 10-20 cm depth versus NTS. No significant differences were found among D s /D 0 between two tillage practices or straw management in sandy soil ( P >0.05). RTS also increased 100-300 μm porosity, total porosity, branch number, and junction number in 10-20 cm depth of both soil types ( P <0.05). Strong positive correlations were observed in clay soil between D s /D 0 , K a and >100 μm porosity, branch number, junction density, branch length density, and total porosity across all tested matric potentials ( P <0.05). However, such relationships were largely absent in sandy soil. The computed D tracked air-filled porosity but did not capture pathway connectivity. In conclusion, under high water content, gas transport is structure-limited in clay and water-limited in sand. Therefore, the optimal strategies derived from our study are the adoption of RT with straw incorporation to enhance aeration in clay soil, and the use of NT with straw mulching to conserve water while preserving soil aeration in sandy soils.