Incubation experiment preparation
Soil samples were taken from a surface soil (0-10 cm) under a productive grassland restored from cropland for about 30 years and a subsoil at a depth of > 2 m in a cropland cultivated by soya/maize in rotation for about 30 years at the Hailun Experimental Station, Chinese Academy of Sciences, Hailun, Heilongjiang Province (47°26′N, 126°38′E). The soil was developped from sedimentary matierals of loess-like materials and classified as a Pachic Haploboroll according to USDA Soil Taxonomy or Phaeozems according to the World Reference Base for Soil Resources. The subsoil consisted of 420 g kg-1 clay (< 2 μm) and 356 g kg-1 silt (2-20 μm), and its clay mineralogy was dominated by vermiulite, illite and a mixed layer mineral of vermiculite and illite43.
The surface soil was used to extract microbial inoculum at a soil: water ratio of 1: 15 (mass: volume)45. The soil suspension was shaken with glass beads for 2 hours to better slake the soil and centrifuged (1000 g ) for 12 minutes to separate soil solution from the soil. The supernatant was then transferred into a new vial and centrifuged at 3470 g for 30 minutes. These procedures were repeated twice, and the mixed supernatants were used as the inoculum. The subsoil was used to extract the silt+clay (< 53 μm) fraction by wet sieving. The silt+clay fraction of the subsoil was heated at 500 oC for 2 hours to remove soil organic carbon and microbes to derive a carbon-free natural soil material that was dominant withvermiulite, illite and a mixed layer mineral of vermiculite and illite. This soil material and other three commercially available pure clay minerals (vermiculite, illite or kaolinite) (Supplementary Table 1) were used to prepare four model soils, which were composed of single soil material or pure clay minerals with silt- (5-6.3 µm) and sand- (200-100 µm) sized quartz at a clay: silt: sand ratio of 5: 4: 1, reproducing the particle size distribution of the original subsoil. We mixed these four model soils with either maize (Zea mays L.) or soya (Glycine max L.) straw, and with one inoculum, then incubated the model soils at constant moisture and temperature for 120 days. Briefly, aliquots of 90 g of each model soil and 3 g of a litter were added into each of 24 Erlenmeyer flasks (240 mL) ensuring a similar volume or soil bulk density by slight packing in each flask, providing three replicates for each gas and soil sampling. Both maize litter and soya litter had been harvested three months after sowing and chopped into < 0.25 mm particles. The C/N ratio was 29 for maize litter and 14.9 for soya litter. Then 20.6 mL distilled water and 5.4 mL inoculum from the surface soil were added and thoroughly mixed by spoon for 5 minutes in each flask. The flasks were sealed with rubber stoppers and incubated in the dark at 25oC for 120 days.