Precise assessment of soil organic carbon (SOC) storage requires understanding how vegetation and soil physicochemical properties differ in SOC fractions. Therefore, we aimed to analyze the dynamics of aggregate-associated, liable organic carbon (LOC) fractions corresponding to depth to clarify the effect of vegetation and soil properties on water stable aggregate (WSA) mineral adsorption in subtropical, red soil with five vegetation restoration regimes. The results showed that the large macro-aggregate fraction dominated the degraded red soil, which had the highest content of dissolved organic carbon (DOC). WSA-associated, easily oxidized organic carbon (EOC) varied from 6.26 to 20.02 g/kg and was not affected by vegetation types. Schima superba pure forest (SP) significantly increased DOC (0.38 g/kg on average) and particulate organic carbon (POC, 7.92g/kg on average), which had the highest biomass. Along with soil depth, WSA-associated POC declined, while exhibiting a growth trend with decreasing particle size, e.g., the highest POC was found in silt + clay fraction. The RDA ordination indicated that soil porosity and TN were the main soil parameters that explained the most variance. Meanwhile, the vegetation biomass, except for litter, were all significantly positively correlated with silt + clay fractions. Leaf biomass played the most important role on DOC in macro-aggregate with a 53.42% contribution. For aggregate-related POC, the largest contribution was from the interactions between branch biomass and pH (47.78%) followed by TN (35.1%) of micro-aggregate-related POC. Leaf biomass, silt + clay fractions, and TN can be used as indicators to evaluate the impact of vegetation restoration on WSA-associated SOC fractions. Broad leaved forest or combined with indigenous coniferous species was a better choice for SOC sequestration improvement in the study area.

Understanding sedimentation and runoff variations caused by land use change have emerged as important research areas, due to the ecological functions of landscape patterns. The aims of this study were to determine the relationship between landscape metrics (LMs), runoff, and sedimentation and explore the crucial LMs in the watersheds on the Loess Plateau. From 1985 to 2010, Grassland (GRA) was the dominant landscape in the Tuweihe (TU) and Gushanchuan (GU) watersheds. Unused land (UNL) and farmland (FAR), respectively, experienced the greatest transformations. The landscape in the study area tended to become regular, connected, and aggregated. The landscape stability of the TU watershed was higher than that of the GU watershed. Annual runoff and sedimentation gradually decreased and significant relationship was found between them (P <0.01). Due to larger FAR area and lower landscape stability in the GU watershed, the sedimentation of the two watersheds were similar, even though the runoff in the TU watershed was greater. The LMs had more significant effects on runoff than that on sedimentation yield. Shannon’s evenness index and the patch cohesion index had the greatest effects on runoff and sedimentation, making them the key factors of influencing water and soil loss. which could provide a scientific basis for the prevention and treatment of soil erosion.