Key factors determining soil organic carbon changes after freeze-thaw cycles in a watershed located in northeast...

The transformation and migration process of soil organic carbon (SOC) could be changed during freeze-thaw cycles (FTCs) and may further affect the SOC distribution in the watershed. In this study, both field investigation and lab incubation combined with geostatistics were used to clarify how environmental factors influence the SOC heterogeneity in Mollisol after FTCs, from a watershed in northeast China. The results showed that after FTCs, SOC decreased in 68.5% of the total watershed area at 0-20 cm soil depth, and the mean value decreased by 7.4%. Spatial autocorrelation (Moran's І) decreased in the 0-5 cm and 10-20 cm soil depths after FTCs (P < 0.01), but did not change in 5-10 cm and 0-20 cm soil depths. SOC increased at the top slope positions, the watershed outlet, and the upper slope position of the intersection area between farmland and forestland. The SOC decrease was 25 times and 14 times greater at the 0-5 cm and 0-20 cm soil depths respectively in the forestland than in the farmland. The SOC decrease was significantly higher on the 6-8° slopes than on the 0-2° slopes in 0-20 cm soil depth. SOC increased in most areas of cross-slope tillage (CST), but decreased in most areas of downslope tillage. The increase of SOC (positive change) decreased with increasing soil depth under soybeans field, while the decrease of SOC (negative change) increased with increasing soil depth under corn. Topographical factors alone, and topographical factors combined with land use types all influenced the SOC change in this watershed. High levels of ferrous minerals tended to reduce the rate of SOC after FTCs. SOC change is positively correlated to soil bulk density, FTC frequency and soil moisture after FTCs at the soil depth of 0-30 cm. Equations based on soil properties before FTCs, topographical factors, sediment transport index, runoff intensity index, and tillage method can be used for coarsely predicting SOC distribution after FTCs (45% < R² < 78%, P < 0.01). Generally, SOC dynamics were mainly determined by topography, land use, and to tillage methods that possibly attributes to soil and water loss during FTCs. Both erosion caused by snowmelt runoff and vertical migration of SOC could be the key factors that changed the SOC spatial pattern. CST was beneficial to conserve SOC during FTCs, while forestland could reduce SOC loss by reducing snowmelt erosion.