Spatial Changes Of Soil Organic Carbon Dynamics Induced By Tillage And Water Erosion On A Cultivated Slope In Chinese Loess Plateau

Soils play a major role in the global carbon cycle and have a huge potential for either sequestering or releasing carbon (C) to the atmosphere. The fate of SOC redistribution caused by soil erosion is a controversial subject. Globally, large amounts of soil organic carbon (SOC) are laterally redistributed on sloped cultivated land by soil erosion. On the toposequence or a field scale, it is generally recognized that loss of soil results in reduction of SOC. However, significant amounts of carbon might be stored in sediments in reservoirs and other deposition sites. Whether this results in a net source or sink of atmospheric CO2is unclear. Global estimates range from a source of～1Pg C per year to a sink of the same magnitude. Against this background, this study investigates impacts of soil redistribution processes on SOC stocks and fluxes and the corresponding C source or sink function in a small agricultural hillslope (0.07ha) in Chinese Loess Plateau. Spatial patterns of SOC stocks were studied by analysis of soil samples of the top layer, including plough layer and subsoil layers (up to a depth of0.6m) taken in a3x3m grid. In same soil samples, spatial patterns of the radionuclide tracer caesium-137(137Cs) and unsupported lead-210(210Pbcx) were used to quantify net erosion and net deposition within the field over a50-year and100-year period, respectively. Results revealed a substantial accumulation and stabilization of SOC with highest activity of the137Cs and210Pbex inventory at lower slope position, stressing the importance of subsoil C for budgets on agricultural land prone to erosion. Predicted maps of SOC,137Cs and210Pbex based on regression kriging showed that reduced SOC concentrations on the slope convexities at the upper slope and an increase SOC concentrations within the slope concave which traverse the centre of the field, along the boundary of bottom of field, indicates spatial distribution of SOC,137Cs and210Pbex are very similar patterns. For a spatially integrated analysis of the impact of soil redistribution on SOC dynamics at the small cultivated field scale, soil organic carbon redistribution was calculated by multiplying SOC concentration by total soil redistribution (TSR) including both tillage and water-induced soil redistribution derived from137Cs and210Pbex inventories and from the tillage erosion prediction model (TEP). It was applied for two periods from1911to1954and1954to2011, covering periods of the equilibrium of annuals SOC associated to soil redistribution estimated from137Cs and210Pbex measurement, and these measurement provide retrospective information on the medium-term (-50years span) and long-term (100-150years span) redistribution patterns of the eroded soil and SOC within the hillslope landscape. Results indicated that a substantial amount of SOC was lost laterally due to the export by water erosion. Tillage-induced soil redistribution increased SOC and compensated for3%and16%of the SOC losses due to water erosion during1911-1954and1954-2011, respectively. During the period1911-1954, the net SOC loss from the entire slope was6.77t C ha-1(0.157t C ha-1yr-1). Within the period1954-2011, the net SOC loss was12.58t C ha-1(0.221t C ha-1yr-1). Significant increase of SOC at the lower field boundary on the convergent footslope and backslope resulted from tillage-induced soil redistribution by moldboard plowing. Deposition on the lower field boundary was successfully validated by depth profiles of137Cs and210Pbex measured by hyper pure germanium gamma spectrometer. Land management had a profound effect on the lateral and vertical SOC. The effect of long term tillage reduced the lateral SOC export and enhanced the vertical SOC fluxes, leading to a reduced negative C balance. In situ measurements of soil respiration were carried out in the most dynamic area with respect to soil redistribution in one consecutive growing period (2011). No universal relation to spatial patterns of soil redistribution or to other parameters (soil properties and terrain attributes) was found for measurement period, underlining the large variability of soil respiration. However, soil respiration was significantly linearly related to total erosion. Hence, a possible C sink at erosion sites might partly be compensated by enhanced mineralization at depositional sites. Overall, this study substantially improves the knowledge and understands about the impacts of soil redistribution on SOC stocks and fluxes at the small cultivated field scale.