Karsten Kalbitz^{1, 2}, Erik Cammeraat², Frédérique Kirkels^{2, 3}, Xiang Wang^{2, 4}

　　¹   Dresden University of Technology, Institute of Soil Science and Site Ecology, Dresden, Germany

　　²   University of Amsterdam, IBED, Earth Surface Science, Amsterdam, The Netherlands

　　³   Utrecht University, Organic Geochemistry, Utrecht, The Netherlands

　　⁴   Department of Soil, Water and Climate, University of Minnesota, St Paul, USA

　　Water erosion influences the redistribution of soil organic carbon (SOC) in landscapes and there is a strong need to better understand these processes with respect to carbon (C) budgets, from local to global scales. Particularly, the fate or eroded soil organic carbon (SOC) after deposition is a large uncertainty in assessing the impact of soil erosion on C budgets. Globally, large amounts of SOC are transported by erosion and a substantial part is transferred into adjacent inland waters, linking terrestrial and aquatic C cycling. Furthermore, agricultural production has to meet the largely increasing demand for food particularly challenging production systems in the Global South. The development and implementation of soil conservation strategies using smart intensification might result in decreasing risks for soil erosion (Govers et al., 2016) but should consider impacts of soil erosion on the C cycle. Therefore, there is strong need to advance our understanding of the role of soil redistribution on C cycling. We have to expand our experimental knowledge about the mechanisms determining whether soil erosion results in decreasing or increasing emissions of atmospheric CO₂.

　　We present a combination of studies using sampling and analysis of soils along a hillslope including eroding and depositional areas, climate controlled pseudo-replicated rainfall-simulation experiments and long-term incubation experiments in the laboratory using a European gradient of agricultural sites to determine the turnover of SOC after simulated deposition on downslope soils or inland waters.

　　Deposition of eroded soil material resulted in C enrichment throughout the soil profile. Both macro-aggregate associated SOC and C associated with minerals increased in their importance from eroding to depositional areas. However, the formation of mineral-associated organic matter seems to be particularly important for the observed stabilization of SOC after deposition. Limited availability of O₂ in subsoils can be excluded as an important control of soil C accumulation. The sediments being eroded, transported and deposited were enriched in C in comparison to the bulk soil at eroding areas. Overall, CO₂ emission is the predominant form of C loss contributing to about 90.5% of total erosion-induced C losses in our 4-month rainfall simulation experiment, which were equal to 18 g C m^-2. Nevertheless, only 1.5% of the total redistributed C was mineralized to CO₂ indicating a large stabilization after deposition.Simulated SOC deposition in aquatic environments resulted in upto 3.5 timeshigher C turnover than deposition on downslope soils. Labile C inputs enlarged total CO₂emissions, with the largest increase for aquatic conditions. Temporal trends in CO₂ emissions clearly differed between soils and inland waters.We established a quantitative model, based on the ten sites of the European gradient, that is capable to describe/predict CO₂ emissions for SOC deposited on soils and in inland waters and upon different levels of labile C inputs. Our findings indicate that deposition conditions (soils vs. inland waters) play a crucial role in determining C turnover. Erosion management to prevent deposition in aquatic environments could therefore serve as carbon saving measure. We envisage that these quantitative resultscan be used to parameterize biogeochemical models and contribute to better estimates of the impact of soil erosion on C budgets and reduce uncertainties in the link between terrestrial and aquatic C cycling.