Identification of the principal mechanisms driving soil organic carbon erosion across different spatial scales.
Soil water erosion is recognized as the principal mechanisms behind soil organic carbon (SOC) losses from soils, a soil constituent essential for ecosystem functions. SOC erosion can thus be far-reaching, affecting the future human welfare and the sustainability of ecosystems. Little research has yet been done to investigate the main mechanisms involved in the lateral translocation of SOC on the landscape. Understanding the effects of the different water erosion mechanisms, which control SOC losses (SOC[L]) at the hillslope level, creates scope for further scientific studies. Empirical data from 357 plots, with a range in slope length from 1 (n=117) to 22.1m (n=240) were analysed to estimate the global variations of particulate organic carbon content (POC[C]), POC losses (POC[L]) and sediment POC enrichment ratio (ER). The global average POC[L] rate was calculated to be 12.1 g C m⁻² y¯¹. Tropical clayey soil environments revealed the highest POC[L] (POC[L]=18.0 g C m⁻² y¯¹), followed by semi-arid sandy (POC[L]=16.2 g C m⁻² y¯¹) and temperate clayey soil environments (POC[L]=2.9 g C m⁻² y¯¹). The global net amount of SOC displaced from its original bulk soil on an annual basis was calculated to be 0.59±0.09 Gt C, making up an approximated 6.5% of the net annual fossil fuel induced C emissions (9 Gt C). POC[L] data for different spatial scales revealed that up to 83% of the eroded POC re-deposits near its origin in hillslopes, and is not exported out of the catchment. The low organic carbon sediment ER obtained from the data of clayey soils (ER of 1.1) suggests that most of the eroded POC remains protected within soil aggregates. Consequently, erosion-induced carbon dioxide (CO₂) emissions in tropical areas with clayey soils are likely to be limited (less than 10%), as the process of POC re-burial in hillslopes is likely to decrease the rate of organic matter (OM) decomposition and thus serve as a potential carbon sink. Water erosion in sandy and silty soils revealed organic carbon sediment ER as high as 3.0 and 5.0, suggesting that in these soils the eroded POC is not re-buried, but is made vulnerable to micro-decomposers, thus adding to the atmospheric CO₂ influx. The results obtained in the review study only reaffirm that large variations of POC[L] are evident across the different pedo-climatic regions of the world, making it a scientific imperative to conduct further studies investigating the link between SOC erosion by water and the global carbon cycle. A field study was designed to quantify the POC exported in the eroded sediments from 1x1m² and 2x5m² erosion plots, installed at different hillslope aspects, and to further identify the main erosion mechanisms involved in SOC erosion and the pertaining factors of control. The erosion plots were installed on five topographic positions under different soil types, varying vegetation cover, and geology in the foothills of the Drakensberg mountain range of South Africa. Soil loss (SL), sediment concentration (SC), runoff water (R) and POCL data were obtained for every rainfall event from November 2010 up to February 2013. Scale ratios were calculated to determine which erosion mechanism, rain-impacted flow versus raindrop erosion, dominates R, SL and POC[L]. Averaged out across the 32 rainfall events, there were no significant differences in R and POC[L] between the two plot sizes but SL were markedly higher on the 5m compared to the 1m erosion plots (174.5 vs 27g m¯¹). This demonstrates that the sheet erosion mechanism has a greater efficiency on longer as opposed to shorter slopes. Rain-impacted flow was least effective where soils displayed high vegetation coverage (P < 0.05) and most efficient on steep slopes with a high prevalence of soil surface crusting. By investigating the role of scale in erosion, it was possible to single out the controlling in situ (soil surface related conditions) and ex situ (rainfall characteristics) involved in the export of SOC from soils. This information will in future contribute toward generating SOC specific models and thus further inform erosion mitigation.