Investigations on Temporal and Spatial Variation of Slope-Scale SOC Erosion and Deposition
Yaxian Hu
DOI: https://doi.org/10.5451/unibas-006327957
2014-01-01
Abstract:The net effect of soil erosion on global carbon cycling, especially as a source or sink for greenhouse gas emissions, has been the subject of intense debate. The controversy arises, to a large degree, from the inadequate understanding of the variation of soil organic carbon (SOC) in eroded sediment, and from the limited information on the fate of eroded SOC whilst in-transit from the site of erosion to the site of deposition. During a slope-scale erosion event, soil fractions and associated SOC will be transported away from eroding sites mainly by overland flow. If by interrill erosion, eroded sediment is often enriched in SOC. While the reported SOC enrichment ratios (ERSOC) are mostly greater than unity, they vary widely. Conservation of mass dictates that the ERSOC of sediment must be balanced over time by a decline of SOC in the source areas material. Although the effects of crust formation on SOC erosion have been discovered, a systematic study on crust formation over time and interrill SOC erosion has not been conducted so far. In addition, the inherent complexity of soil properties and SOC erosion process may inevitably introduce variations between replicates in SOC erosion data. Yet, the significance of such variation has not been systematically investigated. Even after erosion, SOC distribution in eroded soil also can change during transport. Regardless of selective interrill erosion or non-selective rill erosion, eroded soil will be either gradually re-deposited along hillslopes or further transferred to river systems. Under given flow conditions, the site of SOC deposition depends on the transport distances of sediment particles where the SOC is stored. Very often, soil and SOC erosion risk is assessed by comparing the SOC stock on eroding and colluvial depositional sites, or by applying the mineral particle specific SOC distribution observed from either site to estimate the SOC stock of its counterpart. However, soil is not always eroded as dispersed mineral particles, but mostly in form of aggregates. Aggregates possibly have distinct settling velocity from individual mineral particles, which may considerably change the transport distance of the associated SOC. In addition, SOC concentration in different aggregates probably differs from soil average SOC concentration, which also complicates the spatial re-distribution of eroded SOC. Yet, little has been known about the potential effects of aggregation onto the movement and fate of eroded SOC. Mineralization during transport may add an extra risk to SOC loss. Some reports claimed that most of the SOC transfer occurs during large-scale erosion events, rapidly transporting eroded SOC into depositional sites. Mineralization of eroded SOC during such rapid transport, therefore, is of minor importance and thus can be ignored when calculating carbon balances between eroding and depositional sites. Meanwhile, some other reports argued that erosion and transport tend to break down aggregates, expose previously protected SOC to microbes and atmosphere, and hence accelerate mineralization of eroded SOC during transport. To solve this discrepancy, it is required to understand the susceptibility of eroded SOC to mineralization during transport, especially for erosion events that mobilize soil but do not necessarily move it far enough to reach permanent depositional sites. The above-described debate on the fate of eroded SOC highlights four knowledge gaps: 1) how does SOC enrichment in eroded sediment vary with crust formation over rainfall time, and how the accordingly derived systematic variability affects soil and SOC erosion prediction; 2) how does the inherent complexity of interrill erosion processes affect the variability of SOC enrichment in eroded sediment; 3) how aggregation affects likely transport distance of eroded SOC; 4) whether or not erosion and transport induce acceleration of eroded SOC mineralization. In this study, a series of experiments was conducted to address the above-identified knowledge gaps: SOC-Variability experiment, SOC-Settling velocity experiment (SOC-Settling), SOC-Aggregation effects experiment 1 (SOC-Aggregation 1) and SOC-Aggregation effects experiment 2 (SOC-Aggregation 2). The SOC-Variability experiment was conducted to identify the temporal variation of SOC enrichment with crust formation during prolonged rainfall time, by applying a simulated rainfall to two silty loams placed in round flumes for 6 hours. A two-step erosion model was developed, based on the erosional response data obtained from six selected sub-events, to examine the systematic variability derived from crusting evolvement over rainfall time. In addition, the simulated erosion events were repeated ten times, enabling reliable statistical analysis for inter-replicate variability. Key results are: 1) the temporal variation of SOC enrichment ratio shows that ERSOC of eroded sediment cannot be always greater than unity, but varies with rainfall time, in agreement with conservation of mass; 2) the gradually improved systematic variability of SOC erosion prediction over rainfall time shows that observations from short events cannot be directly extrapolated to predict soil and SOC loss over prolonged events and vice versa; 3) the significant inter-replicate variability at maximum runoff and soil erosion rates suggests that variability remains significant even under ideal laboratory conditions. A settling tube apparatus was built up in the SOC-Settling experiment to fractionated soil samples according to the potential transport distances of various fractions. To further examine the aggregation effects onto the likely transport distance of eroded SOC, this settling tube apparatus was then applied in the experiment SOC-Aggregation 1, to fractionate eroded sediment generated from a silty loam. Results show that aggregation of source soil considerably reduces the likely transport distance of eroded SOC, and potentially increases its likelihood to be re-deposited along hillslopes. Based on this observation on a single soil in the experiment SOC-Aggregation 1, SOC-Aggregation 2 was then carried out with two types of soils, a silt loam and silt clay. Furthermore, the fractionated sediments were incubated for 50 days to assess their long-term mineralization potential. Key results from the experiment SOC-Aggregation 1, and SOC-Aggregation 2 show that: 1) Aggregation of source soil and preferential deposition of SOC-rich coarse sediment fractions potentially skew the re-deposition of eroded SOC into the terrestrial system. 2) Erosion and transport tend to accelerate mineralization of eroded SOC, demonstrating their potential to contribute additional CO2 to the atmosphere. Overall, this study demonstrates that both the temporal variation of SOC erosion and the spatial variation of SOC deposition on hillslopes have to be considered when assessing the role of soil erosion on net CO2 emissions. Applying “constant” SOC enrichment ratios in erosion models will lead to bias estimation of SOC loss. Aggregation effects of source soil considerably reduce the likely transport distance of eroded SOC, potentially skewing the re-deposition of SOC-rich coarse sediment fractions towards the terrestrial system. Erosion and transport processes tend to accelerate mineralization of eroded SOC, and hence potentially contribute additional CO2 to the atmosphere. Such findings may profoundly alter our current accounting for erosion-induced lateral SOC transfer, further suggesting that current recognition on deep burial of SOC on long-term depositional sites and the accordingly derived CO2 sink strength would be over-estimated. Significantly accelerated mineralization of eroded SOC during transport adds a further error into current carbon sink balances. Therefore, results from this study suggest that soil erosion is more likely to be a source of atmospheric CO2.