The Mechanisms Of Soil Organic Carbon (SOC) Loss And Its Modeling

Soil organic carbon?SOC?loss accompanying with water erosion can lead to soil degradation and decline of soil productivity.As an important part of carbon cycle,SOC loss also accelerates global warming which is an important environmental problem all over of the world.In prevous studies,the regulation of SOC loss and the SOC enrichment mechanism in sediments have been widely researched,whereas less considered the hydraulic mechanism of SOC loss.Meanwhile,soil erosion models were usually built based on the hydraulic mechanism.Therefore,a total understanding of the internal relationships of SOC loss and the runoff hydraulic characteristic and building a hydraulic based SOC loss model will be of great significance.In our study,four loess soils with clay contents of 26%,21%,16%,and 12%,called Lou soil?Yangling?,Calcic Kastanozem?Changwu?,Yellow Loamy Soil?Ansai?and Yellow Loamy soil?Suide?,respectively,were selected.The sample sites are distributed from south to north across the Loess Plateau of China.By the artificial simulation of rainfall technology,the thesis further research the feature and regulations of SOC loss under the interaction of rainfall intensity,slope and soil texture,the internal relationships of SOC loss and the runoff hydraulic characteristics,and building a new hydraulic-based SOC loss model.The main results were illustrated as follows:?1?The interaction effects of rainfall intensity,slope and soil texture on the festures and regulations of SOC loss were further investigated,and the exact relationships of soil loss and SOC loss were clarified.The experimental results shows that the changing ranges of the enrichment ratios of SOC?ERocs?in sediments of Lou Soil,Calcic Kastanozem,Yellow Loamy Soil?Ansai?and Yellow Loamy Soil?Suide?were1.15-2.36?with a mean of 1.74?,1.18-1.69?with a mean of 1.27?,1.06-1.34?with a mean of 1.12?and 0.97-2.22?with a mean of 1.60?,respectively.Similar to the ERocs,the percentage of each sediment size class changed little within the range of experiment.The soil loss rate of Lou Soil,Calcic Kastanozem,Yellow Loamy Soil?Ansai?and Yellow Loamy Soil?Suide?were 66.59?67.33?22.79 and 152.15 g m^-2min^-1,respectively;their average SOC loss rate were 0.28,0.20,0.06 and 0.14 g m^-2min^-1,respectively.The average SOC loss rate has the similar variation tendency to the soil loss rate.Hence,the rate of soil erosion has an important effect on the SOC loss rate.Meanwhile,a significant correlation existed between the sediment yield and SOC loss after rainfall?P<0.05?,and the link is less pronounced when the SOC and clay content is higher.The SOC was easily enriched in sediments when the SOC content in soils was high.The correlation between soil loss and SOC loss could also become weak with the decrease of rainfall intensity and slope due to the increasing values of ERocs in sediments.In addition,both fine soil texture and high SOC content will make the effect of slope and rainfall intensity on soil loss and SOC loss become weak.Moreover,for Lou Soil and Yellow Loamy Soils,their sediment loss and SOC loss increased with the increasing of slope first;after the slope attached the critical slope gradient,their sediment loss and SOC loss decreased with the increasing of slope.However,for Calcic Kastanozem,the soil loss and SOC loss always increased with slope.?2?The internal relationships between flow velocity,runoff depth,shear stress and stream power and ERocs were detected by in-depth analysis of the changes of runoff hydraulic characteristics.The results shows that,under the interaction of rainfall intensity,slope and soil texture,the flow velocity always had a significant positive linear relationship with the SOC concentration in sediments?R²=0.731,P<0.001?whereas the runoff depth,shear stress and stream power didn't have a significant relationship with SOC concentration in sediments.The studies on the effect of soil texture on the relationship of flow velocity and SOC concentration in sediments find the soil texture has a decisive effect on flow velocity,and the flow velocity always change within an exact range for one soil.Therefore,the flow velocity was significantly positive correlated with the clay content of soil?R²=0.893;P<0.01?.Moreover,flow velocity also has close relationships with the first transported sediment particles with high SOC content,e.g.clay,silt and light aggregates.Hence,the correlation of flow velocity and SOC concentration in sediments is more pronounced under the low rainfall intensity?R²=0.958;P<0.001?.Therefore,under the interaction of soil texture,slope and rainfall intensity,the flow velocity always positively correlated with the SOC concentration in sediments.?3?The effect of flow hydraulic characteristics on the distribution of SOC in sediments was investigated by monitoring the SOC transport process.The study finds that when the rainfall intensity was less than 45 mm h^-1 or slope was less than 5°,the values of stream power and flow velocity were enough low.Hence,the ERocs of each size aggregates were obvious higher than 1.0,which was caused by the first transported of light fine aggregates with high SOC content produced by raindrop peeling.Meanwhile,when the rainfall intensity was less than 45 mm h^-1,no significant differences existed for the SOC concentration of each size aggregates in sediments under different slope?P<0.05?;when the slope was less than 5°,the increasing rainfall energy could lead to SOC enriched in finer and finer particles.Under the rainfall intensity greater than 45 mm h^-11 and on the slope greater than 5°,the effect of the breakup of aggregates on the distribution of SOC in sediments was less due to the large rate of water erosion.Therefore,the ERocs of each size aggregates closed to 1.0.By contrasting the time variation of the ERocs in different size aggregates during the erosion process,the SOC concentrations in several aggregates size classes decreased with time.This is because the increasing of runoff rate with time makes the intensity of raindrop peeling weak.By correlation analysis,the SOC concentration in aggregates significantly negative correlated with the flow velocity,runoff rate,stream power and shear stress,but significantly positive correlated with the clay and silt content in sediments.Flow velocity might have close relationships with the transport of SOC enriched aggregates.Our study will provide important helpful for understanding the fate of SOC and building a physically-based SOC transport model.?4?Referring to the soil erosion model and the SOC enriched hydraulic mechanism,a new hydraulic based SOC loss model was built.The results shows that the relationship between the average flow velocity and average sediment concentration for the Yellow Loamy Soil could be regressed by a linear function?R²>0.94;P<0.05?,and the relationship between slope and runoff could be modeled with a quadratic function?R²>0.94;P<0.05?.For Calcic Kastanozem,both sediment concentration and runoff showed signifcant quadratic function relationships with slope?R²>0.55;P<0.06?.Thus,soil loss can be simply predicted by flow velocity,slope.Then,the SOC loss was calculated by multiplying soil loss,SOC content in original soil and the ERoc values.However,the prediction accuracy of this traditional SOC calculation method mainly depends on the prediction accuracy of soil loss.Moreover,the datas of ERoc are difficult to obtain when predicting the long-time and large-scale SOC loss.To soilve this problem,on the basis of the hydraulic mechanism of SOC transport,by nonlinear regression,the SOC concentration in sediments was predicted by rainfall intensity,slope and flow velocity with R²=0.761 and P<0.005.Meanwhile,on the basis of the water transport capacity function,a hydraulic-based sediment and SOC loss model was built.The model includes six input variables:runoff rate,average flow velocity at 2 m location from the outlet,slope steepness,the difference between the effective and ultimate median diameter??D50?,the SOC content of the original soil and the rainfall time.The hydraulic-based sediment and SOC loss model includes one master equation and three equations used to calculate the parameters of the master equation.The input datas of the new proposed model were easily obtained.However,for the soils with a high?D₅₀ value,the prediction accuracies of the proposed model were low?R²>0.496,d>0.861?.For the soils with a low?D₅₀ value,the prediction accuracies of the proposed model were high?R²>0.903,d>0.974?.In addition,with the increase of?D₅₀,the effect of flow velocity on soil loss and SOC loss become large.With the increase of SOC content,effect of slope on soil loss and SOC loss become large.Considering that our study mainly focused on the loess soils,similar model may be built for other soils or in large scale field condition,which need to be further investigated.