Geomorphology 226 (2014) 217–225

Contents lists available at ScienceDirect

Geomorphology

j ourna l homepage: www.e lsev ie r .com/ locate /geomorph

Effect of soil erosion on dissolved organic carbon redistribution insubtropical red soil under rainfall simulation

Wenming Ma a,b,c, Zhongwu Li a,b,d,⁎, Keyi Ding e,⁎⁎, Jinquan Huang a,b, Xiaodong Nie a,b, Guangming Zeng a,b,Shuguang Wang f, Guiping Liu f

a College of Environmental Science and Engineering, Hunan University, Changsha 410082, PR Chinab Key Laboratory of Environmental Biology and Pollution Control (Hunan University), Ministry of Education, Changsha 410082, PR Chinac College of Tourism Historical Culture, Southwest University for Nationalities, Chengdu 610041, PR Chinad State Key Laboratory of Soil Erosion and Dryland Farming on the Loess Plateau, Institute of Soil and Water Conservation, Chinese Academy of Sciences, Yangling, Shanxi 712100, PR Chinae College of Chemistry & Environment Protection Engineering, Southwest University for Nationalities, Chengdu 610041, PR Chinaf Soil and Water Conservation Research Institute, Shaoyang 422002, China

Abbreviations:NT-H, no-till plot with high intensity; Nsity; CT-L, chisel till plot with low intensity.⁎ Correspondence to: Z. Li, College of Environmental S

University, Changsha 410082, PR China. Tel./fax: +86 731⁎⁎ Correspondence to: K. Ding, College of ChemistEngineering, Southwest University for NationalitiesTel./fax: +86 28 85522503.

E-mail addresses: lizw@hnu.edu.cn (Z. Li), keyiding20

http://dx.doi.org/10.1016/j.geomorph.2014.08.0170169-555X/© 2014 Elsevier B.V. All rights reserved.

a b s t r a c t

a r t i c l e i n f o

Article history:Received 24 April 2014Received in revised form 10 August 2014Accepted 12 August 2014Available online 22 August 2014

Keywords:Soil erosionDissolved organic carbonVertical transportRainfall simulation

Water erosion governs soil carbon reserves and distribution across thewatershed or ecosystem. The dynamics ofdissolved organic carbon (DOC) underwater erosion in red agricultural soil is not clear. To determine the effect oftillage management and water erosion on vertical and lateral transportation of soil organic carbon (SOC) andDOC production under distinct rainfall intensities in the hilly red soil region of southern China, a chisel tillageplot with low rainfall intensity (CT-L) and two no-tillage plots with high (NT-H) and low rainfall intensity(NT-L) studies were conducted. Soil samples were collected from 0–5, 5–10, 10–20, and 20–40 cm soil layersfrom triplicate soil blocks pre- and post-rainfall for determining concentration of SOC and DOC. Runoff sampleswere collected at every 6 min for determining concentration of DOC and sediments during rainfall simulationson runoff plots (2 m × 5 m) with various intensities. No fertilizer was applied in any plots. Results clearlyshow that runoff volumes, sediments and SOC entrainedwith sediment, and laterallymobilizedDOCwere signif-icantly larger on NT-H compared to other plots, coinciding with changes in rainfall intensity; and the extent ofroughness of the plot surface (CT vs. NT) was the variation in runoff DOC concentration. During the simulatedrainfall events, DOC exports average 0.76, 0.64, and 0.27 g C m−2 h−1; SOC exports average 3.52, 1.08, and0.07 g m−2 h−1 in the NT-H, NT-L, and CT-L soils, respectively. The maximum export of DOC was obtainedunder a high intensity rainfall plot, which lagged behind maximum runoff volumes, sediments, and SOC losseswith sediment. Export of DOC was proportional to SOC content of soil loss. The least DOC losses in surface runoffand SOC losses with sediment were observed in CT-L plots. Vertical DOC mobilization achieved its maximumwith low intensity rainfall under CT treatment. The DOC did not accumulate at the soil surface andwas distribut-ed mainly in the second and third soil horizons. The distribution of DOC content down the soil profile increasedcompared to pre-rainfall, except for subplots E at NT-H and NT-L. Results indicate that rainfall significantly in-creased DOC content in experimental plots. The SOC content of sediment leaving the erosion zone was signifi-cantly correlated with overland flow volume and soil loss. These observations lead to the conclusion that soilerosion is an important factor controlling the export of dissolved organic carbon.

© 2014 Elsevier B.V. All rights reserved.

1. Introduction

Soil organic carbon (SOC) storage and transportation across terres-trial ecosystems are strongly affected by water erosion (Lal, 1995).

T-L, no-till plot with low inten-

cience and Engineering, Hunan88640078.

ry & Environment Protection, Chengdu 610041, PR China.

00@aliyun.com (K. Ding).

The earth's surface soil contains large quantities of organic carbon, stor-ing 1462–1548 Pg C (Batjes, 1996), (1 Pg = 1015 g) in the top 1 m androughly 2344 Pg C in the uppermost 3 m (Jobbagy and Jackson, 2000).Any changes in it will therefore cause a series of environmental prob-lems, i.e., eutrophication, nonpoint source pollution (Lal, 1998; Ninget al., 2006), and eventually, land degradation (Singer and Shainberg,2004). Water erosion has the greatest impact on C storage in theworld (Lal, 2003) because of the loss of soil carbon during erosion bytransport of surface runoff and sediments, which carries part of thesoil carbon and, moreover, different sized soil particles may carry differ-ent amounts of soil carbon (Shi et al., 2013;Wang et al., 2014). Principalfactors affecting erosion include soil properties, intensity and duration

218 W. Ma et al. / Geomorphology 226 (2014) 217–225

of rainfall, topography, surface cover, and soil moisture (Nearing et al.,2005). The complex process of soil erosion comprises detachment andtransport of soil particles combined with the influence of raindrop im-pact and surface runoff (Quansah, 1981). The kinetic energy of rain-drops breaks soil aggregates into individual components. The amountof soil available for removal by runoff depends on the strength of aggre-gates to resist the disruptive force of raindrop impact (Wuddivira et al.,2009).

Effect of soil erosion on the C dynamics (Rimal and Lal, 2009; Shiet al., 2013) and attendant emission of greenhouse gas arewidely recog-nized (Gregorich et al., 1998; Lal, 2003). A large proportion of terrestri-ally derived C has been found that may have beenmineralized (Durrieuet al., 2000; Goñi et al., 1998) and estimated that 1.14 Pg C yr−1 is emit-ted into the atmosphere (Lal, 1995). Jacinthe and Polyakov usedmicrorunoff plots and laboratory incubation to quantify erosion-induced CO2 evolution from the soil surface and runoff during rainfallevents (Jacinthe et al., 2002; Polyakov and Lal, 2004). Resultsshowed that CO2 emission from the deposition areas (17.0 g C m−2)during a 60-day period was higher than that from the eroded site(13.80 g C m−2) under the condition of tillage practice, and that 29–35% of the C exported in runoff wasmineralized in 100 days. More spe-cifically, the mineralizable carbon in runoff, eroded site, and depositionsediment should be a majority of labile carbon, in particular DOC (Lal,2003). Previous researcher's studies also indicated that the solubilitycharacteristics of SOCmight be another reason for high C loss associatedwith high surface runoff and soil loss during simulated rainfall (Jin et al.,2008). Lal estimated that depletion of the SOCpool upon cultivationwasattributed to leaching and translocation as dissolved organic carbon(DOC b 0.45 μm) or particulate organic carbon (POC N 0.45 μm) (Lal,2003). Jacinthe also reported that 29–46% of C is transported in surfacerunoff (Jacinthe et al., 2004). Leaching of DOC to subsoil horizons orgroundwater results from water erosion also leads to soil carbon lossin another way. Leaching of DOC increases terrestrial land carbon stor-age and plays a major role in carbon sequestration (Dlugoß et al.,2012; Rumpel and Kögel-Knabner, 2011).

To date, previous reports on the relationship between SOC and soilerosion have largely been confined to the lack of data of dissolved or-ganic carbon (DOC) distribution and to a failure of understanding verti-cal transport of SOC down the soil profile (Bremenfeld and Poesen,2012; Zhang et al., 2006). The literature related to soil erosion and thecarbon cycle is abundant (Lal, 2003; Van Oost et al., 2007). However,data related to the transport rate of DOC with various tillage manage-ment by water erosion are unavailable, particularly the downwardmovement pattern. The purpose of this study were, in the context of di-verse conditions of rainfall intensity and soil surface treatments, (i) todetermine the on-site vertical distribution of DOC and SOC along aslope, (ii) to identify the export of DOC and SOC at field plots, and (iii)to investigate the relationship between soil organic carbon contentand DOC transportation. From this we can further understand the dy-namics of soil carbon under water erosion typical of a subtropical redhilly region.

2. Material and methods

2.1. Site description

Red soil, which covers a total area of 2.18 × 106 km2 in southernChina, is characterized by low SOC and extensive weathering (Shiet al., 2010).Water erosion occurs throughout this region,with raindropsplash and slope runoff erosion being the dominant mechanisms(Zhang et al., 2013). Although the total amount of soil loss is lowerthan that of the Chinese Loess Plateau, the large geographic area of thehilly red soil region also makes it an important component of China'sterrestrial carbon cycle. The current study site (111°22′ E., 27°03′ N.)is located at the Soil and Water Conservation Research Station ofShaoyang in the southwest of Hunan Province, China (Fig. 1). The

average annual temperature is 17.1 °C and the mean annual precipita-tion is 1327.5 mm.

Soil in the region is Quaternary red clay with a clay-to-loam texture,which according to the U.S. Soil Taxonomy is classified as an Ultisol(Zhang et al., 2013). As is typical for the hilly red soil region, it has loworganic matter content. Soil general properties from experimentalplots were tabulated in Table 1.

2.2. Plot setup and rainfall simulation

The study was carried out in 2012 on three plots (5 m long and 2 mwide) with a 10% slope (Fig. 2). Each plot was surrounded with thinmetal frames driven into the ground to shield them from runoff fromadjacent plot areas. A V-shaped runoff funnel was positioned at theend of each plot to collect runoff samples. A plastic sheet was used tocover the runoff funnel at the lower end to prevent direct rainfall fromfalling into the collection system. The plot design was determined ac-cording to the following criteria: (i) minimal soil disturbance when set-ting the plot; and (ii) minimal distance between plots to minimize localvariations in soil between plots.

The simulated rainfall experiments were conducted at the threeplots described above using Spraco cone jet nozzles placed on 4.75-m-long stand pipes, which provided homogeneous rainfall on anarea 2.4 m in diameter. To test for homogeneity of rainfall, a pre-experiment was carried out. The homogeneous coefficient reached89.75, 90.23, and 91.36%, respectively, which was characterized byChristiansen's uniformity coefficient (CU) (Christiansen, 1942).

Types of rainfall intensity with 1.38± 0.06, 0.53 ± 0.08, and 0.52±0.10mmmin−1 representing the high and low intensity storms (lastingfor 60 min) of this region were applied in the three plots. The time atwhich runoff first occurred wasmeasured using a clock. Runoff sampleswere collected every 6 min in covered storage tanks (1 L) and plasticbuckets (30 L) until runoff generation ended. Runoff samples quantifiedthe losses of soil and solution volume and determined the rate of soiland water loss. Portion of runoff samples were taken to the laboratoryto test DOC within 24 h. From the amount of rainfall on the tray surfaceand the total runoff volume, runoff coefficients were calculated. Therunoff from the storage tanks was later air-dried to determine sedimentdelivery and carbon content.

2.3. Sampling

Three plots were divided into five subplots (A to E) at 1-m intervals(Fig. 2). In order to assess the vertical transport of SOC andDOCduring asimulated rainstorm, the sampling strategy involved pre- and post-rainfall samples. Sampling depths of 0–5, 5–10, 10–20, and 20–40 cmwere set. In a pre-rainfall experiment, three arranged gridswere chosenas the three replicates in each subplot. Each grid had dimensions of20 cm × 20 cm. Three replicated samples (0.5 kg) were obtained. Inthis way 360 samples (4 layers × 3 replicates × 5 subplots × 2 pre-and post-rainfall × 3 plots) were obtained. Samples were taken to thelaboratory, air-dried, and sieved through 0.25 mm mesh to determineorganic carbon. During sieving, residual plant matter was removed byhand. All soil samples for determination of DOC were stored in a refrig-erator at 4 °C until tested.

2.4. Sample treatment and analysis

The SOC concentrations were determined with the dichromate oxi-dation method of Walkley and Black (1934). Soil bulk density was ob-tained from three plots with the cutting ring method (Black andHartge, 1986).

Dissolved organic carbon is the fraction of organic substances thatpasses through a filter (0.1–0.7 μm) (Liu et al., 2007). In our study, it isthe fraction of organic substances that passes through a 0.45-μm filter.For determination of DOC content, 15 g of freeze-dried soil was shaken

Fig. 1. Location of study site.

219W. Ma et al. / Geomorphology 226 (2014) 217–225

with 30 mL of distilled water (1:2 w/v soil-to-solution ratios) for 2 h in50-mL Erlenmeyer flasks on a reciprocating shaker (Jingli LD5-10B) at aspeed of 200 rev. min−1 with a constant temperature of 25 °C. A mixedsample was centrifuged at 3000 rev. min−1 for 10min; the supernatantwas recovered and through cellulose nitrate filter (b0.45 μm). The fil-tratewas stored in polypropylene bottles for analysis of DOC (ShimadzuTOC-VCPH). The limit of detection was 0.5 mg L−1.

2.5. Statistical analysis

Statistical data analysis was performed using SPSS 20.0. ANOVA wasused to detect the effects of rainfall intensity, tillage practice, and SOCcontent on lateral and vertical distribution of DOC. Statistical significancewas determined at P b 0.05. Graphics were prepared using Origin 8.5.

3. Results

3.1. Runoff development and SOC loss

The time required for runoff initiation varied with rainfall intensity.Runoff from the experimental plots began 1 min 31 s, 48 s, and 2 min31 s after the initiation of rainfall; and within 12 min had reached asteady state rate of 72.87 and 30.83 L min−1 in NT-H and NT-L, respec-tively (Table 2). Runoff reached a steady rate of 9.61 L min−1 at 66 min

Table 1Properties of the surface soil in the plots.

Plots Tillagepractice

Water content(g g−1)

pH SOC(g kg−1)

Plot I NT-Ha preb 0.12 ± 0.04 4.47 ± 0.12 7.89 ± 2.64NT-H postc 0.20 ± 0.05 4.46 ± 0.13 7.63 ± 2.62

Plot II NT-Ld pre 0.14 ± 0.03 4.46 ± 0.08 7.87 ± 2.39NT-L post 0.19 ± 0.02 4.41 ± 0.11 7.81 ± 2.66

Plot III CT-Le pre 0.13 ± 0.03 4.50 ± 0.11 6.66 ± 2.42CT-L post 0.31 ± 0.22 4.54 ± 0.10 6.93 ± 2.56

a No-till plot with high intensity (1.38 ± 0.06 mm min−1).b Pre-rainfall simulation.c Post-rainfall simulation.d No-till plot with low intensity (0.53 ± 0.08 mm min−1).e Chisel till plot with low intensity (0.52 ± 0.10 mm min−1). The standard deviation of the

in the CT-L plot, that is, runoff from the CT-L plot took longer to developand stabilize. Themaximumrunoff volume (697.7 L)was obtained fromthe NT-H plot and theminimum (83 L) from the CT-L, whichwas fueledby infiltration processes.

Sediment delivery rate from the plots showed a varying pattern(Fig. 3), soil loss rate from trials continued to increase with rainfall in-tensity, reaching equilibrium after 12 min of rainfall. In the two plotsthat received the same rainfall intensity, the soil loss rate of the NT-Lplot was greater than that of the CT-L plot. This soil loss pattern was,perhaps, attributable to the dispersion of soil aggregates on the surfaceby raindrop impact. Low soil loss in CTwas seen in low rainfall intensitywhen detachment forces were small. The SOC concentration in the sed-iment for the NT-H event was initially high but diminished within ashort period, eventually reaching relatively low values with minor fluc-tuations. The SOC concentration in sediment decreasedwith rainfall du-ration and increased with rainfall intensity, reaching a relatively steadystate after 12 min, except for CT-L. The SOC losses were 3.52, 1.08, and0.07 g m−2 h−1 and soil losses were 3.54, 0.76, and 0.059 kg m−2

from the NT-H, NT-L, and CT-L plots, respectively. Soil loss was directlyrelated to the amount of surface runoff, which is consistent with previ-ous studies (Guerra, 1994; Jacinthe et al., 2004). The amount of sedi-ment, runoff volume, and SOC in sediment indicates a relationshipbetween erosion intensity and rainfall intensity: the greater the intensi-ty of rainfall, the more severe the erosion.

Soil bulk density(g cm−3)

Clay (%)(b0.002 mm)

Silt (%)(0.002–0.05 mm)

Sand (%)(N0.05 mm)

1.62 ± 0.15 40.60 ± 3.47 27.72 ± 3.98 33.14 ± 4.061.59 ± 0.17 38.41 ± 3.64 27.91 ± 3.74 33.68 ± 1.221.67 ± 0.14 38.18 ± 2.67 28.23 ± 3.43 33.59 ± 1.521.75 ± 0.18 40.60 ± 3.03 26.24 ± 3.11 33.16 ± 2.131.66 ± 0.17 37.42 ± 1.74 27.52 ± 1.63 35.06 ± 1.271.77 ± 0.07 37.51 ± 1.74 28.54 ± 2.00 33.95 ± 1.46

mean (n = 15).

Fig. 2. Plot design for the rainfall simulation experiments.

220 W. Ma et al. / Geomorphology 226 (2014) 217–225

3.2. Lateral transport of DOC

During the simulated rainfall events, the quantity of DOC exportedfrom the plots was related to rainfall intensity and practice treatment.Losses of 0.76, 0.64, and 0.27 g C m−2 h−1 from the NT-H, NT-L, andCT-L soils were collected, respectively. The DOC concentration in therunoff samples changedwith respect to rainfall duration and rainfall in-tensity (Fig. 3). The highest DOC concentration in runoff samples wasdetected at 48, 42, and 96min in the NT-H, NT-L, and CT-L, respectively.Statistical analysis showed that the effects of rainfall intensity, rainfallduration, and surface treatment on DOC export were significant(P b 0.05).

Although surface runoff and soil loss were high for NT-H and NT-Lplots, the highest loss of DOC concentration was observed in the CT-Lplot, which reached 45.29 mg L−1 and was significantly greater thanDOC concentration from NT-H and NT-L plots, but the loss of DOC waslowest in the three plots because of the low runoff volume in CT-L.

3.3. Vertical distribution of DOC

The DOC contents of soil samples, whichwere taken before the rain-fall simulation, are shown in Fig. 4. The result indicates that DOC hasconsiderable spatial variability in the experimental plots. The highestpre-rainfall DOC concentration values appeared in the 5–10 or10–20 cm soil horizons except for subplot C at all plots combined withsubplot D at the NT-L plot. It reveals that DOC did not accumulate inthe surface layer but in the second or third soil horizon. This result dif-fers from the SOC distribution pattern reported by previous researcherswho determined that SOC was concentrated in the surface layer(Reicosky et al., 1995).

The changes in DOC content down the soil profile post-rainfall areshown in Fig. 5. In general, the trend of increasing DOC content with

Table 2The characteristics of surfaces under different rainfall intensities.

Plots Surface treatment Rainfall intensity(mm min−1)

Initial production flow of time(min)

D(m

Plot I NT-H 1.38 ± 0.06 1′31″ 6Plot II NT-L 0.53 ± 0.08 2′31″ 6Plot III CT-L 0.52 ± 0.10 48′ 10

depth was also observed in all three plots. The highest post-rainfallDOC concentration values were observed in the 5–10 or 10–20 cmsoil layers, except for subplots C and D in the CT-L plot. The highestDOC concentration in the NT plots was observed in the third soillayer (10–20 cm). In other plots it appeared in the second soil layer(5–10 cm). This result indicates that DOC accumulated in NT-Hand CT-L plots after rainfall simulation, except for subplot E in NT-Hin which DOC content decreased in all layers compared to pre-rainfall values. The DOC content was also highly stratified withsoil depth under NT-H and CT-L plots, except for subplot C of the NT-Hand CT-L plots in which DOC concentration obviously decreasedor fluctuated with depth. The DOC concentrations in subplot E soilhorizons of NT-H and NT-L were remarkably attenuated relative topre-rainfall.

3.4. Vertical transport of SOC

Prior to the rainfall simulation, distribution of total SOCwas calculat-ed for the 0–40 cm deep soil horizons (Table 3). The concentrations ofSOC were significantly different among slope positions (P b 0.001);but on average, SOC accumulated at the surface, except in subplot B ofNT-H and NT-L, and in subplot A of NT-L and CT-L. The SOC concentra-tions in subplots A and B were considerably higher than those in sub-plots C, D, and E of trials, that is, significantly lower in toe slopes. TheSOC concentrations in the upper and middle slope positions (subplotsA and B) showed similar vertical distribution patterns, and they signifi-cantly increased with depth for the 5–10 cm layers (P b 0.01, except forthe 5–10 cm of subplot A at NT-H plot).

Post-rainfall redistribution of SOC exhibited fluctuation relative topre-rainfall. Nevertheless, in addition to subplots B and C of NT-H andsubplot A of CT-L, SOC was still substantially concentrated at the soilsurface (0–5 cm), regardless of the rainfall intensities. This surface

uration runoff producingin)

Sediment yields(g)

Runoff volume(L)

Runoff coefficient(%)

0 511.1 697.7 84.260 262.5 292.6 92.018 66 83 26.28

Fig. 3. Effect of various rainfall intensities and tillage practices on runoff and outlet DOC. NT-H: no-till plot with high intensity, NT-L: no-till plot with low intensity, CT-L: chisel-tilled plotwith low intensity.

221W. Ma et al. / Geomorphology 226 (2014) 217–225

accumulation pattern is consistent with previous reports (Jin et al.,2008; Reicosky et al., 1995).

4. Discussion

The field experiments were conducted under simulated rainfall toevaluate the effects of diverse rainfall intensities and soil surface treat-ment on DOC and SOC loss in subtropical red soil agricultural land.The results showed that DOC and SOCwere strongly affected by rainfallintensity and slope treatment. Transport of dissolved organic matter inrunoff involves two processes: (i) leaching movement (Ahuja et al.,

Fig. 4. Concentration of dissolved organic carbon (DOC) pre-rainfall simulation experiment for d1.38± 0.06mmmin−1, NT-L: no-till with 0.53± 0.08mmmin−1, CT-L: chisel till with 0.52± 0intervals. The error bars represent the standard deviation of the mean (n = 3).

1981), and (ii) transport with surface runoff and sediment phasealong the lateral direction (Cauwet and Mackenzie, 1993). Statisticalanalysis revealed significant effects of rainfall intensity on the transportand content of DOC (P b 0.05).

4.1. Soil erosion and SOC loss

Two kinds of plots, chisel-till and no-till plots, were used in the pres-ent study. The output of soil carbon (including SOC andDOC) fromNT-Hand NT-L was higher than that from CT-L. Previous research has shownthat macroaggregates in NT soils can provide an important mechanism

epths 0–5, 5–10, 10–20, and 20–35 cm in theNT-H, NT-L, and CT-L plots (NT-H: no-till with.10mmmin−1). A to E represent subplots from the slope top to the bottom located at 1-m

Fig. 5. Concentration of dissolved organic carbon (DOC) post-rainfall simulation experiment for depths 0–5, 5–10, 10–20, and 20–35 cm in the NT-H, NT-L, and CT-L plots. A to E representsubplots from the slope top to the bottom located at 1-m intervals. The error bars represent the standard deviation of the mean (n = 3).

222 W. Ma et al. / Geomorphology 226 (2014) 217–225

for the protection of SOM (Beare et al., 1994) and that no tillage practicecould relieve water erosion and play an important role in carbon se-questrations (McHunu et al., 2011). Whereas our results were preciselyopposite. One explanation is that the rates of runoff at the chisel tillageplot tended to decrease because chisel tillage contributed to roughnesssurface. Infiltration capacity increases with surface roughness becauseof changes in ponding dynamics (Belnap et al., 2005). Runoff amountis determined by soil infiltration capacity and affected by soil porosity(Barger et al., 2006). Secondly, soil aggregation is also a very pivotal fac-tor, as it controls water infiltration to a greater extent than the amountsof sand, silt, and clay (Wuest et al., 2006). The crucial factor determiningwhether water will soak in or runoff is the ability of the soil surface to

Table 3Vertical distribution of soil organic carbon (SOC) in the 0–40 cm deep soil layer for pre- and p

Subplots Depths (cm) NT-H

Pre-rainfallSOC g kg−1

Post-rainfallSOC g kg−1

A 0–5 10.83 ± 0.93 10.08 ± 1.415–10 9.35 ± 1.16 10.27 ± 0.39

10–20 8.66 ± 2.05 11.03 ± 0.920–40 6.57 ± 1.22 8.42 ± 0.59

B 0–5 10.54 ± 0.92 10.31 ± 0.595–10 10.93 ± 0.91 10.51 ± 0.73

10–20 10.04 ± 1.19 11.93 ± 1.8820–40 6.35 ± 0.57 8.47 ± 2.20

C 0–5 7.06 ± 1.57 7.61 ± 2.465–10 5.79 ± 1.21 8.11 ± 1.57

10–20 5.06 ± 1.34 5.48 ± 0.1120–40 4.57 ± 0.29 4.58 ± 0.65

D 0–5 5.04 ± 0.79 4.94 ± 0.325–10 4.38 ± 0.86 4.39 ± 0.55

10–20 4.65 ± 0.59 4.76 ± 0.3320–40 3.02 ± 0.44 3.28 ± 0.46

E 0–5 5.97 ± 1.22 5.20 ± 0.325–10 5.41 ± 1.04 5.02 ± 1.26

10–20 5.44 ± 0.73 3.89 ± 0.9720–40 4.18 ± 2.11 3.83 ± 1.02

a Vertical distribution of soil organic carbon (SOC) in the 0–40 cmdeep soil layer for pre- andto the bottom located at 1-m intervals. The standard deviation of the mean (n = 3).

resist slaking and reconsolidation or crusting of the soil surface. Whenthe individual sand, silt, and clay particles resulting from slaking arefree to move, they settle into a very compact layer (Wuest et al.,2006). In this experiment, the soils have high clay content (more than37%); and clay helps bind aggregates of sand, silt, and clay particles,making them more resistant to slaking (Wuest et al., 2006). Chisel till-age practice broke down aggregates and released organic carbon mat-ter, which is a bonding agent between mineral soil particles (Mikhaand Rice, 2004). Furthermore, labile C fractions were found to accumu-late preferentially at the soil surface in no-tillage practices (Alvarezet al., 1995). Another explanation is that soil organic carbon releasedfrom aggregates was not transported out of the runoff plot in the chisel

ost-rainfall simulation.a

NT-L CT-L

Pre-rainfallSOC g kg−1

Post-rainfallSOC g kg−1

Pre-rainfallSOC g kg−1

Post-rainfallSOC g kg−1

9.77 ± 0.65 10.19 ± 0.77 7.74 ± 0.39 8.03 ± 0.6310.19 ± 0.57 9.99 ± 0.49 9.64 ± 1.42 8.81 ± 1.697.79 ± 1.34 8.23 ± 0.84 11.03 ± 2.49 11.63 ± 2.105.67 ± 1.08 5.32 ± 1.12 6.43 ± 1.27 5.79 ± 0.61

10.53 ± 0.12 10.57 ± 0.50 10.39 ± 0.47 10.75 ± 1.8210.73 ± 0.51 10.51 ± 0.86 9.31 ± 1.21 10.59 ± 1.168.14 ± 2.79 8.21 ± 1.45 8.47 ± 2.62 8.70 ± 2.026.49 ± 1.69 8.56 ± 2.67 4.93 ± 1.32 6.23 ± 1.957.56 ± 1.79 8.40 ± 1.58 4.75 ± 1.86 6.26 ± 1.696.10 ± 1.95 5.78 ± 0.36 4.13 ± 1.05 5.19 ± 2.464.81 ± 0.18 5.15 ± 0.47 4.27 ± 0.79 5.18 ± 0.733.76 ± 0.91 5.35 ± 1.98 4.26 ± 0.20 4.28 ± 0.735.43 ± 1.23 5.37 ± 0.76 5.80 ± 1.16 5.28 ± 0.745.02 ± 0.98 4.59 ± 0.47 5.46 ± 1.70 5.04 ± 0.794.59 ± 1.02 4.44 ± 0.45 5.30 ± 0.74 4.88 ± 0.493.63 ± 1.21 3.86 ± 1.42 3.31 ± 0.23 4.12 ± 0.826.04 ± 2.14 4.53 ± 0.34 4.61 ± 0.39 4.31 ± 0.085.3 ± 1.24 4.05 ± 0.11 4.15 ± 0.81 3.21 ± 0.27

5.71 ± 0.79 4.25 ± 0.96 3.81 ± 2.10 3.93 ± 2.824.15 ± 1.71 3.92 ± 1.93 3.42 ± 0.57 3.94 ± 2.24

post-rainfall of theNT-H, NT-L, and CT-L plots; A to E represent subplots from the slope top

223W. Ma et al. / Geomorphology 226 (2014) 217–225

practice. The third potential reason is that rainfall intensities of NT-L andCT-L did not achieve the energy threshold that required fully dispersingsoil aggregates.We therefore observed a result of a higher output of soilcarbon under no tillage than under chisel tillage.

4.2. DOC stratified with soil profile

The SOC concentration was generally higher in the surface soil, butthis distribution patternmay not be consistent for DOC that substantial-ly accumulated in the second or third soil horizon. Biological decompo-sition likely strongly affected the dissolved organicmatter in the surfacecompared to sublayers (Post and Kwon, 2000). Hence a large portion oftheDOMpercolated through the soil to the subsoil, but little soaked intodeeper layers. Subsoil DOC absorbed to the soil matrix is retained andconcentrated in the subsoil layer (Qualls and Haines, 1992). Mineralsurfaces in the deeper layer of the soil profile are unsaturatedwith C be-cause little C input from upper horizons occurs in these deeper layers,which therefore have a larger saturation deficiency (Hassink andDalenberg, 1996; Stewart et al., 2007). Chisel tillage brings subsoil tothe surface where it is subject to raindrop impact (Beare et al., 1994).The mixture of soil profiles will cause the burial of soil fragmentationand crop residues and then increase disruption susceptibility of aggre-gates (Gregorich et al., 1998; Paustian et al., 2000). For example, itleads to macroaggregates on the surface prior to being dispersed byraindrops and chisel tillage activities. Becausemacroaggregates are gen-erally less resistant to external disruption compared tomicroaggregates(Rasiah et al., 1992), the occluded light fraction had higher C and N con-centrations than the free light fraction and contained more alkyl C(i.e., long chains of C compounds and peptides) and less O-alkyl C(e.g., carbohydrates andpolysaccharides). Thoughmuchof the occludedfraction does not dissolve in water, the O-alkyl C is generally hydrophil-ic. Besides, chisel tillage greatly increases surface roughness on the CT-Lplot, therefore generated less surface runoff and more percolation withan increase inDOC,which consequentlymobilized downward along soilprofiles and accumulated in second or third soil horizons. Infiltration ofwater into soil subsurface was always greatest under CT-L than NT-H

Fig. 6.Water content pre-rainfall simulation experiment atNT-H, NT-L, andCT-L plots (NT-H: nowith 0.52 ± 0.10 mmmin−1). A to E represent subplots from the slope top to bottom located

and NT-L plots (Fig. 6). This difference in infiltration rates were alsostrongly related to differences in DOC stratification (Fig. 4).

4.3. The increase in DOC concentration after rainfall simulation

TheDOC content was elevated after rainfall, which is consistentwithresults of numerous authors (Kalbitz et al., 1997; Lundquist et al., 1999;Tipping et al., 1999). The primary ways of DOC accumulation duringrainfall simulation could be attributed to four processes: (i) dead biotaon the surface (root residues, plant debris, and organisms) and water-soluble organic matter moving into the subsoil horizons; (ii) the physi-cal protection of SOC afforded by aggregation is removed by water ero-sion when raindrop-impacted aggregates are broken down and labileSOC fractions are released. These parts of the water-soluble organicmatter continued their downward movement to lower horizons. TheDOC content in each soil profile increased after rainfall simulation. Inthe CT plot, chisel tillage may cause a release of C by breaking upthe aggregate structures, thereby increasing C availability. Morespecifically, this leads to a loss of C-rich macroaggregates and an in-crease of C-depleted microaggregates. The carbon content of macroag-gregates was 1.65 times greater compared to microaggregates (Sixet al., 2000). (iii) Water content had a pronounced effect on DOC con-centrations; water content curves showed higher water content athigher DOC concentrations (Fig. 7), whichwas consistentwith previousresults (Falkengren-Grerup and Tyler, 1993); and (iv) inmost cases, an-aerobic conditions caused by water saturation increased DOM releasefrom soils. Fiedler and Kalbitz (2003) analyzed the influence of redoxconditions on DOM dynamics in the field and showed that DOC concen-trations were smallest in the aerobic soil and largest in the most anaer-obic soil, which agrees with our results. The soil surface becameanaerobic as a result of overlandflowandsurface seal and crust develop-ment (Wischmeier andMannering, 1969). If the infiltrationwas limited,water accumulation and flow downslope combined with movement ofsoil particles exacerbated the extent of the anaerobic condition.

The concentration of SOC did not significantly alter after rainfall sim-ulation in this study — a possible reason is that a proportion of SOC re-leased from aggregations were taken away by runoff, another still left

-till with 1.38±0.06mmmin−1, NT-L: no-till with 0.53±0.08mmmin−1, CT-L: chisel tillat 1-m intervals. The error bars represent the standard deviation of the mean (n = 3).

Fig. 7. Water content post-rainfall simulation experiment at NT-H, NT-L, and CT-L plots. A to E represent subplots from the slope top to bottom located at 1-m intervals. The error barsrepresent the standard deviation of the mean (n = 3).

224 W. Ma et al. / Geomorphology 226 (2014) 217–225

in situ, moreover, the latter less than the former. The relationship be-tween SOC and DOC were proportional to growth in experimentalplots. Fitted simple linear regressions of SOC and DOC showedthat SOC and DOC simultaneously increase when aggregations weredisrupted and that they decreased along with runoff occurrence(Table 4).

5. Summary and conclusions

We investigated the effect of soil erosion on DOC redistribution intwo differing intensity rainfall events. The transport of DOC and SOCby water erosion and its accumulation were investigated. We demon-strated that soil erosion is an important factor controlling export ofDOC and SOC. Rainfall intensity influenced DOC lateral transport morethan DOC vertical mobilization. The DOC lateral exports reached amax-imum level under high intensity rainfall. Vertical DOC mobilizationreached amaximum level with low intensity rainfall under the CT treat-ment plot. The least SOC loss with sediment and DOC loss with surfacerunoff were observed from the CT-L plot. The SOC content of sedimentleaving the erosion zone was significantly correlated with overlandflow volume and soil loss. Maximum DOC loss lagged behind those of

Table 4Fitted simple linear regressions of y to x in the form of y = ax + b, x: SOC content, y: DOC con

Tillage practice Depths (cm) a b R2 P value

NT-H pre-rainfall 0–5 1.60 0.64 0.56 b0.015–10 2.21 −2.19 0.78 b0.01

10–20 2.49 −3.52 0.71 b0.0120–40 2.03 −2.31 0.38 b0.01

CT-L pre-rainfall 0–5 0.95 1.19 0.42 b0.015–10 1.59 −1.03 0.61 b0.01

10–20 1.79 −2.45 0.72 b0.0120–40 1.47 −1.16 0.45 b0.01

NT-L pre-rainfall 0–5 2.32 −3.80 0.52 b0.015–10 3.11 −7.63 0.72 b0.01

10–20 4.13 −11.56 0.80 b0.0120–40 1.53 −0.09 0.29 b0.05

a Present no significant.

runoff volume, sediment, and SOC loss with sediment, which was pro-portional to SOC content of soil loss. The DOC content did not accumu-late at the soil surface and mainly distributed to the second and thirdsoil horizons. The distribution of DOC content down the soil profile insubplots of NT-H and NT-L declined compared to the pre-rainfall distri-bution. In addition, the post-rainfall vertical distribution of DOC contentwasmore pronounced than pre-rainfall, indicating throughfall elevatedDOC content in experimental plots. Based on this we conclude that soilerosion is an important factor controlling export of DOC, and rainfall isan important factor affecting the concentration of DOC.

Acknowledgments

We thank the Editor, Richard Marston, and the anonymous re-viewers for their helpful comments and suggestions that contributedto improve the original manuscript. We also wish to acknowledgeProfessor Brain McGarvey for providing language help. This studywas funded by the National Natural Science Foundation of China(41271294) and the Program for New Century Excellent Talents in Uni-versity (NCET-09-330).

tent.

Tillage practice Depths (cm) a b R2 P value

NT-H post-rainfall 0–5 3.81 −4.25 0.47 b0.015–10 5.26 −12.19 0.59 b0.01

10–20 4.04 −2.60 0.27 b0.0520–40 2.23 4.56 0.17 nsa

CT-L post-rainfall 0–5 0.76 2.90 0.31 b0.055–10 2.90 −5.85 0.65 b0.01

10–20 2.14 −0.66 0.69 b0.0120–40 1.91 −0.21 0.38 b0.01

NT-L post-rainfall 0–5 2.91 −6.23 0.64 b0.015–10 4.11 −11.10 0.86 b0.01

10–20 3.70 −8.37 0.72 b0.0120–40 1.52 0.91 0.66 b0.01

225W. Ma et al. / Geomorphology 226 (2014) 217–225

References

Ahuja, L., Sharpley, A., Yamamoto, M., Menzel, R., 1981. The depth of rainfall–runoff–soilinteraction as determined by 32P. Water Resour. Res. 17, 969–974.

Alvarez, R., Díaz, R.A., Barbero, N., Santanatoglia, O.J., Blotta, L., 1995. Soil organic carbon,microbial biomass and CO2-C production from three tillage systems. Soil Tillage Res.33, 17–28.

Barger, N.N., Herrick, J.E., Van Zee, J., Belnap, J., 2006. Impacts of biological soil crust distur-bance and composition on C and N loss from water erosion. Biogeochemistry 77,247–263.

Batjes, N.H., 1996. Total carbon and nitrogen in the soils of the world. Eur. J. Soil Sci. 47,151–163.

Beare, M., Hendrix, P., Cabrera, M., Coleman, D., 1994. Aggregate-protected and unpro-tected organic matter pools in conventional- and no-tillage soils. Soil Sci. Soc. Am. J.58, 787–795.

Belnap, J., Welter, J.R., Grimm, N.B., Barger, N., Ludwig, J.A., 2005. Linkages betweenmicro-bial and hydrologic processes in arid and semiarid watersheds. Ecology 86, 298–307.

Black, G., Hartge, K., 1986. Bulk density. Methods Soil Anal. 1, 347–380.Bremenfeld, F.e.a.C., V., Poesen, J., 2012. Sediment, soil organic carbon and runoff delivery

at various spatial scales. Catena 88, 46–56.Cauwet, G., Mackenzie, F., 1993. Carbon inputs and distribution in estuaries of turbid

rivers: the Yangtze and Yellow rivers (China). Mar. Chem. 43, 235–246.Christiansen, J., 1942. Irrigation by Sprinkling. California Agricultural Experiment Station

Bulletin 670. University of California, Berkeley, CA.Dlugoß, V., Fiener, P., Van Oost, K., Schneider, K., 2012. Model based analysis of lateral and

vertical soil carbon fluxes induced by soil redistribution processes in a small agricul-tural catchment. Earth Surf. Process. Landf. 37, 193–208.

Durrieu, D.E., Madron, X., Abassi, A., Heussner, S., Monaco, A., Aloisi, J.C., Radakovitch, O.,Giresse, P., Buscail, R., Kerherve, P., 2000. Particulate matter and organic carbon bud-gets for the Gulf of Lions (NW Mediterranean). Oceanol. Acta 23, 717–730.

Falkengren-Grerup, U., Tyler, G., 1993. The importance of soil acidity, moisture, exchange-able cation pools and organic matter solubility to the cationic composition of beechforest (Fagus sylvatica L.) soil solution. Z. Pflanzenernähr. Bodenkd. 156, 365–370.

Fiedler, S., Kalbitz, K., 2003. Concentrations and properties of dissolved organic matter inforest soils as affected by the redox regime. Soil Sci. 168, 793–801.

Goñi, M.A., Ruttenberg, K.C., Eglinton, T.I., 1998. A reassessment of the sources and impor-tance of land-derived organic matter in surface sediments from the Gulf of Mexico.Geochim. Cosmochim. Acta 62, 3055–3075.

Gregorich, E.G., Greer, K.J., Anderson, D.W., Liang, B.C., 1998. Carbon distribution andlosses: erosion and deposition effects. Soil Tillage Res. 47, 291–302.

Guerra, A., 1994. The effect of organic matter content on soil erosion in simulated rainfallexperiments in W. Sussex, UK. Soil Use Manag. 10, 60–64.

Hassink, J., Dalenberg, J.W., 1996. Decomposition and transfer of plant residue 14C be-tween size and density fractions in soil. Plant Soil 179, 159–169.

Jacinthe, P.-A., Lal, R., Kimble, J., 2002. Carbon dioxide evolution in runoff from simulatedrainfall on long-term no-till and plowed soils in southwestern Ohio. Soil Tillage Res.66, 23–33.

Jacinthe, P.-A., Lal, R., Owens, L., Hothem, D., 2004. Transport of labile carbon in runoff asaffected by land use and rainfall characteristics. Soil Tillage Res. 77, 111–123.

Jin, K., Cornelis, W.M., Schiette, W., Lu, J.J., Buysse, T., Baert, G., Wu, H.J., Yao, Y., Cai, D.X.,Jin, J.Y., De Neve, S., Hartmann, R., Gabriels, D., 2008. Redistribution and loss of soil or-ganic carbon by overland flow under various soil management practices on the Chi-nese Loess Plateau. Soil Use Manag. 24, 181–191.

Jobbagy, E.G., Jackson, R.B., 2000. The vertical distribution of soil organic carbon and its re-lation to climate and vegetation. Ecol. Appl. 10, 423–436.

Kalbitz, K., Popp, P., Geyer, W., Hanschmann, G., 1997. β-HCH mobilization in pollutedwetland soils as influenced by dissolved organic matter. Sci. Total Environ. 204,37–48.

Lal, R., 1995. Global soil erosion by water and carbon dynamics. Soils Glob. Chang.131–142.

Lal, R., 1998. Soil erosion impact on agronomic productivity and environment quality. Crit.Rev. Plant Sci. 17, 319–464.

Lal, R., 2003. Soil erosion and the global carbon budget. Environ. Int. 29, 437–450.Liu, R., Lead, J.R., Baker, A., 2007. Fluorescence characterization of cross flow ultrafiltration

derived freshwater colloidal and dissolved organic matter. Chemosphere 68,1304–1311.

Lundquist, E., Jackson, L., Scow, K., 1999.Wet–dry cycles affect dissolved organic carbon intwo California agricultural soils. Soil Biol. Biochem. 31, 1031–1038.

McHunu, C.N., Lorentz, S., Jewitt, G., Manson, A., Chaplot, V., 2011. No-till impact on soiland soil organic carbon erosion under crop residue scarcity in Africa. Soil Sci. Soc.Am. J. 75, 1503–1512.

Mikha, M.M., Rice, C.W., 2004. Tillage andmanure effects on soil and aggregate-associatedcarbon and nitrogen. Soil Sci. Soc. Am. J. 68, 809–816.

Nearing, M.A., Jetten, V., Baffaut, C., Cerdan, O., Couturier, A., Hernandez, M., Le Bissonnais,Y., Nichols, M.H., Nunes, J.P., Renschler, C.S., Souchère, V., van Oost, K., 2005. Modelingresponse of soil erosion and runoff to changes in precipitation and cover. Catena 61,131–154.

Ning, S.-K., Chang, N.-B., Jeng, K.-Y., Tseng, Y.-H., 2006. Soil erosion and non-point sourcepollution impacts assessment with the aid of multi-temporal remote sensing images.J. Environ. Manag. 79, 88–101.

Paustian, K., Six, J., Elliott, E., Hunt, H., 2000. Management options for reducing CO2 emis-sions from agricultural soils. Biogeochemistry 48, 147–163.

Polyakov, V., Lal, R., 2004. Modeling soil organic matter dynamics as affected by soil watererosion. Environ. Int. 30, 547–556.

Post, W.M., Kwon, K.C., 2000. Soil carbon sequestration and land-use change: processesand potential. Glob. Chang. Biol. 6, 317–327.

Qualls, R.G., Haines, B.L., 1992. Biodegradability of dissolved organic matter in forestthroughfall, soil solution, and stream water. Soil Sci. Soc. Am. J. 56, 578–586.

Quansah, C., 1981. The effect of soil type, slope, rain intensity and their interactions onsplash detachment and transport. J. Soil Sci. 32, 215–224.

Rasiah, V., Kay, B., Perfect, E., 1992. Evaluation of selected factors influencing aggregatefragmentation using fractal theory. Can. J. Soil Sci. 72, 97–106.

Reicosky, D., Kemper, W., Langdale, G., Douglas Jr., C., Rasmussen, P., 1995. Soil organicmatter changes resulting from tillage and biomass production. J. Soil Water Conserv.50, 253–261.

Rimal, B.K., Lal, R., 2009. Soil and carbon losses from five different landmanagement areasunder simulated rainfall. Soil Tillage Res. 106, 62–70.

Rumpel, C., Kögel-Knabner, I., 2011. Deep soil organic matter—a key but poorly under-stood component of terrestrial C cycle. Plant Soil 338, 143–158.

Shi, Z.-H., Yan, F.-L., Li, L., Li, Z.-X., Cai, C.-F., 2010. Interrill erosion from disturbed and un-disturbed samples in relation to topsoil aggregate stability in red soils from subtrop-ical China. Catena 81, 240–248.

Shi, Z., Yue, B., Wang, L., Fang, N., Wang, D., Wu, F., 2013. Effects of mulch cover rate oninterrill erosion processes and the size selectivity of eroded sediment on steep slopes.Soil Sci. Soc. Am. J. 77, 257–267.

Singer, M.J., Shainberg, I., 2004. Mineral soil surface crusts and wind and water erosion.Earth Surf. Process. Landf. 29, 1065–1075.

Six, J., Elliott, E., Paustian, K., 2000. Soil macroaggregate turnover and microaggregate for-mation: a mechanism for C sequestration under no-tillage agriculture. Soil Biol.Biochem. 32, 2099–2103.

Stewart, C., Paustian, K., Conant, R., Plante, A., Six, J., 2007. Soil carbon saturation: concept,evidence and evaluation. Biogeochemistry 86, 19–31.

Tipping, E., Woof, C., Rigg, E., Harrison, A.F., Ineson, P., Taylor, K., Benham, D., Poskitt, J.,Rowland, A.P., Bol, R., Harkness, D.D., 1999. Climatic influences on the leaching of dis-solved organic matter from upland UKmoorland soils, investigated by a field manip-ulation experiment. Environ. Int. 25, 83–95.

Van Oost, K., Quine, T.A., Govers, G., De Gryze, S., Six, J., Harden, J.W., Ritchie, J.C., McCarty,G.W., Heckrath, G., Kosmas, C., Giraldez, J.V., da Silva, J.R.M., Merckx, R., 2007. The im-pact of agricultural soil erosion on the global carbon cycle. Science 318, 626–629.

Walkley, A., Black, I.A., 1934. An examination of the Degtjareff method for determiningsoil organic matter, and a proposed modification of the chromic acid titrationmethod. Soil Sci. 37, 29–38.

Wang, L., Shi, Z., Wang, J., Fang, N.,Wu, G., Zhang, H., 2014. Rainfall kinetic energy control-ling erosion processes and sediment sorting on steep hillslopes: a case study of clayloam soil from the Loess Plateau, China. J. Hydrol. 512, 168–176.

Wischmeier, W., Mannering, J., 1969. Relation of soil properties to its erodibility. Soil Sci.Soc. Am. J. 33, 131–137.

Wuddivira, M.N., Stone, R.J., Ekwue, E.I., 2009. Clay, organic matter, and wetting effects onsplash detachment and aggregate breakdown under intense rainfall. Soil Sci. Soc. Am.J. 73, 226–232.

Wuest, S., Williams, J., Johlke, T., 2006. Effects of tillage on water infiltration. ColumbiaBasin Agric. Res. Annu. Rep. 1068, 86–92.

Zhang, J., Quine, T.A., Ni, S., Ge, F., 2006. Stocks and dynamics of SOC in relation to soil re-distribution by water and tillage erosion. Glob. Chang. Biol. 12, 1834–1841.

Zhang, X., Li, Z., Tang, Z., Zeng, G., Huang, J., Guo, W., Chen, X., Hirsh, A., 2013. Effects ofwater erosion on the redistribution of soil organic carbon in the hilly red soil regionof southern China. Geomorphology 197, 137–144.