Spatial distribution of soil organic carbon and analysis of related

factors in croplands of the black soil region, Northeast China

Dianwei Liu a,b, Zongming Wang b,*, Bai Zhang b, Kaishan Song b,Xiaoyan Li b, Jianping Li b, Fang Li b, Hongtao Duan b

aCollege of Geo-Exploration Science and Technology, Jilin University, Changchun, Jilin Province 130026, ChinabDepartment of RS and GIS, Northeast Institute of Geography and Agricultural Ecology,

Chinese Academy of Sciences, Changchun, Jilin Province 130012, China

Received 8 January 2005; received in revised form 29 August 2005; accepted 7 September 2005

Available online 2 November 2005

Abstract

Little is known concerning the spatial variability of soil organic carbon (SOC) and the relationship between SOC and landscape aspects, in

the black soil region of Northeast China, at county level. For this reason, the spatial characteristics of soil organic carbon and related factors,

i.e. land use, topography, and soil type, etc., were explored using GIS and geostatistics, taking Dehui County, Northeast China, as a study area.

Soil organic carbon in topsoil samples were taken at 354 locations in croplands of Dehui County. SOC concentrations follow a log-normal

distribution, with an arithmetic mean of 1.61% and geometric mean of 1.55%. The experimental variogram of SOC has been fitted with an

exponential model. Lower SOC concentrations were associated with larger gradient. Chernozems have the highest SOC concentrations, and

those under aeolian soils have the lowest SOC values. The spatial distribution pattern of SOC concentrations interpolated by Kriging,

indicated that after being divided into two parts along the Yinma River, samples in the western part have statistically higher SOC contents than

those in the eastern part. This pattern is approximately consistent with the spatial structure of topography and land use type.

# 2005 Elsevier B.V. All rights reserved.

Keywords: Soil organic carbon; Geostatistics; GIS; Spatial distribution; Black soil region; Northeast China

www.elsevier.com/locate/agee

Agriculture, Ecosystems and Environment 113 (2006) 73–81

1. Introduction

Food security and sustainable development are two

fundamental and strategic goals in China. Among the factors

that may heavily affect these two goals, land degradation is a

crucial one (Huang and Rozelle, 1998; Hubacek and Sun,

2001). China’s food security can be threatened by losses of

cultivated land due to disasters, soil erosion, and chemical

and physical deterioration. Although there have been

controversies over food demand and supply in China for

the next 30 years, there is an agreement that loss of arable

land and land degradation are undermining China’s food

production capacity (Gardner, 1996; Rozelle and Huang,

* Corresponding author. Tel.: +86 431 5542230; fax: +86 431 5542298.

E-mail address: zongmingwang@neigae.ac.cn (Z. Wang).

0167-8809/$ – see front matter # 2005 Elsevier B.V. All rights reserved.

doi:10.1016/j.agee.2005.09.006

1997). Agricultural over-exploitation and industrial pollu-

tion also exacerbate these degradation problems.

In recent years, considerable interests have been

generated in assessment of the physical, chemical, and

biological quality of agricultural soils (Carter et al., 1997;

Haynes et al., 2003). Soil organic carbon (SOC) is a dynamic

component of terrestrial systems, with both internal changes

and external exchanges with the atmosphere and the

biosphere (Zhang and McGrath, 2004). SOC plays an

important role in enhancing crop production (Stevenson and

Cole, 1999) and mitigating greenhouse gas emissions (Lal

et al., 1995; Flach et al., 1997; Post and Kwon, 2000). Like

other soil properties, SOC levels exhibit variability as a

result of dynamic interactions between parent material,

climate and geological history, on regional and continental

scale (Wang et al., 2001). However, landscape attributes

including slope, aspect, elevation, and land use may be the

D. Liu et al. / Agriculture, Ecosystems and Environment 113 (2006) 73–8174

Fig. 1. Soil sampling locations in Dehui County, Northeast China (n = 354).

dominant factors of SOC in an area with the same parent

material and single climate regime (Rezaei and Gilkes,

2005). Landscape attributes affect organic activity, run-off

and run-on processes, condition of natural drainage, and

exposure of soil to wind and precipitation (Buol et al., 1989).

The SOC content in cropland is also strongly dependent

upon crop and soil management practices, such as crop

species and rotation, tillage methods, fertilizer rate, manure

application, pesticide use, irrigation, and drainage, and soil

and water conservation (Batjes, 1998; Bruce et al., 1999; Lal

et al., 1999; Bergstrom et al., 2001; Lal, 2002; Heenan et al.,

2004). All these practices control the SOC input from crop

residue and addition of organic amendments, and the SOC

output through decomposition into gases and transportation

into aquatic ecosystems via leaching, run-off, and erosion

(Turner and Lambert, 2000).

To study the relationship between SOC and these factors

and to quantify the spatial distribution patterns of SOC,

statistics and geostatistics have been applied widely (Van

Meirvenne et al., 1996; Saldana et al., 1998; Chevallier et al.,

2000; Frogbrook and Oliver, 2001; McGrath and Zhang,

2003; Zhang and McGrath, 2004). Based on the theory of a

‘‘regionalized variable’’ (Matheron, 1963; Goovaerts, 1997;

Webster and Oliver, 2001), geostatistics provides advanced

tools to quantify the spatial features of soil parameters and to

carry out spatial interpolation. Geographic information

systems (GIS) are useful to produce the interpolated maps

for visualization, and for raster GIS maps, algebraic

functions can calculate and visualize the spatial differences

between the maps.

A better understanding of the spatial variability of SOC is

important for refining agricultural management practices

and for improving sustainable land use (McGrath and

Zhang, 2003). It provides a valuable base against which

subsequent and future measurements can be evaluated. The

black soil region of Northeast China is located in the central

part of the Northeast Plain, as one of the main agricultural

regions in China, its crop sown area and total yield now

accounting for 12 and 16%, respectively, of the nation’s

total. The black soil before cultivation has a high organic

matter content due to its parent material, climate, and natural

vegetation characteristics. However, croplands in this region

have been affected by human-induced degradation to a

serious degree after about 300 years of tillage. After long-

term reclamation and tillage, SOC in this region has declined

much. Conventional tillage increases aeration and breaks up

aggregates exposing organic matter to microbial attack that

was previously physically protected and it can also favor

losses of soil through erosion (Haynes and Beare, 1996).

In the black soil region of Northeast China, many efforts

have beenmade to analyze the stock of soil organic carbon in

croplands, the distribution of SOC in organic–mineral

complex, the dynamics of SOC, and effects of cultivation on

SOC (Xin et al., 2002; Fang et al., 2003; Liang et al., 2000;

Yu et al., 2004; Zhao et al., 2005). However, in this region,

the spatial variability of SOC and influences of natural and

anthropogenic factors on SOC have received very little

attention. The objectives of this study were to investigate (1)

the spatial distribution characteristics of soil organic carbon

in croplands of Dehui County, a typical agricultural area in

Northeast China, and (2) the possible factors influencing

SOC stocks, such as soil type, slope and topography, with the

help of GIS and geostatistics.

2. Materials and methods

2.1. Study area

Dehui County (1258450–1268230E, 438320–448450N) is

located in the middle part of Jilin Province, Northeast China

(see Fig. 1). The county has an altitude between 149 and

241 m with an area of 3435 km2. The study area is

characterized with a temperate, semi-humid continental

monsoon climate. The mean annual temperature is about

4.4 8C and the average annual precipitation is 520 mm. The

average of sunshine each year is 2695 h and average wind

speed is about 3.2 m s�1. The frost-free period is about 130–

140 days. In this county, The Second Songhua River, the

Yitong River, and the Yinma River flow through the area and

then into the Songhua River. The soils are black soil (Luvic

Phaeozem, FAO), chernozem (Haplic Chernozem, FAO),

meadow soil (Eutric Vertisol, FAO), and aeolian soil

(Arenosol, FAO).

D. Liu et al. / Agriculture, Ecosystems and Environment 113 (2006) 73–81 75

2.2. Soil sampling and analysis of SOC

As an agricultural county, about more than 80% of the

total area of Dehui County is used as cropland. In this study,

the SOC data were collected in a regional soil fertility

investigation. A maximum of two sites were selected at

random from each grid with area of 10 km2. Samples of 0–

20 cm depth from 354 sites were taken in November 2003.

Among 354 points, 272 locations are for dry farming land

under maize, and 82 locations are for paddy fields. The five

replicate samples were homogenized by hand mixing and

were sieved for the determination of SOC after being air-

dried. SOC was determined by the Walkey–Black method

(Schnitzer, 1982). The locations of the cropland sampling

sites are shown in Fig. 1.

2.3. Statistical and geostatistical methods

Some main statistical parameters, which are generally

accepted as indicators of the central trend and of the data

spread, were analyzed. They include description of the

mean, standard deviation, variance, coefficients of variation,

and extreme maximum and minimum values. To decide

whether or not data follow the normal frequency distribu-

tion, it may be sufficient to examine the coefficients of

skewness and kurtosis (Paz-Gonzalez et al., 2000). For a

population that follows the normal frequency distribution,

these coefficients should have values of 0 and 3, respectively.

These statistical parameters were calculated with EXCEL

2000 and SPSS 8.0.

Geostatistics (Matheron, 1963) uses the semi-variogram

to quantify the spatial variation of a regionalized variable,

and provides the input parameters for the spatial interpola-

tion method of Kriging (Krige, 1951). The semi-variogram

is half the expected squared difference between paired data

values z(x) and z(x + h) to the lag distance h, by which

locations are separated (Webster and Oliver, 2001):

rðhÞ ¼ 1

2E½zðxÞ � zðxþ hÞ�2 (1)

For discrete sampling sites, such as sampled in our study,

the function is usually written in the form:

rðhÞ ¼ 1

2NðhÞXNðhÞ

i¼1

½zðxi � zðxi þ hÞ�2 (2)

where z(xi) is the value of the variable z at location of xi, h the

lag and N(h) the number of pairs of sample points separated

by h. For irregular sampling, it is rare for the distance

between the sample pairs to be exactly equal to h. That

is, h is often represented by a distance band.

The experimental variogram is calculated for several lag

distances. This is then generally fitted with a theoretical

model, such as spherical, exponential, and Gaussian models.

The models provide information about the spatial structure

as well as the input parameters for the Kriging interpolation.

Kriging is considered as an optimal method of spatial

prediction. It is a theoretical weighted moving average:

zðx0Þ ¼Xn

i¼1

lizðxiÞ (3)

where zðx0Þ is the value to be estimated at the location of x0,

z(xi) the known value at the sampling site xi and n is the

number of sites within the search neighbourhood used for

the estimation. The number n is based on the size of the

moving window and is defined by the user. Kriging is

different from other methods (such as inverse distance

weighting), because the weight, li, is no longer arbitrary.

The weights depend on the parameters of the variogram

model and the sampling configuration and are decided under

the conditions of unbiasedness and minimized estimation

variance.

The geostatistical analyses were carried out with GS+

(Version 3.1a Demo), and maps were produced with GIS

software ArcView (Version 3.2a) and its extension module

of Spatial Analyst (Version 2.0).

3. Results and discussion

3.1. Descriptive statistics

The histograms of SOC with a normal distribution curve

of both the raw and the logarithmically transformed data are

shown in Fig. 2. It is clear that the statistical distribution of

the raw data of SOC is positively skewed, but the log-

transformed data are near normal.

In general, a P–P plot shows the SOC concentration’s

cumulative proportions against the cumulative proportions

of the normal distribution. These probability plots are

generally used to determine whether the distribution of a

variable matches the normal distribution. If it does, the

points cluster around a straight line. In our study, results

showed that there is a significant deviation from the straight

line for raw data of SOC. However, the log-transformed data

are close to the straight line. This implies that the SOC

concentration in croplands of the study area generally follow

a log-normal distribution.

We also analyzed the quantitative parameters of the

probability distribution and the significance level of the

Kolmogorov–Smirnov test for conformance to a normal

distribution for the variables. The probability distribution of

SOC is positively skewed (skewness = 2.60) and has a sharp

peak (kurtosis = 9.82). The log-transformed data have rather

small skewness (1.03) and kurtosis (3.07), and pass the K–S

normal distribution test at a significance level of higher than

0.05 (P = 0.000).

The coefficient of variation, standard deviation, and basic

statistical parameters of percentiles and means are shown in

Table 1. SOC has a relatively higher C.V. (29.7%), which

could be linked to the heterogeneity of land use pattern,

D. Liu et al. / Agriculture, Ecosystems and Environment 113 (2006) 73–8176

Table 1

Coefficient of variation (C.V.), standard deviation (S.D.), and basic statistic

parameters of SOC concentrations (n = 354)

Variables SOC (%)

C.V. 29.7

S.D. 0.48

Min 0.780

5% 1.11

25% 1.35

Median 1.52

75% 1.72

95% 2.47

Max 4.43

Mean 1.61

GeoMean 1.55

Fig. 2. (a and b) Histograms of SOC concentration in soils of croplands in Dehui County, Northeast China.

fertilizer, or soil erosion. We can see that SOC concentration

ranges from 0.78 to 4.43%, with the arithmetic mean of

1.61%. The geometric mean and median of SOC are 1.55

and 1.52%, respectively. Because SOC has the log-normal

feature here, we argue that the geometric mean and median

are more representative for the mean value of SOC than

arithmetic mean.

3.2. Analysis of spatial dependence of SOC

For SOC, the semivariogram model and best-fit model

parameters are shown in Table 2. SOC shows a positive

nugget, which can be explained by sampling error, short

range variability, random, and inherent variability. In

general, the nugget-to-sill ratio can be used to classify the

spatial dependence of soil properties (Cambardella et al.,

1994). The variable is considered to have a strong spatial

Table 2

Parameters for variogram model for LnSOC

LnSOC

Model Exponential

Range (km) 632

Nugget 0.036

Sill 0.224

Nugget/sill (%) 16.1

dependence if the ratio is less than 0.25, and has a moderate

spatial dependence if the ratio is between 0.25 and 0.75;

otherwise, the variable has a weak spatial dependence. In

our study, the nugget-to-sill ratio showed a strong spatial

dependence for soil organic carbon (the ratio is 0.16), which

might be attributed to intrinsic (soil-forming processes)

and extrinsic factors (soil fertilization and cultivation

practices).

3.3. Analyses of factors affecting SOC content

3.3.1. Effect of elevation

Since elevation was also able to explain much of the

variability in soil carbon concentrations (Powers and

Schlesinger, 2002). We used GIS software ArcView# to

analyze the spatial distribution of SOC with different

elevation. The samples were assigned to three elevation

groups: 160 (140–180), 200 (181–210), and 240 m

(>210 m), based on which contour line the sampling

location is close to.

The box-plot (Fig. 3) shows the difference of SOC

concentrations among these three groups. In each box-plot,

Fig. 3. Box-plot of SOC in different elevation groups (n = 181, 141, and 32,

respectively).

D. Liu et al. / Agriculture, Ecosystems and Environment 113 (2006) 73–81 77

Table 3

Results of post hoc tests in ANOVA with Duncan’s method (with mean

values of LnSOC in each evaluation group)

Elevation

group (m)

n S.D. Subset 1 Subset 2

160 181 0.255 0.467

200 141 0.251 0.431 0.431

240 32 0.186 0.369

0.141 0.221

Table 5

Results of post hoc tests in ANOVA with Duncan’s method (with mean

values of LnSOC in each soil type group)

Soil type group n S.D. Subset 1 Subset 2 Subset 3

Black soil 99 0.264 0.468

Chernozem 24 0.249 0.566

Meadow soil 155 0.246 0.461

Aeolian soil 76 0.199 0.336

Significant level 1.000 0.875 1.000

the lower boundary of the box shows the 25th percentile, and

the upper boundary shows the 75th percentile. The whiskers

are lines extending from the box to the highest and lowest

values, and the line across the box indicates the median. This

figure showed that the SOC concentrations in the 240 m

(>210 m) elevation group are the lowest, and those in the

160 m (140–180 m) group are higher.

To find whether the differences of SOC concentrations

among the elevation groups are significant, analysis of

variance (ANOVA) was applied, and analysis of the post hoc

test with Duncan’s test (Duncan, 1955) was conducted, as

shown in Table 3. The Kolmogorov–Smirnov test has shown

that all the groups have passed the test for normality

(P > 0.05). The Levene test showed that the variances

between the groups of the data set are homogenous at the

significance level of 0.222 and thus Duncan’s test can be

applied. The three groups of samples can be separated into

two subsets. The first subset contains samples with the

elevation groups of 160 and 200 m, while the second subset

consists of 200 and 240 m elevation groups. The differences

within either of subsets are not significant at the level of

0.05, with significance levels of 0.402 and 0.141,

respectively. And the difference between the subsets is also

not statistically significant at the level of 0.090 (with an F-

value of 2.421). This result indicates that in our study area,

elevation is not a main determinant of spatial distribution of

SOC concentrations.

3.3.2. Impact of slope on SOC

In the black soil region of Northeast China, soil erosion

is considered one of the important factors affecting the

SOC decline (Tang, 2004). To explore this, the difference

between SOC concentrations in coplands under different

slope was carried out. ArcView software was used to

assign the samples to two slope groups: 1.58 (0–38) and4.08 (3–88), and a t-test was done for the log-transformed

data sets to compare mean values of the two groups

(Table 4).

Table 4

Results of Levene’s test and t-test between LnSOC under 0–3 slope degree and

Levene’s test for equality of varia

F Signific

Equal variances assumed 0.137 0.712

Equal variances not assumed

Results indicate that the variances between the two data

sets are homogenous at the significance level of 0.712, based

on results of the Levene test, a test for equality of variances.

Thus, the t-value of 2.893 with ‘‘equal variances assumed’’

was used. The significance level of 0.004 for the two-tailed

test shows that there is a significant difference between SOC

concentrations under 0–3 slope degree and those under >3

slope degree. In explanation, it can be assumed that

relatively steeper slope might result in more soil erosion,

which led to SOC decline.

3.3.3. Differences among SOC concentrations of soil

samples from different soil types

To find the effect of soil type on SOC concentrations,

comparison of SOC among soil samples from different soils

were conducted, as shown in Table 5.

Analysis of variance (ANOVA) indicates that the

variances between different soil type groups are homo-

genous at the significance level of 0.566 and thus Duncan’s

test be applied. The four soil type groups of samples can be

divided into three subsets, as shown in Table 7. The first

subset contains samples from the aeolian soil, the second

subset consists black soil and meadow soil groups, and the

third subset contains the chernozem soil type group. The

differences within either of subsets are not significant at the

0.05 level, with significance levels of 1.000, 0.875 and

0.141, respectively. However, the difference between the

subsets is statistically significant at the level of 0.000 (with

an F-value of 7.655). It can be found that in this area,

statistically samples under Chernozem have the highest SOC

concentrations, and those under aeolian soil have the lowest

SOC value. This result reflects the effect of soil parent

materials on condition of soil organic carbon.

3.3.4. Comparison of means of SOC concentration

under different land use types

The soil samples were classified into two land use types,

dry farming land and paddy field, to analyze effects of land

those under >3 slope degree

nces t-Test for equality of means

ance t Significance (two-tailed)

2.893 0.004

2.956 0.011

D. Liu et al. / Agriculture, Ecosystems and Environment 113 (2006) 73–8178

Table 6

Mean values of SOC content in western part and eastern part (%)

Western part Eastern part

n 180 174

Minimum 1.07 0.78

Maximum 4.43 3.63

Average 1.77 1.45

Median 1.61 1.40

Geometric mean 1.71 1.42

S.D. 0.53 0.35

Fig. 4. Spatial distribution map of SOC concentrations in Dehui County.

use pattern on SOC concentrations. Results of the Levene’s

test show that the variances between the two data sets are

homogenous at the significance level of 0.528. Thus, the t-

value of �1.126 with ‘‘equal variances assumed’’ was used.

The significance of 0.261 for the two-tailed test indicates

that there is no significant difference between SOC

concentrations under different land use types.

The effect of different land use type on organic matter

status is dependent on a balance between organic matter

inputs and the degradative effect of the way of tillage and

reaping. In general, paddy field has a higher SOC

concentration due to greater dry matter production than

dry farming land. The lower organic matter content under

maize reflects the lower organic matter inputs for the maize

crop. This occurs because of the wide spacing of maize

plants, their sparse root system and the removal of

substantial amounts of dry matter at harvest (Haynes and

Francis, 1993). However, in the study area, peasants

removed substantial amounts of dry matter at harvest,

which led to lower organic mater inputs for the paddy crop.

Meanwhile, from the early 1990s, more manure was applied

for maize crop because more attention was paid to the

importance of organic manure application for protection of

soil fertility. Furthermore, turnover of the maize root is

another notable source of organic matter except the

inorganic manure for the dry farming land growing maize

(Yang et al., 2003b, 2004).

3.3.5. Impact of fertilizer practices on SOC status

The level of soil organic matter could be explained by the

input of litter and organic manure, decomposition and loss

with soil erosion of soil organic carbon. Former researches

showed that addition of chemical fertilizer and manure could

increase SOM (Biederbeck et al., 1994; Campbell et al.,

1991). Chemical fertilizer can increase shoot and root

production of crop, subsequently increasing residue input

into soil, while manure contains material that is a precursor

to SOM and also provides available nutrients. This effect has

been realized and used for nearly 4000 years in China, Japan,

and Korea to restore soil fertility and get a satisfactory yield

(Dormaar et al., 1988).

The increase in SOM has been attributed to greater dry

matter production under inorganic fertilization leading to

higher inputs of carbon to the soil through increased root

mass, root turnover, stubble, and crop debris (Glendining

et al., 1996). The chemical fertilizer alone may have had

the effect of stimulating microbial activity in soil and

enhancing the decompostion of SOM in this area. However,

manure plus chemical fertilizer had larger effects than

chemical fertilizer only. Manure application and return of

straw residue appears effective in helping maintain the

organic carbon of cultivated land of this region (Wang

et al., 2004; Shi et al., 2005). Applying organic and

inorganic fertilizer together had a positive effect on organic

carbon over fertilizer alone. As a result, return of residue

cover is expected to be more populous in this region, which

should be of benefit in maintaining or improving soil

productivity. Application of manure and crop residues,

either alone or in combination with chemical fertilizer,

should have a positive benefit to maintain and restore soil

organic matter quality and quantity for croplands of black

soil region (Yang et al., 2003a; Wang et al., 2004; Zhuge

et al., 2005). The need for adoption of soil conservation

practices to reduce soil erosion and organic carbon loss is

increasing in this region.

3.4. Spatial distribution of SOC content

The parameters of the exponential model were used for

kriging to produce the spatial distribution map of SOC

content in soils of the study area. A search region of 12

nearest-neighbours was applied. For the spatial interpola-

tion, a cell size of 100 m � 100 m was chosen to divide the

study area into a grid system containing 243 rows and 200

D. Liu et al. / Agriculture, Ecosystems and Environment 113 (2006) 73–81 79

Table 7

Results of Levene’s test and t-test of SOC content for western part and eastern part

Levene’s test for equality of variances t-Test for equality of means

F Significance t Significance (two-tailed)

Equal variances assumed 10.748 0.001 6.548 0.000

Equal variances not assumed 6.592 0.000

Fig. 6. Land use map of Dehui County.

columns. The final result of this spatial interpolation process

was shown as Fig. 4.

From the spatial distribution map of SOC content, we can

see that SOC values are higher in western part than those in

eastern part, and we can also find that this trend is along with

the Yinma River. To explore the spatial difference of SOC in

the whole area, we divided the study area into two parts:

westernpart andeasternpartwith a linealong theYinmaRiver.

The soil sampleswere then separated into twosubsets too.The

sample sizeswere 180 and 174, respectively. Themeanvalues

of SOC in these two areas are shown in Table 6. Analysis

indicates that the average value of 1.77 in western part was

obviously higher than that of eastern part, 1.45. The relative

difference was 18.1%. Meanwhile, the relative difference of

the geometric mean of SOC was 16.9%. The t-test was then

used to further compare the differences of SOC in two parts of

the study area and the results are given in Table 7.

Levene’s test indicates that the variances between the two

data sets are not homogenous because the significance was

0.001 (<0.05). Thus, the t-test value of 6.592 with ‘‘equal

variances not assumed’’ was used. The significance level of

0.000 for the two-tailed test shows that there is significant

Fig. 5. Hillshade map indicating topography of Dehui County.

difference between SOC content of western part and eastern

part. It can be concluded that statistically samples of western

part have higher SOC content than those of eastern part.

Comparison between spatial distribution map of SOC and

slope map and land use map of the study area can give us the

information that the spatial distribution of SOC is generally

consistent with the topography (Fig. 5) and different land use

types (Fig. 6). In eastern part, water and tillage erosion

maybe the main reason for soil organic carbon decline due to

relatively bigger water and soil loss and little protection

tillage management.

Acknowledgements

This research was jointly supported by the Knowledge

Innovation Program of Chinese Academy of Sciences (No.

KZCX3-SW-356), the National Natural Science Foundation

of China (No. 40401003), the Knowledge Innovation

Program of Chinese Academy of Sciences (No. KZCX1-

SW-19), and the Foundation Item of Key Laboratory of

Ecological Restoration and Ecosystem Management of Jilin

Province (No. DS2004-03). We thank the editor of the

journal and two anonymous reviewers for their useful

comments and suggestions on an earlier draft. Special thanks

D. Liu et al. / Agriculture, Ecosystems and Environment 113 (2006) 73–8180

also go to Dr. Y.Z. Zhang for his zealous help in improving

English of the manuscript.

References

Batjes, N.H., 1998. Mitigation of atmospheric CO2 concentrations by

increased carbon sequestration in the soil. Biol. Fertil. Soils 27, 230–

235.

Bergstrom, D.W., Monreal, C.M., St. Jacques, E., 2001. Influence of tillage

practice on carbon sequestration is scale-dependent. Can. J. Soil Sci. 81,

63–70.

Biederbeck, V.O., Janzen, H.H., Campbell, C.A., Zentner, R.P., 1994. Labile

soil organic matter as influenced by cropping practices in an arid

environment. Soil Biol. Biochem. 26, 1647–1656.

Bruce, J.P., Frome, M., Haites, E., Janzen, H., Lal, R., Paustian, K., 1999.

Carbon sequestration in soils. J. Soil Water Conserv. 54, 382–389.

Buol, S.W., Hole, F.D., McCracken, R.J., 1989. Soil Genesis and Classi-

fication. Iowa State University Press, Ames, IA 50010.

Cambardella, C.A., Moorman, T.B., Novak, J.M., Parkin, T.B., Karlen,

D.L., Turco, R.F., Konopka, A.E., 1994. Field-scale variability of soil

properties in Central Iowa soils. Soil Sci. Soc. Am. J. 58, 1501–1511.

Campbell, C.A., Biederbeck, V.O., Zentner, R.P., LaFond, G.P., 1991. Effect

of crop rotations and cultural practices on soil organic matter, microbial

biomass and respiration in a thin Black Chernozem. Can. J. Soil Sci. 71,

363–376.

Carter, M.R., Gregorich, E.G., Anderson, D.W., Doran, J.W., Janzen, H.H.,

Pierce, F.J., 1997. Concepts of soil quality and their significance. In:

Gregorich, E.G., Carter, M.R. (Eds.), Soil Quality for Crop Production

and Ecosystem Health. Elsevier, Amsterdam, pp. 1–19.

Chevallier, T., Voltz, M., Blanchart, E., Chotte, J.L., Eschenbrenner, V.,

Mahieu, M., Albrecht, A., 2000. Spatial and temporal changes of soil C

after establishment of a pasture on a long-term cultivated vertisol

(Martinique). Geoderma 94, 43–58.

Dormaar, J.F., Lindwall, C.W., Kozub, G.C., 1988. Effectiveness of manure

and commercial fertilizer in restoring productivity of an artificially

eroded Dark Brown Chernozemic soil under dryland conditions. Can. J.

Soil Sci. 68, 669–679.

Duncan, D.B., 1955. Multiple range and multiple F tests. Biometrika 11, 1–

42.

Fang, H., Yang, X., Zhang, X., 2003. Organic carbon stock of black soils in

Northeast China and its contribution to atmospheric CO2. J. Soil Water

Conserv. 17 (3), 10–13 (in Chinese).

Flach, K.W., Barnwell Jr., Crosson, P., 1997. Impacts of agriculture on

atmospheric carbon dioxide. In: Paul, E.A., Paustian, K., Elliott, E.T.,

Cole, C.V. (Eds.), Soil Organic Matter in Temperate Agroecosystems,

Long-Term Experiments in North America. CRC Press, Boca Raton,

FL, pp. 3–13.

Frogbrook, Z.L., Oliver, M.A., 2001. Comparing the spatial predictions of

soil organic matter determined by two laboratory methods. Soil Use

Manag. 17, 235–244.

Gardner, G., 1996. Shrinking Fields: Cropland Loss in a World of Eight

Billion. Worldwatch Paper Worldwatch Institute, Washington, DC.

Glendining, M.J., Powlson, D.S., Poulton, P.R., Bradbury, N.J., Palazzo, D.,

Li, X., 1996. The effects of long-term applications of inorganic nitrogen

fertilizer on soil nitrogen in the Broadbalk wheat experiment. J.

Agricult. Sci. 127, 347–363.

Goovaerts, P., 1997. Geostatistics for Natural Resources Evaluation. Oxford

University Press, New York, USA.

Haynes, R.J., Francis, G.S., 1993. Changes in microbial biomass C, soil

carbohydrate composition and aggregate stability induced by growth of

selected crop and forage species under field conditions. J. Soil Sci. 44,

665–675.

Haynes, R.J., Beare, M.H., 1996. Aggregation and organic matter storage in

meso-thermal, humid soils. In: Carter, M.R., Stewart, B.A. (Eds.),

Advances in Soil Science. Structure and Organic Matter Storage in

Agricultural Soils. CRC Press, Boca Raton, pp. 213–262.

Haynes, R.J., Dominy, C.S., Graham,M.H., 2003. Effect of agricultural land

use on soil organic matter status and the composition of earthworm

communities in KwaZulu-Natal, South Africa. Agricult. Ecosys.

Environ. 95, 453–464.

Heenan, D.P., Chan, K.Y., Knight, P.G., 2004. Long-term impact of rotation,

tillage and stubble management on the loss of soil organic carbon and

nitrogen from a Chromic Luvisol. Soil Tillage Res. 76, 59–68.

Huang, J., Rozelle, S., 1998. China’s Grain Economy to the Twenty-First

Century. Agriculture Press, Beijing, China.

Hubacek, K., Sun, L.X., 2001. A scenario analysis of China’s land use and

land cover change: incorporating biophysical information into input–

output modeling. Struct. Change Econ. Dyn. 12, 367–397.

Krige, D.G., 1951. A statistical approach to some basic mine valuation

problems on the Witwatersrand. J. Chem. Metall. Mining Soc. South

Africa 52 (6), 119–139.

Lal, R., Kimble, J.M., Levine, E., Whitman, C., 1995. World soils and

greenhouse effect: an overview. In: Lal, R., Kimble, J.M., Levine, E.,

Stewart, B.A. (Eds.), Soils and Global Change. CRC Press, Boca Raton,

FL, pp. 1–8.

Lal, R., Follett, R.E., Kimble, J., Cole, C.V., 1999. Managing U.S. cropland

to sequester carbon in soil. J. Soil Water Conserv. 54, 374–381.

Lal, R., 2002. Soil carbon dynamics in cropland and rangeland. Environ.

Pollut. 116, 353–362.

Liang, W., Wen, D., Li, W., Guan, Z., 2000. Effects of cultivation on soil

organic carbon dynamics in agro-ecosystems in black soil region. Syst.

Sci. Compr. Stud. Agricult. 16 (4), 241–244 (in Chinese).

Matheron, G., 1963. Principles of geostatistics. Econ. Geol. 58, 1246–1266.

McGrath, D., Zhang, C.S., 2003. Spatial distribution of soil organic carbon

concentrations in grassland of Ireland. Appl. Geochem. 18, 1629–1639.

Paz-Gonzalez, A., Vieira, S.R., Castro, T., 2000. The effect of cultivation on

the spatial variability of selected properties of an umbric horizon.

Geoderma 97, 273–292.

Post, W.M., Kwon, K.C., 2000. Soil carbon sequestration and land-use

change: processes and potential. Global Change Biol. 6, 317–327.

Powers, J.S., Schlesinger, W.H., 2002. Relationships among soil carbon

distributions and biophysical factors at nested spatial scales in rain

forests of northeastern Costa Rica. Geoderma 109 (3–4), 165–190.

Rezaei, S.A., Gilkes, R.J., 2005. The effects of landscape attributes and

plant community on soil chemical properties in rangelands. Geoderma

125, 167–176.

Rozelle, V., Huang, J., 1997. The impact of environmental degradation on

grain production in China: 1975–1990. Econ. Geogr. 73 (1), 44–66.

Saldana, A., Stein, A., Zinck, J.A., 1998. Spatial variability of soil proper-

ties at different scales within three terraces of the Henare River (Spain).

Cantena 33, 139–153.

Schnitzer, M., 1982. Total carbon, organic matter, and carbon. In: Page,

A.L., Miller, R.H., Keeney, D.R. (Eds.), Methods of Soil Analysis. Part

2, 2nd Edition. Agronomy Monograph, vol. 9. American Society of

Agronomy, Madison, WI, pp. 539–577.

Shi, Y., Zhang, L., Chen, X., Yu, W., 2005. Effect of different management

patterns on the fraction of water stable aggregates and accumulation and

distribution of particulate organic matter in black soil. Chinese J. Eco-

Agricult. 13 (2), 122–124 (in Chinese).

Stevenson, F.J., Cole, M.A., 1999. Cycles of Soil, second ed. Wiley, New

York.

Tang, K.L., 2004. Soil and Water Conservation in China. Science Press of

China, Beijing, China (in Chinese).

Turner, J., Lambert, M., 2000. Change in organic carbon in forest plantation

soils in eastern Australia. Forest Ecol. Manag. 133, 231–247.

Van Meirvenne, M., Pannier, J., Hofman, G., Louwagie, G., 1996. Regional

characterization of the long-term change in soil organic carbon under

intensive agriculture. Soil Use Manag. 12, 86–94.

Wang, J., Fu, B., Yang, Q., Chen, L., 2001. Soil nutrients in relation to land

use and landscape position in the semi-arid small catchment on the loess

plateau in China. J. Arid Environ. 48, 537–550.

D. Liu et al. / Agriculture, Ecosystems and Environment 113 (2006) 73–81 81

Wang, T., Yan, L., Guan, L., Wang, J., Lu, Y., 2004. Variations of organic

matter with long-term location monitoring in black soil. J. Agro-

Environ. Sci. 23 (1), 76–79 (in Chinese).

Webster, R., Oliver, M., 2001. Geostatistics for Environmental Scientists.

John Wiley & Sons, Chichester, UK.

Xin, G., Yan, L., Wang, J., Guan, L., 2002. Changes of organic carbon in

black soils with the different reclamation year. Chinese J. Soil Sci. 335

(5), 332–335 (in Chinese).

Yang, X., Zhang, X., Deng, W., Fang, H., 2003a. Black soil degradation

by rainfall erosion in Jilin, China. Land Degradation Dev. 14, 409–

420.

Yang, X., Zhang, X., Fang, H., 2003b. Simulating effects of long-term

fertilization on changes of soil organic carbon under continuous cultiva-

tion of corn in Northeast China. Scientia Agricltura Sinica. 36 (11),

1318–1325 (in Chinese).

Yang, X., Zhang, X., Fang, H., Liang, A., 2004. Changes in organic matter

and total nitrogen of black soils in Jilin Province over the past two

decades. Scientia Geographica Sinica. 24 (6), 710–714 (in Chinese).

Yu, J., Liu, J., Wang, J., Liu, S., Qi, X., Wang, Y., Wang, G., 2004. Organic

carbon variation law of black soil during different tillage period. J. Soil

Water Conserv. 18 (1), 27–30 (in Chinese).

Zhao, L., Wang, J., Liu, J., Liu, S., Wang, Y., Wang, H., Zhang, Z., 2005.

Humus composition of black soil and its organic-mineral complex under

different fertility level. Chin. J. Appl. Ecol. 15 (1), 93–99 (in Chinese).

Zhang, C.S., McGrath, D., 2004. Geostatistical and GIS analyses on soil

organic carbon concentrations in grassland of southeastern Ireland from

two different periods. Geoderma 119, 261–275.

Zhuge, Y., Zhang, X., Lu, Q., 2005. Effect of long-term fertilization on

respiration process of black soil. Chin. J. Soil Sci. 36 (3), 391–394 (in

Chinese).